Field of Science

Knocked Off the Perch (Taxon of the Week: Percidae)

The rainbow darter, Etheostoma caeruleum, a representative of the North American radiation of small and often colourful freshwater fish known as darters. This species breeds on fast gravel riffles, where pairs mate with the female half buried in the gravel so the eggs are automatically covered over (Reeves, 1907). Photo by Jim McCormac.

Okay, this post has been delayed again. It's been an unusual week, is all I can say. I'd tell you all about, but I have very good reasons to believe that that would be extremely dull.

In earlier posts, I have ranted in a rather esoteric manner about my distaste with the commonly recognised fish order "Perciformes", really a random multi-paraphyletic assemblage of the more generalised members of the clade Percomorpha. In the recent partial reclassification of the Percomorpha by Li et al. (2009), the name "Perciformes" was ditched entirely, and the clade containing the family Percidae was instead called Serraniformes (suggesting, offhand, that some sort of taxonomic karma is dooming this taxon to be associated with confusing names - the family Serranidae as commonly recognised itself seems likely to be polyphyletic, and a number of "serranids" are not guaranteed Serraniformes). But even before the Perciformes of common use were recognised as a wastebasket assemblage (if, indeed, there ever really was such a time), Percidae was always a slightly odd choice for the type family. The Percidae, the perches and darters, are not particularly average Perciformes.

Among the percomorphs, percids are unusual for one main reason - they're almost entirely freshwater (a few European species stray into brackish waters, but only one species - Sander marinus, the estuarine perch of the Black and Caspian Seas - is a permanent resident in them). While the percomorphs have achieved true world dominance in the upper parts of the ocean, including the vast majority of coastal and surface-pelagic fish species, they have never made such significant inroads into fresh water. A few percomorph lineages have been very successful in fresh water, such as the Cichlidae, the Anabantiformes and various members of the Smegmamorpha*. But in contrast to their surface-marine monopoly, percomorphs have to share dominance of the fresh-water environment with members of the clade Otophysi - Cypriniformes, Characiformes and Siluriformes.

*No, honestly, it's a real name.

The zander, Sander lucioperca, a much larger Eurasian percid. Photo from EoL.

The Serraniformes also include the Trachinidae (weevers), the circum-Antarctic notothenioids and the majority of what were the Scorpaeniformes. Relationships within the Serraniformes are yet to be hammered out, but the Percidae probably divide from the others reasonably basally. Ten genera of living Percidae are currently recognised, with more than two hundred species. Phylogenetic analysis of the family by Sloss et al. (2004) recognised three main clades of unresolved relationships - the Holarctic genus Perca, the mostly Eurasian clade of Gymnocephalus plus Luciopercinae (genera Romanichthys, Sander and Zingel, with three species of Sander in North America), and the North American clade of Etheostomatinae (Ammocrypta, Crystallaria, Etheostoma and Percina). [The tenth genus includes the single uncommon species Percarina demidoffi of rivers running into the Black Sea, and was not analysed by Sloss et al. due to lack of material. Percarina was previously classified in the possibly non-monophyletic Percinae* with Perca and Gymnocephalus and differs from most other Percinae in spawning in brackish waters, so establishing its relationships would be very interesting.] While the greater phylogenetic disparity of Percidae is concentrated in the western Palaearctic and the family is believed to have originated in that area, the greater diversity of species is definitely found in North America. Well over two-thirds of percid species belong to the Etheostomatinae, with the greater part of those in the genus Etheostoma (which, however, may not be monophyletic).

*Though the non-monophyly of Percinae found by Sloss et al. is in contrast to their breeding behaviour - Percinae differ from other percids in laying their eggs encased in long gelatinous strands, while Luciopercinae and (ancestrally) Etheostomatinae are broadcast spawners.

The European perch, Perca fluviatilis. A very similar species, Perca flavescens, is found in North America. Photo from here.

Human interest in the Percidae (as with most matters, really) has usually been related to one of two things - eating or sex. The larger percids of the "Percinae" and Luciopercinae are widely caught for food, and the European perch Perca fluviatilis has been introduced to many localities outside its native range such as New Zealand for the amusement of anglers. Some percids, such as the walleye Sander vitreus, have been recorded reaching lengths of over a metre (though such sizes are, of course, exceptional - a more average walleye would be about twenty centimetres). Species of the Etheostomatinae, known as darters, are not targets of fishing - members of this subfamily (as well as some species of Luciopercinae) are smaller than other percids, less than ten centimetres in length*, and wouldn't offer much in the way of eating. Still, darters more than make up their interest in the other regard of sex. They show a wide diversity of breeding behaviour, from broadcast spawners to some that bury their eggs in sediment or gravel to species that lay their eggs safely hidden on the underside of rocks. Other species may glue their eggs to vegetation (Winn, 1958a, b). During the breeding season, most (but not all) darters move from deeper to shallower waters (many species favour riffle areas) where the males usually establish a breeding territory (as reported by Winn, 1958b, the presence of other males seems to be required to incite the successful establishment of a territory - solitary males tended to lose interest in a potential territory and wander off). Some darter species are fairly relaxed about their territories and only fend off males of their own species, but other darters may be decidedly pugnacious and attack just about anything that moves. Challenging males approach each other with fins held high, and their colours will often become brighter. They may circle each other and butt or bite at each other's tail regions. After a male has mated with a female and she has laid her eggs, he may or may not remain in the area to guard them. Experiments have shown that if the eggs are removed or replaced, the male continues to guard the same spot, so it is the territory that induces guarding behaviour rather than the presence of eggs. Hybrids have been recorded between a number of darter species and seem to be not uncommon, especially where species have been spread outside their native range (Stauffer et al., 1995).

*As a corollary of their smaller size, it is worth noting that darters (and the smaller Luciopercinae) also lack swim bladders.

Percarina demidoffi as illustrated by N. Kondakov in a 1957 Russian textbook. For some reason, I find a certain whimsy in this illustration of what is perhaps one of the more mysterious percids. Image via NOAA Photo Library.


Reeves, C. D. 1907. The breeding habits of the rainbow darter (Etheostoma cœruleum Storer), a study in sexual selection. Biological Bulletin 14 (1): 35-59.

Sloss, B. L., N. Billington & B. M. Burr. 2004. A molecular phylogeny of the Percidae (Teleostei, Perciformes) based on mitochondrial DNA sequence. Molecular Phylogenetics and Evolution 32 (2): 545-562.

Stauffer, J. R., Jr, J. M. Boltz & L. R. White. 1995. The fishes of West Virginia. Proceedings of the Academy of Natural Sciences of Philadelphia 146: 1-389.

Winn, H. E. 1958a. Observation on the reproductive babits of darters (Pisces-Percidae). American Midland Naturalist 59 (1): 190-212.

Winn, H. E. 1958b. Comparative reproductive behavior and ecology of fourteen species of darters (Pisces-
Percidae). Ecological Monographs 28 (2): 155-191.


  1. I have a problem with this post: not that it's not clear in itself-- to the degree I can understand it, it's very clear-- but that I don't know enough ichthyotaxonomy to have an intuitive feel for what you are describing. A problem I have with other things I have read about teleost clssification/phylogeny: I see a bunch of technical names, and have no idea which taxa include fish I know by appearance and "common" name. Are you willing to indulge me by putting a few of my "acquaintances" in the proper slots?
    You say the Perco-whatsises share freshwater dominance with a number of other groups. Cypriniforms-- that includes carp, minnows, goldfish? Siluriforms I know: catfish. ... Now for some other stereotypical (Nearctic) freshwater fish. Trout (Salmoniformes) and Pike (Esoci...) are (is this right?) much more basal teleosts than any of the groups you mention? And what are the critters North American English-speakers call Bass and Sunfish? The mental image the word "Perch" calls up for me has strong fin-rays in the dorsal (so: zig-zag profile), so I'm going to guess that Sunfish and Bass are Perciwhatsises: is this right?
    (Sorry to be so ignorant: ignore me if you are annoyed by such questions!)

  2. I certainly have no problem with the question, though it brings up an interesting point about communication. Because I personally deal with organisms using technical names more often than vernacular names, I'm often more familiar with the former than the latter. My favorite story to illustrate that point involves a case where I asked a friend to identify a bird we were looking at out in the field, and he replied that it was a spinifex-bird. I'd never heard of such a thing. It wasn't until after I arrived home and looked it up that I discovered that a "spinifex-bird" was Eremiornis. Now if my friend had just said Eremiornis in the first place... The flipside of this, of course, is that I can forget that other people have the opposite bias. It's my fault if they don't understand me as a result, not theirs.

    That said, I have some very good reasons for avoiding vernacular names when it comes to fish - for some reason, fish names are just awful (I've complained about this before). The same names have been re-used all over the world for different fish. So when you ask about "bass" and "sunfish", it's good that you specified North American, because the fish that would come to mind for you when you say those names would not be the fish that come to mind for a New Zealander such as myself.

    Anyway, enough ranting. "Percomorpha" is a clade including the greater part of "Acanthomorpha", spine-finned fishes (and almost all the freshwater acanthomorphs). So as you mention, these are the fish that have hard spines rather than soft rays in the front of the dorsal fin. North American freshwater "bass" and "sunfish" belong to a family called Centrarchidae, and yes, they're also percomorphs.

    You're right about the contents of Cypriniformes and Siluriformes. Characiformes, the other major order of freshwater fish I referred to, includes the tetras and related fish. Not very significant in North America (if they're even found in that area), but pretty important in South America and Africa. These three orders belong to a clade called Otophysi whose members have a bony apparatus connecting the swim bladder to the inner ear. Salmoniformes (salmon and trout) and Esociformes (pike) are closer to Percomorpha than are Otophysi, but they're separate from either. You can use the tree at ToLWeb as a basic reference, if you like.

  3. Characiformes, the other major order of freshwater fish I referred to, includes the tetras and related fish. Not very significant in North America (if they're even found in that area)

    They are found there - barely. The Mexican tetra Astyanax mexicanus - the species that also has a blind cave form - reaches as far north as southern Texas. And I think that a few other tetra species are found in northern Mexico (i.e., in 'North America', biogeographically speaking). But yeah, tetras are not particularly significant components of the Nearctic fish fauna.

  4. Thanks for this post. Undergrad turned me off to fish mostly (even though my degree is in fisheries biology) but I do have my pet groups, including the darters. Since your former post on the Cypriniformes, I've come to realize just how screwed teleost taxonomy has become, what will all these junk taxa floating around in paraphyletic assemblages. Does this mean all these groups are going to be split off, like the disintigration of the formerly Mycetophilidae in Diptera (yet another group that needs major taxonomic work)?


  5. 'Smegmomorpha'!? Was this clade named by a red dwarf fan?

  6. That's interesting - while looking up the derivation of "Smegmamorpha", I discover that it doesn't refer to the clade that I thought it did. I thought that it included the Mugilidae (mullets) and Atherinomorpha (killifishes, livebearers, halfbeaks, flying fish, etc.), but it turns out that it was originally named to cover a larger group, and it was named after its contents - "Smegma" stands for "Synbranchoidei-Mastacembeloidei-Elassomatidae-Gasterosteiformes-Mugilidae-Atherinomorpha". Mugilidae and Atherinomorpha are probably closely related, but this broader Smegmamorpha is probably polyphyletic. Pity.

    Hopefully, the teleosts will be properly divided into monophyletic taxa over time. It's already pretty much happened for non-acanthomorphs (we're past the days when almost every marine non-acanthomorph teleost was included in the Clupeiformes, thank goodness). Unfortunately, relationships within the acanthomorphs (particularly the percomorphs) are still a mess of polytomies, and there's a lot of work yet to do sorting them out.

  7. Nice Etheostoma caeruleum photo. A few weeks ago my staff and I collected some examples of this species for a live museum display. I posted a decent habitat photo from the same outing a year ago. I'm always surprised at how small a stream can be and yet support a good population.

  8. You might want to take a look at a couple of papers from 2003 in Bulletin of Fisheries Sciences, Hokkaido University and 2007 in Copeia that formally addressed these issue before several of the papers you discuss. These studies also include diagnostic features, which some people prefer, although not really necessary because these are outside the scope of the ICZN.

    The solution has to be a restricted Perciformes that includes just a few hundred fishes (Percidae, Notothenioidei, and a few other enigmatic forms).

    Serraniformes is never going to be a reasonable solution for any group that includes Percidae or Serranids (sensu lato or stricto) because it will be forced to include 1000s of scorpaeniform species, which will be a more logical group to name that whole assemblage after when things get worked out.

    Imamura, H. and Yabe, M. 2003. Demise of the Scorpaeniformes(Actinopterygii: Percomorpha): An Alternative Phylogenetic Hypothesis. Bulletin of Fisheries Sciences, Hokkaido University 53: 107-128.

    Smith, W.L. and Craig, M.T. 2007. Casting the percomorph net widely: The importance of broad taxonomic sampling in the search for the placement of serranid and percid fishes. Copeia 2007:35-55.

  9. Doug, I hasten to stress that it is not my own photo. It is a good one, though, and the site linked has a number of good darter photos, plus a detailed description of how they were taken.

    I was aware of the Smith & Craig paper, but not of the Imamura & Habe one - thank you! As I said at the previous post, I suspect that Li et al. chose the name Serraniformes to avoid the baggage associated with the names Perciformes and/or Scorpaeniformes. I'm just happy to watch and see how this one pans out, though.

  10. Chris--
    Thank you for your indulgent (and helpful) reply! ... Common names for FISH aren't the only mess, of course: ask an Australian, an American and a Brit to identify a ROBIN!


Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="">FoS</a> = FoS