The general climate of Iran is based on Bobek (1952), Ganji (1960, 1968), Taha et al. (1981), "Aquastat" from the Food and Agriculture Organization, Rome (www.fao.org/ag/agl/aglw/aquastat/iran.htm) and www.bibliothecapersica.com/articlenavigation/index.html, under ab (= water) and climate, downloaded 24 December 2004. Kouchoukos et al. (1998) give an overview of climatology for Southwest Asia based on satellite datasets. Precipitation, its amount, nature and seasonality, is important in determining the water regime and thus the habitats for fishes. Iran is sparsely vegetated, both naturally and through the agency of man, and the air temperature and amount of insolation has a direct effect on water temperatures. Insolation is continuous through summer days when clouds are a rarity over much of Iran and the weather remains settled for weeks at a time.

In general, the climate of Iran can be classified as arid to semi-arid, with more than 80% of the country characterised by less than 250 mm annual rainfall. Mountain ranges block off the interior of Iran and give extremely continental conditions except for the narrow littoral zones on the Caspian shore and the Persian Gulf. Summers are hot and dry with little change from day to day. Three main climatic types are found: warm, temperate and rainy with a dry summer in the Caspian coastal area, dry, hot desert in the central plateau, and dry, hot steppe in the rest of the country. Humidity is generally low because of the altitude, much of Iran being over 1000 m average height. Coastal regions along the Persian Gulf have a high humidity, especially in summer. Wind patterns are deflected by the Zagros and Alborz ranges in the west and north. Summer winds are mainly north and northwest over much of northern and central Iran and are hot, dry, and strong for long periods. The Sistan "Wind of 120 Days" from the northwest blows from the end of May to September continuously and is very hot, dry and sand-laden. The "shamal" blows from the northwest over Khuzestan and coastal regions of the Persian Gulf from February to October, most intensely in summer. These summer winds undoubtedly contribute to the desiccation and, in some cases, filling-in of water courses. In the south the winds are west and southwest.

Temperature varies greatly over Iran with latitude and altitude, as well as with the seasons. Winter lows are found in January and summer highs in July in general, with the Zagros and Alborz mountains and the Caspian shore having maximum temperatures in August as a result of the influence of altitude and the sea. The mean monthly temperatures for January at 15 selected stations across Iran (Ganji, 1968) had a range of -1°C to 20°C, average about 8°C. For July these figures are 25 to 37°C, average 30°C. The annual range is 14C° at Jask on the Sea of Oman and 30.5C° at Mianeh in East Azarbayjan. Outside the coastal areas of the Caspian and Gulf, the annual range is considerable, and daily ranges also are large. Nights can be very cold in the northeast, less so on the plateau. Some areas, like the Khuzestan plains, have maximum temperatures over 50°C (53°C at Gatvand near Dezful; possibly over 55°C in the interior, hotter than anywhere else on earth) in summer while in the northwest in winter the temperature can fall below -30°C (to a low of -36°C at Bijar in Kordestan). Five temperature provinces have been delineated for Iran: the Caspian zone along the littoral which has a low annual temperature range; the Persian Gulf zone which has a low annual range but high values; the Zagros zone with a much higher range than the first two zones and a very low January mean; the Alborz zone which is similar to the Zagros but has higher temperatures and a greater range; and the interior zone with the greatest annual range coupled with relatively high values.

Precipitation falls in winter as snow on the mountains of the north and west. The highest mountains remain snow-covered year round. The plateau also receives snow but it does not last long and there is no snow along the Persian Gulf coast. Rain falls mainly in November to May with a mean annual of 416 mm, although the Caspian littoral is much higher and the interior plateau much less. Rain is uncommon from May to October over most of Iran. Maximum rain is found on the outward slopes of the Alborz and Zagros ranges where the mean annual rainfall is more than 1200 mm, 1950 mm at Anzali. The plateau has less than 120 mm annually, Sistan less than 70 mm, and Mirjaveh on the Pakistani border only 48 mm annually. The Caspian littoral has rain in every month at some localities. The plateau receives most of its rain in spring, the Caspian in autumn, and the Gulf coast in winter. The result of this pattern of rainfall is heavy runoff in spring with silt-laden floods and erosion a feature. Many streams marked on maps are actually dry for much of the year. Even a major, interior basin river like the Zayandeh which flows through Esfahan does not reach its terminal basin for much of the year.

A review of modern and historical floods in Iran can be found in Mazra'eh, News, Analytical and Educational Monthly, No. 10, January 1998 at www.netiran.com/Htdocs/Clippings/DEconomy/980100XXDE05.html. Devastating floods occurred in 2001, after several years of drought, in Gilan, Golestan and Khorasan provinces (IRNA, 11 August, 14 August, 4 September 2001).

Droughts occur and can be devastating for fish habitats. The drought years 1999-2001 were the worst in 30-40 years and resulted in a United Nations Technical Mission (see ReliefWeb, 22 August 2000, UN Office for the Coordination of Humanitarian Affairs (OCHA) at www.reliefweb.int; Foghi, 2004). Various effects were noted including the drying of 2500 qanats in Yazd, in southern Fars groundwater became saline, the Latian, Lar and Karaj dams near Tehran had water reserves of 51 million cu m, down from 173 million cu m for the same period in the previous year and were within about 2 months of drying up, several lakes and wetlands of international importance dried out (Bakhtegan-Neyriz and surrounding wetlands, Hamun-e Saberi, south end of Hamun-e Puzak and Gav Khuni), rivers dried completely (Hirmand River and its terminal lake), the Dez and Karkheh rivers in Khuzestan were depleted by 70% in 2001, water rationing was implemented in Tehran and 30 other cities, and lower water levels in rivers that retained flow had reduced oxygen affecting fish (IRNA, various news reports, 2001). In East Azarbayjan, 190 ha of 220 ha used for fish breeding were useless through drought (IRNA, 29 August 2001). Marshes south of Lake Orumiyeh near Mahabad encompassing 30,000 ha dried up (IRNA, 25 August 2001). Water reserves behind dams in Khorasan were depleted by 65% in 2001, the precipitation rate having declined by 40% in the period November 2000-August 2001 (IRNA, 3 September 2001).

Abbaspour and Sabetraftar (2005) reviewed Iranian drought cycles and found arid conditions were experienced for 13 of the previous 23 years. Drought affected fishes in the drying of wetlands where hundreds of thousands of fish died, in Sistan 8-12,000 tons of fish were lost as the lakes dried up, in Fars fish losses were reported from the Kor River, in East Azarbayjan 174 ha of fish culture farms were damaged, and rivers draining to the Persian Gulf lost fishes including migratory species.

The nature of the drainages of Iran is directly related to climate. The Alborz Mountains in the north block movement of moisture to the south while the Zagros Mountains in the west block moisture from that direction. The southeast monsoon is almost completely dry before it reaches eastern Iran. In consequence the best watered parts of Iran lie on its northern and western fringes and the interior becomes drier from west to east and north to south. Interior rivers exist in large part because of mountain ranges which store water as snow, in the case of the Hirmand River and the Sistan lakes, far removed from Iran.

There has been many studies on past climates in Iran and neighbouring countries, attempting to link climate with past environmental conditions in the Late Pleistocene-Holocene. The Early to Middle Pleistocene, however, is practically unknown for the Middle East and is not dealt with here (Butzer, 1978). Past environments have significance for fish habitats, distributions and zoogeography. The brief summary below is based on Butzer (1957, 1958a, 1958b, 1961, 1975, 1978), Bobek (1959), Whyte (1961), Hutchinson and Cowgill (1963), van Zeist and Wright (1963), van Zeist (1967), Wright et al. (1967), Krinsley (1970), Diester-Haass (1973), Turnbull and Reed (1974), Nützel (1976), van Zeist and Bottema (1977, 1982), Wright (1977; 1983), Ganji (1978), Neumann and Sigrist (1978), van Zeist and Woldring (1978), Woosley and Hole (1978), Farrand (1979), Storch (1980), Coad (1980c), Kay and Johnson (1981), Lamb (1982), Neumann (1993), Qin and Yu (1998); Griffiths et al. (2001); Stevens et al. (2001); Snyder et al. (2001); this being by no means an exhaustive listing of the studies in this field nor is the below a critical assessment of conflicting views. Evidence for these past environments is taken from a number of studies in different fields. The Pleistocene ice has been gradually withdrawing from its last maximum at 20,000 B.P. and the remains of ice fields and glacial moraines can be used to determine former conditions such as the snowline. The advance and retreat of deserts and the use and abandonment of settlements are indicative of changes. Such erosional physical features as dry riverbeds and other riverine structures, alluvial fans, sand dunes, and aeolian deposits all give clues to environmental change. The extent and level of lakes and playas have been widely studied as indicators of climatic fluctuations. Pollen and other organisms associated with lake sediments can be used to trace changing conditions and finally historical records can be analyzed.

Glacial deposits in the outward slopes of the Zagros and Alborz mountains indicate that the snowline was 600-800 m lower than today, perhaps as much as 1800 m in some areas, and as much as 1500 m at Shir Kuh near Yazd and Kuh-e Jupar near Kerman in south-central Iran. Lowered snowlines cannot be explained by temperature alone but were probably due to much greater precipitation. Winter would have been longer and colder in the Pleistocene, more snow would accumulate and summers may have been cloudier. The runoff period would have been longer and river habitats could have been less prone to desiccation in late summer.

The climate in the Zagros Mountains of the late Quaternary in Iran has been examined by means of sediment analyses from lakes Zaribar and Mirabad and for nearby Turkey at Lake Van. Pollen, chemistry, sediments, diatoms, cladocerans, ostracods and palaeobotany all confirm geological studies. The last glacial maximum (the Würm) at about 20,000 B.P. led to local glaciation, a depression in the snow line and absence of trees. The climate was cool and relatively dry, with less precipitation than today. The cooler temperatures meant less evaporation, more runoff and filling of intermontane lakes. The Caspian Sea and Lake Orumiyeh were much larger than today, being 78 m and 55 m higher. As the glaciers receded, the land environment or life zones moved up the mountains. The significance of this for fishes is unknown; there were few trees and the environment may have resembled modern denuded conditions. There may have been a higher flow than later when trees developed to hold runoff and before man chopped them down. However bushes could have retained water and reduced silt load in rivers. By 12-14,000 B.P the evidence from Zaribar and Mirabad indicates a warming climate but without increased precipitation. Indeed rainfall may have been less than today, reducing river flows and perhaps habitats for fishes. This arid period was succeeded by a more humid period. An increase in precipitation at Lake Van did not take place until 6500 B.P., about 4000 years later than in western Iran. Climate changed not only through time but also geographically, just as today. Regional variations mask general statements about earlier climate for Iran and the outline given here is perhaps best seen as indicative that change occurred. The humid period was followed by a period of less rainfall, and then in the late Holocene by an increase in rainfall. The last 3000 years have been humid with perhaps two, short, arid episodes. Southern Iran may have been cool and comparatively moist when the highlands were moderately cold and relatively dry. Climate probably changed markedly over short periods. Short cold phases are recorded for Europe in the last several thousand years, e.g. from about 1400 to 1230 B.C., associated with rises in lake levels. Similar events may occurred in Iran. Barley harvest dates in Babylonia derived from clay tablets indicate they were 10-20 days earlier in the period 1800-1650 B.C. and 10-20 days later in 600-400 B.C. It is concluded that the former period was warmer and the latter cooler than today.

Pluvial conditions as recognised for more northerly areas of Europe probably did not occur in Iran during the Pleistocene although summers may have been less dry because of greater cloudiness and lower temperatures and evaporation. Lake levels were probably higher 18,000-20,000 years ago (Roberts and Wright, 1993). Krinsley (1970), in his study of playas in Iran, concluded that the climate was semi-arid rather than pluvial in the period of maximum cold during the Pleistocene. Lakes, which occupied endorheic basins and could have facilitated local fish movements, dried up as the climate warmed with the retreat of ice sheets and glaciers and evaporation exceeded precipitation. These shallow lakes were found along the inner mountain front or within basins which received greater discharges. As distance from the mountains increased, there were only intermittent lakes and finally playas. An immense lake filling much of central Iran, as proposed by earlier authors, seems unlikely. Generally conditions over Iran appear to have varied as much, if not more, in the Pleistocene as they did in recent centuries through the agency of man. Conditions 9000 years ago were probably drier than today (Roberts and Wright, 1993). The fishes may have been selected for an ability to survive highly variable conditions in terms of stream flow, temperature, silt load, local fluctuations in lake levels and salt content, etc.