Freshwater Fishes of Iran

Introduction - Geography

Revised:  26 June 2007

Acknowledgements     Purpose     Materials and Methods     History of Research     Fisheries     Geography     Climate     Habitats     Environmental Change     Drainage Basins     Scientific Names     Fish Structure     Collecting Fishes     Preserving Fishes     Quotes

Iran is the second largest country in Southwest Asia (after Saudi Arabia with less than 20 freshwater fish species), has an area of 1,648,000 sq km and ranks fourteenth in the world in size, nearly as large as the British Isles, France, Italy and Spain combined (Firouz et al., 1970). It lies between latitudes 25N and 40N and longitudes 44E and 63E. Its northern border is shared with the former U.S.S.R. (Armenia (35 km long) and Azerbaijan (611 km) in the west opposite Iranian Azarbayjan, and Turkmenistan (992 km) in the east opposite Mazandaran, Golestan and Khorasan) and includes the southern part of the Caspian "Sea", by far the world's largest lake (436,284 sq km) and one of the deepest (1025 m). The Iranian coastline extends for 740 km. The eastern border is shared with Afghanistan (936 km) and Pakistan (909 km). The southern border fronts on the Sea of Oman and the Persian Gulf, a coastline of 2440 km. The western border is with Iraq (1458 km) in the south and Turkey (499 km) in the north. Much of Iran lies at an average altitude of about 1000 m, a feature found only in a few countries world-wide. Only Khuzestan, the Caspian Sea coast and the Persian Gulf coast form lowlands. These lowlands are quite narrow, often less than 20 km wide. Mountains are the most prominent feature of the Iranian landscape. The two major chains are the Alborz or Elburz, which rim the Caspian Sea basin in the north, and the Zagros which form a chain down the western side of the country. Inland of these chains lies the Iranian plateau, which is flanked on the east and south by lesser chains of mountains. The country has been likened to a bowl or saucer. This central plateau has extremely high summer temperatures and often very cold winters. The deserts of this plateau are barren and among the driest in the world. Rain falls only in winter. The terminal basins for streams and springs may be dry for years. There are extensive salt crusts, known as kavirs, over black, slimy mud and large areas are composed of hard, gravel plains known as dashts, prominently the Dasht-e Kavir and the Dasht-e Lut. Water is scarce in these regions, often restricted to small streams and springs. Larger rivers have their source in distant mountains. Between the Tigris and the Indus, only the Hirmand River on the Afghanistan border is large enough to be a river on a world scale - various "rivers" in the intervening area are really small streams easily fordable on foot for much of the year.

The total renewable water resource of Iran is estimated as 137.5 km3/year. 9 km3/year are through transboundary rivers such as the Hirmand, Tedzhen and Aras and about 10 km3/year is surface runoff to other countries notably Iraq. More than 1900 km or 22% of Iran's borders are rivers (Chavoshian et al., 2005).

Fisher (1968) gives a general, physical geography and Breckle (1983) gives a general account of the features and life (excepting fishes) of deserts and semi-deserts in Iran. Barthold (1984) gives an historical geography of Iran and Yarshater (continuing) has many articles on geographical features. Geological literature is summarised in Dürkoop et al. (1979) and Davoudzadeh (1997).

It is pertinent here to interject a note on geographical names. Transliteration of Farsi place names into English is possible by more than one system. This results in variant spellings for geographical features in articles and on maps of Iran. For convenience, I have followed the official standard names approved by the U.S. Board on Geographic Names. The Board publishes a gazetteer for Iran with a designation of the geographical place (e.g. lake, populated place, stream, spring, etc.) and its latitude and longitude. The latest gazetteer is available from the Defense Mapping Agency, Combat Support Center, Washington, D.C. 20315-0010. Some literature localities could not be identified from maps or gazetteers. They are placed in quotes (".....").

I have not included the diacritical marks used in the Board's system. They would be of little help to those unfamiliar with Farsi and perhaps unnecessary to those who are. Needless to say, there are variant diacritical marking systems and in any case pronunciation varies throughout Iran.

The situation is further complicated by transliterations into other European languages and readers should be aware of this when reading non-English papers on Iran or Iranian fishes, e.g. the English Shiraz is Chiraz in French, and Genu, the type locality of Aphanius ginaonis, has such variants as Ginau, Genow, Gueno, Geno, and finally Ginao from the German transliteration, hence the trivial name. As if this were not enough, the vagaries of political fortune are writ large upon the face of Iran (which used to be Persia). Bandar-e Pahlavi has reverted to its older name of Anzali (often spelt Enzeli on older maps), Reza'iyeh to Orumiyeh (= Urmia in older English literature), and Shahreza to Qumisheh after the fall of the Pahlavi Dynasty in 1979. Other variants are Bandar-e Khomeyni (formerly Bandar-e Shahpur), Bakhtaran (formerly Kermanshah), and Khuninshahr or City of Blood (formerly Khorramshahr or City of Joy, and again Khorramshahr). I have retained names current for the years 1976-1979 recorded in the Board's gazetteer (1984). One exception is the province of Hormozgan (or Hormozdgan) which I have preferred for its brevity over the older name on some maps of Saheli-ye Jazayer va Banader-e Khalij-e Fars va Darya-ye Oman! The province of Mazandaran is now split into two with the eastern part termed Golestan. Iranian governments have a distressing tendency to change the names and borders of provinces. The provinces are as follows:-


Azarbayjan-e Bakhtari (= Azarbayjan-e Gharbi or West Azarbayjan)

Azarbayjan-e Khavari (= Azarbayjan-e Sharqi or East Azarbayjan)


Mazandaran (now split to include Golestan in the east)





Kermanshahan (or Bakhtaran)


Markazi (= Central or Tehran; sometimes split into Tehran and a southeast part called Markazi)




Ilam (or Ilam va Postkuh)



Chahar Mahall va Bakhtiari

Bovir Ahmadi va Kohkiluyeh (or Boyer Ahmadi-ye Sardsir va Kohkiluyeh)




Bushehr (or Khalij-e Fars)

Hormozgan (or Hormozdgan or Saheli-ye Jazayer va Banader-e Khalij-e Fars va Darya-ye Oman)

Sistan va Baluchestan

Another complication is the tendency for long rivers to have several names along their course, sometimes taken from the nearest population centre, and for locally used names to be different from map or gazetteer names. Names also vary with language and through time. One of the major rivers of Fars Province appears on maps as the Mand River, but near Shiraz it is called by its Turkic name Qarah Aqaj (also transliterated Qara Aghach, Qareh Aghaj, Qara Agach, Qareh Aqaj, Qareh Aqach, Kara Agach, and Kara Agaj). The Kor River, also in Fars, is known in older papers as the Araxes River which is not the same as that forming the border between Iran and the former U.S.S.R. (which anyway is often spelt Aras or Araks!).

The early geological history of Iran and neighbouring areas has necessarily affected the distribution of fishes, facilitating dispersal or hindering it, isolating or joining species. Some historical features are discussed under the appropriate drainage basin descriptions below or under the relevant genus or species but others are more widespread and are briefly outlined here. Sources include in particular Wolfart (1987) but also Harrison (1968), Takin (1972), Falcon (1974), Stöcklin (1968, 1974a, 1974b), Krinsley (1970), Stoneley (1974), Kashfi (1976), Shearman (1976), Booth (1977), Jackson and Wood (1980), Berberian and King (1981a, 1981b), Haynes (1981), Rögl and Steininger (1984), Šengör (1984), Oosterbroek and Arntzen (1992), Rögl (1998; Rögl, 1999), and Adams et al. (1999). There have been no cladistic analyses of taxa on which history can be determined. Zoogeographical analyses are based on present day distribution and suppositions on relationships. During the Cretaceous and through the Early Oligocene the Tethys Sea, several thousand kilometres wide, extended from the Mediterranean Sea to the Indian Ocean, separating the Afro-Arabian and Eurasian continents. Afro-Arabia was part of Gondwanaland. The usual assumption is that Iran belongs to Eurasia, perhaps with Central Iran a microcontinent or island or as a northern continuation of Arabia, and with East Iran a microcontinent or peninsula of Eurasia. Förster (1976), however, maintains that Central Iran, and probably North Iran, were part of Gondwana. The Tethys covered much of what is now Iran and was a barrier to the movement of freshwater fishes. The ocean regressed during the Late Oligocene except for a Euphrates-Persian Gulf furrow and the Zagros and Makran troughs. Continental sediments were deposited in endorheic basins of Iran. The Tethys closed in the Middle to Late Miocene as evidenced by mammal migrations between Asia and Africa. The establishment of continental conditions over Iran has been continuous since the Late Miocene except for an inundation in the Late Pliocene in the Zagros trough and the Makran coastal region. There may also have been an early Miocene connection between Arabia and Iran/Iraq allowing movements of freshwater fishes (Adams et al., 1999). Iran is therefore composed of parts of Gondwana, which was the continent south of the Tethys, welded to the northern continent and parts of the Eurasian plate (such as the central and eastern Iranian microcontinent). The northeastward movement of the Arabian Plate caused the closure of the Tethys and led to the folding which in the Miocene/Pliocene orogenies formed the Zagros Mountains, a prominent feature of western Iran important in zoogeographic studies of fishes (see Kashfi (1976) for an opposing view). The Zagros orogeny is related to the opening of the Red Sea which formed a barrier to fish dispersal. The Alborz Mountains are a northern part of the Alpine-Himalayan orogen of which the Zagros are a southern part and started to rise in the upper-lower Pliocene (Krinsley, 1970; Stöcklin, 1974). A continuous land-bridge between Eurasia and Africa has been in existence since the upper Miocene, facilitating freshwater fish dispersal. Hora (1937) and Menon (1957) refer to wet, marshy, tropical conditions and headwater captures along the whole southern face of the Himalayas and westwards during the Pliocene and early Pleistocene facilitating the spread of fishes from the east to Iran. Hora (1937) and Briggs (1987) consider that cyprinids entered Africa from southeast Asia 18-16 MYA, in the early Miocene, while other groups moved through Iran and the Arabian Peninsula beginning in the early Eocene. Kosswig (1951; 1952; 1955a; 1955b) notes the similarity at the generic level between Indian and African fishes, e.g. the cyprinids Barilius, Garra and Labeo, indicating that these fishes arrived in Africa from India after the desiccation of the Syrian-Iranian Sea in the Pliocene. The primary route, according to Kosswig and to Por (1987), was a northern one around the barrier of the Persian Gulf and Sea of Oman via northern Arabia, Syria and the Levant. Cooling conditions in these areas during the Pliocene and especially the Pleistocene glaciations, and arid climates at times, were unsuitable for tropical forms. These movements left a selection of fishes in what is now Iran including the cyprinid Garra, the sisorid catfish Glyptothorax and the spiny eel Mastacembelus.

The Pleistocene fore-deep of the Himalayas may have had connections with the Tigris-Euphrates basin which extending down the Persian Gulf as a river valley. The Tigris-Euphrates basin formed during the Pliocene and was colonised by primary freshwater fishes no earlier than the late Pliocene (Krupp, 1983). Movements of fishes into Iran from the west and north were also affected by the presence of the Tethys Sea and a brief account is given under the genus Barbus which has been studied in this regard.

The present picture of the Arabian peninsula is of an arid desert unsupportive of fish life. The presence of fishes in Arabia and the Levant, and even Africa, with apparent relationships to fishes from Iran and the east indicate that fishes must once have traversed this area. Movements of fishes are thought to have been in a northern arc around the Fertile Crescent or its earlier version. However this modern picture is perhaps illusory as there is evidence of a more hospitable environment in the Arabian Peninsula at various times in the past. Wadis were active during "pluvial" periods of the Pleistocene as evidenced by deposition of fluvial material (Al-Asfour, 1978). One of these wadis drained much of central Arabia to the Kuwait area. The "Kuwait River" once ran from the Hijaz Mountains in western Saudi Arabia northeastwards for about 850 km to drain into the Persian Gulf via a vast delta occupying much of modern Kuwait. The river was 8 km wide and over 15 m deep along most of its length (Hamblin, 1987; Anonymous, 1993b). This river last ran between 11,000 and 6,000 years ago and could have provided a highway for fish dispersal. Earlier rivers of this nature dating to the Late Miocene (Forey and Young, 1999; Hill and Whybrow, 1999; Friend, 1999), the Pliocene (Gerson, 1982), and others like it in other parts of the peninsula, as well as shallow lakes (e.g. Lake Mundafan in the Rub' al Khali at 36,000-17,000 B.P. and again at 9000-6000 B.P.) would have facilitated transfer of species across the Arabian Peninsula, today an impassable desert for fishes, e.g. at the height of the Würm glaciation 40,000 years ago (Chapman, 1971; McClure, 1976; Al-Sayari and Zötl, 1978; Brice, 1978; Jado and Zötl, 1984; Wagstaff, 1985). A freshwater connection between Iran and Arabia was almost continuous from 70,000 to 20,000 years B.P. (Krupp, 1983). However no fish remains have been found in the late Pleistocene lakes although freshwater molluscs are frequent, Hippopotamus remains are reported and Neolithic fish hooks have been found in Al Hasa in eastern Saudi Arabia. Incomplete Miocene freshwater fish fossils are reported from the Jizan basin in the Tihama north of the Saudi Arabian-Yemen border (Brown, 1970). One was identified as a Barbus and the other as a Tilapia. Both these identifications are of such a general nature (see account on the genus Barbus for example) as to throw little light on past history or relationships with modern taxa. The Lower Miocene fauna of Al-Sarrar at 15-17 MYA, northwest of Dhahran in eastern Saudi Arabia, contains pharyngeal teeth thought to be Barbus, and more interestingly several thought to be Labeo (Thomas et al., 1982). This latter genus is not now found in the Middle East but occurs in the Indian subcontinent and Africa. The Late Miocene Baynunah Fauna of Abu Dhabi in the United Arab Emirates contains Clarias, Bagrus shuwaiensis and Barbus in a river connected with an ancestral Tigris-Euphrates system (Forey and Young, 1999). These fossils tend to confirm the hypothesis that fishes of Asian origin reached Africa through the Middle East and could have taken what may be termed a southern route across the Arabian Peninsula. However Forey and Young (1999) point out that the modern Arabian fauna may not have a history stretching back to the Miocene but is due more to a re-invasion after a loss of an earlier fauna. The modern Iranian fauna, in part, may be a remnant of movements at various times yet to be resolved in the absence of species-level phylogenies.


Brian W. Coad (