Freshwater Fishes of Iran
Introduction - Materials and Methods
The descriptions in this work are founded on original observations of material and a consideration of the literature. The sources of this material are various museums which house a scattering of Iranian species including in particular the Natural History Museum, London (formerly the British Museum (Natural History)), the Naturhistorisches Museum Wien, and the Zoological Institute, St. Petersburg which are depositories for older type material, but the bulk of the research has been based on four collections. The first of these was made by V. D. Vladykov during 1961 and 1962 when he was an Inland Fisheries Biologist under the Expanded Programme of Technical Assistance of the Food and Agriculture Organization, UN. This material was deposited in the National Museum of Natural Sciences, Ottawa (now the Canadian Museum of Nature) and consists mainly of specimens from the Caspian Sea basin. The second collection was made by employees of the Department of the Environment, Tehran, and N. B. Armantrout and R. J. Behnke. Half this collection was placed in the National Museum of Natural History, Tehran (Muze-ye Melli-ye Tarikh-e Tabi'i) and half was retained by R. J. Behnke and formed the basis of Saadati's (1977) thesis at Colorado State University, Fort Collins. This collection covered the whole of Iran except the Caspian and Sistan basins. Through the courtesy of Dr. Behnke I have been able to examine this material in Fort Collins and make extended loans for study in Ottawa. The Muze-ye Melli-ye Tarikh-e Tabi'i collection is small (examined in 1995; catalogue 2000) and not as diverse as the Fort Collins material. Oregon State University contains a collection of fishes made by W. Kinunen, S. Bullock, R. RaLonde and P. Walczak, who were members of the Peace Corps in Iran (some of this collection was deposited at the Smithsonian Institution, Washington, which helped to fund the collection and transport of specimens). Dr. Carl Bond kindly loaned me much of this material for long periods. This collection was from all parts of Iran. The last collection, comprising the bulk of the material, was made by me from 1976 to 1979 while I was teaching at Pahlavi (now Shiraz) University in Shiraz. This collection is housed in the Canadian Museum of Nature, Ottawa (formerly NMC, now CMNFI), and covers all of Iran except the extreme northeast and northwest. Field trips were funded by the Research Council of Pahlavi University. Subsequently various Iranian colleagues have sent me specimens and these too are incorporated in the present work. Principal among these were materials collected by Asghar Abdoli (then based in Golestan) and Nasser Najafpour and associates of the Iranian Fisheries Research Organisation (IFRO), Ahvaz. These collections together effectively cover all the major drainages of Iran and provide the best foundation yet assembled for a study on this ichthyofauna.
All material stored at the Canadian Museum of Nature, Ottawa was examined in 45% isopropyl alcohol. Preservative was later changed to 70% ethanol. The Canadian Museum of Nature also stores extensive field records including slides, numerous data sheets on most species (counts and measurements including x-ray plates), an extensive literature base including translations from foreign languages, and comparative specimens and literature from other countries in Southwest Asia.
Specimens collected by me were caught by any means that presented themselves. Gear used included seines of various lengths and mesh sizes (much repaired and patched!), gill-nets of various stretch meshes (sometimes used as seines), cast-nets of several diameters (thrown skilfully by others and poorly by me), by hand, and by purchase from small boys and anglers using a variety of techniques (of angling on their part and of persuasion on mine to extract catches from their possession). The object was to sample any water body for all the kinds of habitat found there within the limitations of a hasty schedule and the available equipment. Most habitats were visited for less than one hour, but in the small springs and streams, which comprise the bulk of Iranian fresh waters outside the large rivers and lakes of Khuzestan and Sistan and the deep waters of the Caspian Sea, this was more than adequate to catch a good and varied sample of most species. This was borne out by repeated visits of longer duration to certain localities near Shiraz. Pools and flowing sections were seined, gill-netted or cast-netted. Riffle areas were also attacked in this fashion or seines were used to block off sections of riffle and upstream rocks disturbed by kicking to scare secretive species like loaches into the fixed net. In small streams a dip-net was placed downstream of individual rocks which were kicked over and the net scooped along the stream bed. Cast-nets proved particularly useful in rocky streams which had little open water. Draped over the rocks and only partly in the water, they nevertheless caught large and fast specimens which were unobtainable by seining. The available fishing gear was less effective on large rivers and on the Caspian Sea. Here boats, long gill-nets and trawl gear would have been most useful. The collections are poor in inhabitants of the main current of large rivers and in the deep water species of the Caspian Sea. Larger specimens in major water bodies undoubtedly evaded my nets with ease; some samples of larger individuals were available from other collections and by purchase from commercial fisheries.
Several criteria were used to select specimens for counts and measurements. Where few specimens were available, all were counted and measured. Where several hundred specimens were available selection was by size (usually larger fish; sometimes much smaller fish as well for comparison with adult values), by sex to ensure an adequate representation of males and females, and by locality where geographical variation was examined. Badly damaged or grossly deformed specimens were excluded but there was no (conscious) selection for "ideal" specimens.
Wherever a putative species was collected from more than one drainage basin and material diversity permitted, a comparison was made between the drainage basins. This work is continuing and details of methods and materials are to be seen in published results. Students of Iranian fishes should note that the application of sufficient statistical "weight" will reveal differences between drainage basin samples and this is especially true of a desert and semi-desert country like Iran. Springs and streams may have been colonised by only a few founders. A small population sampled in the lower reaches of a stream may not have had any contact with conspecifics higher up in the stream for many generations. Conversely, several seasons of heavy rain may have afforded recent opportunities for contact and gene exchange. A one-time sample from a stream may therefore give a quite inaccurate picture of the character suite of that population. Whether any of the differences detected have systematic significance requires careful consideration. For example, Balletto and Spano (1977) described 9 subspecies of Garra tibanica in the southwest of the Arabian Peninsula using Principal Components Analysis. This has been termed "statistical overkill" by Alkahem and Behnke (1983). Also Krupp (1983) has observed that samples of Garra rufa from the same locality collected in different years or seasons varied in several characters. Description of subspecies based on limited material requires a great deal of care therefore.
There are various methods of measuring and counting anatomical features of fishes. The ones I have used are outlined below. They are based on Hubbs and Lagler (1958) and Trautman (1981). Some particular characters are outlined in papers by me in the Bibliography.
The method of counting fin rays differs from that in use in North America since unbranched and branched rays are counted separately. A "III,8" count in the European literature would be "9" in the system advocated by Hubbs and Lagler (1958), i.e. the soft ray count is increased by one to convert from the "European" to the "American" system. The bulk of the work on fishes of southwest Asia follows the European system and I have adopted this methodology to facilitate comparisons, although eschewing Roman numerals.
A) Meristic characters
In this book, scale counts, number of gill rakers and of vertebrae are usually expressed as ranges based on literature sources since frequency counts are rarely given. A separate section gives counts on Iranian fish examined by me followed by a frequency in parentheses (..). Fin ray counts often show strong modes, but citing the mode alone would be misleading. Pharyngeal tooth formula is often a modal value from the literature; loss of or incomplete development of major or minor row teeth is not uncommon, so counts may vary quite markedly.
Scale counts and paired fin ray counts were made on the left side of each fish. In some instances, such as a badly deformed fin or where scales on the left were mostly missing, counts were made on the right. These instances were rare and restricted to species with low sample sizes.
Not all meristic characters had equal sample sizes; some material from other museums was not available for x-rays, large series of pharyngeal tooth counts was not often available because removal of arches damages specimens, some specimens were damaged in certain characters, time did not always permit all characters to be counted, some species are well-known and additional data from Iran is clearly a subset of widely gathered data, some species were examined in detail to address systematic problems, and so on.
All vertebrae were counted including the hypural plate as one vertebra. In Cypriniformes and Siluriformes, the four Weberian vertebrae were included in the count. Almost all counts were made from radiographs.
2) Gill rakers
All rakers on the first gill arch were counted. A lower limb count in the literature includes any raker at the angle of the upper and lower limbs. Gill raker counts presented something of a problem when comparing specimens of disparate sizes. The smaller fish often had very small rakers at each end of the arch. These were easily missed or torn off when cleaning a debris-encrusted arch. Removal of arches for a more careful examination may also damage or destroy the finer rakers which are intimately associated with the tissues adjacent to the arches. Alizarin preparations can be of assistance, but the finer rakers may have no bony content and thereby be omitted. Counts of juvenile fish may therefore give lower values than counts for larger fish, whether this be due to an increase in gill raker number with age or because rakers are more easy to count in larger fish. This kind of variation is only critical where this character is being used in species identification or in analyses meant to define and relate species.
3) Pharyngeal teeth
The teeth of the modified fifth gill arch in Cyprinidae were counted in each row and given as a formula from left to right. A count of 2,5-4,2 consists of two teeth in both the outer left and outer right rows, five teeth in the inner left row and four teeth in the inner right row. Pharyngeal teeth rows in Iranian cyprinids varied from one to three on each side. In certain cases, it was evident from the presence of a socket that a tooth had been lost. The count then included that tooth.
4) Fin rays
a) Dorsal and anal fins
Fin ray counts were divided into two types. One count is of spines or hardened soft rays or any unbranched, unpaired unsegmented rays and this is usually given in Roman numerals in the literature. In deference to some Iranian unfamiliarity with Roman numerals, the spine count is given in Arabic numerals in this text. Spine count included rudimentary rays which, at the anterior dorsal and anal fins, may be obscured by flesh or scales requiring some probing or dissection. Radiographs were often useful to confirm counts made under a microscope. The second count is of soft rays and is also indicated by Arabic numerals. These rays are usually branched, flexible, segmented and laterally paired. The last two unbranched rays often arise from a single internal base and were then counted as one. This is generally the case in Cyprinidae. The branched ray count is the most diagnostic and variable in such fishes. Some families contain species with more than one dorsal fin. The first dorsal fin may be composed of spines and the second dorsal fin of spines and soft rays. In such species the count is given separately for each fin.
b) Caudal fin
The branched caudal fin rays only were counted. Dorsal and ventral to these central rays are a series of unbranched rays which become progressively smaller and may be obscured by flesh and scales where the caudal fin attaches to the caudal peduncle. Counts in other works often comprise the branched rays plus one dorsal and one ventral unbranched ray. Caudal fin ray counts are remarkably uniform within families. In Cyprinidae the count is almost always 17, except for occasional variants. Garra persica was unique in having a strong modal count of 16 branched caudal fin rays.
c) Paired fins
Paired fin ray counts can be separated into unbranched and branched rays. A small splint in some species at the origin of the paired fins was excluded from the count. There is usually one unbranched ray which is not included in counts cited here. The branched ray counts were the most important and are the ones given here. However, in the pectoral fin the innermost rays were often difficult to discern and may increase with age.
a) Lateral line count
The first scale counted was that scale contacting the pectoral girdle. The count continued along the flank following the pored scales and including small, additional scales lying between the large, regular scales as well as any unpored scales. The small, additional scales were relatively rare occurrences and any obviously abnormal fish - those with healed injuries for example - were not counted. The count terminated with the scale lying over the end of the hypural plate as determined by flexing the caudal fin. Some works recommend inclusion of a scale overlying the flexure only if most of its exposed field is closer to the body than to the caudal fin. Since the flexure of the caudal fin produces a relatively broad groove, this is difficult to judge in smaller fish. Therefore, the most posterior scale whose exposed surface touched the groove was the last scale counted. I have also continued the count onto the caudal fin in some species for a total count as this sometimes proved useful in comparison with counts in older literature.
b) Scales above the lateral line
This count commenced with the scale at the origin of the first dorsal fin and continued down and back to, but not including, the lateral line scale. Any scale partially or wholly straddling the dorsal fin origin was counted as one scale. The count followed the natural scale row and included any small or irregular scales in the row.
c) Scales below the lateral line
This count commenced with the scale at the origin of the anal fin, followed the natural scale row up and forward to, but not including, the lateral line scale and included any small or irregular scales. In this, and the previous count, it sometimes proved necessary to shift the counting row because of the scale arrangement. This was always a backward shift. In some instances there were several scales at the anal fin origin which overlapped each other very closely. All these were counted and account for the large degree of variation in counts between individuals of some species.
d) Scales between the lateral line and the pelvic fin origin
This count was made as in the above count.
e) Predorsal scale rows
All rows of scales between the origin of the dorsal fin and the head were counted just below the mid-line of the back on the upper flank. The final "row" at the occiput may consist of a single scale. This method was used because scales on the mid-line may be small and irregular, obscured by heavy pigment, or absent.
f) Caudal peduncle scales
This was the lowest count of the scale rows around the caudal peduncle, usually at its narrowest point. Both lateral line scales were included. Scale rows were counted even when the scale arrangement was such that occasional alternate rows touched. This count may be quite consistent between individuals of a species, but it may also vary markedly. The variation depended on the presence of large scales dorsally and ventrally on the caudal peduncle connecting the flank scale rows. When such large scales were present bridging over the top and bottom of the caudal peduncle, the total count could be, e.g. 12, but in some individuals two or more smaller scales occupied their positions so that the scale count jumped to 16.
B) Morphometric characters
All measurements were to the nearest 0.1 mm using dial calipers. Measurements were taken on the left side unless a left fin, for example, was badly deformed or broken. Badly deformed specimens were not measured. Distortions due to preservation, such as a gaping mouth or expanded gill covers, were gently adjusted to as natural a position as possible. The following list explains how the various measurements were taken. All measurements were taken in a straight line and not over the curve of the head or body.
1) Total length
From the anteriormost part of the head to the tip of either lobe of the caudal fin when that fin is normally splayed.
2) Standard length
From the anteriormost part of the snout (even when the lower jaw projects) to the end of the hypural plate (the end of the plate is found by flexing the caudal fin; in small fish it may be seen by shining a strong light through the caudal region). Standard length can be an inaccurate measurement. The end of the hypural plate is obscured by scales, flesh and caudal rays. Its position is determined by flexing the caudal fin; this flexure is taken to be the end of the hypural plate. Small fish have thin, delicate bones and the flexure may be at the anterior base of the hypural plate, at the origin of the caudal fin rays which articulate with and overlap the end of the hypural plate, or even between the last whole vertebra and the hypural plate. Large fish have a broad flexure which can give a variety of measurements by independent observers. Fortunately, in this study most fish were comparatively small and strong illumination helped to discern the end of the hypural plate. For larger fish I can only plead an attempt at consistency.
3) Head length
From the anteriormost part of the snout to the bony margin of the opercle (excluding the opercular membrane).
4) Body depth
Maximum straight line depth excluding fins or fleshy and scaly structures at fin bases
5) Body width
Maximum distance from one side of the body to the other.
6) Head depth
From the occiput vertically to the breast or lower head surface.
7) Head width
The distance between the opercles when in their normal, closed position. The opercles are gently pressed into a closed position if greatly dilated.
8) Snout length
From the anteriormost part of the snout or upper lip at the mid-line to the bony front margin of the orbit.
9) Orbit diameter
Greatest diameter between the bony rims of the orbit. This distance is not always horizontal.
10) Postorbital length
Greatest distance between the posterior bony orbit margin and the bony opercular margin.
11) Interorbital width
Least bony width between the orbits over the top of the head in a straight line.
12) Predorsal length
From the base of the anteriormost dorsal fin ray to the tip of the snout or upper lip.
13) Prepelvic length
From the base of the anteriormost pelvic fin ray to the anteriormost point on the head (snout or upper lip).
14) Preanal length
From the base of the anteriormost anal fin ray to the anteriormost point on the head (snout or upper lip).
15) Length of caudal peduncle
The oblique distance from the insertion of the anal fin to the mid-point of the end of the hypural plate.
16) Depth of caudal peduncle
The least depth of this structure from the mid-line of the ventral surface.
17) Length of the longest dorsal and anal fin rays
From the structural base of the ray to its tip.
18) Length of the dorsal and anal fin bases
From the anteriormost ray base (the origin of the fin) to the point where the fin membrane contacts the body behind the last ray (the insertion of the fin).
19) Length of the pectoral and pelvic fins
From the extreme base of the uppermost, outermost or anteriormost ray to the tip of the fin.
20) Distance between pectoral and pelvic fin bases
Used principally in Cyprinidae and Cobitidae, this and the following measurement are from the extreme base of the anteriormost, uppermost or outermost ray of the appropriate fin to the anterior base of the next fin.
21) Distance between the pelvic and anal fin bases
22) Length of fin spine
From the base of the spine to its tip. In pungent spines, as in catfishes, this excludes soft rays or membranes distal to the sharp tip, but in more flexible spines, which may taper gradually as in Cyprinidae, this measurement includes the soft tip.
© Brian W. Coad (www.briancoad.com)