Classification is the practice of arranging items into groups or categories, and a classification is the resulting arrangement. Taxa (singular taxon) are groups of organisms recognized in a classification and given biological names (e.g., Salmoniformes, Salmonidae, Oncorhynchus, Oncorhynchus nerka). A category is the level or rank at which the taxon is placed (e.g., order, family, genus, species). Generally, the objective in constructing a classification of a group of organisms is to show in a hierarchical system the relationships of the various taxa. We may agree that the kind of relationship we wish to show, as best we can in a listing of names, is an evolutionary one. However, there have been differences of opinion as to what evolutionary relationship means and how it should be determined, and there have also been different ways of expressing evolutionary relationships or phylogeny in a classification. Students working with older literature must be aware of these differences. Classifications are now based, when possible, on postulated genealogical branching points (the cladistic methodology), as attempted in the previous edition (Nelson, 1994) and in this edition. The classification in Nelson (1976, 1984) also considered degrees of divergence.
The fundamental unit in a biological classification is the species, and those involved with constructing classifications must deal with species definitions. I prefer the biological species concept (as a concept, not usually a working definition) (e.g., Nelson, 1999) and regard the species as the only taxonomic unit with evolutionary reality. It is inferred to represent an irreversible evolutionary discontinuity. Of course, in any work such as this, it is not possible in giving estimates of species numbers to expect uniformity of species concepts between workers on different families. Some taxonomists in their revisionary work may adhere to the biological species concept and some may not. Definitions that recognize a species as any terminal clade or as any genetically distinct population would, of course, result in a marked but artificial increase in species numbers that I would prefer not to recognize. Such definitions may result in potentially relatively unstable evolutionary units, and this seems to me to be undesirable to employ in management, systematics, and biology in general (other effective means exist for recognizing differences within a species for any given objective—see also above under "Numbers").
The science of systematics, in studying the relationships of species, is the study of the diversity of organisms in order to understand the evolutionary history of life (practice, methods, and principles thereof). Biological classification is based on systematic studies. Taxonomy is that part of systematics dealing with the theory and practice of describing diversity and erecting classifications. During the past few decades there has been an impressive accumulation of information on extant and fossil material and on morphological and molecular-based phylogenies. More work is needed on species diversity and on analyzing various characters to determine homologies before we reach a sound understanding of how evolution has produced the diversity of fishes that exists. Numerous families of fishes are very poorly classified. In addition, cladograms produced by employing molecular characters and their comparison with morphologically based cladograms promise to give us new insights to aid in our understanding of relationships. Although there is general agreement on many aspects of fish classification, there is also much disagreement. Conflict is especially prominent between some morphological and molecular-based phylogenies, although it is encouraging to see so much agreement in some areas. Reference is made throughout the classification to differences in some of the morphological and molecular-based classifications (see also, for example, under Acanthopterygii). It was not possible to refer to all relevant literature, and in any event, it behooves nonsystematists relating their findings to systematic work to refer back to the primary literature.
The study of fish systematics—ichthyology in the limited sense of the word—has had a long and interesting history. The history of Canadian and American ichthyology is reviewed by J. R. Dymond, G. S. Myers, and C. L. Hubbs in Copeia of 1964 (No. 1). In 1997, T. W. Pietsch and W. D. Anderson, Jr., edited the book Collection Building in Ichthyology and Herpetology, published by the American Society of Ichthyologists and Herpetologists, which gave good insights into some of the giants of our past. Some of the history of collection building and the challenges facing natural history museums and biodiversity research in Asia are discussed by Matsuura (2000) and Akiyama et al. (2004). During the history of ichthyology, numerous classifications of fishes have been proposed throughout the world. Although our present classifications and methods are improvements over past ones, we should not forget that our current efforts are made far easier by the contributions of past biologists, often working under great difficulty, such as P. Artedi (considered by many as the "Father of Ichthyology"), J. Müller, L. Agassiz, M. E. Bloch, G. Cuvier, A. Valenciennes, P. Bleeker, T. N. Gill, B. A. Boulenger, A. Günther, D. S. Jordan, C. T. Regan, S. Tanaka, K. Matsubara, G. S. Myers, C. L. Hubbs, and D. E. Rosen. Thankfully, there are many active masters still with us. Many younger contemporary ichthyologists are making important contributions, but the field will remain rich in problems for future generations of researchers. Unfortunately, while there is a growing need for a young generation of taxonomists/systematists, there are concerns that this need will not be met unless there are changes in government and public support for future positions.
Students of ichthyology should know the principles and methodology of cladistic (= phylogenetic) systematics and classification, where, in simple terms, the systematist seeks to resolve which two taxa of a group of three or more are each other's closest genealogical relatives. A dichotomously branching cladogram (diagram) is constructed in which paired lineages, called sister groups, are recognized on the basis of sharing derived character states (termed synapomorphies, with a particular derived character state being termed apomorphic; plesiomorphies are the primitive states and do not indi cate the existence of sister groups). The sister group possessing more apo-morphic character states relative to the other is the derived group, while the other is the primitive one; each is given or is understood to have the same tax-onomic rank. A common source of disagreement is over which character states are apomorphic; consequently, a good understanding of the distribution of character states and homology is essential to a cladistic analysis. As with any approach, one must take care that characters are not arbitrarily chosen or their states arbitrarily polarized, consciously or subconsciously, for the express purpose of either producing a change in existing classification or supporting preconceived ideas of relationships (perhaps to provide systematic evidence to support a favored biogeographic hypothesis). In identifying sister groups, cladograms allow the systematist and nonsystematist to focus clearly on questions and test hypotheses concerning the evolution of given traits, whether morphological, behavioral, or physiological.
In a cladistic analysis, there is usually a clear presentation of the character states employed (but, unfortunately for those wishing to appraise the work, characters discarded from analysis are generally not given). Polarity of mor-phoclines or of character states is determined by evidence from ontogeny or, more usually, by reference to what is called the out-group (the nearest presumed related taxon or taxa—a character state widely distributed in related taxa is taken to be primitive) with the group under consideration being called the in-group. Computer programs assist in analyzing data to construct phylo-genetic trees (cladograms).
It is important for nonsystematists who rely on classifications in their studies to remember that in some cladistic studies new classifications are constructed on the basis of only a few weak synapomorphies. In addition, often not all species are examined, resulting in a poor knowledge of character distribution. Such practices are not likely to produce a sound and stable evolutionary classification (certainly not a utilitarian one), any more so than is a synthetic study based on ill-chosen characters or a phenetic study based on overall similarity. Apart from methodological problems or problems resulting from poor practice, there appears in some groups to be such a mosaic of character states of uncertain polarity that a stable cladistic analysis may be difficult to establish.
There are many problems in translating a phylogeny into a classification. Ideally, the classification is based solely on the hypothesized genealogical relations such that one is faithfully derivable from the other. Each taxon is strictly monophyletic, in that all groups sharing a common ancestry and only those groups, including the common ancestor itself, are included in the taxon. In this book, a cladistic classification is employed wherever I feel that there is reasonably sound phylogenetic information to present such a classification, whether based on molecular or morphological data. Where the evidence seems uncertain, I maintain the status quo. There are a great many groups that we know to be paraphyletic, but we lack sufficient evidence to erect monophyletic taxa.
I consider fossils to be critical in understanding evolutionary relationships. Unfortunately, the fossil record in fishes is very incomplete, and many decisions must be made without any evidence from fossils. However, we can answer many critical questions of interrelationships of higher taxa only with a study of new fossils and not, conclusively at least, from extant material. Fossils are ranked along with extant taxa in the classification of this book.
Was this article helpful?