From Jean H. Huber
Private address: 7 Bd Flandrin, 75116 Paris, France
M.N.H.N., Ichthyology, PARIS, France.
Paris, February 28. 2013
Dear Colleague, dear Aquarist!
No mystery… major overall reviews of North-American Killifish species and genera have been rare recently (probably because in the American culture, both dynamic and competitive, full reviews must correspond to real milestones which necessitate large ressources in terms of material, time, and people-funding).
But each is equal to a breackthrough in our knowledge.
The first paper was issued in 2005 by Tony Echelle's team (op. cit.) and concerned the molecular review of nearly all species in the genus Cyprinodon and the results were solid and stabilized positively the systematics of the genus (still today !), and only the valuable contributions by Ulrike Strecker for the Chichancanab pupfishes brought something new in the landscape (and [October 2013] the description by Martin and Wainwright of another species flock in lakes of San Salvador island in the Bahamas has sharpened our curiosity with 2 new species Cyprinodon desquamator, Cyprinodon brontotheroides, along a form of Cyprinodon variegatus, sympatrically).
Today a new work has been published by Michael J. Ghedotti and Matthew P. Davis (op. cit.) for a complete revision, in 65 pages, of the family Fundulidae that includes the following genera, Leptolucania, Lucania, and Fundulus, and synonymizes Adinia with Fundulus, while recognizing 4 subgenera within the genus Fundulus: Plancterus, Zygonectes, Fundulus s.s., and a newly described (longly awaited) subgenus, Wileyichthys, dedicated to the renown specialist of fundulids and professor emeritus of ichthyology at Kansas University, E.O. Wiley.
The fishes in the family Fundulidae, commonly called topminnows, are widely distributed in freshwater, brackish, and coastal marine environments of North America, northern coastal areas of the Yucatan Peninsula and northern Cuba, and in Bermuda ; most fundulids inhabit shallow habitats that usually are less than 2 m deep ; these habitats also are often marginal with respect to quality of environmental conditions, undergoing substantial variation in temperature, oxygen saturation, and salinity ; fundulids are small, ranging in size from the diminutive pygmy killifish, Leptolucania ommata, with a maximum standard length (without Caudal fin) of 27 mm, to the so called giant killifish, Fundulus grandissimus, which can reach 180 mm in standard length.
Despite this long history and widespread standard usage of the generic name Fundulus for most fundulid fishes, authors have either not found clear evidence supporting its monophyly (e.g., Wiley, 1986) or have found evidence suggesting that the genus is not monophyletic (Parenti, 1981, with osteology, Bernardi, 1997, with a limited DNA sample) ; most authors have though followed the subgeneric classification of Wiley (1986), sometimes slightly modified, and have subdivided Fundulus into 5 subgenera: Fontinus, Fundulus, Plancterus, Xenisma, and Zygonectes, even if Wiley (1986) did not include the Pacific coastal species of fundulids, Fundulus parvipinnis and F. lima, thus leaving them unclassified at the subgeneric rank (while a rumour of scientific results, never published, has placed these 2 species closer to Goodeins or to Profundulus, a genus in its own family presently not thought to be related to Fundulus, despite its name).
In the present study, the authors use a range of available evidence (skeletal, myological, visceral, external morphological, color pattern, behavioral, karyological characteristics, and nucleotide-sequence data from 2 mitochondrial and 2 nuclear genes) to assess phylogenetic relationships within the family, and the relationships of the family to other Cyprinodontiformes.
The major strength is that the study includes 41 of 43 known as solidly valid species of fundulids (already, Whitehead in 2010 had proposed a molecular tree of many species with similar results, but that work was not intended for systematics and taxonomy) ; as a matter of fact, F. philpisteri and F. saguanus have not been included (because no material has been available for F. philpisteri that is strongly related to F. grandis and because the authors, based on the analysis of publications, treat F. saguanus as conspecific (synonym) with F. grandis) ; on the other hand, the endemic Fundulus of Bermuda (sometimes recognized as the separate species F. bermudae and F. relictus) have been considered and have not been studied but this is minor (Grady et al., in 2001, hypothesized that these populations may be the result of multiple, possibly recent, colonization events of Bermuda by F. heteroclitus and could be either most closely related to or conspecific with F. heteroclitus).
The second major strength of the study is that the authors use several species (precisely, 20 additional cyprinodontiform species from 8 families) as outgroups (hence if one of them is unstable at the upper level reconstruction it does not jeopardize the whole tree).
The third major strength of the study is that the authors use a longer set of sequences of genetic data (including for the outgroups) and a larger number of morpho-osteological characters (precisely, 154 anatomical, 19 color pattern, 7 behavioral) in their matrix (with the provision that the larger is better, as per Huber, 2012, for Rivulids).
The major limit of the study, a limit that is commonly encountered in all trees combining morpho-osteological and molecular data, is that the resulting tree with morpho-osteological data alone is substantially diferent from the tree with molecular data ; for example, Adinia xenica is isolated and much primitive in the morpho-osteological tree while it is clustered among other Fundulus species and closest to F. luciae, another aberrant morphospecies (over-slender and smaller, while Xenica is over-deep and smaller) in the molecular tree (and some analysts would conclude that the molecular data are the best because they tackle genes and other analysts would conclude that the molecular data include too limited sequences and too few populations to reflect reality).
The results of this study are strong at the upper level because Fundulidae form a monophyletic assemblage ; among them, some genera are solidly positioned, i.e., Leptolucania, Lucania, and Fundulus, as a whole ; the issue of the synonymization of Adinia into Fundulus is uneasy (see also the above remarks on the morphological tree) because the combined tree shows a branch (a clade) made of chrysotus, luciae, xenicus not as strongly supported as the other branches (and chrysotus, at least, could be unstable there)… and the authors could have opted for a less radical solution than the synonymization (e.g. a downgrading of Adinia as a subgenus of Fundulus with 3 species) ; similarly, due to the effect (weight) of the molecular data (the nucleotide-sequence data), the combined analyses do not recover any of the subgenera as monophyletic and suggest some novel relationships ; most notably, Fundulus jenkinsi, in this study, is a sister taxon to the F. heteroclitus species group, and as a consequence other species are moved from their previous subgeneric assignment ; the authors rightly comment that most of the other members of the larger super-clade containing the subgenera Fundulus, Xenisma, and Fontinus require revision as to their subgeneric classification : again they opted for a radical solution that expands the subgenus Fundulus s.s. to include the previous members of subgenera Fundulus, Fontinus, and Xenisma (i.e. synonymizing all these subgenera into Fundulus), as well as F. jenkinsi (removed from Zygonectes) because according to them «it is a simpler taxonomic solution than re-arranging the membership in all 3 subgenera», but… they could have alternatively kept these names (with their present value and Killi-Data will keep a conservative solution at the subgeneric level until further evidence) ; however, the main breackthrough in the branches within Fundulus is the grouping of the 2 species, Fundulus parvipinnis and F. lima into a new subgenus Wileyichthys (note : the systematic situation of F. brevis, sometimes seen as to be ressurected by some authors, is not discussed here), their close but distinctive relationship with the 2 cryptic species (zebrinus, kansae) of the subgenus Plancterus and their distinctive position, as a super-clade from all other species of the genus Fundulus.
Important note ; in Killi-Data, the synonimization of Adinia and the valid status of Wileyichthys is acknowledged, like the move of F. jenkinsi from the subgenus Zygonectes to Fundulus s.s., but the current subdivisions of the subgenus Fundulus s.s., as the subgenera Fundulus, Xenisma, and Fontinus are conservatively maintained, but this is a minor deviation from the results of the present study because it concerns subgeneric levels only.
Family Fundulidae (fossils are quoted between )
Genus Leptolucania Myers, 1924
Leptolucania ommata (Jordan, 1884)
Genus Lucania Girard, 1860
Lucania goodei Jordan, 1880
Lucania parva (Baird & Girard in Baird, 1855)
Lucania interioris Hubbs & Miller, 1965
Genus Fundulus Lacépède, 1803 [note : the name of Lacépède is also spelt Lacepède, following a detailed study by the senior, now deceased, French ichthyologist, Jacques Daget… but it seems that French scientists could not reach a common conclusion after his death]
Subgenus Wileyichthys subgen. nov.
[Fundulus eulepis Miller, 1945 —Pliocene Pleistocene, Death Valley, CA]
Fundulus parvipinnis Girard, 1856
Fundulus lima Vaillant, 1894
Subgenus Plancterus Garman, 1895
[Fundulus detillae Hibbard & Dunkle, 1942 —Pliocene, western KS]
Fundulus kansae Girard, 1859
Fundulus zebrinus Jordan & Gilbert, 1883
Subgenus Zygonectes Agassiz, 1854
Fundulus chrysotus (Günther, 1866)
Fundulus luciae Baird, 1855
Fundulus xenicus Jordan & Gilbert, 1882
Fundulus cingulatus Valenciennes in Cuvier & Valenciennes, 1846
Fundulus rubrifrons (Jordan, 1880)
Fundulus sciadicus Cope, 1865
The Fundulus notatus species group: Fundulus notatus (Rafinesque, 1820), Fundulus olivaceus (Storer, 1846), Fundulus euryzonus Suttkus & Cashner, 1981
The Fundulus nottii species group: Fundulus dispar (Agassiz, 1854), Fundulus blairae Wiley & Hall, 1975, Fundulus lineolatus (Agassiz, 1854), Fundulus escambiae (Bollman, 1887), Fundulus nottii (Agassiz, 1854)
Subgenus Fundulus s.s. Lacépède, 1803
The Fundulus majalis species group: Fundulus majalis (Walbaum, 1792), Fundulus persimilis Miller, 1955, Fundulus similis (Baird & Girard, 1853) (uncertain status)
Fundulus seminolis Girard, 1859
Fundulus rathbuni Jordan & Meek in Jordan, 1896
Fundulus diaphanus (Lesueur, 1817)
Fundulus waccamensis Hubbs and Raney, 1946
Fundulus albolineatus Gilbert, 1891 —extinct
Fundulus julisia Williams & Etnier, 1982
The Fundulus catenatus species group: Fundulus stellifer (Jordan, 1877), Fundulus catenatus (Storer, 1846), Fundulus bifax Cashner, Rogers & Grady, 1988
Fundulus jenkinsi (Evermann, 1892)
Fundulus confluentus Goode & Bean in Goode, 1879
Fundulus pulvereus (Evermann, 1892)
The Fundulus heteroclitus species group: Fundulus heteroclitus Linnaeus, 1766 (note : macrolepidotus is not mentioned), Fundulus bermudae Günther, 1874 (uncertain status), Fundulus relictus Able & Felley, 1988 (uncertain status), Fundulus philpisteri Garcia, Contreras & Lozano, 2007, Fundulus grandis Baird & Girard, 1853, Fundulus grandissimus Hubbs, 1936
Genus Fundulus incertae cedis, unclassified within a subgenus
[Fundulus curryi Miller, 1945 —Pliocene Pleistocene, Death Valley, CA]
[Fundulus davidae Miller, 1945 —Pliocene Pleistocene, Death Valley, CA]
[Fundulus lariversi Lugaski, 1977 —Miocene, central NV]
[Fundulus nevadensis (Eastman, 1917) —Pliocene, northwestern NV]
Even if the above title is (footballistically) anecdotal and provocative, it says well what it means.
The regular reader of recent systematic publications with morpho-osteological and molecular trees, or trees with their combined data, may well be perplex and uncertain as to how appropriately interpret the huge differences in naming taxa or not between authors and according to their geographical origin.
Clearly the US authors of this present major study, Ghedotti and Davis, present their results in a lumping way (to make it short, with few generic names and levels and with many species, conservatively seen, i.e. with doubts first, per genus, except monotypic aberrant units) and here several subgeneric names are synonymized and several subgeneric levels are fused and vanished.
Clearly also, the Brasilian researcher, Costa is acting in an opposite way, in a splitting way (to make it short, with many generic names and levels and with many species, explosively seen, i.e. with descriptions first, per genus), all the more that he publishes many papers yearly with incremental knowledge, and for exemple, Melanorivulus that he considers as a distinct genus (and not as a subgenus of Rivulus) from his other genera Atlantirivulus and Prorivulus is now moving close to 50 species named.
One could be tempted to see one attitude as good and the other as bad… that would be an error and, according to us, that would be pure caricature.
We'll just have to live with these 2 extreme options, until maybe one day, maybe never, a better knowledge is reached or even until both sides of the pendulum are tipped, at one time or another, for all groups of Killifish, without certainty of the outcome (the only certainty is that the pendulum is swinging and we get older !).
Anyway our knowledge is still very limited and it should be stressed that even with the (promising, but not ultimate) emergence of molecular techniques, the understanding of the generic and subgeneric levels for many large assemblages in groups related to Aphyosemion, Epiplatys, Nothobranchius, Rivulus, Simpsonichthys and to the lampeyes (etc.) is unsatisfactory with no or poor diagnosis that is congruent with the molecular tree, itself with still limited species included (in the present study, Ghedotti and Davis give a fine diagnosis of their new unit Wileyichthys, but, alas, not for their new definitions of Fundulus s.s., Zygonectes).
Yes, these 2 extreme options are a fact.
Killi-Data will attempt to follow a median reasonable path in-between, not because it is better than lumping or splitting, but because it respects the added value by previous researchers (if not erroneous !) and because it avoids many and rapid changes or reversals of names.
Even if we (following Killi-Data) keep a reasonable and conservative median strategy towards the splitting or lumping tactics in using/creating names, we are still facing with the embarrassing situation of having each of the 3 methods (morphology, osteology, molecular biology) currently used in systematics (not only for Killifish, obviously) that yield significantly distinct results (i.e., phylogenetic trees).
Would this lead to be pessimistic for today or the future ? Not at all… just have a look back to the impressive progresses that have occurred over long periods (say 15 - 20 years) in our systematic knowledge of killifishes.
Yes, that embarrassing situation is also a fact, but we can live with it, can't we ?
For today, let's celebrate a very good work by 2 US professional researchers to whom we wish a rewarding future.
In total a very important and sensitive newsletter !
Hopefully a boost to our community and a spur to speed up knowledge progress on our (beloved) fishes !
Take care and enjoy the scientific or aquaristic complexity of killifish !
Do not hesitate to ask questions for future Newsletters.
Visit frequently the website www.killi-data.org!
Thank you for your support over the years.
With my kindest regards.
Echelle, A.A., E.W. Carson, A.F. Echelle, R.A. Van Den Bussche, T.E. Dowling & A. Meyer. 2005. Historical Biogeography of the New-World Pupfish Genus Cyprinodon (Teleostei: Cyprinodontidae). Copeia, (2): 320-339, 5 figs., 1 tab.
Ghedotti, M.J. & M.P. Davis. 2013. Phylogeny, Classification, and Evolution of Salinity Tolerance of the North American Topminnows and Killifishes, Family Fundulidae (Teleostei: Cyprinodontiformes). Fieldiana Life and Earth Sciences, 7 :1-65, 22 figs, 2 tabs., 4 append.
I am interested in reading other Newsletters, click INFOWEB.