From Jean H. Huber
Private address: 7 Bd Flandrin, 75116 Paris, France
Laboratoire d'Ichtyologie, M.N.H.N., 43 rue Cuvier, 75231 PARIS Cedex 05.
e-mail : (NEW!) or [both today inactive]
S.F.I. : Societe Francaise d'Ichtyologie (same address).

Paris, February 21., 2003.

Dear Colleague, dear Aquarist!

Let's, unfortunately, start by very sad news. 

Prof. Robert R. Miller passed away on February 10., 2003, at 6 am. 

Bob Miller was an important figure in American ichthyology and conservation from 1940 to the present. He was a graduate student and collaborator with Carl L. Hubbs and married Frances Hubbs in 1940. The Miller children, Gifford, Francis, Roger, Ben, and Lawrence were also important contributors to the field work along with many graduate students who were welcomed into "his" ichthyological family over the years.

Miller was a faculty member in the Department of Zoology (later Biology) and a Curator in the Museum of Zoology from 1948 to his retirement in 1987. He published over 300 contributions to ichthyology and conservation, beginning with systematic and experimental studies of the pupfishes, Cyprinodon, in Death Valley, and continuing with studies of fishes of western U.S. and Mexico. He actively worked on trout, minnows, suckers, live-bearers, and sculpins, and pioneered work in karyology, hybridization, fossil fishes, and experimental studies. His call for conservation in 1961: "Man and the changing fish fauna of the American Southwest," and his list of "Extinct, rare, and endangered American freshwater fishes," were the beginnings of the American fish conservation movement.

In details, Robert R. Miller was born April 23., 1916, in Colorado Springs, Colorado. He earned his Ph.D. at Michigan in 1944. The most important collaborations in his career began in the summer of 1938 when he joined Carl L. Hubbs and his family to conduct field work in eastern Nevada and met Francis Hubbs. From the time of their marriage in 1940 until her death in 1987, Bob and Fran worked as an inseparable team, creating a research enterprise that is unmatched, as it joined the energetic program of Frances' parents, Carl and Laura Hubbs. Robert Miller's first five papers as a student, 1936-40, established the zoogeographical region and taxonomic direction of his interests with small papers on Mohave/Southern California minnows and Fundulus.  In the 1941-44 period, he published 11 papers on his dissertation work covering Cyprinodon and other SW fishes. The Hubbs influence appears stronger in the years 1945-49, with 20 wide ranging taxonomic papers and reviews, including Miller's landmark dissertation monograph on Death Valley Cyprinodon and the classic Great Basin paper with Carl Hubbs. Miller worked as Associate Curator of Fishes at the U.S. National Museum of Natural History from 1944 to 1948 and in 1948 became Associate Curator of Fishes at the Museum of Zoology, University of Michigan (UMMZ). In 1950-55, as a new faculty member at Michigan, he published 65 reviews and short papers on cyprinodontoids (Killifish and live-bearers), trouts, Dorosoma, gars, eels, post-larval fishes, hybrid suckers, and identification of archaeological remains, mostly from the southwest U.S. and Mexico.  Miller's signature period was 1956-60, when he published 21 wide-ranging, but deeper, creative papers and reviews including the ground-breaking paper on the "Origin and affinities of the freshwater fish fauna of Western North America". In this period, an increasing concentration on the freshwater fishes of Mexico was marked by the beginning of the biennial field trips to Mexico. The period 1961-65 began with the conservation call to arms: "Man and the changing fish fauna of the American Southwest," and continued with a paper on speciation rates in western fishes, four ground-breaking papers on Cenozoic fossil fishes with Teruya Uyeno, his summary review of "Quaternary freshwater fishes of North America," the Ichthyology textbook with Karl Lagler and John Bardach, the book on Fishes of Utah with William Sigler, and the first list of "Extinct, rare, and endangered American freshwater fishes."  Thirty-four papers were published in this productive and influential period.  In 1966-70, 16 papers focussed especially on the diversity of the fauna of Mexico and the status of threatened and endangered fishes. The years, 1971-75, were a period of enormous productivity, with 53 papers and reviews, including descriptions of new recent and fossil fishes of Mexico, creative work on fish karyology with Teruya Uyeno, conservation papers. In 1976-80, Miller published 47 papers, reviews, and abstracts on Mexican fishes, fish chromosomes, and conservation. In 1981-85, he published 43 papers, reviews, and abstracts on fishes of Mexico and southwest U.S., conservation, biogeography, and systematics. After Fran's death, work became more difficult, and Bob's contribution became more irregular.  In 1986-90, 17 papers and abstracts were published, including the two important zoogeography chapters on Mexico and the Rio Grande Basin, with Michael L. Smith and additional works on the fishes of Mexico. W.L. Minckley became an important partner in fieldwork in Mexico during this period. In 1991-94, 17 abstracts and papers appeared, mostly aimed at tying up loose ends on the Freshwater Fishes of Mexico.  An important contribution to history of fish study appeared in 1992, written with Clark Hubbs and Frances H. Miller - "Ichthyological exploration of the American West, the Hubbs-Miller era." In the later 1990s, work on the Freshwater Fishes of Mexico was carried forward by W.L. Minckley, Steve Norris, and Martha Gach, with Miller's participation when his health permitted. The manuscript was closed a few days before Minckley's death in July, 2001. Bruce Turner organized a Symposium on Cyprinodon, held in July, 2001 at the meetings of the American Society of Ichthyologists and Herpetologists at State College Pennsylvania, in honor of Robert R. Miller.

Through a long and distinguished career, Bob Miller has contributed to our knowledge of the systematics and evolutionary biology of many groups of fishes, but his work on pupfishes---his first love and "favourite" group---is seminal. One simply cannot study pupfishes, be they from Death Valley, central Mexico, the Atlantic coast, or the Caribbean, without encountering original papers by Bob Miller and or his students and collaborators. And they are papers that remain significant, and often central, many years after they were first published.

He himself described or co-described not less than 34 Killifish taxa. He is the most important describer of Killifish from North America, notably in the genus Cyprinodon.

Three Killifish taxa have been described after his name: Cyprinodon salinus milleri, Cyprinodon bobmilleri and the Mexican genus Millerichthys.

This Newsletter titled "an underlying crisis?" is dedicated to Bob with our deepest and friendly souvenir. 
No doubt that he would have liked to participate to this debate in a passionate and responsible way, with a tutorial touch. 

Bob Miller is remembered as a person of formidable creative energy and outstanding dedication to the science of ichthyology and the conservation of fishes. He was a major source of inspiration to his colleagues and students. He promoted "his" ichthyological family, i.e. those who feel part of a tradition of lasting importance to natural history, to whom we all belong or aim to belong.



An underlying crisis ? Here is the major issue that is tackling us for 2003, and further, regarding the species concept for Killifish and zoology in more general terms.

The reasons of the problem are simple.

Nothing new in terms of definitions, except the fact that along time many diverse and partly contradictory definitions of the species in biology have been proposed.
For those who want to dig in this difficult subject, please read the recent synopsis and most updated review (2000): Species Concept and Phylogenetic Theory. Q.D. Wheeler & R. Meier (Eds), Columbia University, New York.

The problem for us is in fact a consequence -an unexpected effect- of recent scientific progress.
Or to say it differently an outcome of the discrepancy between the results with new techniques and their usefulness for human observers. 
Or between the biological objective of the species as a subject of study and the systematic objective of the species as a subject of practical classification.

Let's go back in the 1950's to 80's, the only techniques that were available to describe, diagnose a species were morphology and global osteology (eventually lumped together), with a taste of ecology or ethology in rare cases.
Cryptic species were underlined as strange examples, not rare, but still strange.
And karyology, especially with Killies, gradually helped to make these strange things to become more common. 
On the whole, the researcher's experience, opinion and notoriety were prevalent. And if not, they were challenged on the same grounds, opinion, experience, notoriety as to what were the significant characters that would prevail.
The picture was already complex, but manageable. And instead of describing say 12 new names for karyotypically distinct populations of Aphyosemion ahli, it was agreed to keep a single one with at least one diagnostic character (barred pre-peduncular region) from its nearest relative A. australe. But the consensus was not prevailing and for example, it was clear that for some recognised species of Cyprinodon, for example, the single diagnostic character was missing (or very minor).

In recent years, thanks first to the hypothetical polarity of characters initiated by cladism, then to computer sciences (and software, such as PAUP) and finally to a brand new tool, the molecular biology, also using polarity of characters and the same IT software, but this time with genetic data, the species has been based on a more objective - less subjective - definition because the disclosure of characters is traceable, reproducible and open to anybody's control, i.e. more scientific.

A perfect world ? Far from it. The more scientific rationale of the species ends up in a myriad of potential and actual naming of species.
Let's take some recent examples of the new difficulties and questions that are facing us all (by no means, there is an underlying criticism in the following sentences ; a simplistic style is used, just to communicate clearly the issues):

* Scriptaphyosemion banforense and nigrifluvi have been considered by many as junior synonyms of guignardi, because their initial diagnosis has been shown to fall within the latter's global variability and crossings were fully fertile. Bad luck, molecular biology has shown that their genes (exemplified by mitochondrial DNA portions) were so different that all 3 must correspond to distinct valid species (and similarly for Austrofundulus sp. from Colombia and Venezuela, or, for Northern and Southern populations of Cyprinodon variegatus in Eastern USA). Then what to do ? Revalidate them automatically without a renewed pertinent diagnosis ?
* Various populations of the elegans superspecies which are assignable to the species Aphyosemion cognatum by phenotype (according to the accepted diagnosis) have been DNA-studied and the results split them in various positions of the phylogenetic tree (the evolution tree), sometimes much closer to distinctly colored phenotypes. Then what to do? Do not consider the DNA techniques for species systematics, at least in those special regions of complex palaeo-history?
* Species with matching external, internal and behavioural characters have been put together with evidence within a genus. For example, the large and cannibalistic, cheradophilus and wolterstorffi have been assigned to Megalebias, together with elongatus, monstrosus and alike. Bad luck, molecular biology has shown that their genes (exemplified by mitochondrial DNA portions) placed them nested between 2 groups of Austrolebias that lives nearby geographically. And not at all with elongatus and alike. Then what to do? Do not consider the morphological and osteological techniques that might be "polluted" by convergence?
* Last years, several cryptic species have been described on the basis of minor stable osteological and/or morphological characters, such as Simpsonichthys picturatus vs. magnificus, Nematolebias papilliferus vs. whitei, Cynopoecilus fulgens, intimus, multipapillatus, nigrovittatus vs. melanotaenia, Austrolebias jaegari vs. gymnoventris, Aphanius iberus vs. baeticus. In all cases, differences in live color patterns of both sexes are little if any, and even, in aquarium further generations they may be washed up (e.g., with papilliferus or with picturatus). Then what to do ? Forget about color patterns that are so important in species recognition (for breeding female, at least… and human researcher, too) ?
* And finally in a review of Brasilian adloffi-like species, Austrolebias charrua, minuano and nigrofasciatus have been described on the basis of stable osteological and morphological characters of Brasilian populations. But another author had studied in-depth Uruguayan populations of adloffi and concluded that the variation in color pattern was huge, even in a single population. And his samples were sometimes only a few kilometers away from those in Brasil. Then what to do ? Forget about the analysis of color pattern variations after in depth field studies that have been useful in species diagnosis since long ?

Those are some of the thorny, provocative and confusing questions (without answers) in front of us. And the list is not limited and will be extended in the future.
What are these questions addressing ?
In general terms, not the value of a given researcher, not the value of the used techniques. 
On the contrary, that given researcher has faced reviewers and advisory boards with often strict policies before his paper may be published. And those used techniques have shown their merits on other grounds and the experiments can be redrawn to check if the chosen characters are significant and reliable and if the results are acceptable. 
No. Simply, these new techniques and this new evidence are forwarding lots and lots of differentiating characters and in the end possibly lots and lots of new cryptic species and new names.
Is it necessary ? Is it worthwhile ? Is it useful to the classification (systematics) ?
Time will tell ! But for sure, our future will be made of more uncertainties. Those who thought that science progress is to clear up their understanding of biological complexity will be disappointed.

But there may be a modern consensus way of tackling the problem, case by case !
And I am pleased to mention the first example of problem solving by the board of advisors of Killi-Data online for a similar issue : the possible revalidation of Fundulus kansae, which had been considered as a junior synonym of F. zebrinus and which had been shown by DNA samples to be distinctive. Several external (scale counts) and internal (length of intestine) characters, among others, had been shown different between the northern (namely, kansae) and the southern (namely, zebrinus) populations, but these differences were subsequently proved to belong to a continuum of variation… before the molecular evidence!

Please read hereafter the digest of Brian Kreiser's molecular study, then the vote of the members of the board of advisors, with some anonymous qualitative (and diverse) comments.

Digest of Brian Kreiser's molecular study (2 papers)
Am. Midl. Nat. 146:199-209 (2001)
Evolution, 55(2). 2001, pp. 339-350 (with JEFFRY B. MITTON AND JOHN D. WOODLINO).

"The current literature treats the plains killifish as a single species (Fundulus zebrinus). However, a recent range-wide survey of genetic variation detected two distinct groups. The presence of the two groups suggests that the plains killifish may actually represent two species (F. zebrinus and F. kansae), as has been proposed in prior taxonomic treatments. Previous molecular studies within the genus Fundulus have relied upon the mitochondrial cytochrome b gene. A set of cytochrome b sequences has been generated in order to compare the phylogenetic support for the putative species F. zebrinus and F. kansae with that found in other accepted sister species of Fundulus. Both neighbour-joining and maximum parsimony analyses recovered two clades of plains killifish that were strongly supported by a variety of measures including branch lengths, decay indices and bootstrap proportions. The strength of support for the putative F. kansae and F. zebrinus clades was comparable to, or better than, similar measures for six other sister species pairs of Fundulus. While the molecular data supported the recognition of two species of plains killifish, the groups recognized were not in agreement with the range descriptions previously proposed for the two species. 
Drainage systems of the Great Plains and western Gulf Slope underwent substantial changes through diversions and stream captures during the Pleistocene, either as the result of the glacial advances or through independent geologic processes. The distributions of a variety of fishes that range across west-central North America, such as the plains killifish (Fundulus zebrinus), are thought to be the product of this Pleistocene influence.
The phylogeographic patterns were concordant between the allozyme and mitochondrial data with the exception of the population in the North Canadian River. The populations fell into three geographic assemblages, which we designated as northern, central, and southern. The northern region was likely colonized sometime during the mid-Pleistocene. Fish in the Brazos and Pecos Rivers probably reached these drainages through stream captures of the Red River. The large phylogenetic break between the northern/central and southern clades supports previous attempts to recognize two species of plains killifish: F. zebrinus and F. kansae."

Vote: the problem is not simple and the 16 advisors were split between revalidation and status quo. But after counting actual voices (kansae, valid: yes, 8 votes, no, 3 votes, neutral, 3 votes), the final vote of the members of the board of advisors is then to declare kansae as valid.  
But this is not the end of that story of zebrinus / kansae
Research continues with new characters and new techniques. A renewed detailed morphological study of zebrinus should be undertaken and a precise type locality should be allocated to kansae. This detailed study will start on different grounds than previous studies and, notably, should focus on the intermediate populations from the Red river region.
This is not the end of the full story. Many new critical cases are in front of us and not only those listed ahead. 
The mountain is in front of us. Let's hope that Killifishes do not transform us into Sisyphus !


Comments : in details, all comments that were attached to the votes are presented (anonymously !). Sometimes, with a good sense of humour to communicate better. These comments are very valuable because they give a good picture of the situation through their diversity. They should  stimulate your thinking: they show the difficulty and complexity of the problem, the pressure to consider issues one way or another, the difference of cultures and reasoning, all being of course respectable.

PRO's (sometimes with limitations)

"Yes, Fundulus kansae should be revalidated. There is at least some evidence of diagnosable characters, albeit DNA and not "accessible" without a laboratory study. The author concludes they are separate and this can be regarded as a testable hypothesis. New morphological data can be gathered based on the newly suggested ranges to test the hypothesis."

"I vote to treat the two populations as separate species. I believe that it is better to call attention to differences rather than similarities, especially with the conservation concerns that face all species today. So my vote is to recognize both F. zebrinus and F. kansae to reflect the genetic differences revealed by the DNA work."

" It looks like there are two species involved. I would then be inclined to speak of two species names. However, the putative range of the two species is not clear. Let us not jump the gun here."

"Evaluation of the specific status of allopatric forms is always difficult, and the use of some vaguely arbitrary criteria is likewise difficult to avoid. There are clearly two phylogenetic lineages in this group, well separated by DNA sequence information, each coherent within itself. I think that science will be better served by considering the two as different species, for it will be more efficient in the future to retrieve information by accessing two species names.  If only one name is used, the fact of the divergence may be lost. An analogous situation exists in Cyprinodon variegatus. The northern populations are a distinct clade and probably merit recognition as a separate species. They correspond fairly well to Cyp. var. ovinus Hubbs, but that subspecific name is almost never used, and many biologists who use Cyp. variegatus as an experimental animal are unaware of the differentiation. Making them into two species, which is rumoured to be proposed, will call attention to this distinction. There is another issue here. Traditionally, within killifish, many of European researchers have been reluctant to recognize taxonomically entities that are genetically but not morphologically distinctive. I believe that this stance is archaic, and the F. zebrinus example will perhaps contribute meaningfully to this debate."

"My opinion is that there are two species, in process of creation of a new recent species or, for they inhabit similar biotopes, they didn't suffer significant morphologic variations."

"My answer to the zebrinus/kansae problem is : subspecies zebrinus zebrinus in Red river and south of it; zebrinus kansae north of Red river. If this is not possible to vote for that compromise solution, then I vote for 2 species. I received the complete article from Kreiser. After reading it, it is obvious that genetically there is quite a gap between north and south of Red river. I asked Kreiser if data is available about biological differences  between the two taxa. Unfortunately he didn't answer yet. It is clear that Red river captured a more northern river in the past and received the zebrinus population in this way. After that moment, the southern population changed genetically considerably. The fact that there is little or no genetic exchange between southern and northern populations could however be a geographical threshold and not a biological one. DNA sequencing tells a lot (if not all) about phylogenetic relationships, but very little (or nothing) about species rank. Genetically and statistically a good paper… but with not enough biological information. So subspecies … no problem… maybe even a classical example of subspecies. But species? "

CON's and Neutral

"No, Fundulus kansae should not be revalidated. The author's decision is based only on mitochondrial DNA. There may be two species in what we now call F. zebrinus, but the mitochondrial DNA data disagree with both morphology and the author's own results for allozymes. The boundary between zebrinus and kansae will have to be based on congruence of more than one gene locus or some other character, such as breeding colors or other aspects of morphology. Level of divergence for a single gene (or for mtDNA) is not a valid basis for recognition of two species. The mitochondrial DNA results might reflect past hybridisation and lineage replacement or lineage sorting from a polymorphic ancestor."

"About the matter of Fundulus zebrinus and F. kansae, the article by Kreiser et al. did not convince me that the two are distinct species. The differences found are rather small and, for all, they found a population (Canadian River) that can be regarded as an intermediate between the two. I think that they are typical products of the American thinking on taxonomy: related species that occupy different drainages or different bodies of water that, under normal circumstances, cannot meet, are viewed as different species, e.g. the desert fishes of the genus Cyprinodon. They have no genetic or a very low genetic differentiation but are regarded as different species.

"That Fundulus issue directly parallels what is going to happen with Austrofundulus, while, basically 25 years ago, the various Austrofundulus populations were suspected to be different species because of color differences in the males, but they interbred readily, and only very minor morphometric differences could be found. Now they show different DNA sequences, and are sufficiently distinct to warrant description as different species. I guess the question is eventually going to boil down to a matter of convenience, as well as correctness. If our species differ only in DNA, with no outward sign of difference, it will be extremely difficult for the average biologist to know which one he has. I guess the fish are laughing at us, they after all "know" which females to court! and who´s who.
In most cases of the annual fishes, there is never a choice, since their rain pool is their universe."

"As a non-US and a non-European citizen, I would like to push a compromise solution. I propose to change the meaning of the subspecies level for this type of examples likezebrinus/kansae. Did the researchers specializing in Barbs, for example, find big DNA differences for their subspecies (these are frequent in that group, not to speak of botanists who add the variety level) ? Otherwise, I would like to criticize the US system of grants that promotes indirectly new names and the European conservatism that limit new thinking (isn't it significant to mention that cladism is a German concept that was broadly sent to iconoclasm, until the US school rediscovers it and its benefits?)."

"A single species name should be maintained and kansae should be kept as a synonym, for the time being. Despite the merits of Kreiser's study, the two sets of populations are not externally distinguishable and there is no geographical barrier between them. To accept kansae as valid would be a door open to an endless list of new species in Killifish systematics for very little added value. But we must accept science progress and Kreiser's work and findings are undoubtedly interesting and should not be discarded because they make us feel uncomfortable. There must be an alternative solution to acknowledge valuable works like Kreiser's. That should be offered to the ICZN decision power with several options: to end up in dropping the Linnean binominal system? in adding to the binominal system a new name (but not the subspecies one that is valueless, here, because of the genetic separation), such as a code, a number (already, a gene bank code does exist) ? in separating the species in systematics and the species in biology ? in something else ? In a way, the system of population naming that is rightly used by aquarists to avoid unwanted crossings of sterile populations would be worth considering without ICZN new rules. But can new (young) researchers accept that their works do not materialize into new official names ? And also, many of them may argue that a name of a population or a new species name, it is not so different in essence, and that already many species names, already accepted, correspond to externally not separable fishes, in Cyprinodon and in other families of fishes (e.g. Characoids, Cichlids, Catfishes). Many articles deal with the concept of species, but none addresses those human issues like the ICZN committee composition, the necessity of new names by young researchers and new techniques, the real objective of classification in face of biological complexity. Yes, a major problem is in front of us. Inevitably ! But discussing widely that issue may well improve our thinking for the future and ultimately may let us take wiser decisions."

"I have read over the two abstracts that you sent and conclude that there is not enough information to make a decision. Therefore, I suppose my vote is neutral. My reasons for this are that the recognition of one or two species must be made in the context of a phylogenetic analysis of all Fundulus species, or at least the complex to which F. zebrinus belongs. Both papers make the assumption that F. zebrinus and F. kansae are sister species, but this is a hypothesis that must be tested. It is never proven right or wrong, just confirmed or rejected. That is also how I feel about recognizing one or two species; we can vote on it, but our vote is irrelevant to the question being asked."

My warmest thanks to the advisors for their first cooperation. They will be asked for their opinions from time to time, while respecting their work load. 
In such cases as kansae / zebrinus where objective distance is available. 
Not for those brand new cases raised above : for them the policy of Killi-Data online is to accept latest available data, until conflicting results arouse. For example, all the new names, as charrua, fulgens, papilliferus, baeticus, jaegari (… ) stated above are accepted as valid in Killi-Data online, until opposite findings if any.

With this problem-solving example, the principles of a cooperative platform that had guided the conception of Killi-Data online are in good tracks !

Finally, 3 short news:

1- Please note that a petition has been launched to the ICZN commission to keep the genus Aphanius as valid and reject Lebias as a senior synonym. Killi-Data is in coherence with this conservative thinking among scientists and palaeontologists until the ICZN decision. That decision is not anticipated soon and until then, everybody is free to use either Aphanius or Lebias.

2- A short word special to aquarists to inform them of a very attractive new venture on the Internet. Julian Haffegee, as an IT expert, has structured a Data Base aimed to nest all aquarium information on Killifish, based on data provided by the Internet breeders themselves. It is extremely well conceptualized and built for everybody's benefits, without "ego" or country flagship. According to us, it should become a major international project as a mirror of aquarists' knowledge and contribution. Today it is hosted by the BKA at (link is ianctive, now). Please visit and bring your own contributions.

3- Let's end up with wonderful news: the full genomic map of Rivulus marmoratus [present name Kryptolebias marmoratus] is on its way… To understand, one day, the mechanisms of its unique hermaphroditism. Please, have a look to the first milestone:
Lee, J.S., M. Miya, Y.S. Lee, C.G. Kim, E.H. Park, Y. Aoki & M. Nishida. 2001. The complete DNA Sequence of the mitochondrial Genome of the self-fertilizing fish Rivulus marmoratus (Cyprinodontiformes; Rivulidae) and the first Description of Duplication of a control Region in Fish. Gene, 280 (1-2): 1-7, figs.


Take care and enjoy the scientific or aquaristic complexity of killifish!

Do not hesitate to ask questions for future Newsletters.

Visit frequently the website!

Thank you for your support over the years.

With my kindest regards.


I am interested in reading other Newsletters, click INFOWEB.