The Art and Science of Aquaria
Evolution of Fishes

This chart shows the ages of the rise of fishes. Nearly all aquarium fish are the bony fishes on the far right.

Evolution of Cichlids

By 250mya there were around 1200 species of fish. Between then and now around 27,000 species of fish evolved.

killifish geo
killifish ref killifish similar

Evolution of Killifish.

Arthur Cronquist, 1968 wrote in "On the Relationship between Taxonomy and Evolution":

"In the hundred plus years since Darwin, it has become practically axiomatic that a taxonomic group must be monophyletic, and that a polyphyletic group is per se unnatural and must be abandoned. Some definitions of taxonomy even make the discovery and exposition of phylogeny the basic goal of taxonomy, rather than a means to an end, or a related but separate objective, as I believe it to be. Too rigid an adherence to the rule of monophylesis, however, would wreak havoc with our system, without providing anything useful to replace it. Thus, on the basis of comparative morphology and physiology it is perfectly clear that the protozoa, and through them eventually the whole animal kingdom, arose in several more or less parallel lines from different groups of algal flagellates. There was no original protozoan, from which all other protozoans evolved, and no original animal that was ancestral to all others.

Must be protozoan phylum and the animal kingdom therefore be abandoned as unnatural? Not at all. The assumption of the holozoic mode of nutrition by certain members of several different groups of algal flagellates had such an all-pervasive influence on their subsequent evolutionary development that we are justified in regarding them and their descendants as a coherent natural group. All of the other features which we think of as characterizing animals as opposed to plants were evolved in response to or in association with the holozoic mode of nutrition, and many of these characters persist even in animals such as tapeworms, which have abandoned holozoic nutrition and taken to absorbing their food.

We need not further pursue the still knotty problem of where to draw the line between the plant and animal kingdoms. Our point is that the monophyletic requirement must be considerably stretched in order to permit the very natural and useful grouping of organisms into plants and animals. As an aside, it might be mentioned that efforts to avoid this problem by recognizing three or more kingdoms create at least as many difficulties as they avoid. And to refer all the unicellular flagellates to the protozoa would deprive the animal kingdom of its essential characters, as well as making the plant kingdom polyphyletic.

The exact same problem exists with killifish. The Rainbow Fish evolved before killies while the sleeper gobies, the Eliotridae evolved after. Thus, in the clade that comproses the rainbows, killies and sleeper gobies has the distinction of havie a clade of single dorsal fishes in with two others, at least one of which has a specues wirh a single dorsal ray in the first dorsal - if it loses that, what then? Similarly so we classify the "live bearing tooth carps" - the livebearers of guppy, molly and swordtail fame.

So, it's a fair point,taxonomy can't be a slave to phylogony but simiarly so phylogeny can't be a slave to taxonomy and any attempt to show both is bound to have to make tradeoffs between the two. So be it, but let us perhaps focus more phylogey at the micro level but defer more to phylogoney on the macro level.

Thus we may find our methods of classifying fishes difficult with our current definitions of species and genus made more complicated by the fact there are not good definitions of either, opinions vary and one mans genus is another mans species and the eternal friendly fied between "splitters" and "lumpers". One would like to think members of a genus were very closely related and each species differed from any other by one thing being different.

This works very well with linear evolution with a single path and the evolution of the horse is a good example. But as with the evolution of man, the evolution of killifish and cichlids and probably most fish is not so simple as that and information systems and ontological schemas laid down before Linneas are now very poor at expressing real world evoltuionary trends. That is if you want to show the evolution of the horse it works very well but when you consider untold billions and trillions of herring in the oceans around primordial Africa living in the intertidal shore zone and swimming up a brackish river and getting stuck eventually in a freshwater river system.