Fish do not belong to a single genus. The word “fish” is an informal label covering an enormous range of vertebrates spread across thousands of genera, hundreds of families, and multiple classes that diverged from one another hundreds of millions of years ago. A salmon, a shark, and a hagfish are all called fish, but they are about as closely related to each other as a lizard is to a frog. Understanding fish classification means understanding why biologists abandoned the idea of one tidy “fish” category long ago and how modern tools keep redrawing the lines between groups.
Why There Is No Single Group Called “Fish”
In everyday language, a fish is anything with fins and gills that lives in water. In evolutionary terms, that description picks up several lineages that split apart deep in the vertebrate family tree. The group we casually call fish is what biologists call paraphyletic: it includes some descendants of a common ancestor but leaves out others (notably, land-dwelling vertebrates like us, which evolved from lobe-finned fish). Because a genus is a narrow rank just one step above species, asking “what genus are fish?” is a bit like asking “what genus are mammals?” The answer is that there are thousands of them, each containing a cluster of closely related species.
FishBase, the largest online database of fish species, currently lists well over 35,000 described species. Those species are organized into roughly 5,000 genera, which themselves sit within more than 500 families and dozens of orders. New species and new genera are still being described every year, particularly from deep-sea surveys and tropical rivers that remain poorly sampled. So the classification of fish is not a settled map but a work in progress.
The Major Lineages
The living animals we call fish fall into a handful of deeply separated branches. The broadest split is between jawless fish and jawed fish. From there, jawed fish divide into cartilaginous fish (sharks, rays, and chimaeras) and bony fish, and bony fish further divide into ray-finned fish and lobe-finned fish. Each branch contains its own set of orders, families, and genera.
Jawless Fish
Lampreys and hagfish are the only surviving jawless fish, collectively called cyclostomes. They diverged from other vertebrates before the evolution of hinged jaws or paired fins, making them a kind of living window into very early vertebrate life.1PubMed. Evolutionary crossroads in developmental biology: cyclostomes (lamprey and hagfish) Despite both being jawless, lampreys and hagfish are so anatomically different from each other that reconstructing what their common ancestor looked like has proven difficult. No fossil has been found that sits between them on the tree without clearly belonging to one group or the other.2Zoological Science. Early Jawless Vertebrates and Cyclostome Origins Together they account for only a small number of genera and around 120 species, a tiny fraction of fish diversity.
Cartilaginous Fish
Sharks, rays, skates, and chimaeras make up the class Chondrichthyes. Their skeletons are made of cartilage rather than bone. This group includes roughly 1,200 described species across several hundred genera. Though far less species-rich than bony fish, cartilaginous fish occupy a wide range of ecological roles, from filter-feeding whale sharks to bottom-dwelling electric rays. DNA barcoding studies have confirmed that major clusters within cartilaginous fish (chimaeras, rays, sharks) form distinct groupings that can be reliably separated by genetic distance.3PubMed Central. DNA barcoding Australia’s fish species
Ray-Finned Fish
This is where the real numbers live. Ray-finned fish (Actinopterygii) account for about half of all living vertebrate species. Nearly all of them are teleosts, the group that includes everything from tuna and trout to seahorses and eels.4PubMed Central. Resolution of ray-finned fish phylogeny and timing of diversification A time-calibrated family tree of ray-finned fish shows that much of their diversification into today’s groups happened between the late Mesozoic and early Cenozoic, a burst of evolutionary activity sometimes called the “Second Age of Fishes.”4PubMed Central. Resolution of ray-finned fish phylogeny and timing of diversification When someone asks about the genus of a bass, a clownfish, or a catfish, they are asking about a tiny corner of this enormous radiation.
Lobe-Finned Fish
Only a handful of lobe-finned fish species survive today: two species of coelacanth and six species of lungfish. The coelacanth was thought to have gone extinct roughly 70 million years ago until a living specimen was hauled up by a fishing trawler off South Africa in 1938. Genome sequencing of the African coelacanth (Latimeria chalumnae) later showed that lungfish, not coelacanths, are the closest living fish relatives of land-dwelling vertebrates.5PubMed Central. The African coelacanth genome provides insights into tetrapod evolution That finding underscores a recurring theme in fish classification: outward appearance can be deeply misleading about evolutionary relationships.
Extinct Lineages That Shaped the Tree
Living fish represent only a fraction of the lineages that once existed. Placoderms, the armored jawed fish of the Paleozoic, were the dominant vertebrates of the Devonian period (roughly 420 to 360 million years ago). They first appeared in the Early Silurian, with their main burst of diversification in the Early Devonian, eventually spanning at least 335 known genera before going completely extinct.6Annual Review of Earth and Planetary Sciences. Placoderms (Armored Fish): Dominant Vertebrates of the Devonian Period Understanding these lost groups matters because they fill in the branching points on the vertebrate family tree. Without placoderms, for instance, the relationship between cartilaginous and bony fish would be much harder to reconstruct.
How DNA Is Redrawing the Map
For most of the history of fish taxonomy, scientists sorted species into genera and families based on what the animals looked like: tooth shape, fin-ray counts, scale patterns, skull geometry. That approach worked well enough for many groups, but it hit a wall whenever unrelated fish evolved similar features in response to similar environments. Over the past two decades, DNA-based methods have upended parts of the classification that seemed settled.
DNA barcoding, which identifies species by comparing short, standardized gene sequences, has become a routine tool. In a large Australian study, every fish species tested could be distinguished by its barcode sequence, and species within the same genus consistently clustered together.3PubMed Central. DNA barcoding Australia’s fish species Similar results have been confirmed elsewhere, including validation of scad species in the genus Decapterus and fish assemblages in India’s Chilika Lagoon, where barcoding discriminated species with high accuracy and flagged cases where look-alike fish turned out to be genetically distinct.7Malaysian Journal of Science. DNA Barcoding in the Validation of Scad Species Identification (Genus: Decapterus) in Ambon8Genetics of Aquatic Organisms. DNA Barcoding for the Assessment of the Taxonomy of Fish from Chilika Lagoon, India
Barcoding works well for identifying known species, but resolving deeper relationships, such as which orders and families are truly related, requires whole-genome data. A recent genomic study tackled one of the longest-running controversies in teleost classification: the branching order among the deepest lineages. By assembling high-quality genomes of tarpons and eels (Elopomorpha), researchers showed that eels and the bony-tongued fish (Osteoglossomorpha, which includes arapaima) form a group that is sister to all remaining teleosts, a finding that resolved more than 50 years of debate.9PubMed. Genome structures resolve the early diversification of teleost fishes Even more recently, a massive genomic project covering 464 species has challenged the traditional grouping of catfish and electric knifefish, suggesting that knifefish may actually be more closely related to characins (tetras and piranhas) than to catfish.10The Innovation. Large-scale genomic sequencing and comprehensive analyses provide insights into teleost evolution and phylogeny Findings like these mean that some genera, families, and even orders that appear in current textbooks will need to be rearranged.
When Body Shape Misleads
One of the biggest headaches for fish taxonomy is convergent evolution: unrelated species independently evolving similar body plans because they face similar ecological pressures. A streamlined, torpedo-shaped body is useful for fast open-ocean swimming, so it shows up in tuna, mackerel, and billfishes, groups that look superficially similar but sit on very different branches of the teleost tree. Among open-ocean fishes in the clade Pelagiaria, species from many different families fall into just three distinct skull-shape clusters, with high convergence within each cluster and low evolutionary signal from ancestry.11Evolution. How to tuna fish: constraint, convergence, and integration in the neurocranium of pelagiarian fishes In other words, the skull shapes reflect lifestyle, not family ties.
A striking example involves billfishes (marlins, swordfish) and flatfishes (flounders, soles). Billfishes were long grouped near tunas on the basis of body form and warm-bloodedness, but molecular analysis places them as close relatives of flatfishes and jacks, a grouping nobody would have guessed from looking at the animals. Their common ancestor apparently gave rise to both the elongated bills and heater organs of marlins and the bizarre asymmetric skulls and eye migration of flounders.12PubMed. Evolutionary affinity of billfishes (Xiphiidae and Istiophoridae) and flatfishes (Plueronectiformes)
Convergence has misled genus-level classification too. In the high-altitude lakes of the Tibetan Plateau, schizothoracine fish (snow trout relatives) were traditionally split into genera based on whether they had a terminal mouth with dense gill rakers (open-water feeders) or an inferior mouth with a sharp horny jaw sheath (bottom scrapers). Molecular work showed that these feeding-related traits evolved multiple times independently, meaning the old genera did not reflect actual evolutionary groupings.13PubMed. Convergent evolution misled taxonomy in schizothoracine fishes (Cypriniformes: Cyprinidae) A parallel story played out with piranhas and pacus: reliance on tooth shape to distinguish genera turned out to be confounded by repeated dietary convergence, leaving several piranha genera non-monophyletic, meaning the genus contained species that were not each other’s closest relatives.14Systematic Biology. Phylogenomics of Piranhas and Pacus (Serrasalmidae) Uncovers How Dietary Convergence and Parallelism Obfuscate Traditional Morphological Taxonomy
Rapid Speciation and Genus Boundaries in Cichlids
If convergent evolution makes classification harder by making unrelated fish look alike, explosive speciation makes it harder in a different way: producing hundreds of genuinely distinct species so recently that the genetic differences between them are tiny. The cichlids of Africa’s Great Lakes are the textbook example. In Lake Tanganyika, roughly 240 endemic cichlid species evolved within the lake itself, diversifying in successive pulses that affected body shape, jaw structure, and pigmentation at different times.15Nature. Drivers and dynamics of a massive adaptive radiation in cichlid fishes In the Lake Victoria region, a single cichlid lineage radiated into every available lake, with the largest burst happening in a lake less than 16,000 years old. That radiation was fueled by repeated cycles of lineage mixing through hybridization and splitting through speciation.16PubMed. Cycles of fusion and fission enabled rapid parallel adaptive radiations in African cichlids
Drawing genus boundaries within these radiations is genuinely difficult. Two cichlid species that occupy completely different ecological roles and look nothing alike may be closer genetic relatives than two species in what was traditionally treated as the same genus. Taxonomists working on cichlids regularly propose new genus assignments, then see those assignments challenged by the next round of genomic data. The cichlid situation is extreme, but it illustrates a broader truth about fish genera: they are hypotheses about relationships, not fixed truths, and new evidence can change them.
Hybridization Blurs the Lines
Fish hybridize more readily than most vertebrates. When two species interbreed and produce viable offspring, the resulting hybrids can have intermediate features that muddle species boundaries and confuse genus assignments. In the Palearctic genus Rutilus (roaches and related cyprinids), genomic work has uncovered extensive hybridization both within the genus and between Rutilus and other genera, producing widespread mismatches between the mitochondrial DNA a fish carries and what its nuclear genome says about its ancestry.17Diversity and Distributions. Genomic Insights Into Diversity, Phylogeny, Hybridization and Evolutionary History of Palearctic Fish Genus Rutilus (Leuciscidae) This kind of mismatch, called mito-nuclear discordance, means that a simple DNA barcode drawn from one gene can place a fish in the wrong genus.
Hybridization is not just an academic nuisance. It has practical conservation implications. If a narrow-range endemic species is being swamped by genes from a widespread relative, the endemic’s distinct genetic identity can be eroded. Recognizing when this is happening requires accurate genus and species-level taxonomy in the first place.18FishTaxa – Journal of Fish Taxonomy. Exploring the Impact of Hybridization on Fish Taxonomy and Conservation Strategies
Cryptic Species Hiding Inside Known Genera
Sometimes the problem is not that we have placed a fish in the wrong genus, but that what we thought was one species turns out to be several. Cryptic species look alike but are genetically distinct enough to qualify as separate species. This is especially common in environments where natural selection strips away visible differences. The southern cavefish, Typhlichthys subterraneus, was long treated as a single widespread species. Multilocus genetic data revealed that it is actually a complex of several cryptic lineages, including support for resurrecting T. eigenmanni as a separate species.19PubMed. Delimiting species using multilocus data: diagnosing cryptic diversity in the southern cavefish, Typhlichthys subterraneus (Teleostei: Amblyopsidae) Because subterranean fish converge on the same suite of traits (reduced eyes, pale coloration), morphology alone cannot distinguish them. Many of these cryptic lineages remain formally undescribed and have not been assessed for conservation status, leaving them vulnerable to extinction before they are even officially named.20Biodiversity and Conservation. Doomed before they are described? The need for conservation assessments of cryptic species complexes using an amblyopsid cavefish (Amblyopsidae: Typhlichthys) as a case study
Cryptic diversity is not limited to caves. Deep-sea anglerfishes (order Lophiiformes) are another group where novel features like male dwarfism, bioluminescent lures, and bizarre reproductive strategies allowed rapid diversification in the food-poor deep ocean, producing lineages that can be difficult to separate by appearance alone.21PubMed Central. Evolutionary history of anglerfishes (Teleostei: Lophiiformes): a mitogenomic perspective Anywhere selective pressures push different fish toward the same body plan, or where the environment offers little variation to drive visible divergence, cryptic species are likely waiting to be discovered.
Why Accurate Classification Matters for Conservation
Getting genera and species right is not just an exercise in academic tidiness. Legal protections for endangered fish flow from taxonomy. The Endangered Species Act, the IUCN Red List, and international trade agreements like CITES all operate at the level of named species and recognized evolutionary units. If two populations that need separate protection are lumped under a single species name, one can quietly vanish while the other’s healthy numbers mask the loss. Bonefishes illustrate the problem well: lingering taxonomic confusion within the group has hampered conservation prioritization, because it is hard to protect a species whose boundaries have not been settled.22Fisheries. Lingering Taxonomic Challenges Hinder Conservation and Management of Global Bonefishes
Animals Called “Fish” That Are Not Fish at All
Adding to the confusion for non-specialists is the long list of animals with “fish” in their common name that are not fish by any biological definition. Jellyfish are cnidarians. Starfish are echinoderms. Cuttlefish are mollusks. Silverfish are insects. Shellfish is a culinary term applied to mollusks and crustaceans alike. Even “dolphinfish” (mahi-mahi) is a real fish, but dolphins themselves are mammals. These names are holdovers from an era when sailors and fishers classified anything hauled from the sea as some kind of fish.23Fisheries. Red Fish, Blue Fish The persistence of these common names is a reminder that everyday language and scientific classification serve different purposes, and they often land on different answers to the question “what is a fish?”
For the actual fish, the picture is clearer in principle but constantly shifting in detail. Every genus is a working hypothesis, supported by the best available anatomical and genetic evidence but subject to revision as new data arrive. The pace of that revision has accelerated sharply, with whole-genome sequencing now capable of reshuffling relationships that stood for decades. For anyone looking up the genus of a particular fish, the practical advice is to check a current database like FishBase or the Catalog of Fishes, and to expect that the answer you find today may have a different genus name attached to it tomorrow.