Subterranean fishes of the world
Barbatula undecribed species from Germany.

A history of investigations and significant landmarks in the study of subterranean fishes

The oldest known association of fishes and caves are a wall painting of a flatfish in Cuevade la Pileta, Spain, dating from 20000 years ago and from a trout scrached in a clay floor at Cave Niaux, France,16 000 years ago (Nellen and Dulcic 2008).

But these are not cave fishes of course and the first known published account of a possible subterranean fish dates from 1436, in China, when Mao Lan mentioned the troglophile now known as Sinocyclocheilus grahami.

The second known reference to a subterranean fish is perhaps one published in China in 1541 (Jie Yijing 1541). In this publication mention is made of a transparent fish which was seen in the Alu Limestone Cave in Yunnan Province. This may be the species now known as Sinocyclocheilus hyalinus (see also Chen, Yang and Zhu 1994; Weber, Proudlove and Nalbant 1998).

The year 1569 saw the publication of “L’Art et science de trouver les eaux et fontaines cachees sovbs terre” by the French engineer Jacques Besson (Besson, 1569; see also Romero and Lomax 2000, 2000). He had observed small eels in various subterranean waters in France. These are clearly not true cave fishes but show that people were noticing such phenomena.

In 1584 the Dinaric karst fish Delminichthys ghetaldii (originally placed in the genus Paraphoxinus) was discussed in letters from Jakov Sorkocevic to Ulisse Aldrovandi (Grmek and Balabanic 2000, Nellen and Dulcic 2008, Tvrtkovic and Franicevic 2002, Gottstein-Matocec et al. 2001).

The fifth published account of a fish from a cave may be that of Kircher (1665) who reported on the presence of cave fish in an intermittent lake (Lavoie 1998).

Another Frenchman, the Marquis de Montalembert, recorded pike (Esox lucius: Esocidae) from a spring near to Gabard, Angoumois, western France. They were commonly blind or one eyed (Montalembert 1748; see also Romero and Lomax 2000).

Between 1799 and 1803 the German explorer Alexander von Humbolt visited South America. During this time he visited Ecuador (which he referred to as The Kingdom of Quito) and at some point in this visit there is the possibility that he was shown a blind cave-dwelling fish. In his account he describes the new species Pimelodus cyclopum (Humboldt 1805). This species is included by Packard (1888:106) in his list of cave animals and he comments: “Eyes more or less rudimentary”. It seems likely that Humboldt received only a description of a possible blind fish from someone else. Whatever the truth of this, he was of the opinion that the animal was ejected from the mouth of a volcano. We will probably never know if Humboldt saw a cave-dwelling fish but it is possible that he did. Hanson (1981) has suggested that Humboldt’s account may refer to Astroblepus pholeter. He thinks that the illustration of Pimelodus fissidens in Humboldt (1805) most closely resembles this species. Nico (2001) also discusses this case and suggests that P. cyclopum may be a synonym of A. pholeter. Girard (1889a) also discusses the subterranean fauna of this region. (See also Romero and Paulson 2001).

The next published account of a blind cave dwelling fishes is that of James Flint, who travelled in America in the early 1820s In his account (Flint 1820, reprinted as Thwaites 1904) he states that:

“... a Colonel C--- of Indiana told me that a settler in his neighbourhood, on digging a well, penetrated into a stream of water, and found blind fishes in it.”

The Amblyopsid fish Amblyopsis hoosieri is found in Indiana so we can be reasonably certain that this is an accurate account. Flint also mentions Mammoth Cave in his writings, although not by that name (he says that it is “known by the name of the great cave”), but since fishes were not seen here until 1838 or later (see below) he does not mention them. A most interesting foot note to this account says:

“Since the above was written, a notice of blind fishes has appeared (if I mistake not) in the memoirs of the Wernerian Society of Edinburgh.”

In the British Library in London are copies of the Memoirs of the Wernerian Natural History Society, Edinburgh from 1808 to 1838, published in Edinburgh between 1811 and 1839. I have examined all of these and am sad to say that Flint must have been mistaken as there is no mention of cave fishes anywhere in their pages.

The first cave‑dwelling fish to be properly described was discovered in the Echo River of Mammoth Cave, Kentucky, USA between 1838 and 1840. After crossing a deep shaft (The Bottomless Pit) the cave‑guide and explorer Stephen Bishop found a large deep river in which he observed small white fishes (Sides 2017). It is reported  that he would catch some to exhibit and possibly even give them to visitors to the cave: "One remarkable discovery was that the new river contained blind white fish a few inches long. Stephen caught some for exhibit" (Brucker and Watson 1980).

The first published account was that by the Reverend Robert Davidson in a small book entitled "An Excursion to the Mammoth Cave, and the Barrens of Kentucky" (Davidson 1840). He says:

 “After some time we approached a low arched passage, not more than a yard high, but a quarter of a mile in length, through which we crept on our hand and knees; a very fatiguing operation. The bottom was composed of a fine dry yellowish sand, instead of the stiff clay found elsewhere. After ploughing the sand, we emerged into a more roomy space; and a few more turns and descents brought us at last to The River.

This is a stream of water twenty feet wide, and they said as many deep. It was discovered only about a year ago. Its current is very sluggish, as has been proved by launching a piece of wood bearing a lighted candle, on its bosom. We were informed that a species of  white fish were found here without eyes, and the keeper of the hotel assured us he himself had seen them, but that their other senses were so acute, the slightest touch of water overhead was sufficient to alarm them, and make them dart off like lightening.

...Two of my acquaintances a week afterwards ... saw some of the white fish ... As for us, on our visit, we were not favoured with a site of these natural curiosities ...”

Thus we have the first sensible biological observations of  a cave fish.

The next mention of this animal is in volume 1 of the Proceedings of the Academy of Natural Sciences of Philadelphia where it is reported (Anonymous 1842:175) that:

"...a small white fish, also eyeless, (presumed to belong to a sub‑genus of Silurus) ... taken from a small stream called the 'River Styx,' in the Mammoth Cave, Kentucky, about 2 1/2 miles from the entrance. Presented by W. T. Craige, M. D."

It is very likely that this individual was collected by Stephen Bishop himself (Romero and Woodward 2005, Cohen 2021). The animal received a name the same year when James E. De Kay (De Kay 1842) gave a description of it and called it Amblyopsis spelaeus (see also Romero 2002). Whether the individual described by De Kay is the same as that presented to the Philadelphia Academy is not known.

Dr H. Smith (1867) saw some fishes in 1867 and formed to opinion that the were deaf as well as blind, but this is not the case (see also Anonymous 1910 and Olson 2017).

In 1852 the French explorer M. de Saint-Amant published an account of his explorations in what is now Washington State, USA (Saint-Amant 1854; relevant pages reprinted in The Journal of Spelean History 5:39-42). In a cave which he marks on a map of the region, but which is still unidentified, he reports that there were blind fishes. One of these was collected and exhibited in Vancouver but this is all we know of it. It seems unlikely that this was a true blind fish as claimed by Saint-Amant as the cave is located at about 47oN, well outside of the known latitude for cave dwelling fishes which are mainly tropical and subtropical. However it is not impossible that one day this cave, and its fishes, will be rediscovered and Saint-Amant proved correct.

The next fifty years revealed only a further four species. The first two in the genus Lucifuga from Cuba (L. subterraneus and L. dentatus; Poey 1858, see also Romero 2007), and the second and third members of the family Amblyopsidae. Typhlichthys subterraneus was found in a well at Bowling Green, Kentucky, USA, only thirty miles from Mammoth Cave (Girard 1859), and Typhlichthys rosae [subsequently Troglichthys, now Amblyopsis] (Eigenmann 1898). All of these species were studied in some detail by Carl H. Eigenmann during these years, his primary interest being in the eyes of Amblyopsis spelaea although he examined several aspects of their biology (distribution, variability, habitat, evolution etc.). Numerous papers resulted from these studies, culminating in the magnificent book "Cave vertebrates of America : A study in degenerative evolution" (Eigenmann 1909). This is still the most detailed summary of the structure and development of the eyes of a cave‑dwelling fish, and a delight to read. (The first subtitle of the chapter 1 is particularly interesting: "Caves in their relation to the rest of the universe" !).  A brief but interesting biography of Eigenmann has been written by Romero (1986).

The cave population of Poecilia mexicana from Cueva del Azufre (also known as Cueva de las Sardinas and Cueva de Villa Luz) has been known about since at least 1896 (R. R. Miller pers. comm.). For its modern rediscovery see below.

The Americas continued to dominate in the first twenty years of the present century. The first decade saw the description of two species from Brazil, the enigmatic and rare Phreatobius cisternarum from an island in the mouth of the Amazon (Goeldi 1905), and Typhlobagrus kronei (now Pimelodella) from São Paulo State (Miranda Ribeiro 1907). The fourth North American species to be described, Trogloglanis pattersoni, was collected from the strangest of all cave‑like habitats so far found. A series of very deep wells (up to 700m) had been drilled to penetrate an aquifer near San Antonio, Texas and the fishes appeared at the well outflows, usually dead as a result of the massive pressure changes involved in the ascent. T. pattersoni was described by Eigenmann (1919) in his last contribution to cave‑biology. These contributions, over twenty‑five publications in twenty years, are an achievement equalled by few other biologists in this field.

The first non‑American cave‑dwelling fish was a little cyprinid discovered by M. Geerts in the Belgian Congo (now Democratic Republic of Congo), Africa. Described by Boulenger (1921) as Caecobarbus geertsii this species has subsequently been studied in detail by several biologists (see below). Africa retained its prime position by revealing the next three cave‑dwelling fishes, all of them from Italian Somaliland (now Somalia). One fish now placed in its own unique family, Uegitglanis zammaranoi (Gianferarri 1923), and two cyprinids, Phreatichthys andruzzii (Vinciguerra 1924) and Barbopsis devecchii (Di Caporiacco 1926), still the only three species from that country, were discovered and described by Italian biologists. In subsequent years they discovered another two species (Eilichthys microphthalmus Pellegrin 1929 and Barbopsis stefaninii Gianferrari 1930) but these have since been shown to be synonyms of B. devecchii. Italian interest has remained strong and these species have been extensively studied by them in recent years (for a review see Ercolini, Berti, Chelazzi and Messana 1982). Caecorhamdia urichi, from the Guacharo Cave, Trinidad, was described by Norman (1926) in a paper which also contained a list of the cave‑dwelling fishes known at that time (see also Romero and Creswell 2000 and 2000 and Romero et al. 2002). Another work to contain a list of known cave fishes was published in the same year by Spandl (1926). In addition to this summary he also provided information on the non-troglobitic fishes from various areas of former Yugoslavia. These “Paraphoxinus”  are most interesting. The last discovery of the decade was of Caecorhamdella brasiliensis (Borodin 1927) from Sao Paulo, Brazil. This has subsequently been shown to be a synonym of Pimelodella kronei (Trajano and Britski 1993). 1927 also saw the publication of Reichel’s (1927) monumental anatomical study of Phreatobius cisternarum which, apart from a few subsequent notes, is the last work of significance on this species which was not rediscovered until 1963 (Thines and Proudlove 1986:732).

By far the most significant of early workers was Carl H. Eigenmann (1863-1927). He produced 31 publications on blind fishes in a 29 year period, 1890-1919, including the magnificent monograph Cave Vertebrates of America: A study in degenerative evolution. His life and work are recorded by Stejneger (1937). A much briefer summary of his work is provided by Romero (1986) and obituaries of Eigenmann are provided by Henn (1927) and Myers (1928). His wife, Rosa Smith Eigenmann, was a significant ichthyologist in her own right and indeed, according to another very significant ichthyologist, Carl Hubbs "Rosa Smith was ... the first woman ichthyologist of any accomplishments." (Hubbs 1964, cited in Maisel, Smart, Bamber and Hart 1997). One of the most significan blind fishes known, not a cave fish, is the Californian blind goby, Typhlogobius californiensis. It is fitting that Rosa Smith Eigenmann should have discovered this enigmatic little fish, which she did. She collected it from Point Loma near San Diego and named it Othonops eos (Smith 1881). Unfortunately for her, and for a fitting bit of history, the fish had been described two years before by the Austrian ichthyologist Franz Steindachner (Steindachner 1879) who named it Typhlogobius californiensis. Under the rules of the International Code of Zoological Nomenclature Sreindachner's name takes prescedence over Smith's name.

The 1930s revealed five species. The first, Typhleotris madagascariensis, was discovered on the island of that name and described by Petit (1933), it was the first of two species in this cave‑restricted genus. The second is now the most famous and intensively studied of all cave‑dwelling animals. Anoptichthys jordani (now Astyanax jordani) was discovered by Salvador Coronado in La Cueva Chica, reported by Basil Jordan to Carl Hubbs and William Innes, and descibed by them in 1936. At this time only one population was known but it is now known from over 30 sites (Mitchell, Russell and Elliott 1977). The first person to study this animal in detail was Charles Breder. Romero (1986) provides an interesting biography of Breder.

Clarias cavernicola from Namibia was described in the same year as A. jordani but, in contrast, has been studied hardly at all and to date less than ten publications are available. We still know very little about this animal although this is perhaps not too surprising given its location.

In the 1930s the Mexican state of Yucatan was visited by A.S. Pearse and colleagues. They collected two species of cave dwelling fishes which were described by Carl Hubbs (1938). A Brotulid (now Bythitid) Typhlias pearsei has more recently been studies by the German worker Wilkens who, together with his colleagues Parzefall, and the Peters', has done more research on cave fishes than anyone since Eigenmann. The second of Pearses' fishes was Pluto infernalis (subsequently Furmastix and Synbranchus now Ophisternon), the first of three known species of troglobitic synbranchid eel. Throughout the century there have been reports of other eel‑like fishes in caves and it is likely that further species will turn up. A history of Hubbs's interest in cave fishes is provided by Romero and Moseley (2023). Note that there is an error in their Table 1 and the genus Typhliasina was coined by Whitley (1951) as replacement of Hubbs's original Typhlias. See the entry for Typhlias pearsei as to why there was no need to provide a replacement name!

Tomiyama (1936) provided an account of the Gobiid fishes of Japan. In it he described and figured two subterranean fishes, one from a well the other from a cave. Although his descriptions are clearly of distinct fishes he considered them as forms in the polymorphic subspecies Luciogobius guttatus guttatus. It was four years later that Regan (1940) elevated them both to specific status as Luciogobius albus (cave) and L. pallidus (well). It was therefore one hundred years after the discovery of Amblyopsis that the east was known to have cave‑dwelling fishes.

In 1937 the Belgian colonial authorities in the (then) Belgian Congo placed Caecobarbus geertsii (Cyprinidae) on a list of protected animals (together with the more prominent Gorilla and Chimpanzee) (Frechkop 1941).This was the first African fish to receive some statutory protection.

Garra typhlops (originally as Iranocypris) was discovered in 1937 and publication was due in 1944. The Second World War stopped this and it was not until 1948 that the paper appeared. It carries the date 1944 and this is the official date of publication.

The “modern”, post-war, era was marked by the discovery of the first (of two) Australian species, Milyeringa veritas, from a man-made well. Whitley (1945) erected a unique family the Milyeringidae for this animal but it was based in most part on troglomorphic features and is not now considered useful. Milyeringa veritas has spent most of its taxonomic life in the Eleotridae and the Gobiidae but is now placed in its own Family, Milyeringidae, with the Madagascar subterranean species in the Genus Typhleotris.

The second species from the deep aquifer of Texas, Satan eurystomus was described by Hubbs and Bailey (1947). This was the second cave fish to be described by Carl Hubbs, probably the twentieth century’s most prolific ichthyologist.

The first Indian species to be discovered was Horaglanis krishnai (Menon 1950 1951, 1951). This is a remarkable animal which appears to be intermediate between two catfish families.

A paper by Gordon and Rosen in the journal Copeia in 1962 rediscovered the cave population of the common molly Poecilia sphenops (Poecilia mexicana herein). They descibed the cave and its biota but it was left to Walters and Walters (1965) to look at the behaviour of the fishes. Since the 1970s this animal has received a lot of attention from the German group in Hamburg, and particularly from Jakob Parzefall and his students who have studied it in great detail. It is surprising that it took until the 1960s for this rediscovery to take place as it was known in 1896 (see above).

1969 saw the publication of the most comprehensive, and arguably the most important, book ever published on cave fishes. Georges Thines’ “L’evolution regressives des poissons cavernicoles et abyssaux” was based on his 1955 paper published in the Annales Societe Royale Zoologique de Belgique. It provided an essential summary of knowledge and is still very valuable. Its main fault is the emphasis on regressive evolution as a central theme and the concomitant of including all eyeless fishes. This results in fishes from caves, from the deep sea, and from other habitats (e.g. mud burrows, crevices in river beds etc.) being lumped together when in reality they share very little other than the lack (or reduction) of an eye and, possibly, melanin pigment. A study of the ecologically coherent hypogean habitat (and including fishes with eyes if they are deemed part of the habitat), or the ecologically coherent deep sea (and including fishes with eyes if they are deemed part of the habitat) etc. would seem to be a more biologically sensible approach. Nevertheless, Thines (1969) is the reference source.

In 1981 the cyprinid Caecobarbus geertsii from Democratic Republic of Congo, west Africa was listed in Appendix II of the CITES agreement (CITES 1981). This was as a result of it being a popular aquarium novelty. It was recognised that protection was required if the caves from which it was collected were not to be totally depleted of fishes. Now it is necessary for an export permit from the Democratic Republic of Congo government to be granted before export is allowed. Whether this actually provides any protection is debatable. Although the system allows specimens to be tracked after removal from the cave it does not stop the removal. It seems unlikely that fishes removed will ever go back even if export is disallowed. Nonetheless it was an important recognition that cave fishes are very vulnerable. At the time it was suggested that all cave fishes then known, with the exception of “Anoptichthys jordani” ( = Astyanax jordani herein), should be included in CITES Appendix II but “The second part of this proposal (inclusion of all other species of blind cave fishes ...) was withdrawn by the UK delegation. (Note from the Secretariat)” (footnote to CITES 1981). Why this second part was withdrawn is not known to me.

In the period 1977-1996 the International Union for Conservation of Nature and Natural Resources (IUCN) have produced five Red Lists of threatened animals. The number of cave fish appearing in these lists has altered dramatically during this time. In the first (1977) only 7 were listed, 8% of all known at the time. Subsequent lists contain: 1988: 8 (16%); 1990: 31 (51%); 1994: 31 (47%); 1996: 61 (85%). This demonstration that cave fishes are among the most vulnerable of all fishes is welcome. Another indication of their threatened status is the special section in the WWF sponsored book “Global Biodiversity” (Banister and Groombridge and 1992). The authors considers cave fishes to be particularly vulnerable and mark them down for particular consideration.

Subterranean species represent only a small proportion of all known fishes. These animals will always be a very insignificant part of the total piscine biodiversity but because of their relatively limited distributions they will always be very high up any list of threatened biodiversity. So far there have been no known instances of extinction but it may not be long before one has to be recorded.

Romero (2001 and 2009) also provides histories of research on subterranean fishes.

The increase in described species and in publications relating to them

Figures 1 to 4 show how the numbers of described species has increased over time. The curve shows two clear trends. From 1920 to 1980 the gradient was constant with the addition of 5 species per decade. After 1980 the gradient increased and four times as many species (20 per decade) have been descibed since. This increase is attributable to two factors. The number of caving expeditions has increased and the number of caves explored in the past 20 years is much greater than previously. Second, the Chinese have discovered fishes in their extensive limestone areas and 14 species have been described since 1979. Further extrapolation is less certain but by 2020 there may be 135. The curve for the number of publications relating to cave fishes closely parallels that for described species. The major difference is that it appears to steepen in about 1960, 20 years before that for species. This is probably as a result of the increase in the size, and in the funding, of Universities since the beginning of the 1960s.

Significant landmarks in the study of subterranean fishes

Year Event
1436  Cave fishes first noted in south Yunnan, China, by Mao Lan, it may have been Sinocyclocheilus grahami.
1541 A blind fish observed in a Chinese cave by Yingjing Xie, it may have been Sinocyclocheilus hyalinus.
1569 French engineer Jacques Besson reported small eels in various underground waters in France.
1665 Athanasius Kircher reported on a cave fish in an intermitant lake where.
1748  The Marquis de Montalambert recorded Pike (Esox lucius) emerging from a spring in western France.
1787 Brown trout (Salmo trutta) recorded in the subterranean Demanovka river in Vyvieranie Cave, Slovakia by Buchholtz.
1805 Alexander von Humboldt reported on a blind fish in Ecuador which he thought had been thrown out of a volcano. It may have been Astroblepus pholeter.
1820 James Flint observed blind fishes, almost certainly in the Amblyopdidae, in Indiana, USA.
1840 The famous Mammoth Cave guide Stephen Bishop collected blind fishes within the cave and this was recorded by R. Davidson.
1842  Description by DeKay of Amblyopsis spelaea, the first described subterranean fish, from Mammoth Cave, USA.
1858 Description by Poey of Lucifuga subterranea and Lucifuga dentata from Cuba, the second and third described species.
1899 First year when total references gets to 10.
1900 101 references in 58 years.
1909 Eigenmann’s major review of North American cave fishes.
1920 The last year with zero publications.
1936 Discovery of Astyanax jordani by Salvador Coronado, the best studied subterranean fish.
1940 10% of the current known total species described.
1942 The first 100 years: 22 species described and 238 references published.
1945 260 publications which is 10% of the current total.
1955 The first review of the world subterranean fish fauna by Georges Thines (xx species).
1966 Barr provides commentary on the evolution of cave biology in the USA 1822-1965
1969 Thines update of world species (xx species but including non-cave troglomorphs and abyssal species).
1977 First review of the distribution and biology of the Mexican cavefish Astyanax jordani by Mitchell, Russell and Elliott.
1979 The first subterranean fish from China described.
1980 20% of the current known total species described.
1982 The species discovery curve steepens dramatically, probably resulting from an increase in exploratory caving expeditions and interest by Chinese biologists. Publications reach 1000 after 140 years.
1985 The last year in which no new species were described.
1986 The first cave Sinocyclocheilus described in China. Thines and Proudlove review the world fauna (92 species, but including non-cave troglomorphs).
1988 Second major review of the biology of Astanax jordani, concentrating on the genetics, by Wilkens.
1998  Weber, Proudlove and Nalbant review of he world fauna (xx species).
2000 Weber review of the world fauna (xx species).
2001 Romero and Paulson review of the world fauna (86 species).
2002 50% of the current known total species described.
2004 First year in which the number of described species reached double figures (10).
2006 Proudlove review of the world fauna (103 species)
2009 Zhao and Zhang review of the Chinese species of Sinocyclocheilus.
2010  Proudlove review of the world fauna (164 species) in the book Biology of Subterranean Fishes by Trajano, Bichuette and Kapoor.
2013 Largest number of species described in one year (22), which is equivalent to the first 98 years of known subterranean species.
2013 Review of cave fishes in Guangxi, China by Lan, Gan, Wu and Yang.
2016 Third major review review of the biology and evolution of Astyanax jordani by Keene, Yoshizawa and McGaugh. Total known species is 217.
2017 Fourth major review of the biology and evolution of Astyanax jordani by Wilkens and Strecker.
2017 A troglomorphic fish discovered in a German cave at 47oN and in glaciated terrain by Behrmann-Godel et al.
2018 Fifth major review of the biology of Astyanax jordani by Elliott.
2019 Aenigmachanna gollum is the 250th known subterranean fish.

The bibliography of works relevant to subterranean fishes (begun in 1979) passed 4000 entries


Some statistics

The genera which are restricted to subterranean habitats

The following Table lists the genera of fishes which are only found in subterranean habitats, i.e they are hypogean-restricted (HR).

Genera Number of species
Barbopsis 1
Caecieleotris 1
Caecobarbus 1
Caecocypris 1
Caecogobius 1
Cryptotora 1
Draconectes 1
Longanalus 1
Pseudosinocyclocheilus 1
Satan 1
Speolabeo 1
Speonectes 1
Speoplatyrhinus 1
Stygichthys 1
Troglichthys 1
Troglocobitis 1
Troglocyclocheilus 1
Trogloglanis 1
Typhlias 1
Typhlobarbus 1
Uegitglanis 1
Amblyopsis 2
Eidinemacheilus 2
Horaglanis 3
Kryptoglanis 2
Lucifuga 7
Milyeringa 2
Phreatobius 4
Prietella 2
Protocobitis 3
Troglonectes 7
Typhleotris 3
Typhlichthys 12
33 genera hypogean-restricted 42 species in HR genera (18% of all genera)