The freshwater goby (Awaous aeneofuscus) in the Wadi Hatta watershed (U.A.E./Oman)

The freshwater goby (Awaous aeneofuscus) in the Wadi Hatta watershed (U.A.E./Oman)

A small population of the freshwater goby Awaous aeneofuscus (Peters, 1852) has been identified in the northern Hajar Mountains in an intermittent watercourse that empties into the Gulf of Oman. The population is suspected to reproduce in situ. A. aeneofuscus occurs in Indian Ocean watersheds from South Africa to Pakistan, but is rare in Arabia and has not been publicly reported from Gulf of Oman watersheds for more than a century.

A small population of gobies was recognized in late 1997 in an Omani tributary of Wadi Hatta, not far from the U.A.E. town of Hatta, during a survey of the freshwater fish of the northern Hajar Mountains. This was noteworthy since only four other fish species, three native and one introduced, occur in this area. These are the freshwater cyprinids Garra barreimiae, endemic to the Hajar Mountains, and Cyprinion microphthalmum, found on both sides of the Gulf of Oman; the euryhaline cyprinodont Aphanius dispar or Arabian killifish, a circum-Arabian species; and the introduced tilapia Oreochromis spp. (Feulner, 1998).

A specimen was initially identified as a goby (Family Gobiidae) with the assistance of the Breeding Centre for Endangered Arabian Wildlife in Sharjah and the fish was described and depicted, pending identification, in Feulner (1998).

Gobies constitute one of the largest families of fish, containing almost 200 genera and more than 800 species (Nelson, 1976; Moyle, 1993). Most inhabit coastal marine environments, where they may occupy relatively specialized niches, but a number of species enter or live primarily in estuaries or freshwater rivers and streams (Nelson, 1976; Berra, 1981). Gobies are typically small but distinctive, being elongated and characterised by bulging eyes, two dorsal fins (one spiny, the other soft) and pectoral fins placed anteriorly and fused to form a sucker or grasping organ. Most are flat-bottomed and live on the substrate, often sheltering in crevices in rock or coral or burrows in sand (Berra, 1981; Moyle, 1993). Some, like the tropical mudskippers, can breathe air and regularly forage out of water for extended periods (Nelson, 1976; Berra, 1981).

The Breeding Centre, and in particular Christian Gross, made arrangements for expert identification of the Hatta goby through the good offices of Gordon McGregor Reid of the North of England Zoological Society (Chester Zoo). Peter Miller of Bristol University identified it as belonging to the genus Awaous Valenciennes, 1837 (Reid, pers. comm.). Ronald E. Watson, then at the Forschungsinstitut Senckenberg in Frankfurt, provided the final determination of A. aeneofuscus (Watson, pers. comm.).

According to Watson, the identification of A. aeneofuscus is problematic because some characteristics vary with overall size, and A. aeneofuscus can grow to be relatively large. A specimen from Madagascar in the Musee de l'Histoire Naturelle in Paris is said to measure 40-50 cm in length, making it reputedly the largest known goby. This, plus the disjunctive distribution of the species, has bedeviled taxonomists. Watson points out that A. aeneofuscus has been described as a new species at least eight times in the scientific literature (Watson, pers. comm.). A specimen from Muscat was described as Gobius jayakari by Boulenger in 1887.

A. aeneofuscus is well known from watersheds along the Indian Ocean coast of South Africa (Skelton, 1993) as well as Madagascar, East Africa, the Gulf of Aden and southern Arabia (Yemen and Oman) (Watson, pers. comm.; Beech, 1999). It occurs eastwards at least as far as Pakistan (Watson, pers. comm.). Krupp (1983) mentioned the occasional presence of gobies in wadis of the Arabian Peninsula but without identifying species or locations and without conveying the impression that these might constitute resident populations. A. aeneofuscus was observed and some were collected at a few sites in northern Oman in the late 1970s and early 1980s, including Wadi Jizzi, approximately 55 km south of Wadi Hatta. The specimens were tentatively identified as A. jayakari (now synonymised with A. aeneofuscus) by Brian Coad, then Associate Curator of Fishes in the National Museums of Canada, but these determinations were never published (Gallagher, pers. comm.).

Apart from proper taxonomic identification, the discovery of an isolated goby population in the Wadi Hatta watershed raises questions about the ecology and lifestyle of this fish and the history of its arrival and persistence, since Wadi Hatta is an intermittent stream that flows continuously to the sea only after exceptional rains -- an event which occurs only every several years at best. Normal pluvial discharge is dispersed and absorbed as it transits the coarse gravels of the Batinah coastal plain, which here is some 12.5 km wide.

A. aeneofuscus is present for several kilometers in the middle reaches of the Wadi Qahfi tributary of Wadi Hatta, a few kilometers downstream from the popular "Hatta Pools," in an area of mixed bedrock and coarse gravel substrate with occasional large boulders and scattered permanent pools of a meter or more in depth. There it is most abundant in deeper pools that are somewhat isolated from the main surface flow. The site is centered at 24 deg. 45'23"N, 56 deg. 11'36"E, and is approximately 25 km inland, 15 km within the mountain front and 35 km upstream from the Gulf of Oman coast, at an elevation of about 250 meters. The area is relatively difficult to reach by vehicle but it is nevertheless regularly visited by local people, who use traditional techniques to harvest the endemic cyprinid fish Garra barreimiae which is eaten (Feulner, 1998). A smaller population has been found in a more limited section of a second, downstream tributary of Wadi Hatta, but has not been extensively observed and is not discussed here.

Although rare in Arabia, A. aeneofuscus is better known in South Africa, where it is called the Freshwater Goby (Varswater-dikkop) and is featured in several comprehensive popular guides on freshwater fish (Skelton, 1993; Smith and Heemstra, 1986). There it is said to be found in estuaries and rivers, in both "pools and running water, usually over sandy bottoms into which it may bury itself with only the head and eyes exposed. Preys on invertebrates." (Skelton, 1993). A. aeneofuscus has been reported up to 400 miles inland in the Transvaal (Jubb, 1967, cited in Hyslop, 1991).

Watson states more generally that "A. aeneofuscus is well adapted to a variety of habitats that range from large, muddy rivers with silty bottoms to fairly clear desert streams. It is an omnivorous feeder and is able to feed on small invertebrates. In desert streams it is known to ingest large amounts of fine sand and the detrital and algal material surrounding each grain is digested and the remainder passed." (Watson, pers. comm.).

A. aeneofuscus has long been suspected to breed in freshwater (Pienaar, 1978, cited in Hyslop, 1991). Watson (pers. comm.) says that "Reproduction has never been observed but may be similar to other large species of Awaous in moving downstream to the mouth of streams, in freshwater, where males guard large rocks, stumps, etc., and numerous females lay eggs. Males do most guarding of eggs . . . . Eggs hatch in about 24 hours, becoming part of the oceanic plankton community for an unknown amount of time. Recruitment of fry into streams occurs during lunar high tides and after a significant amount of seasonal rains."

The population of A. aeneofuscus in Wadi Qahfi has been observed intermittently since April 1998 and is estimated at 100-200 individuals. Actual counts of 48 and 78 individuals were made, respectively, during visits for other purposes in January and March 2000. The population appears, impressionistically, to be stable or even increasing slightly, except in the deepest of the pools, where numbers seem to have declined. The largest individuals observed were approximately 22 cm (n = 6) and the smallest were approximately 9 cm. The presence of gobies having a range of sizes suggests either multiple recent episodes of marine recruitment or the occurrence of breeding in situ.

Wadi Hatta is not known to have flowed to the sea since the gobies were first observed in late 1997, but it might possibly have done so during wetter years such as 1992-93 and 1995-96. The winter months are statistically the rainiest in the U.A.E. but rainfall is highly variable from year to year, both in amount and time of occurrence. Mountain regions are susceptible to heavy but highly localised showers. The frequency with which Wadi Hatta discharges to the sea has undoubtedly been diminished by the construction of three dams in the watershed within the U.A.E., above the Wadi Qahfi tributary. Two of these have been constructed within the past decade. As a result, the existing goby population may now be unusually isolated and dependent on self-reproduction rather than recruitment.

Tilapia (Oreochromis spp.) were first noticed in Wadi Qahfi in the Hatta Pools area, a few kilometers upstream from the sites occupied by A. aeneofuscus, in April 1998. Their introduction may have been relatively recent, since no tilapia were seen in the goby area when A. aeneofuscus was first discovered in October 1997 and only one tilapia was seen there when goby specimens were first collected in April 1998. In January 2000, however, tilapia were present in essentially all of the pools containing gobies. A few individual tilapia reached an estimated length of c.30 cm - significantly larger than the largest gobies. A. aeneofuscus co-exists with the notoriously adaptable and resilient O. mossambicus in South African rivers, but it seems reasonable to be concerned that the presence of tilapia at such close quarters may have an adverse effect on the survival of A. aeneofuscus.

Why is A. aeneofuscus confined to the few kilometre stretch of Wadi Qahfi where it is found, whereas the other three resident fish species are found both above and below the goby area? Downstream the wadi is broader and flatter, and the deeper pools favoured by A. aeneofuscus are uncommon. Upstream, however, many seemingly suitable pools exist but few of these are as isolated from the main current as the favoured goby pools downstream. Physical barriers could also be responsible, especially if the gobies are relatively recent arrivals, since the gorge immediately below the Hatta Pools constitutes the most abrupt change in gradient from the head of Wadi Qahfi to the sea. There the wadi drops some 15 meters over less than 100 meters horizontally, through a system of falls and chutes in a narrow channel cut in bedrock.

Human disturbance could be another determining factor, as the pools upstream are accessible by a good graded track and are regularly visited. This is reflected in various measures including the amount of litter. Water samples taken by Peter Cunningham at the Hatta Pools and analyzed by Farouk Al Basit of the Water Department, Laboratory Unit, Abu Dhabi Water and Electricity Department, showed a coliform organism count of 38 per 100 ml versus nil for samples from the goby area downstream (Al Basit, pers. comm.). In addition to considerations of water quality, a relatively large and distinctive fish might suffer selectively from human attention.

Based on accounts of its observed habitats elsewhere, A. aeneofuscus must be presumed to have a broad tolerance for most measures of water quality (Whitfield, 1996; Watson, pers. comm.), although it is said in South Africa to favour water having a high concentration of dissolved solids and generally to breed in ephemeral pools (Hoffman, pers. comm.). It is also known to be intolerant of hypersalinity, i.e., salinity in excess of normal seawater (35 ppt) (Whitfield, 1996). Hypersalinity in the wadi environment would in any case exclude Garra barreimiae, which cannot withstand even 50% seawater concentration (Haas, 1982).

The authors wish to thank the individuals and institutions mentioned in the text for their contributions to the identification of A. aeneofuscus from the Wadi Hatta watershed and for the unpublished information provided by them in connection with the preparation of this note.