The Killifish Master Index is organized into three main divisions:
the Annotated Checklist of Genera and Species;
the Literature Cited;
the Concordance of Synonyms, a cross-referenced tabulation of synonyms and their valid names.

The KMI is supplemented by three Appendices, and a Chart showing graphically the phenomenal increase in the number of published descriptions during the last quarter century. Appendix A is a ready reference guide to all the names appearing in the Annotated Checklist. Running counts are supplied for the genera and for the number of species in each genus. Appendix B lists all genera and their type species. Appendix C shows genera and species chronologically by the year in which they were described and is accompanied by a graphical analysis of the frequency of killifish descriptions.

The Annotated Checklist is arranged alphabetically by genus and species with each genus initiating its own section. All available genera and subgenera not sharing the same type species are considered potentially valid names. Genera and species appear in italics. Other uses of Latin are underlined, except for the titles of articles, as in the following example, “Bodenheimer, F. S. 1937. Prodromus faunae Palestinae. Mémoires Présenté a l'Institut d'Égypte 33: ii + 286 pp.” Original spellings of genera and species are used unless there are multiple spellings in the original description, in which case the spelling chosen by the first reviser is used. Abbreviations have not been used for genera, species, or in the Literature Cited, although in the latter case exceptions have been made for books and serial publications that use abbreviations in their official titles. Annotations appear between square brackets. Annotations which make original contributions to taxonomy and/or nomenclature appear in boldface type.

An example of a citation of a genus follows:

Adinia Girard, 1859.

Girard, C. 1859. Ichthyological notices. Proceedings of the Academy of Natural Sciences of Philadelphia 11: 113-122. [description: pp. 117-118]

Type species: Adinia multifasciata Girard, 1859b, by monotypy.

Adinia Girard, 1859b: Myers, 1931; Parenti, 1981; Wiley, 1986; Bernardi, 1997.

As general rule, the entry for a genus is arranged as follows: generic name; describer and year; publication; type species; synonyms; subgenera and their synonyms, all accompanied by pertinent references. In the case of Adinia, there being no synonyms or subgenera, its situation is fairly simple. Sometimes situations are complicated, as indicated by this partial excerpt for Xenisma, a subgenus of Fundulus:

Xenisma Jordan, 1877.

Jordan, D. S. 1877. Xenisma, X. stellifera, X. catenata, and footnote. p. 142 In: D. S. Jordan and H. E. Copeland, Checklist of the fishes of the fresh waters of North America (Concluded). Bulletin of the Buffalo Society of Natural Science 3 (4): 137-164. [Cyprinodontidae: pp. 141-143] [The first part of Jordan and Copeland's article, covering pp. 133-136 appeared in the Bulletin of the Buffalo Society of Natural Science vol. 3 no. 3 and is dated 1876 on the cover of no. 3. The conclusion of the article in which the section on the Cyprinodontidae appeared is dated 1877 on the cover of vol. 3 no. 4.]

The dating of the description of Xenisma requires a comment. The citations of the publication in which Xenisma was made available give 1876 as its date. Without the accompanying annotation, the correct date, 1877, would most likely be considered an error.

Genera sharing the same type species are objective synonyms. If, however, an author places the type species of a genus (or subgenus) into the species group of a recognized subgenus, then that genus (or subgenus) becomes a subjective synonym. This is the case in the following excerpt.

Synonym of Xenisma:

Gambusinus Jordan & Evermann, 1896.

Jordan, D. S. and B. W. Evermann. 1896. The fishes of North and Middle America. A descriptive catalog of the species of fish-like vertebrates found in the waters of North America, north of the Isthmus of Panama. Part I. Bulletin of the United States National Museum 47 (1): lx + 1240 pp. [name with type species p. 633; description: p. 635]

Type species: Fundulus rathbuni Jordan & Meek, 1889, by original designation. [According to Wiley (1986), F. rathbuni is a member of the subgenus Xenisma, thus making Gambusinus a subjective synonym of Xenisma.]

Gambusinus Jordan & Evermann, 1896.

The following represents a typical genus-species entry of the non-complicated sort.

Aphyolebias wischmanni (Seegers, 1983).

sandy longfin

Seegers, L. 1983. Pterolebias wischmanni nov. spec. aus dem Ucayali-Einzug in Peru (Pisces: Atheriniformes: Rivulinae) Deutsche Killifisch Gemeinschaft Journal 15 (5): 67-74, 7 photos, 3 figs., table, map. [color photo of P. wischmanni preceding article on p. 65]

Type locality: in a small brook between the rivers Chipiria and Amaquiria, upper Río Ucayali, about 120 km from Pucallpa, Peru.

Austrofundulus U-67 "Sandy Longfin": Goldstein, 1969a; Ricco, 1969a,b.

Pterolebias wischmanni Seegers, 1983b: Busse, 1984; Seegers, 1984a, 1987; Wischmann, 1984; Hutchings [L.], 1985a; Ortega & Vari, 1986; Hutchings & Hutchings, 1990b; Schindler & Staeck, 1993b; Brousseau, 1994a [egg incubation time, distribution map]; Costa, Sarmiento, & Barrera, 1996.

Aphyolebias wischmanni (Seegers, 1983): Costa, 1998d.

The sequence of information is genus, species, describer(s), common name (if any), original description [in a longer work, the specific pages on which the original description appears are given], type locality, and, lastly, a chronologically arranged nomenclatural history with pertinent references. According to the prevailing convention, “Seegers, 1983” appears within parentheses because the species wischmanni was originally described in a different genus. Should wischmanni revert to Pterolebias then the parentheses would be dropped. This species has a common name. When first discovered it was called, informally, Austrofundulus U-67 and also given a common name, the “sandy longfin,” in aquarium literature. In order to have nomenclatural standing a species name must appear as a two part name in Latin, genus first, then species, i.e., Linnaean binomial nomenclature. Austrofundulus U-67 does not qualify as a zoological name because the U-67 is a form of common name. Fourteen years after its introduction to the aquarium hobby, the sandy longfin was described as a new species, Pterolebias wischmanni, by Seegers in 1983. In 1998 Costa placed wischmanni into his newly erected genus Aphyolebias. Pertinent references are cited, no special annotations appear.

A number of complicating factors are liberally laced throughout the nomenclatural history of the killifishes, as with any other group.

The International Code of Zoological Nomenclature requires that the rules of Latin grammar must be applied to genus and species. If these rules are violated an annotation will be made, as in the case of “Aphyosemion cameronensis (Boulenger, 1903) [incorrect Latin ending]” which is found under the entry for Aphyosemion cameronense. Aphyosemion is neuter in gender and species names used as its adjectives must have neuter endings.

There are 706 species and subspecies in the Annotated Checklist. More than 400 misidentifications are listed. This should give pause for thought about the identity of any species appearing in any professional paper or hobbyist article. Under Aphyosemion cameronense, the following entry appears, “Aphyosemion striatum [misidentification]: Radda, 1970a; Seegers, 1988a…” Aphyosemion striatum is a valid species. According to Seegers (1988a), however, the Aphyosemion striatum appearing in Radda (1970a) is a misidentified Aphyosemion cameronense.

Also appearing under Aphyosemion cameronense is “Panchax microstomus Ahl, 1924d: Holly, 1930 [as a synonym of Panchax cameronensis]; Scheel, 1968a; Paepke & Seegers, 1986.” Panchax microstomus is a subjective synonym of Aphyosemion cameronense.

Finding a species can be a problem if it is currently considered a synonym. Aphyosemion melanopteron, long considered a valid species, will not be found alphabetically tucked between Aphyosemion meinkeni and Aphyosemion mimbon. However, if the Concordance of Synonyms is consulted, the species under which A. melanopteron may be found is seen to be A. congicum. The same possibility exists if one had only the common name of a species such as sandy longfin, or a non-Linnaean provisional name, such as Austrofundulus U-67, or Nothobranchius species Ruhoi. These may be found in the Concordance under sandy longfin, U-67, and Ruhoi, respectively. The best method for discovering the value of the Concordance of Synonyms is to browse its contents.

The conventions used in the Literature Cited are fairly standard but perhaps some aspects are worth noting. For example:

Boulenger, G. A. 1902d. Additions à la faune ichthyologique du bassin du Congo. Annales du Musée du Congo, Zoologie (1) 2 (2): 58 pp., plates 12-16. [no killifishes figured] [later known as Annales du Musée du Congo belge]

Volume numbers appear in bold face, series and issue numbers appear within parentheses. Series numbers precede volume numbers, while issue numbers follow them.

Brill, J. [S., Jr.] 1984. Profundulus punctatus– A rarely-seen New World killifish. Tropical Fish Hobbyist 32 (10) (June): 59-63, color photos.

Some serial publications have volume numbers which straddle two calendar years so that issue numbers do not correspond to the month in which they are issued. In those cases the month of the issue is given, as in the above Tropical Fish Hobbyist citation. In this way, a reader can see that issue 10 was not issued in the 10th month.

Canadian Endangered Species Fish Website. 1998.

http://www-nais.ccrs.nrcan.gc.ca/schoolnet/issues/risk/fish/efish/bndkilli.html

[GIF in color]

Web site addresses have been included in the Literature Cited, the date being the year in which the citation was first located. Since punctuation is part of a Web address, such citations do not end with a period. Square brackets and their contents are not part of the address.

The basic orientation of the literature cited is toward systematics, ecology, and zoogeography.

On The Origin Of The Term "Killifish"

According to the Oxford English Dictionary, killifish is a composite word "commonly supposed" to have been created by the suture of "kill," the Dutch word for stream, and “fish.”

As a boy growing up in New York City, killifish meant only one thing, a bait fish, the mummichog, and a word from which I was protected until I reached my majority– a fish, common though it may be, with a tortuous nomenclatural career.  That fish is either Fundulus heteroclitus heteroclitus or F. h. macrolepidotus, for both are found in the waters of Long Island.  The former never forms freshwater colonies, the latter does.  Only recently, a freshwater population was discovered in the Bronx River, Samaritan & Schmidt (1982).  Fundulus heteroclitus macrolepidotus is the fish that gave killifishes their name.

The earliest reference to the term killifish I have found is Schoepf (1788) spelled "killfish.”  Schoepf's "killfish" was specifically applied to what we refer to today as F. h. macrolepidotus.  In 1792 Walbaum described Cobitis killifish (= Fundulus h. macrolepidotus).  By 1815, killifish appears as a common name in Mitchill's "The fishes of New York."

But why does our etymological resource, the Oxford English Dictionary, hedge a little when it tells us the "commonly supposed" origin of the term killifish?  Well, as it turns out, there is another theory for its origin, one no more or no less plausible, related to the use of killifishes as bait fish.
According to the Oxford English Dictionary, based on documented usage from 1681 until 1880, a very effective bait was referred to as "killing," as in As killing a Bait as any whatever, and Fishing with the young frog is a very killing method of fishing for chub.  The transition from "killing bait" to "killing fish" to "killfish" to "killifish" is believable.

Geographically, the term "killifish" arose in the New York City area where killifishes are not stream fish.  In the south, Fundulus heteroclitus heteroclitus was called the "mudfish."  The term "kill" was not always used with great precision, however.  The water between Staten Island and New Jersey, known as the Arthur Kill, is not a stream, while Great Kills Harbor opens only into the Atlantic Ocean and these are areas where Fundulus heteroclitus macrolepidotus would abound.

The often complicated nomenclatural history of the two subspecies of Fundulus heteroclitus may be found under their respective entries.  Recognition of the subspecies macrolepidotus has been checkered until recently.

To some, the bait fish theory is bound to be a more interesting explanation, if only because it is so different, so unexpected an alternative.  Either way, those who have used the New York saltwater killifish, Fundulus heteroclitus macrolepidotus, to catch blue claw crabs or fluke or bluefish know it to be a killing bait indeed.
 

The higher order relationships within the Cyprinodontiformes are, in general, in a great state of flux. The following systematic overview is based primarily on Costa (1998f), representing his revision of the Cyprinodontiformes on the family level. There have been no challenges to the major division of the order into two suborders, the Aplocheiloidea and the Cyprinodontoidea. That is a stability of sorts.

In the Aplocheiloidea, the hypotheses of Murphy and Collier (1997) produced a very satisfying concordance with the geology of plate tectonics and the origins of annualism, but nomenclaturally this yielded a situation in which the family Aplocheilidae, if retained, would include Rivulus and Cynolebias and their allied genera as well. In harmony with Murphy and Collier (1997) and in order to retain the family Rivulidae and its division by Costa (1990a,b) into the Rivulinae and Cynolebiatinae, I have incorporated a coordinated rank structure using four families.

In the Rivulidae, proposed relationships are not stable. This may be seen in the mitochondrial DNA analysis of Hrbek & Larson (1999) versus the osteological studies of Costa (1998d). In the former, Rivulus is not monophyletic, while in the latter it is. As Costa, the major investigator of cyprinodontiform relationships, refined his studies of the sub-families Rivulinae and Cynolebiatinae some of the genera changed subfamilies and some may follow suit or revert. This is the nature of systematic research. Numerous papers representing the taxonomic work that followed Parenti's (1981) major revision of the Cyprinodontiformes were consulted in putting together this systematic overview: Wiley (1986); Costa (1990a,b, 1991b, 1992c, 1996a,b, 1997, 1998d,f); Grant & Riddle (1995); Parker & Kornfield (1995); Murphy & Collier (1996, 1997, 1999); Bernardi (1997); Parker (1997); Hrbek & Larson (1999); Murphy, Nguyen, Taylor, & Collier (1999).

The two genera with uncertain placement, Episemion and Foerschichthys, are indicated by "??." The family Cyprinodontidae has not been divided into two subfamilies, the monotypic Cubanichthyinae and the Cyprinodontinae, because "Cubanichthys and the Cyprinodontinae may not be sister taxa" according to Meyer and Lydeard (1993). Cubanichthys, also accompanied by the indicator ??, is tentatively placed in the Cyprinodontidae until more definitive proposals are advanced. Genera are listed in the same order as in Costa (1998d), paralleling his systematic hypotheses. For subgeneric compositions, see entries of the respective genera in the Annotated Checklist.