Subterranean Fishes of the World

Country

Types

Syntypes: MHNG 1213.97 (labelled "TYP?", Buckup, 1988), MHNG 1505.91 (65 slide preparations), FMNH 58580 (ex CM 7603). This species is the type species by monotypy of the genus Phreatobius. This set of specimens includes the two individuals collected by Goeldi in 1903 plus a further four discovered two years later at the same site by Miranda. All six were given to Fuhrmann who later gave one (now FMNH 58580) to Eigenmann. Three of the remaining five were made into slide preparations (now MHNF 1505.91), one preserved in alcohol (now MHNG 1213.97) and the last (now lost) preserved in tetralin. It is no longer possible to determine which two were used by Goeldi in the original description (Mahnert 1976:475).

Distribution

Type locality: A cistern (= well) at an unknown location on the Ilha de Marajó, Pará,  Brazil (Goeldi 1905). Recorded also from Belém city, Para (discovered there in 1963 by R. Arle (Thines and Proudlove 1986:732)) and Macapa city, Amapa to the south and north of Marajo respectively and 320 km apart.(Carvalho 1967,Trajano 1997). The distribution of this species appears to be restricted to an area in the mouth of the Amazon but is not known in detail (01^oS, 50^oW). A specimen collected in 1984 came from a well in Ananindeua, Pará (Mario de Pinna, pers. comm.).

The distribution of the genus Phreatobius is highly distinctive (see map in van der Sleen and Albert 2018:322). P. cisternarum is found only around the mouth of the Amazon at the very eastern side of Brazil, P. dracunculus is found only in the very west of Brazil in the Amazon headwaters and P. sanguijuela is found in Bolivia and in western Brazil both in Amazonian headwaters. There are no molecular data for P. sanguijuela but P. cisternarum and P. dracunculus share a clade with an undescribed species, Phreatobius sp. A (Sulivan et al. 2013). 

Habitat

The true habitat of this species is not known but it seems to be an inhabitant of underground water sheets around the mouth of the Amazon. All of the specimens so far collected were caught accidentally during the excavation of wells (Mario de Pinna pers. comm.).

Systematics

The relationships of the genus Phreatobius have been subject to debate since its description. In the original description of P. cisternarum Goeldi (1905) suggested that: "In total several features indicate that the fish belongs partly to the Trichomycteridae and partly to the Cetopsidae. Still other features are dominant enough to indicate an isolated position between these two groups". Fuhrmann (1905, 1905) disagreed and placed it amongst the Clariidae. Eigenmann (1918) stated that the species was a member of the subfamily Pimelodinae, and therefore the Heptapteridae, and that it was related to Heptapterus. Norman (1926) suggested that the relationships were uncertain. After a monumental anatomical study Reichel (1927) placed Phreatobius in the Siluridae but decided that it was sufficiently different from other members of this family to merit its own subfamily, the Phreatobiinae (originally formed as Phreatobinae Reichel 1927:383, corrected to Phreatobiidae by Fowler
1954:9). Myers (1944) retained the subfamilial status but considered it to reside within the Trichomycteridae. Myers and Weitzman (1966) stated that the species “is different from the Trichomycteridae in lacking a nasal barb” and that it was included in the Trichomycteridae “only on general appearance”. Chardon (1968) examined many siluriform fishes and after considering all of the evidence in front of him replaced Phreatobius into the Heptapteridae. Thines (1955, 1969) returned it in the Trichomycteridae. Carvalho (1967) followed Eigenmann (1918) in pacing the species in the Pygidiidae but within the monotypic subfamily Phreatobiinae. More recently Buckup (1988) has suggested that Phreatobius cisternarum be placed in the genus Heptapterus as it shares an anal fin ray count of 16 or more with other species in this genus. This action has not been wholly accepted and has not been adopted here. Lundberg, Bornbusch and Mago‑Leccia (1991) who placed the genus in the subfamily Rhamdiinae of the Heptapteridae. De Pinna (1998) postulates a relationship between Phreatobius and Horaglanis based on similar skull morphology. A recent photograph of Horaglanis krishnai shows that it is blood red in colour, a very rare condition for fishes (De Pinna 1998), as is Phreatobius and this may be a synapomorphy for a group containing the two genera. Clearly there is much to be learned about the exact systematic position of this interesting genus.

Sullivan et al. (2013) placed the genus in the Family Phreatobiidae and this is followed here and at the Catalog of Fishes. This is not followed in the book by van der Sleen and Albert (2018) who place Phreatobius as incertae sedis.

Biological Notes

Perez et al. (2020) found the following: "Histological analysis showed that the vestigial eyes of Phreatobius cisternarum contain a rudimentary lens. Transcriptome analysis revealed a repertoire of eleven visual and non-visual opsins and the expression of 36 genes involved in lens development and maintenance. In contrast to other cavefish species, such as Astyanax mexicanus, Phreatichthys andruzzii, Sinocyclocheilus anophthalmus and Sinocyclocheilus microphthalmus, neuromast staining patterns did not show an increase in the number of sensory hair cells" (from their Abstract). Further important results on this very significant subterranean fish, which has hardly been studied, are expected from this research group.

Museum Holdings

As above plus: MZUSP 28309 (Lundberg, Bornbusch and Mago‑Leccia 1991), MNRJ 11569 (De Pinna 1998).