Printer Friendly

Re-discovery of the mudskipper Periophthalmus waltoni Koumans, 1941 in the United Arab Emirates.

INDEX
Introduction: Mudskippers, Periophthalmus and P. waltoni
Historical records of Periopthalmus waltoni in the UAE
Current field observations
      Summary of field visits
      Summary description of the UAE study population
      * Physical setting
      * Location within the tidal regime
      * Total area of occupation
      * Population size
      * Size of animals
      * Breeding cycle
      Burrows and burrow distribution
      Surface activity in general
      Activity period
      Locomotion
      Comfort behaviour
      Perambulation
      Hunting, feeding and diet
      Territoriality
      Threat display
      Courtship and mating behaviour
      Other social behaviour
      Juveniles and recruitment of juveniles
      Associated species
      Predation and predator avoidance
Selected aspects of the biology and ecology of P.
 Waltoni
      Adaptations to life on the mudflats: physical,
physiology and
         behavioural
      The typical mudskipper life cycle, including
        a planktonic larval phase
Re-Discovery of P. waltoni in the UAE:
Why here? Why now?
Were two species of mudskippers present in the
UAE in the early 1970s?
An East Coast record, too
Epilogue
Acknowledgements
References


Introduction: Mudskippers, Periophthalmus and P. waltoni

Mudskippers are amphibious fish that inhabit mudflats, muddy creeks and lagoons and mangrove forests in the Old World, mostly in the tropics and subtropics; a few species extend to temperate regions of South Korea and southern Japan where they hibernate in winter (Baeck et al. 2008). They all belong to the goby family (Gobiidae). The common name "mudskipper"' is a descriptive term that is generally applied to the several most amphibious genera of the subfamily Oxudercinae, in particular the genera Periophthalmus, Periophthalmon, Boleophthalmus and Scartelaos, which number in total about 30 species (Murdy 1989, 2011, Lee & Graham 2002, Lee et al. 2005, Polgar 2010, Ishimatsu & Gonzalez 2011). Mudskippers of one species or another are distributed widely around the Indian Ocean and West Pacific (Polgar 2010). A single species is found on the Atlantic coast of Africa (Murdy 1989, 2011, Polgar 2010).

Mudskippers are of scientific interest because they are the most thoroughly amphibious fish. They have been studied in order to understand their physiological and behavioural adaptations to terrestrial life and to the extreme variations of temperature, salinity and dissolved oxygen levels often found in the tropical and subtropical mudflat environment. Those studies have given initiative and support to theories of the evolutionary origins of air-breathing and colonisation of the land by vertebrates (Shultze 1999, Ishimatsu & Gonzalez 2011, Polgar 2010).

They are of popular interest because of their distinctive, comical appearance and their novelty as a "walking" fish"--sometimes a very active one. Their prominent, closely-spaced bulbous eyes, set atop the head, effectively give them a 360[degrees] view of their environment. They breathe air and feed on the damp mud surface when the tide is out, frequently interacting with their conspecifics. Hogarth (1999) says truly, "Anyone who has visited mangroves or tropical mudflats in the Indo-Pacific region will have been captivated by these fish."

Periophthalmus is the largest genus of mudskippers, comprising about 18 species (Murdy 1989 and 2011, Polgar 2010). The name is derived from the Greek peri (around), and ophthalmon (eye), and refers to the wide visual field of these species. The genus is distributed throughout the Indo-West Pacific, with one species found on the coast of tropical West Africa. Most species have limited geographic ranges, but the ranges of individual species may overlap (Taylor et al. 2005, Polgar 2010). The centre of diversity is in the eastern Indian Ocean (Polgar 2010). All are primarily carnivorous predators and most occupy the upper intertidal zone (Polgar 2010).

Periophthalmus waltoni (Fig. 1) is found in the Arabian Gulf and along the Indian Ocean coast of Iran and Pakistan to extreme NW India (the Gulf of Kutch) (Polgar 2010; see also Encyclopedia of Life, Fishwise). The type locality is Iraq and Pakistan (Polgar 2010, Fishwise). Within the Arabian Gulf P. waltoni is locally common on mudflats in northern Kuwait (Clayton & Wells 1987), Iraq (Mhaisen & Al-Maliki 1997) and along the Iranian coast of the Strait of Hormuz (Askari et al. 2010, Kooseg et al. 2011). A single site was known at Al-Khor in north-eastern Qatar from at least the mid-1980s until the early 1990s, when the land was reclaimed (Fran Gillespie, pers. comm.), and P. waltoni was once locally common in the Northern Emirates (see discussion below).

In Kuwait, P. waltoni is sympatric with two other species of mudskippers, Boleophthalmus dussumieri Valenciennes, 1837 (discussed in much of the earlier literature from Kuwait as B. boddarti (Pallas, 1770)) and Scartelaos tenuis (Day, 1876), as well as a fourth Oxudercine species, Apocryptodon madurensis (Bleeker, 1849) (Clayton & Wells 1987; taxonomic nomenclature updated per Murdy 1989, Froese & Pauly 2012b and 2012c, and Polgar 2010). The four species are recognised to segregate themselves within the mudflats environment on the basis of different 'preferences' of habitat and diet, i.e., they occupy different ecological niches (Clayton & Wells 1987). The carnivorous P. waltoni occupies the highest intertidal environments of the four, and is correspondingly the most fully amphibious. B. dussumieri is a somewhat larger herbivorous species that grazes microalgae and cyanobacteria. Its habitat range overlaps with that of P. waltoni but overall it occupies a slightly lower intertidal range. The omnivorous S. tenuis and the much smaller, herbivorous A. madurensis occupy a much lower intertidal zone of permanently wet, oozing mud and do not interact with P. waltoni.

P. waltoni is included in Coastal Fishes of Oman (Randall 1995), along with the mudskippers Boleophthalmus dussumieri and Scartelaos tenuis. Randall's stated range for all of those fish is given as the Arabian Gulf to Pakistan or NW India, but no provenance is given within Oman by Randall or any other authors. One strong possibility is (or was) the extensive mudflats in Khasab Bay, on the north-west of the Musandam peninsula, opposite current Iranian sites in the Strait of Hormuz, but those mudflats have been largely destroyed in the past decade by dredging and reclamation. Suitable sites for any mudskipper species along the Gulf of Oman coast are few and limited, even more so for the lower intertidal B. dussumieri and S. tenuis than for P. waltoni (see Feulner 2000), but an unpublished Gulf of Oman record from the UAE was brought to our attention prior to publication of this paper and is discussed below, under "An East Coast record, too".

Historical records of Periopthalmus waltoni in the UAE

Mudskippers identified as Periophthalmus sp. were known from at least two Arabian Gulf sites in the UAE in the 1970s and had been mentioned to the senior author (GRF) in passing, mostly as a curiosity, by researchers active in that era (Michael Gallagher, pers. comm., Kenneth W. Glennie, pers. comm.). In fact, newly elicited information makes it more likely than not that the herbivorous mudskipper Boleophthalmus dussumieri was also present in the UAE at that time (John Stewart-Smith, pers. comm.; see discussion below).

Historically, the principal UAE mudskipper site, to which all of the foregoing sources referred, was Khor Madfaq, the mouth and estuary of Wadi Lamhah, on the border between Umm al-Qaiwain and Ra's al-Khaimah emirates. Although Wadi Lamhah is normally dry, it drains a very large inland watershed originating in the Hajar Mountains, extending from Manama and Masafi in the north to Shawkah and Fili in the south; before reaching the coast, it crosses ca. 30 km of coastal sands. Khor Madfaq has always been relatively accessible in the modern era, being traversed by a bridge on the original coast road.

An April 2012 reminiscence by John Stewart- Smith (pers. comm.) gives the flavour of the site in the early 1970s:

"During my time in the Abu Dhabi Defence Force / UAE Air Force, my children visited from UK during school holidays. One of their most enjoyed outings was to go to the coast between Umm al-Qaiwain and Ras al- Khaimah to see the funny fish that walked on the sand. These were particularly plentiful around a muddy area at Al Hamra(?). Maps were less than detailed in the early 1970s so I cannot be precise about the location." Subsequent correspondence confirms that the site in question was Khor Madfaq and Stewart-Smith adds that the mudskippers were "widespread and numerous".

A second early mudskipper site was Khor Hulaylah, north of Rams, a broad, shallow lagoon with a sinuous channel, situated inshore of a 9 km barrier island. Attention had been called to Khor Hulaylah in the early 1970s by the work of malacologist Kathleen Smythe and by Michael Gallagher and other naturalists, who remarked on the unique character of the Dhayah salt marshes, which are fed by fresh water springs debouching within the more extensive khor (lagoon). Colin Richardson (1994) mentioned mudskippers at this site in an entry in a compendium of Middle East wetlands sites (now online) with a primary emphasis on avian fauna. He wrote: "The shallows around the mangrove areas are one of the best places in the UAE to find mudskippers Periophthalmus sp." Richardson (pers. comm.) has elaborated in correspondence with the authors that he saw mudskippers regularly at both the Dhayah marshes and Khor Hulaylah generally, beginning in the mid-1980s and continuing at the Dhayah marshes into the early 1990s, but always in small numbers (typically one or two fish per visit). A page about the Dhayah marshes in a current UAE general interest website also mentions the presence of mudskippers (Periophthalmus sp.) (UAE Interact). The source of that information could not be determined (Peter Vine, pers. comm.) but it is likely to be derived from Richardson (1994).

The authors have independently confirmed the former presence of mudskippers at Khor Hulaylah through conversation on site with Mohammed Salem, a middle-aged resident of a nearby village who is a frequent visitor to the khor for bird photography and other activities. He understood our question about the "walking fish" immediately and was able to give a correct indication of its size. He agreed, however, that it has not been seen there in recent times.

Both of the known former mudskipper sites have been substantially altered since the reports cited above. Khor Madfaq was bridged in the 1970s and was re-bridged by a double span in the late 1990s. The beach there today is popular with kite-surfers. At Khor Hulaylah, since the late 1990s the northern end of the khor has been progressively developed as a commercial and industrial free zone, and since 2007 most of the Dhayah marsh area has been covered with landfill for residential development.

The mudskippers themselves, meanwhile, were all but forgotten. They were gone from Khor Madfaq by the early 1980s, if not sooner, and have not been mentioned in any of the now voluminous books and articles on UAE natural history that have been published since the mid-1980s. Richardson's (pers. comm.) records from the early 1990s are the last known for Khor Hulaylah; they are preserved in memory in a corner of cyberspace (Richardson 1994, UAE Interact) but there appear to be no fish on the ground. Several visits by the senior author (GRF) to Khor Hulaylah from 1998 to 2007 resulted in records of rare air-breathing intertidal molluscs, but no mudskippers, and a thorough examination by both authors in March 2012 and September 2013 also found no evidence of mudskippers. This is consistent with the experience of local resident Mohammed Salem (pers. comm.), cited above.

Another UAE resident, a local government conservation official, immediately applied the Arabic name buleghlegh (pronounced BU-legh-EGH) to P. waltoni on seeing it, although he may possibly have been conflating mudskippers with a smaller and more common burrowing, non-amphibious subtidal goby species, since he tried (unsuccessfully) to show us more of them by taking us to the downshore edge of the mangrove forest, in the mid to lower intertidal zone.

Other experienced naturalists who arrived in the UAE as long ago as the early or mid-1990s, and who have worked at one or both of the known historical sites, have proved to be unaware that mudskippers ever existed here, notwithstanding professional interests and field studies encompassing coastal ecology. No mudskipper species have been included in recent lists or accounts of UAE fish species (Beech et al. 2005, Froese & Pauly 2012a, ARKive). Efforts were made in the late 2000s to limit the scope of proposed residential development at Khor Hulaylah and at the Dhayah marshes in particular. Surveys were conducted and reports prepared, highlighting the unique aspects of the habitat, flora and fauna (Llewellyn-Smith et al. 2007, Llewellyn-Smith 2011), but none of those reports mentioned the former presence of mudskippers.

It was a surprise, therefore, when in February 2012 the authors encountered the population reported here.

Current Field Observations

The story of the fortuitous re-discovery of mudskippers at a site in the Northern Emirates (Fig. 2) was reported in the UAE press in March 2012 (Todorova 2012), but it should be noted that Richardson's sightings in the early 1990s (Richardson 1994 and pers. comm.) and the unpublished East Coast record from 1997 (Kevin Budd, pers. comm., Chris Stuart, pers. comm.) were not then known to the authors, or to the several other UAE naturalists whom they consulted. The actual interval since mudskipper populations were last known to occur in the UAE is therefore closer to 15-20 years than to the 35 years initially reported.

The identification of Periophthalmus waltoni could be made confidently on the basis of photographs because of its distinctive appearance and markings, the existence of regional records, and the availability of a number of authoritative print and online photographic references (e.g., Clayton & Wells 1987, Randall 1995, Polgar 2010). In the ensuing weeks the authors attempted to locate additional populations, both nearby and at other broadly similar sites elsewhere along the Arabian Gulf coast of the Northern Emirates, but none were found.

Because of the small size of the known

population, its apparently unique status and its proximity to human access (Figs. 3 and 4), the determination was made that the site location should not be publicly disclosed. Those same factors make the population unsuitable for collection of specimens or for study other than by discreet and non-invasive means. Our own field work has relied exclusively on visual inspection, with or without binoculars, and on photography. Care has been taken to disturb the fish as little as possible during the course of our observations. In the text below, certain details have been omitted in order to prevent indirect disclosure of the site location.

For most of the past decade the area of the study site has been under a regime of protection by local authorities, who have kindly facilitated our periodic visits, but major infrastructural development has continued nearby and recreational amenities are now planned on immediately adjacent land. The presence of the mudskipper population and its significance have therefore been highlighted to the concerned local authorities and the maintenance of a buffer zone has been encouraged.

Summary of field visits

The authors monitored the site for almost 20 months, from mid-February 2012 until early October 2013, under diverse conditions of tide, daylight and weather, visiting at irregular intervals of 24 hours to 6 months.

The territorial nature of the fish, their construction of relatively conspicuous main burrows, and the fact that they were active only when the site was emergent, made it relatively easy for us to obtain a count of the population and to associate burrows with individual fish, to which we assigned field names for convenient reference.

* February 2012:

Preliminary investigation consisted of two periods of observation in mid and late February 2012, the first in midday sunshine following a high tide at 0900 hrs which puddled the site, and the second in late afternoon on a falling tide under breezy and overcast conditions. We observed a minimum of 8 or 9 individual fish and estimated that there were at least 14 occupied burrow complexes. Courtship and mating behaviour was observed in late February.

* June 2012:

The authors' circumstances prevented follow-up observation for several months, until June 2012. Then an evening visit was made in early June, less than two hours after a moderate high tide (which had only barely reached the site). 9 individuals were observed but none ventured beyond their burrows after dark. Two consecutive morning visits (commencing at first light) were made in late June, in each case within two to three hours after inundation by a spring high tide. Only 5 fish were active each day but feeding and an unsuccessful courtship interaction were observed.

* July 2012:

In late July 2012, another early morning visit was made after an exceptionally high overnight spring high tide. A population of 15 fish was confirmed, all but one of which were seen, and two territorial confrontations were observed. The last fish was observed the next day, during an early evening visit following an intermediate afternoon high tide.

Throughout our visits in 2012, the mudskipper population at the study site appeared stable and we were able to recognise specific individuals by their burrow locations.

* January 2013:

Two early morning visits were made in mid-January 2013, after low overnight spring high tides. Both were on relatively cool mornings, one very windy and cloudy with air temperatures steady at about 20.5[degrees]C (69[degrees]F), the other sunny but cooler, with air temperatures in the range of 14[degrees]-16[degrees]C (57[degrees]-61[degrees] F). It was disappointing to find that only 8 burrow complexes appeared to be active, and only 5 fish were actually seen, none of which were active. All but one of the active burrows had been shifted from their positions in July 2012, although it was possible to associate at least half of them with the burrows that they had replaced.

* March 2013:

Another morning visit was made in mid-March

2013, following a low overnight spring high tide. The weather was sunny with estimated air temperatures between 22[degrees]-27[degrees]C (72-80[degrees]F). This visit confirmed a count of 8 active burrows. Burrow distribution was similar to January 2013. 6 fish were seen; 2 were cohabiting and exhibited courtship and mating behaviour.

* August 2013:

Two early morning visits were made in mid-August 2013, after overnight spring high tides. Air temperature at dawn (0545 hrs) was 29[degrees]C (85[degrees]F), reaching 37[degrees]C (99[degrees]F) by 0930 hrs. These visits confirmed a continuing adult population of 8 fish but also revealed--to our great surprise--the recent recruitment of juvenile fish, resulting in a significant population increase within the study area (to ca. 30+ fish) as well as the presence of additional juvenile fish in a nearby satellite location where no mudskippers had been found before.

* October 2013:

A final pre-publication visit was made in early October 2013 to review the status of the juvenile fish. The late afternoon visit followed a spring high tide and confirmed a population and distribution largely unchanged from what was seen in August 2013.

Following our recognition of juvenile recruitment in mid-August 2013, we spent two days in late August and early September visiting other potential mudskipper sites in the Northern Emirates, in an attempt to determine whether newly arrived juveniles could be found more widely. Again, we found no evidence of this.

Summary description of the UAE study population

Basic details of our study population can be summarised as follows:

* Habitat:

Perimangal mud with small to medium mangrove shrubs (Avicennia marina) and scattered clumps of low halophytes (Arthrocnemum macrostachyum), within 15 metres landward of well-developed mangrove forest; not remote from human activity and readily susceptible to potential observation, disturbance and/or destruction (Figs. 2 through 3).

* Location within the tidal regime:

Upper intertidal, inundated by ca. 56% of all high tides.

* Population size:

An estimated 15 individuals, all considered to be adult fish, were present from discovery (in mid-February 2012) through late July 2012. 15 individuals were confidently observed in July 2012 and their burrow systems (sometimes multiple) were accounted for. Possible independent burrow systems would permit an estimate of 16 to 17. Any higher number posits the presence of additional individuals for which no evidence was actually seen. Numbers had fallen by January and March 2013, when only 8 adults were reckoned to be present. By mid-August 2013, juvenile recruits had swelled the population to ca. 30 or more individuals.

* Area of occupation:

For most of the period of observation, all of the adult mudskipper burrows except one (which is ca. 15 metres from any other) could be roughly circumscribed by an elongated rectangle approximately 30 metres by 12 metres (ca. 360 square metres), with its long dimension parallel to the local shoreline. From a mudskipper's perspective, that rectangle is effectively closed to expansion on three sides by either ecological or physical barriers-r -the shoreline in the upshore direction, the mangrove forest in the downshore direction, and a man-made physical barrier in one of the longshore directions. The average density of mudskippers within this area is therefore ca. one per 26 square metres. If we include the most isolated adult, the rectangle of occupation would become approximately 45 metres by 12 metres (ca. 540 square metres) and the average density would fall to ca. one fish per 36 square metres.

All of that expanded rectangle, and somewhat more, was in fact used by the juvenile recruits seen in August and October 2013, adding perhaps an additional 120 square metres to the total area of occupation for ca. 30 fish, equivalent to a density of one fish per 22 square metres.

In mid-August through early October 2013, during or following the recruitment period, ca. 25 juvenile fish were found in a discrete satellite area some 400 metres away from the study site, as further described under "Juveniles and juvenile recruitment", below.

* Size of the fish:

Adult length was estimated very roughly at 11.0 to 17.5 cm (ca. 4.5 to 7.0 inches) total length (TL). Our estimate for the largest animals was twice confirmed by indirect measurements. In several references the maximum total length for P. waltoni is given as 15 cm, but a recent study records a 20.8 cm individual from southern Iran (Sarafraz et al. 2011). Juvenile length was estimated at ca. 9.0 to 11.0 cm (ca. 3.6 to 4.5 inches) total length (TL). Our estimate for the smallest animals was twice confirmed by indirect measurements.

* Breeding cycle:

Breeding appears to occur in late winter to early spring. Courtship and mating behaviour were observed in at least five individuals, including two pairs, in mid-February 2012 and mid-March 2013 and may continue into June (see "Courtship and mating behaviour", below). The arrival of juveniles was not observed in 2012, despite field visits spanning early June through late July (although with hindsight, one of the fish then present may have been a juvenile). In 2013, however, the recent arrival of a significant number of juveniles was evident in mid-August (see "Juveniles and recruitment of juveniles", below).

Burrows and burrow distribution

Adult mudskippers are territorial and construct burrows from which they exploit and defend a surrounding territory. The following account of burrows at our UAE site is based primarily on observations of a stable adult population of ca. 15 fish from February until July 2012, and a residual population of 8 adult fish in January until August 2013. In the final paragraphs we discuss the distinctive features of the burrows constructed by the influx of juvenile mudskippers observed in August and October 2013.

At the study site, the burrows of neighbouring adult fish were at least two metres apart, and generally three to five metres apart, but, with one exception, each burrow was within eight metres of another. The exception was a burrow located approximately 15 metres from any other. This is a much looser concentration than at the site studied by Clayton and Snowden (2000) and could reflect the fact that the UAE site constitutes sub-optimal habitat that cannot support either a large or a dense population of mudskippers.

Most mudskipper burrows were located in areas without significant growth of mangrove pneumatophores (Fig. 5), but two of the more conspicuous burrows and one nondescript one (belonging to one of the smallest fish) were within areas of moderate pneumatophore density (Fig. 6). One main burrow platform enjoyed shade from an erect mangrove sapling and one auxiliary burrow was constructed in the dense shade of a small but spreading mangrove shrub.

A typical P. waltoni burrow at the UAE site has multiple nearby entrances which are joined below the surface. Several of the UAE fish were observed to move between entrances, either on the surface or from within the borrow. The main entrance is most often on a broad-rimmed low turret or volcano or a flat-topped platform (Fig. 7), but those may fall into disrepair by erosion or collapse. Sometimes one burrow entrance opens into a small, circular 'pond' which the fish itself constructs (Fig. 8). Some fish also maintain one (or sometimes two) auxiliary burrows, the entrance to which may be smaller and more directionally oriented (Fig. 9), making them difficult to distinguish from certain crab burrows (Fig. 10). One auxiliary burrow at the site was approximately 4 metres from the corresponding main burrow. The literature indicates that crab burrows may sometimes intersect with those of P. waltoni (Clayton & Wells 1987); at the UAE site this could possibly be the case for the crabs Metapograpsus messor and Eurycarcinus orientalis.

The lifetime of individual adult burrows can evidently be several months or more. One burrow (the most isolated one) appeared to remain intact and in use for more than one year. Eroded or damaged burrows may be repaired, remodeled or abandoned; these results were observed over a 16-day interval in June 2012 and even more so over a 28-day interval from late June to late July 2012. During the latter interval, five fish seemed to have moved to newly constructed main burrows, although still within their original territory and without clearly abandoning older burrows; the longest relocation was ca. 3+ metres, by a small fish on the periphery. During the same period one fish greatly improved its auxiliary burrow, located at the edge of a puddle within its territory, by building a ringed pond at the entrance (Fig. 11). A few conspicuous burrows seemed to be in the same locations in June and July as when they were observed in February, although no attempt was made in February to record the exact position of individual burrows.

We observed only three instances of active burrow maintenance, each sufficiently minor to be better characterised as housekeeping. Twice, adult mudskippers were seen to emerge and deposit a mouthful of mud and water on a low pile near the entrances to their burrows. One of the same fish also deposited a mangrove leaf it had removed from its burrow.

Male and female mudskippers are identical in appearance (Clayton & Wells 1987) although males can grow to be somewhat larger (Sarafraz et al. 2011). The only sure way to distinguish the sexes in the field is by observing breeding behaviour (Clayton & Wells 1987). For this reason we are unable to confidently ascribe differences in burrow architecture, or in the presence or absence of courtship behaviour, to sex differences. Nevertheless it is reasonable to hypothesise that breeding adult males have the greatest incentive to build a substantial burrow platform burrow as a podium for sexual display. Two large individuals which we measured (indirectly) at 17.5 cm (7 inches), above average for P. waltoni, both had prominent platforms, whereas two of the most nondescript main burrows belonged to the two smallest-appearing mudskippers (estimated ca. 11. cm or ca. 4.5 inches) (Fig. 12).

One of the larger adult fish remaining in August 2013, presumably a male, had by that time become the proprietor of three large burrows, each with a platform or turret, arrayed in a triangle with sides of approximately 7.0, 4.5 and 4.0 metres, and we watched the fish travel between as well as beyond the three burrows on two mornings. This was the largest territory we recorded. The principal burrow was a durable but dilapidated platform at the extreme corner of the study site that, on all previous visits, we had considered abandoned. The other two burrows were pre-existing and had been the property (we think) of two different fish; we speculate that with the demise of some of the former adult population, the fish we observed was able to expand his territory and adopt the vacant burrows as his own.

The burrows constructed by newly arrived juvenile fish, first seen in August 2013, were generally less conspicuous than typical adult burrows, but they were also variable. Most were slightly asymmetrical low mounds, resembling crab burrows; they were distinguished, if at all, by a more well-defined entrance groove. Some juvenile burrows, mostly on the firmer upshore margin of the study area, appeared to be no more than finger holes and would have passed unrecognised had we not seen the resident fish (Fig. 13). Even these rudimentary burrows, however, sometimes proved to have two entrances.

The simplest burrows may often be temporary or auxiliary ones, and newly-arrived and metamorphosed juveniles may be essentially nomadic, at least on a small scale, until they find the 'right' place to settle more permanently. We observed at least two instances in which we could be confident that juvenile fish had constructed new burrows in the 24 hours between our visits, each in the immediate vicinity of their previous burrows. In August, only one juvenile burrow, in a satellite area previously uninhabited by mudskippers, approximated the 'platform and pool' arrangement of the most elaborate adult burrows. By early October, however, some juveniles had burrows consisting of a relatively well-formed platform or turret. In at least one instance, a very simple burrow seen in August in a firm, upshore location, had been maintained and improved by early October (Fig. 14). The active juvenile occupant was observed in each instance.

More generally, it is our impression that P. waltoni does not simply choose its preferred habitat and burrow site, but also actively modifies it. In particular, burrowing disrupts the cyanobacterial cover that is found in adjacent upper intertidal areas without mudskippers, and covers the surface with somewhat finer and less consolidated mud (Fig. 15). Impressionistically, it seemed to us that in January and March 2013, when mudskippers at the study site were fewer in number and less active, the cyanobacterial cover within the area was more extensive and more cohesive. Mudskippers sometimes excavate 'swimming pools' in association with their burrows, but we suspect they may also play a role in maintaining shallow feeding pools within their territories. In this way, the presence of a few pioneer mudskippers may make the habitat more attractive for colonisation by others. That hypothesis is arguably exemplified by the nascent satellite colony of juveniles discovered in August 2013, in an area primarily characterised by moderate cyanobacterial cover (see "Juveniles and recruitment of juveniles", below).

Surface activity in general

Periophthalmus species feed and interact with each other primarily on the surface of their mudflat environment when the tide is out. Because they live in the upper intertidal zone, they and their burrows are emergent for longer than they are inundated. They have gills but they can also breathe air, respiring through the skin and through the linings of the mouth and pharynx (cutaneous respiration), assisted by modifications of the gill chambers (Graham 1997). Air-breathing is considered to be an adaptation to the very low dissolved oxygen levels of the seawater remaining at low tide in burrows in anoxic mud. Many and perhaps all Periophthalmus species maintain an air chamber in their burrows (Ishimatsu et al. 1998, Polgar 2010), at least when brooding eggs (Takeda et al. 2012), and some species will reportedly "drown" if forced to remain submerged for prolonged periods (Hagen Aqualab 2005).

In the course of our field work we observed P. waltoni engaged in various behaviours including locomotion, feeding, "comfort behaviour" (rolling and fin waving), threat display, male courtship display and the paired mating ritual, as well as less readily decipherable behaviour. Our observations are summarised and compared under the various headings that follow.

Activity period

Mudskippers generally, including P. waltoni, have been considered as diurnal species (Polgar 2010), although Columbini et al. (1996) concluded that activity patterns of Periophthalmus sobrinus (= P. argentilineatus or P. kalolo) depended more on environmental factors such as air temperature and relative humidity than on the daily light cycle,

and nocturnal activity has also been recorded for the forest-dwelling P. wailailakae. A conspecific, P. minutus from NW Australia, is expressly reported to retreat to its burrow at night (Takeda et al. 2012).

Clayton & Snowden (2000) found that in Kuwait, P. waltoni was active by day when the tide was out, independent of the weather and the exact state of the tidal cycle, with three exceptions: (1) there was little or no activity on dry mud above the level reached by the previous high tide; (2) activity was inhibited at surface temperatures below 15[degrees]C (59[degrees]F) (a temperature generally exceeded by day in the UAE, year-round); and (3) activity was inhibited by high wind speeds. Clayton & Snowden point out that these parameters correspond to those limiting the activity of the mudflat crabs on which P. waltoni preys in Kuwait. Otherwise, Clayton and Snowden (2000) found that, when active, P. waltoni spent more time on the surface than in their burrows, by a ratio of almost 9:1.

Consistent with Clayton and Snowden's findings, our study population was not generally active when the site had not been wet by the previous high tide, or when it had dried out thereafter, or at night. Using local tide tables in conjunction with our field observations, we calculate that, on average, 36 high tides each month, or about 56% of all high tides, are high enough to wet the study site to a greater or lesser extent (Fig. 16). Likewise, on average, about every two months a period of 3 to 4 days may pass when no high tide is high enough to inundate the site.

Independent of physical activity, however, the majority of the mudskippers at the study site seemed disposed to perch out-of-water during most of our observations, either in the mouth of a burrow entrance, atop the entrance platform, or perched on a nearby mound of mud, although they retreated to their burrows periodically, most likely to wet their skin to enhance respiration (Ikebe & Oishi 1996, cited in Clayton & Snowden 2000), or if disturbed. This was true in both warm, sunny midday conditions and cool, breezy late afternoon conditions in February, as well as early evening and early morning conditions in June and July.

The same was generally true at night. Observed after dark in early June 2012 (from 1930 until 2100 hrs), the majority of the population (and 8 of 9 individuals seen that night) could be found

repeatedly by flashlight, perched in the mouths of their burrows (Fig. 17) or atop their platforms. They were tolerant of the light but withdrew if approached to within 2-3 metres. None were observed to travel from their burrows. In late July, 10 of the 15 known fish were emergent during the two hours after sunset. Several were active during twilight, but after darkness fell only a single fish, which had become habituated to flash photography, was seen more than 30 cm from its

burrow, and 5 of the 10 were no longer seen at all.

Takeda et al. (2012) observed seasonal variation in the surface activity of Periophthalmus minutus in Darwin Bay. They found that these mudskippers were active by day in February, the austral summer and breeding season, whether or not they had been inundated by the preceding high tide. In August, however, the same mudskippers were active by day only following inundation. Takeda et al. (2012) also noted that temperatures above 40[degrees]C (104[degrees]F) suppressed the daytime surface activity of the fish.

In the UAE, we found that although adult mudskippers were active at midday in February, they became sedentary by ca. 0930 hrs in June, July and August--this in a summer regime of consistent daytime high temperatures of 40[degrees]C or more (daytime highs during our July visits were reportedly 43[degrees]C (109[degrees]F)). However, our observations do not allow us to disentangle the effect of temperature from that of surface desiccation. All of our daytime visits in summer were in early morning following an overnight high tide; both temperature and desiccation increased rapidly and in parallel during the course of the morning. Consistent with the generalisations of Clayton & Snowden (2000), we would expect that P. waltoni is habituated to the high temperatures of the Arabian Gulf summer and that surface wetness is likely to be more important than temperature as a constraint on surface activity.

The juvenile P. waltoni observed at the study site in mid-August 2013 appeared to be somewhat more active than adults and also somewhat more tolerant than adult fish of surface desiccation and/or high temperature. A number of juveniles continued to move about and to interact on the mud surface into late morning, after all adult fish had settled into their burrow mouths. The same was true in early October 2013, when surface temperature was somewhat less extreme (35[degrees]C (95[degrees]F) at 1400 hrs). However, this could be less a matter of tolerance than of the newcomers' need to explore their surroundings and establish territories of their own.

Juvenile fish were also seen further upshore, on firmer, drier substrate than any adults. But again, this could have been more a matter of necessity than of choice, reflecting their exclusion from the more desirable territories of other fish, including larger juveniles.

Locomotion

Normal locomotion in P. waltoni, as in all mudskippers, is by 'crutching' with the modified pectoral fins and resting the forward part of the body on the fused and forward-shifted pelvic fins (cover photo) (Fig. 17A). The tip of the tail is held in a distinctively semi-erect posture (see Fig. 1). Rarely, mudskipper locomotion is captured in a trail of 'footprints' (Fig. 18); more normally, only the central groove is preserved.

The modification of the pectoral fins, which consist of two muscular segments--a proximal basal lobe and a distal, modified ray fin (Fig. 19)--constitutes a kind of retrofitting for amphibious life, since mudskippers and other gobies are descended not from primitive lobe-finned fishes but from highly developed ray-finned fishes, specifically the perch family (Order Perciformes) (Polgar 2010). Modification of the pectoral fins for burrowing may have been a pre-adaptation for their use in locomotion.

The modification of the pelvic fins, on the other hand, is characteristic of the goby family as a whole (see, e.g., Feulner & Cunningham 2000), and in many species forms an organ capable of suction, allowing the fish to cling to hard substrates. Some Periophthalmus species are even able to climb the aerial roots of mangroves in this way (Hogarth 1999, see photo at p. 113).

When alarmed, P. waltoni can also move rapidly for at least several metres by hopping (and flopping), flexing its muscular body and tail to propel itself forward through the air.

'Comfort' behaviour

Comfort behaviour is a term that has come into use to describe behaviour whose function is to make the animal physically more comfortable. At least three such behaviours have been identified in mudskippers and discussed in the literature, all of which were seen in the UAE population: (i) periodic burrow visits, (ii) rolling, and (iii) pectoral fin waving. To that list we would add two behaviours involving orientation or position: shade-seeking and orientation away from the sun.

P. waltoni burrows are filled with water, except for the air pocket deep within, so the fish get wet when they descend beyond the entrance area. So-called "shuttling" between the burrow and the surface has the effect of promoting evaporative cooling and this behavioural adaptation is important to thermoregulation at high air and water temperatures (Taylor et al. 2011). Rolling, whereby a mudskipper rolls briefly to one or both sides in a shallow puddle, similarly promotes evaporative cooling. But P. waltoni and other species of Periophthalmus are active on tropical mudflats at air temperatures that frequently approach or exceed 38[degrees]C (100[degrees]F) (Taylor et al. 2011, Takeda et al. 2012), suggesting that temperature per se is not the only factor involved. Routine burrow visitation and rolling have been persuasively interpreted as equally or more important to facilitating cutaneous respiration, which requires a wet skin (Clayton & Snowden 2000 and references cited therein). Studies of mudskipper species other than P. waltoni have shown, for example, that the frequency of burrow visits and rolling is unchanged over a spectrum of temperatures (Yang Ka Yee 1996, cited in Clayton & Snowden 2000) but increases with increased wind speed (Tytler & Vaughan 1983, cited in Clayton & Snowden 2000), and that rolling behaviour is increased to the side towards which a stream of drying air was directed (Ip et al. 1991).

Pectoral fin waving is a more subtle and enigmatic behaviour. Sedentary fish will occasionally 'wave' the rayed portion of one or both pectoral fins, bringing it forward from its normal position (directed backwards and pressed back flat against the body) to a position more or less perpendicular to the body and the ground. The movement is brief but deliberate, consisting of forward stroke, brief pause and backward stroke. The entire sequence occurs in about 2 seconds. It does not seem likely to be a signal of any sort, since it does not depend on the presence of conspecifics in the vicinity, but it is sufficiently brief that it also does not seem very likely to contribute significantly to either thermoregulation or respiration. We did not notice that fin waving was associated with opening of the opercular cavity, as has been implied (Stebbins & Kalk 1961, cited in Clayton & Snowden 2000).

Shade is available within the study site from scattered small to medium sized mangrove seedlings and shrubs. Most mudskippers were observed in the sun, most of the time, typically at the mouths of their burrows. But some adult mudskippers had built main or auxiliary burrows with entrances that were shaded or partly shaded by mangroves and they were sometimes found at these locations, especially in summer. We interpret the choice of shade by fish which had the ready alternative of sunshine as a "comfort" choice--one which would obviously protect against overheating and skin desiccation.

It was our impression that the fish tend to orient themselves facing away from the direction of the sun, when it is not directly overhead. We did not attempt to quantify this phenomenon rigorously, but our observations at 0900 hrs on a morning in late June 2012 are illustrative. Of seven fish then in view, four were facing directly away (180[degrees]) from the sun and the other three had orientations of between 90-180[degrees] away. We hypothesise that such an orientation is more likely to be for reasons of improved vision than for thermoregulation. It is probably easier for these visually sensitive fish to see clearly, just as it is for humans, when they are not looking directly "into" the sun, or into its reflection off puddles on the mud surface.

Perambulation

This section attempts to give an account of the frequency and extent of movement by individual fish within the study site. P. waltoni is a relatively phlegmatic species overall. Movement was extremely limited during our visits in January 2013, in breezy and cool weather, when only a few fish even made an appearance in the mouths of their burrows. Movement was also very limited in March 2013, when only two fish left their burrows, notwithstanding that mating behaviour was observed on that occasion. On our two night visits, as previously indicated, only one fish ventured from its burrow mouth or platform.

Even when the fish were generally most active by day, not all fish were equally active. For example, on our two consecutive morning visits in late June 2012, lasting 3.5 to 4.0 hours each, only five fish each day were seen to travel more than one metre from their burrows, and of those, the same four fish were active on both days. This was evidently not a response to the presence of an observer, since the four fish active on both days were those closest to the main points of observation. Nevertheless, they withdrew towards their burrows if approached too closely.

Of the six most active fish, three spent much of their time away from their burrows. One large fish made repeated extended sorties through a chain of shallow puddles on open ground, travelling and feeding up to ca. 5 metres from its well-formed platform burrow. A second, smaller fish, with a nondescript main burrow and an auxiliary burrow ca. 4 metres away, was seldom in either burrow but perched on mud lumps within a trapezoidal area of ca. 10.5 sq. metres consisting of shallow puddles, crab burrows and a few mangrove sprigs.

The third, another large fish, was the champion in terms of distance and diversity of travel. This fish travelled repeatedly between its main burrow and a small clump of mangroves 2 metres away, where it had an auxiliary burrow. It also travelled ca. 8 metres in the same direction, almost reaching another prominent burrow, from which it was chased by the owner, who was coincidentally(?) returning. The first fish also travelled ca. 4 metres in a different direction, where it was surprised by an observer in the shade of another small mangrove clump and hopped back towards its burrow. Finally, it travelled 2-3 metres in yet a third direction for a liaison with a neighbouring fish (described under "Other social behaviour", below). The three main directions of travel were approximately 120 degrees from each other.

A fourth fish made occasional sorties from its burrow, each in the same direction, apparently for foraging. On one occasion it stopped and seemed to show awareness of another fish ahead in its path, about 1 metre away, causing it to stop for several minutes before returning toward its burrow.

The same fish later made a 1.5 metre journey to sit briefly in the shade of a small mangrove shrub.

A fifth fish was seen in liaison with a smaller neighbour (the second fish described above) in early morning on our first visit (as described under "Other social behaviour" below), after which it withdrew to a spot near its burrow, ca. 4-5 metres distant. On the following morning, it never left its burrow.

The sixth 'traveller' was a neighbour of the third fish described above. This fish was normally seen perched in the volcano-like mouth of its burrow or on larger adjacent mud lumps, but it had also travelled ca. 2-3 metres away, where it was approached by the third fish mentioned above (as described under "Other social behaviour", below), beating a hasty retreat in the end.

What might account for the observed differences in individual activity levels? Food and sex are, as always, among the primary hypotheses. On each morning in June, two of the mudskippers that travelled more than one metre engaged briefly in ambiguous but probably amorous physical contact, although we saw no unequivocal male sexual display or courtship behaviour in June 2012 as we had in February 2012 and in March 2013 (see below under "Courtship and mating behaviour" and "Other social behaviour").

On the other hand, at least three of the 'travellers' hunted and fed while they were at large. Were the more sedentary fish already sated and not motivated to hunt? If not, why not? Or was hunting and feeding by the travellers opportunistic and ancillary to travel for other purposes? Three of the fish that travelled farthest visited shallow, flat-bottomed puddles localised within the site, where polychaete worms were successfully hunted (see "Hunting, feeding and diet", below). Were those puddles, within and marginal to the territories of the concerned fish, the main goal of the excursions?

We have mentioned above, under "Burrows and burrow distribution", that one of the largest remaining adults in August 2013 had by then succeeded to a territory encompassing three conspicuous burrows, to which the fish circuited on the two mornings we watched, making a round trip of nearly 20 metres.

The newly arrived juveniles seen in August and October 2013 not only greatly outnumbered adult fish at the study site; they also tended to be more active and were somewhat less wary of approach. Possibly that is because, unlike adults, they did not yet have their own territories and needed to explore their surroundings to establish both geographic and social relationships. (See also "Juveniles and juvenile recruitment", below.)

Hunting, feeding and diet P. waltoni is an opportunistic carnivore and is known to eat small crabs, barnacles, shrimps, snails, worms (and probably other types of marine invertebrates) and even small fish and flying insects that land on the mud (Clayton & Wells 1987, Clayton & Snowden 2000, Mhaisen & Al-Maliki 1997, Polgar 2010). Although P. waltoni forages for food, it is ultimately an ambush hunter that lunges or pounces on its prey. Clayton & Snowden (2000) have described the capture sequence as follows: "Supported on pectoral and pelvic fins, the mudskippers flex their tails to one side forming their bodies into a variably acute "J" shape. By quickly straightening the tail, the fish leap rapidly at their prey." In Kuwait, where P. waltoni has been best studied, small crabs are considered to be their main prey, particularly the Ocypodid species Tylodiplax indica, Nasima dotilleformis and Ilyo-plax stevensi (Clayton & Snowden 2000). Polgar et al. (2009) likewise report the presence of numerous Ocypodid crabs within the zone where P. waltoni is present near Bandar Khamir, Iran, in the Strait of Hormuz.

At the UAE study site, in contrast, small crabs are effectively absent within the area of the mudskipper habitation. Macrophthalmus depressus is present in modest numbers at the downshore periphery, among mangrove pneumatophores. Only a few Uca annulipes, the UAE's most common fiddler crab, were ever observed near the study site. Two were seen on the firm upshore, above the mudskipper range; two others were seen adjacent to the site but separated by a physical barrier; one was seen in the jaws of a large mudskipper. Another small, unidentified Ocypodid was observed about 15 metres alongshore from the site.

Except in winter, the dark purple Metapograpsus messor (Fig. 27) is abundant and the somewhat larger, lavender-coloured Eurycarcinus orientalis (Fig. 10) is common, each in direct association with P. waltoni. Adults of both of those species are probably too large to constitute reasonable prey for P. waltoni (Clayton and Wells (1987) estimate that a carapace dimension of about 1 cm is the practical limit) but juvenile crabs are probably at risk. Large and medium sized M. messor occasionally ventured into the mouth of mudskipper burrows, although the crabs were invariably wary in the presence of actual mudskippers, always facing the fish, and retreating if approached closely.

We were able to observe eight successful food capture events and one unsuccessful attempt, involving six different fish. In three instances the prey was small flies, which the fish, sitting on mud, pounced on or lunged at, and in one instance caught in mid-air.

In three instances in June 2012, mudskippers were hunting in shallow, flat-bottomed puddles and the prey taken was a long (12cm+), reddish-pink polychaete worm. In each such instance, after a lunge, the fish could be seen with part of the worm trailing from its mouth on the mud alongside its head (and in one case alongside the full body length of the fish). The remainder of the worm was gobbled up in successive gulps. The one unsuccessful capture was a lunge by a fish hunting in a shallow puddle (where it was later successful); the intended prey is likely also to have been the polychaete worm. In five of the six foregoing instances in which we could determine the concerned fish and its burrow, prey was captured (or not) at a distance of approximately 4-5 meters from the main burrow.

The identity of the polychaete worm prey is unknown. The fact that it was evidently soft-bodied suggests that it is most likely a member of the Family Nereididae (Richard J. Hornby, pers. comm.) but no species fitting its description, particularly the large size, is known from surface observations of UAE mudflats (pers. obs., R.J. Hornby, pers. comm.). The worm is tentatively presumed to be a shallow subsurface dweller that P. waltoni can somehow sense.

In the seventh successful instance, a fish perched on the berm of its ringed pond (see Fig. 11) lunged into the surrounding shallow pool and caught a prey item that appeared to require several seconds of mastication; the prey was not seen well but could possibly have been a small crab.

The last instance was arguably the most dramatic, and also "the exception that proves the rule". In August 2013, a large adult caught and ate a fiddler crab Uca annulipes. We missed the instant of capture but we were alerted by movement and saw the fish in the mouth of its burrow with the crab in its jaws (Fig. 20). Watching, we were uncertain whether the fish would be successful in ingesting the relatively large crab. The process proved to be slow, but relentless and effective. The crab was repeatedly taken into the oral cavity and then exserted, each time with fewer legs, claws and/or eye stalks than before. Those were undoubtedly severed and crushed by the sharp teeth arrayed in the jaws of Periophthalmus spp. and other carnivorous mudskippers (Sponder & Lauder 1981, Murdy 2011).

Since we had never seen any other Uca annulipes within the mudskipper site, we tentatively infer that P. waltoni predation on Uca annulipes is relatively efficient, and that the mudskippers themselves may be responsible for the absence of these fiddler crabs.

Clayton & Snowden (2000) found that P. waltoni in Kuwait conducted successive hunting forays in different directions and reasoned that this increased their foraging efficiency because their prey, small crabs, remained in their burrows longer than usual following a hunt in their immediate vicinity. In contrast, in the UAE population, individual mudskippers tended to journey to the same areas repeatedly, whether foraging or otherwise, often favouring shallow puddles. Possibly this reflects the difference in prey species available at the UAE site, being worms and flying insects which would not be sensitive to prior visitation.

Also in contrast to Clayton & Snowden (2000), the UAE mudskippers did not return to their burrows soon after a successful prey capture. In the case of small flies, this might be because those prey do not constitute a "meal", but a fish-length polychaete worm undoubtedly approximates or exceeds the body mass of the small crabs on which Kuwait mudskippers typically feed.

In October 2013, we observed an enigmatic feeding behaviour that we had not noticed previously. Two fish, feeding in adjacent shallow puddles, appeared to be 'snuffling' through the muddy substrate in the manner of a dugong eating sea grass. Possibly this was an attempt to find and uncover food items, or to ingest small food items in a slurry of mud.

Territoriality

P. waltoni is a territorial species (Clayton & Wells 1987), like most if not all Periophthalmus species. All of the adult mudskippers that we observed at the study site maintained one or more burrows that were the centre of their activities and that were not trespassed upon by other adult mudskippers. In some cases the burrows or their surrounding areas were actively defended. Clayton & Wells (1987) made the point that, in common with many other species, territorial contests among mudskippers are usually decided in favour of the resident.

We saw four clear instances of intra-specific territoriality. Two, in late July 2012, were classic confrontations at close range, with both fish at or near what we knew to be the margins of their respective territories. In each case the fish alternately turned sideways and then faced each other at centimetre range, sometimes opening their mouths to reveal the bright pink interior (Fig. 21). In one interaction, both fish partially raised their rear dorsal fins and one of the fish repeatedly flashed its forward dorsal fin. In the other, both fish raised their dorsal fins on a few occasions, once for as much as 10 seconds. In no case, however, did the fish make physical contact.

The confrontations were interspersed with intervals when both fish remained motionless for up to ca. 30 seconds at a time, and were occasionally interrupted while one fish or the other would roll briefly. In each case the interaction ended after ca. 6-7 minutes, when the fish gradually separated and ceased to take note of each other. To the human eye, there was no obvious winner or loser, but it may be significant that in each case it was the fish that was most distant from its main burrow that withdrew first and farthest from the site of the contest.

These encounters lacked the drama of some mudskipper accounts. There was little displaying, no leaping, no vocalising and no ignominious retreat. Possibly this was because in both instances there had been no actual transgression of the territorial boundary, or possibly territorial defence is more relaxed outside the breeding season, or possibly P. waltoni is simply a less demonstrative species; however, these do not exhaust the possibilities.

A third instance of territorial defence took place at longer range. In that instance the resident fish was returning towards its prominent burrow. When it was still in a shallow puddle at a distance of about one metre from the burrow, it evidently spied another mudskipper approaching from nearly the opposite direction, also about one metre from the burrow. The resident became agitated and skipped ahead more quickly, whereupon the interloper turned and retreated at a rapid 'walk' in the direction from whence it had come, which was the direction of its own burrow ca. 7 metres distant. These fish were sensitive to each other's presence at a distance of more than 2 metres, despite their low vantage point, the uneven ground between them, and their camouflage colouration.

In a fourth instance, in August 2013, when juvenile fish were also present at the site, an adult emerged from its burrow to clear a mangrove leaf and spied a neighbouring adult that had approached to within ca. 0.5 metres. The burrow owner immediately sallied forth and the interloper beat a retreat. The proprietor remained in the confrontation area for only a couple of minutes before returning to its burrow mouth.

The arrival of large numbers of juvenile fish, first observed in mid-August 2013, increased the population density at the study site by a factor of 2 to 3 from what we had previously observed, but the resident adult fish appeared to tolerate this, at least within the limits observed. On one morning in

August an observer was surrounded by 5 juvenile fish within a 2 metre radius. Two of them were perched on the margins of an abandoned(?) adult burrow (Fig. 22) and three were within one metre of an adult fish sitting in the mouth of its burrow--a distance that would not have passed without notice among adult fish.

This may indicate that juveniles are recognised as such, and specifically as being sexually immature. It does not explain, however, why juvenile fish would be tolerated if an adult territory is maintained for resource protection purposes. A degree of resource-based exclusion is suggested by the fact that, as previously noted, a number of newly-arrived juveniles (and their burrows) were observed in areas at the margins of the study site, which had not been occupied by the earlier adult population.

In any case, we never saw adult fish actively chase juveniles, although that was perhaps partly because juveniles would typically distance themselves from any adults in motion nearby. We did, however, see larger juveniles actively chase smaller ones, for distances up to ca. one metre (see also "Juveniles and juvenile recruitment", below). We tentatively relate this to their need to establish and maintain territories, and perhaps also social relationships, of their own.

Threat display

We saw only a single full-blown threat display, which was directed not at another mudskipper but at a medium-sized crab, the lavender crab Eurycarcinus orientalis. On our second visit in February 2012, a crab had ascended to the burrow platform of a mudskipper, which at the time was ca. 0.4 metres away. The fish returned immediately to the base of the platform and the crab descended to flat ground, where the fish positioned itself between the crab and the burrow and raised both dorsal fins (Fig. 23), the front dorsal fin being sail shaped and the elongated rear dorsal fin being marked with a distinctive black and white stripe. The crab stood its ground for a few seconds before sidling away at a measured pace; the mudskipper lowered its fins but continued to monitor the retreat.

In August 2013 we again saw an adult mudskipper, on open ground near its burrow, briefly flash its dorsal fins at an equivalent sized E. orientalis crab, which likewise moved slowly away.

Courtship and mating behaviour

We saw in February 2012 the first of only a few instances of a male courtship display. A large fish perched atop its well-formed platform burrow, to one side of the central entrance hole. It stood high on its pectoral fins, raised its dorsal fins and turned slowly through ca. 90 degrees, facing the periphery of the platform. Then it suddenly flipped into the air and landed on the other side of the entrance hole, almost 180 degrees opposite and a fish length or more away. It again erected its fins, but only briefly, before settling down atop the platform. We did not notice any other fish in the vicinity towards whom this display might have been specifically directed, but in the circumstances our attention was focused primarily on the displaying male.

We also saw in February 2012 a single instance of the mating or pre-nuptial ritual, which appeared to conclude successfully. We did not observe the male display or the initiation of contact, but our attention was attracted to two fish, one following the other diligently for about 4 metres across the mud, mostly splashing through very shallow puddles, to a well-formed platform burrow. The lead fish mounted to the flat top of the platform and briefly flared its dorsal fins. The second fish, instead of being intimidated, also mounted the platform and they engaged, face-to-face and open-mouthed in a prolonged mudskipper 'kiss', heads elevated and dorsal fins lowered, lasting for perhaps half a minute. Then the first fish turned and popped down the central entrance hole, and the second followed immediately. Egg laying and fertilisation would have followed within the burrow. But for the ending, however, the observed sequence might have been (mis)taken for a territorial or aggressive interaction.

Broadly similar courtship rituals have been described by Brillet (1980, 1984) for Periophthalmus sobrinus (= P. argentilineatus or P. kalolo) and by Lee and Graham (2002) for P. modestus (= P. cantonensis). The procession of the two fish, led by the male, to the burrow built by the male, has been colourfully christened the "nuptial promenade" (promenade nuptiale) by Brillet (1984). In P. modestus, Lee and Graham (2002) did not distinguish the nuptial promenade as a phase distinct from the male's continuous effort to display and draw an attentive female towards the breeding burrow, and neither Brillet nor Lee & Graham described mutual physical contact at the breeding burrow.

Our observation permits a tentative summary of the "full" courtship sequence for P. waltoni as follows: male display (erection of the dorsal fins, arching and jumping), the nuptial promenade, display at the breeding burrow, mutual contact, descent, cohabitation, egg laying and fertilisation. Because we did not observe the beginning of the sequence, we cannot say whether the sequence may also have included some act or contact to confirm the female's interest prior to the nuptial promenade. It is worth noting specifically that the existence of the nuptial promenade seems to entail that the male's display will often, if not always, take place at a location other than the breeding burrow itself.

A somewhat less romanticised and less rigorously sequential version of courtship and mating was observed in March 2013. On a day when the reduced population of adults as a whole was relatively inactive, one pair was evidently cohabiting. The two fish were first seen together at about 0845 hrs, one at each of the entrances atop two large, closely adjacent burrow platforms in a relatively secluded location (Fig. 24). The fish on the left withdrew underground. The fish on the right platform (slightly darker, believed to be the male and the proprietor of the burrow) then emerged on top of its platform, elongated itself and made two jumps in place, reversing its direction. Then it made 2-3 more jumps, with its dorsal fins flashing, towards the left platform, and went immediately down the hole. Both fish re-emerged in ca. 2-3 minutes, one at each entrance, proving the underground connection. Both then descended again, but surfaced within another minute.

Five to ten minutes passed without further action. A fish then emerged on the left and briefly laid out atop its platform, before descending again. The fish on the right emerged onto its platform but immediately took several short leaps towards the left platform and went down. Within a minute, fish had returned to the mouths of both platforms. The time at this point was about 0930 hrs. The fish on the left now laid out on its platform, then hopped towards the fish on the right, making physical contact. Both fish "fell" together down the edge of the platform, but only one (the proprietor and presumed male) returned to the top, where it laid out flat and performed a few low jumps with its dorsal fin flashing. It then went underground, but for less than a minute, returning to lay out on the right hand platform and make two more half-hearted jumps. It descended once more, but reappeared almost immediately at the left hand entrance, where it took up an extended vigil.

The second fish, meanwhile, returned slowly to the vicinity of a neighbouring burrow, probably its own, where earlier a fish (most likely the same one) had been surprised, about 45 minutes before the cohabiting pair was noticed. We tentatively interpret these interactions as a visit by the second fish, presumably a female, to the burrow complex maintained by the first fish, presumably a male, for mating purposes. The observed interactions between the two fish had occupied approximately 1.0-1.5 hours in total. The absence of more structured courtship behaviour could possibly be explained by the fact that the two fish were already acquainted.

Despite studies in Kuwait, the breeding season for P. waltoni in the Arabian Gulf has apparently not been described previously (Polgar 2010); for the sympatric Boleophthalmus dussumieri, the breeding season in Kuwait is said to extend from April through August, and in Pakistan into September (Polgar 2010). In P. modestus of Japan and South Korea, the late spring and summer breeding season has been shown to be somewhat longer at lower latitudes (Ishimatsu et al. 2007, Baeck et al. 2008, Polgar 2010).

The limited observations reported here for the study site suggest that in the southern Arabian Gulf the breeding season for P. waltoni includes the period from mid-February through mid-March, and probably continues through May and perhaps into late June (see "Other social behaviour", below). This is supported by direct evidence of courtship and mating as well as by evidence of the recruitment of seagoing juveniles in late July or early August (see "Juveniles and juvenile recruitment", below). The authors hope that targeted observations in 2014 will elucidate the local breeding and recruitment periods more precisely.

Other social behaviour

We also observed, in late June 2012, what is perhaps best interpreted as an unsuccessful courtship attempt. Two mudskippers, one noticeably smaller than the other, were perhaps a metre apart but were evidently aware of each other when we first noticed them. The larger fish followed the smaller one for another metre or so, into what we later learned was the foraging territory of the smaller fish. When they stopped, both postured at close range (less than a fish length), each one opening its mouth to display the pink interior and then turning sideways and showing its flank to the other, but without any display of fins (Fig. 25).

Finally, the smaller fish approached the larger and held its open mouth obliquely against the side of the larger fish's head--a 'cheek kiss' seemingly identical to the courting behaviour depicted for Periophthalmus modestus in Polgar (2010)--for perhaps half a minute. Then, without any further posturing or interaction, the smaller fish moved ca. 30 cm away and the larger fish began a slow return, over 10-15 minutes to its own burrow ca. 5 metres away. Because of the non-hostile nature of the encounter, the 'following' behaviour and the unilateral close open-mouth contact, we tentatively interpret this interaction to have been an amorous one between a small male and a larger female.

The reasons why this particular encounter was aborted are speculative. The smaller presumed male did not maintain a well-formed burrow and was perhaps not in full breeding condition. He did not display and his approach was demure in comparison to mudskipper courtship descriptions in the literature. Perhaps his small size made a difference, too.

A similar incident was observed in October 2013, involving a small (juvenile) darker fish and a distinctly larger, pale-headed adult. The two fish were seen at the edges of adjacent shallow, flat-bottomed puddles, each with its second dorsal fin up. The smaller fish flashed its first dorsal fin but the two continued feeding. Then the smaller fish advanced to the intervening low ridge of mud. After an interval the larger fish approached and positioned itself obliquely next to the head of the smaller fish, with its mouth open. After a few seconds, the smaller fish lunged unexpectedly at the larger one, which backed off to its puddle and continued its foraging activity. The smaller fish then did the same, and after a few minutes perched on the slopes of its nearby burrow, each fish apparently ignoring the other.

A more dramatic but indecipherable interaction occurred in late June 2012. Two mudskippers who proved to be from neighbouring burrows were observed at a distance, positioned more or less side by side in a trough among mud lumps, at a distance of about 2.5 metres from each of their respective burrows. They remained motionless for perhaps 20-30 seconds. Then, suddenly, one of the fish flared its dorsal fins and turned; there was a splash and the other fish hopped quickly away to a distance of about 0.5 metres, in what turned out to be the direction of its burrow. After a short interval the dominant fish also moved slowly back toward its own burrow. It is possibly significant that the dominant fish was one of the three most active 'travellers', who was observed to explore in three different directions from its burrow.

Juveniles and juvenile recruitment

Mudskipper fry do not develop in situ. Instead, after incubation and hatching, mudskippers spend their early lives as pelagic (seagoing) larvae, after which they return to the intertidal zone on a suitable high tide (this process is called "recruitment") and rapidly metamorphose into smaller versions of adults. See "The mudskipper life cycle, including a planktonic larval phase", below.

Juvenile P. waltoni were first seen at the study site in August 2013, when they outnumbered adults and increased the local population from 8 to ca. 30+ fish. Juveniles are distinguishable from adults by their smaller size and more vivid black markings on their flanks, gill covers and lower spine (Fig. 26).

Juveniles at the study site were also distinguishable behaviourally from adults, as mentioned in discussion of various points above.

Among other things, they tended to occupy more upper intertidal habitats, to build smaller and more nondescript burrows, and to be active on the surface for longer periods.

It should be noted, however, that our description of the smaller fish newly observed in August 2012 as "juvenile" is not meant to preclude the possibility that some of them were sexually mature, or became sexually mature by the time they were observed again in October. The maturation time for P. waltoni is unknown but it has been said of the similar-sized P. barbarus that it can reach sexual maturity within a year, and at half of maximum adult size (Polgar 2010). In two cases, interactions at the study site that were possibly amorous in nature involved fish of markedly different sizes, although in each case those interactions ended inconclusively (see "Other social behaviour", above).

Modest size differences were noted among juveniles, which were estimated to range from 9 to 11 cm in total length. This can reasonably be interpreted to reflect two different episodes of larval recruitment, most likely on successive spring high tides. On several occasions larger juveniles were observed to chase smaller juveniles, for distances up to ca. one metre.

The timing of recruitment can be used in a rough way to reconstruct the breeding period for P. waltoni in the UAE. Data is scarce for mudskipper incubation times, which probably vary with temperature, but an estimate of "about one week" is given for P. modestus (Ishimatsu et al. 2007, Polgar 2010). The only estimate we have found for larval maturation time in mudskippers is 30-50 days for P. modestus (Kobayashi et al. 1972, Lee & Graham 2002, Ishimatsu et al. 2007, Borges et al. 2011), a species which is 50-65% smaller than P. waltoni (Polgar 2010). The exact timing of recruitment is inevitably influenced by the tidal cycle.

If we assume for P. waltoni a minimum larval maturation time equal to or slightly exceeding the high end of the range for P. modestus, say 50-60 days, and if we add a week or two for incubation, then the presence of apparently recently arrived P. waltoni juveniles in mid-August implies that breeding is likely to have continued until at least mid to late May and possibly into early June. A similar calculation implies that the juvenile progeny of mid-February breeding could return to the upper intertidal zone by latest April or early May.

With hindsight, based on the criteria set out above, one of the 15 fish recognised at the study site in June and July 2012 was probably a recently arrived juvenile (see Fig. 12). If so, it would likely have been the progeny of the earliest mating activity observed, in February 2012. But this begs the question why, in 2012, only a single juvenile was seen through late July, whereas in 2013 at least two dozen juveniles were present by mid-August (and continued to thrive in October). What unknown factor(s) made the July-August 2013 recruitment period so much more successful than earlier recruitment episodes in 2012 and 2013?

The recruitment period that preceded our visits in mid-August 2013 not only increased the population of our study site; it also resulted in the colonisation of a nearby area, about 400 metres distant, similarly positioned above the thick mangrove forest in the uppermost intertidal zone, but on largely open ground covered by a thin layer of cyanobacteria (Fig. 15). In August, some two dozen mudskipper burrows could be readily distinguished over a distance of about 100 metres alongshore, although only three fish were actually seen. Nondescript burrows could have boosted that total by up to 50% or more. A repeat visit to this satellite area in October 2013, less than two hours after inundation, again revealed about two dozen burrows, although not all could be confirmed as active and only two or three fish were seen. Several burrows had relatively well-formed platforms and one had a 'swimming pool'. We hope to be able to continue to monitor the satellite population in the future.

The discovery of a second colony of mudskippers, even one in proximity to our study site, raised the possibility that, at least during the recruitment period, juvenile mudskippers could be found at other locations where permanent colonies did not exist. We therefore spent two days in early September looking for evidence of mudskippers at additional likely sites along the Arabian Gulf coast of the Northern Emirates, but we found none.

Associated species

Other plant and animal species present in the immediate environment include:

* Plants. Mangrove trees Avicennia marina (Avicenniaceae) have thrived over the past decade in many parts of the khor area in question, except where they have been physically obliterated. A relatively thick forest with trees two to four metres high begins immediately adjacent to the mudskipper site. A few spreading clumps of the mostly supratidal Arthrocnemum macrostachyum (Chenopodaceae) are found within the site.

* Crabs. Metapograpsus messor (Grapsidae), a dark purple diurnal crab, is a generalist species common in many intertidal environments in the UAE. It was abundant within the study site in summer (Fig. 27) but was absent or scarce in mid-winter; none were observed on our initial visit in mid-February 2012, and only a single M. messor was recorded on the chilly and breezy late afternoon of our second visit in late February.

The pale violet-coloured Eurycarcinus orientalis (Xanthidae) (Fig. 8) is a mudflat specialist. It is common within the study site and its burrows can resemble those of P. waltoni. It is the largest crab species present at the site, although many smaller than average individuals were seen in July 2012.

Macrophthalmus depressus (Ocypodidae), a small (fiddler-sized) mudflat and mangrove specialist, was occasional in damp, puddled mud on the downshore edge of the mudskipper site, adjacent to the more mature mangrove forest.

Uca annulipes (Ocypodidae), the most common fiddler crab in the UAE, was rare at the study site, with fewer than half a dozen seen, in June 2012 and August 2013, all on firm, sandy mud above the area occupied by mudskippers or separated from it by a physical barrier.

A single, unidentified small crab, probably one of the Ocypodidae Tylodiplax indica, Nasima dotilleformis or Ilyoplax sp., was seen in damp, puddled mud adjacent to the mudskipper site, in association with M. depressus. All three of those Ocypodidae are prey species of P. waltoni in Kuwait.

Portunus pelagicus (Portunidae), the Blue Swimming Crab, is common in the main channel and smaller individuals are occasional within the mangroves. Some of the latter might possibly venture into the mudskipper area at highest tides.

* Molluscs. The small, brown-and-white banded mud creeper Cerithideopsilla conica (formerly Potamides conicus) (Gastropoda) is common in damp, puddled mud in the intertidal zone just below the mudskipper site. We saw no evidence that P. waltoni preys on this mollusc, although it is all but totally absent within the area inhabited by mudskippers. No other surface-dwelling molluscs are present within or immediately adjacent to the study site, nor did we find there the shells of burrowing bivalves.

* Polychaete worms. These prey items are apparently present at shallow depth in the substrate (see above under "Hunting, feeding and diet").

* Insects. The tiger beetle Cicindela (Calomera) aulica (syn. Lophyridia aulica) (Carabidae: Cicindelinae) was seen in small numbers in June and July 2012. The same and another tiger beetle, Salpingophora hanseatica, were seen in small numbers in August 2013, mostly on firmer ground above the mudskipper habitat. Small flies (Diptera spp.) were also moderately common. At least one resembled a housefly with wings spread in a wide triangle. Another was smaller and elongated, with wings folded directly back over the abdomen, parallel to the body; this is possibly a species of Ephydridae, the larvae of which are known from intertidal samples in the UAE (Richard J. Hornby, pers. comm.).

* Fish. The Arabian killifish Aphanius dispar is present at the site in large numbers when it is inundated by high tides. This is the most common fish in UAE lagoons and can survive stranding in shallow intertidal puddles, at least for single daily tidal cycles.

* Birds. Individual Kentish Plovers (Charadrius alexandrinus) occasionally fed on the mud surface within the mudskipper site. House Sparrows (Passer domesticus) fed on mangrove flowers in June 2012. Individual Little Green Bee-Eaters (Merops orientalis) briefly hunted over the site in June 2012, August 2013 and October 2013. Also in August 2013, a short-necked heron, probably a Striated Heron Butorides striata, was seen stalking at the edge of the mangrove forest. A few Purple Sunbirds Cinnyris asiaticus and White-Eared (formerly White-Cheeked) Bulbuls Pycnonotus leucogenys leucotis fed in adjacent mangroves in October 2013. Also in October 2013, a few Indian (formerly Western) Reef Herons Egretta gularis schistacea hunted at high tide in the area where a satellite population of juvenile mudskippers was observed, but only while it was inundated.

* Mammals. Tracks of red fox (Vulpes vulpes) and of a large feral dog (Canis domesticus) were found in the immediate vicinity; the dog tracks were headed directly towards the mudskipper site. Despite the current regime of protection, small numbers of humans (Homo sapiens) and their vehicles still pass close by on an occasional basis for fishing, hunting and family recreation. Three adult mudskipper burrows on the periphery of the site were constructed over or adjacent to earlier tyre tracks; at least one platform burrow was partially damaged by vehicle traffic during the course of the study.

Predation and predator avoidance

No predation on mudskippers was observed at the study site, nor (with one exception, discussed below) were any potential predators seen in the immediate vicinity, although large waders and gulls feed in good numbers in and along the adjacent arm of the khor (including Western Great Egret Egretta alba, Grey Heron Ardea cinerea, Purple Heron Ardea purpurea, Indian (formerly Western) Reef Heron Egretta gularis, Striated Heron Butorides striata and 'Yellow-Legged' Gulls Larus sp.). The observers' presence undoubtedly discouraged approach by these larger shorebirds known to feed on mudskippers elsewhere (Clayton & Wells 1987), but it is also likely that the very small and localised population of mudskippers, living in marginal and relatively exposed habitat within an otherwise biologically rich area, has simply not attracted the attention of resident avian predators as a resource worth exploiting. Certainly the site would not support specialised predation techniques of the sort described by Clayton & Wells (1987) from the extensive mudflats of Kuwait, where P. waltoni is sympatric with Boleophthalmus dussumieri:

"Shore birds are [the mudskippers'] worst enemies, and we have already described the hunting methods of herons [and egrets and other shore birds]. Terns [probably the Gull-Billed Tern Gelochelidon nilotica (Richard J. Hornby, pers. comm.)] are another menace to the mudskipper which they catch by flying along about three to five metres up, then dive-bombing their prey with their bayonet-like beaks. Sea gulls use a craftier approach, skimming very low, just above the mud walls* and trying to catch the Boleophthalmus unawares." [*Authors' note: At high densities B. dussumieri builds low mud walls to segregate the polygonal territories of neighbouring fish (Clayton 1987). See discussion below, "Were two species of mudskippers present in the UAE in the early 1970s?"]

The one potential predator we saw at the study site was a short-necked heron, probably a Striated Heron Butorides striata, seen stalking on the mud in August 2013 at the edge of the site, beside the mangrove forest, where it retreated when approached. Although the heron was not actually within the area of mudskipper burrows when observed, its presence was perhaps not coincidental, since the study population was swelled by juvenile mudskippers at that time. Indian (formerly Western) Reef Herons Egretta gularis schistacea were seen hunting at high tide in the area where a satellite population of juvenile mudskippers was observed, but only while the area was inundated.

Sea snakes, fish and crabs have also been identified as predators of mudskippers elsewhere (Clayton & Wells 1987). The mudskipper's principal defence against marine predators is a retreat to its burrow during high tide, and the main burrow typically has multiple entrances. A number of species of sea snakes are found in UAE waters and many are known to prey on burrowing or hole-dwelling gobies, eels and similar species (Egan 2007), but the upper intertidal habitat of P. waltoni makes it less susceptible to regular sea snake predation. The relatively large size of adults may confer additional protection. The two largest burrowing crabs present at the UAE site, M. messor and E. orientalis, are too small to prey on adult mudskippers but could potentially eat eggs or young (Clayton & Wells 1987). Small numbers of the Blue Swimming Crab Portunus pelagicus could possibly reach the site with the highest tides, but it is questionable whether these could effectively prey on mudskippers in their burrows, and in any case the most likely arrivals would be smaller individuals.

P. waltoni is regularly eaten by fishermen in southern Iran (Polgar 2010). It is sufficiently abundant and has been regarded as an important enough part of the local food chain that in both Kuwait and in southern Iran it has been studied as a bioindicator for the accumulation of heavy metals (Bu-Olayan & Thomas 2008, Askari et al. 2010, Kooseg et al. 2011).

The authors found P. waltoni to be well camouflaged when away from its burrow and not in motion. Not only does the body colouration and patterning blend with the damp intertidal mud and its shadows and reflections; the head and prominent dark eyes were found to closely resemble small mud lumps topped with the abundant flotsam of dark, dried mangrove flowers (Avicennia marina) (Fig. 28).

In our experience, mudskippers surprised or approached too closely by human observers withdrew into their burrows, or, if encountered on open ground, they would move away, usually towards their burrows if possible. Speed of movement was variable, depending on the degree of surprise, and ranged from a steady "crutching" away to a series of rapid hops covering three metres or more. One juvenile fish that retreated slowly from an approaching observer took shelter not in a burrow but under a small, spreading mangrove shrub, in exactly the same way that many desert lizard species take refuge under scattered vegetation.

Juveniles seemed to be more tolerant of human observers, or at least were more likely to be active when observers were present, even at a close distance, but this may largely reflect the fact that the juveniles we observed in August and October 2013 were more active than adults overall.

Selected aspects of the biology and ecology of P. waltoni

Adaptations to life on the mudflats: physical, physiological and behavioural Mudskippers are the most thoroughly amphibious fish. They are also well adapted to the rigours of life in tropical and subtropical intertidal mudflats. Their amphibious behaviour itself can in fact be understood as an adaptation to that environment. This recognition has given rise to the hypothesis that the exploitation of the land (and air) by vertebrates was driven in part by the stresses imposed by the muddy intertidal zone (Shultze 1999, Ishimatsu & Gonzalez 2011, Taylor et al. 2011).

The ability to "walk" on land and to breathe air through the skin and bucco- pharyngeal mucosa, described above, are only the most obvious of mudskipper adaptations. Other physical adaptations include not only the positioning of the eyes but also the anatomy of the eye, which is better adapted to vision in air than underwater. Mudskippers have good vision both at close range and at a distance, and close vision seems to be especially sensitive to horizontal movement, presumably an advantage in sensing and catching prey (Polgar 2010).

A major physiological adaptation of mudskippers is an increased tolerance to ammonia (NH3), a natural metabolite which underwater is excreted through the gills, as in most fish, but which cannot be excreted in air. At least some mudskipper species have evolved chemical pathways to detoxify accumulated ammonia (Polgar 2010).

Mudskippers living on exposed tropical mudflats are also tolerant of changes in water temperature and salinity, particularly high temperatures and high salinity, relative to most other fish, and of low dissolved oxygen levels in the seawater that remains in their burrows between inundations (Taylor et al. 2005, Polgar 2010). Exposure to the direct rays of the tropical or subtropical sun may cause extreme warming of surface mud and warming and evaporation of shallow surface water, increasing its salinity to as much as twice that of normal seawater (Takeda et al. 2012). Underground, the seawater in mudskipper burrows becomes depleted in oxygen through reaction with the substrate of anoxic, organic rich mud as well as from increased temperature and salinity; for upper intertidal species like P. waltoni, burrow water may be refreshed by inundation only at extended intervals within the monthly tidal cycle. A complementary behavioural adaptation is the maintenance of an air pocket within the burrow (Ishimatsu et al. 1998, Lee et al. 2005), although in at least one species, P. minutus of northern Australia, this may be only a breeding season or austral summer phenomenon (Takeda et al. 2012).

Amphibious behaviour helps mudskippers to alleviate the rigours of their mudflats environment. In particular, air-breathing frees them from the threat of oxygen-depleted waters and amphibious locomotion allows them to take advantage of evaporative cooling, even at relatively high temperatures.

To assess the adaptive contribution of amphibious behaviour, Taylor et al. (2005) compared the physiological and behavioural responses of a mudskipper (Periophthalmus kalolo, an air-breathing, amphibious goby) and two non-amphibious burrowing gobies (Bathygobius spp.) found sympatrically on mudflats in Sulawesi, Indonesia. Taylor et al. considered that their site experienced some of the most demanding daily temperature fluctuations seen in an aquatic habitat, but there is no reason to think that upper intertidal conditions in the Arabian Gulf are much less rigorous. They found by experiment that in both the mudskippers and the non-air breathing gobies, the increase in metabolic rate as water temperature rose was much lower than expected, thereby reducing increased oxygen demand. Nevertheless, both non-air breathing gobies, who could not leave their intertidal burrows, had significantly higher physiological tolerances than the mudskippers for elevated water temperature and low dissolved oxygen. Consistent with this finding, when temperature and [O.sub.2] parameters were manipulated experimentally, mudskippers would react by exiting their burrows, if possible, when their tolerance limits were reached.

The mudskipper life cycle, including a planktonic larval phase

The life cycles of most individual mudskipper species have not been studied in detail but it is possible to generalise from studies to date of some of the better known species. Adult male mudskippers construct a breeding burrow containing an egg chamber with an air pocket maintained by air gulping (Kobayashi et al. 1972, Lee & Graham 2002, Ishimatsu et al. 2007, Ishimatsu & Gonzalez 2011, Takeda et al. 2012). The air pocket is important to development of the eggs due to the low dissolved oxygen content of burrow water in the anoxic mud, which can be less than one-tenth that of air-equilibrated seawater (Ishimatsu et al. 2007, Ishimatsu & Gonzalez 2011). In at least some species the eggs actually develop within or near the air pocket, embedded in the mud walls (Lee & Graham 2002, Ishimatsu et al. 2007, Ishimatsu & Gonzalez 2011). After mating, the male guards the eggs (as in many non-air breathing goby species), usually alone but sometimes in co-operation with a female (Lee & Graham 2002, Polgar 2010).

Eggs numbers are very high, probably thousands per brood in most cases. Studies for three northerly species from Japan and South Korea report an average of more than 5,000 eggs per burrow for Periophthalmus modestus (Ishimatsu et al. 2007), 5,000-20,000 eggs per fish for Boleophthalmus pectinirostris (Kim et al. 2011) and ca. 1,300-18,300 eggs per fish for Scartelaos gigas (Kim et al. 2011).

Ishimatsu et al. (2007) showed that in P. modestus, when the eggs are ready to hatch, the guardian male causes them to be inundated on a nocturnal flood tide by actively gulping and removing air from the breeding chamber. This triggers hatching and the newly released larvae, which are planktonic (only millimetres long and with a yolk sac that will last them for a few days), disperse to find their way seaward on the ebb tide (Ishimatsu et al. 2007, Kim et al. 2011). In P. modestus, the planktonic larval stage lasts for 30-50 days, after which the larvae re-enter the intertidal zone, become benthic and rapidly metamorphose into miniature adult mudskippers (Kobayashi et al. 1972, Lee & Graham 2002, Ishimatsu et al. 2007, Borges et al. 2011).

P. waltoni has been estimated to reach maximum size in about 1.5 years for males and 0.9 years for females; its lifespan has been estimated at 4+ years (Sarafraz et al. 2011). This is roughly consistent with estimates for the similar-sized P. barbarus from West Africa, which reaches maturity in less than one year and at about half maximum size (Polgar 2010); P. barbarus has an estimated lifespan of 6 years (Polgar 2010).

The existence of a planktonic larval phase in mudskippers parallels the life cycle of many intertidal molluscs, including two of the most common gastropods on the mudflats of the UAE, the Potamidids Cerithideopsilla cingulata (formerly Cerithidea cingulata) and C. conica (formerly Potamides conicus). Planktonic larval dispersal preserves the genetic homogeneity of such species throughout their ranges, despite their often highly localised occurrence as adults (David G. Reid, pers. comm.). The existence of a planktonic larval phase is of particular significance in considering the origin of the UAE study population.

Re-Discovery of P. waltoni in the UAE: Why here? Why now?

The origin of the newly discovered UAE population of P. waltoni is speculative. It seems almost equally improbable that it has survived unnoticed for two decades or more, and/or that it has somehow re-colonised this single small site during a time of massive 'development' and disruption of UAE coastal environments.

The senior author (GRF) has explored many areas of the khor in question in the course of at least 15 prior visits from 1998 through 2010. He has devoted special attention to intertidal molluscs (including the recognition of two air-breathing upper intertidal species new to the UAE) but has also observed there a dozen or more fish species, cuttlefish, spoonworms (Phylum Echiura), an agama lizard and a tiger beetle new to the UAE, all in the intertidal or shallow subtidal zones. On two or three occasions from 2006 to 2010 he passed within easy view of the mudskipper site, in one case having stopped to look for small, air-breathing molluscs under adjacent saltbushes. But it is impossible to say whether mudskippers were then absent or were overlooked.

The population encountered in 2012 consisted of adult fish, some of them evidently males of essentially maximum size, so it is reckoned to have been a minimum of at least 1.5 years old (Sarafraz et al. 2011) when discovered, therefore having existed since at least mid-2010. On the other hand, if the lifespan of individual P. waltoni is a maximum of 4+ years (Sarafraz et al. 2011), then the fish observed in early and mid-2012 must have arrived during or after the second half of 2007.

The distribution of P. waltoni within the Arabian Gulf is patchy, so its requirements are evidently not indiscriminate, but it is not immediately clear what exactly those requirements might be. The principal one appears to be a 'suitable' mud substrate in the uppermost intertidal zone. But what is suitable? Presumably it needs a substrate having the 'right' consistency for burrowing (muddy, but a certain fraction of sand is tolerated), the presence of sufficient animal prey of some sort (crabs are evidently not essential, at least not year-round), and the absence of 'excessive' levels of negative influences such as pollution, predation and habitat disturbance.

Mangrove trees or shrubs are not a requirement for P. waltoni. Mangroves are absent altogether in the northern Arabian Gulf (Kuwait and Iraq) and were absent at the historical UAE site of Khor Madfaq. The P. waltoni populations in the Strait of Hormuz are also not closely associated with mangroves (Polgar 2010).

One UAE popular account gives the impression that freshwater input may be a requirement, noting the (former) occurrence of mudskippers at Khor Hulaylah in Ra's al-Khaimah, where freshwater springs fed by aquifers flowing from the mountains of the Musandam peninsula have created the only brackish water salt marsh environment in the UAE (UAE Interact). In fact, both historical mudskipper sites in the UAE are associated with freshwater input, since the other, Khor Madfaq, constitutes the UAE's principal modern estuary, draining a very large inland area. Similarly, in the northern Arabian Gulf, the mudskipper populations in Iraq, at Khor Al-Zubair (Mhaisen & Al-Maliki 1997), and at Kuwait Bay are associated with the influence of freshwater outflow from the Shatt al-Arab (Swift & Bower 2003) and its surrounding marshes (Zajonz et al. 2002), the modern terminus of the Tigris and Euphrates river systems. However, the Iranian populations in the Strait of Hormuz do not generally receive significant freshwater input (although the population sampled for leeches by Polgar et al. (2009) near Bandar Khamir "received freshwater input from the small town nearby") and the UAE study site is not known or suspected to receive freshwater input, at least not to any greater extent than other khors in Dubai and the Northern Emirates.

Nevertheless, the distribution of the known and historical populations prompts the hypothesis that offshore seawater salinity may play a determining role and that P. waltoni may be indisposed to areas of extreme hypersalinity. All of the known mudskipper sites (Kuwait and Iraq in the northern Arabian Gulf, the Northern Emirates and the Strait of Hormuz) are situated in areas where surface water mapping shows that the characteristic hypersalinity of the Arabian Gulf is mitigated, either by input from the Shatt al-Arab, which favours the

Kuwaiti coast (Swift & Bower 2003), or by input from the Indian Ocean via the Strait of Hormuz, where incoming ocean water of normal salinity favours the Iranian shore (Swift & Bower 2003, George & John 2005). In these areas, the salinity of surface water (0-3 metres) in the open Gulf does not normally exceed 39 psu (Swift & Bower 2003, Fig. 9). In comparison, surface water salinity in the north-eastern Gulf exceeds 39 psu for about half the year, sometimes reaching 40-41 psu, and surface water in the south-western Gulf ranges from 40-41.5 psu for most of the year, and is never lower than 38.5 psu (Swift & Bower 2003, Fig. 9). Steeper nearshore physiographical gradients along the Iranian coast immediately north-west of the Strait of Hormuz (Swift & Bower 2003) probably inhibit the development of mudflats there, where salinity might remain suitable for much of the year.

If, in fact, offshore surface water hypersalinity is a determinant of mudskipper distribution within the Arabian Gulf, its influence must be exerted at the pelagic larval stage, since adult P. waltoni are limited to the upper intertidal zone, where they are necessarily tolerant of increased salinity created by the interaction of local climatic conditions and the tidal cycle.

Consideration of the origin of the UAE population is rendered more complicated by the history of the khor in question. As early as the mid-1990s the khor area as a whole had already experienced dredging and filling in anticipation of actual and prospective development. In 1998 it was noted that many specimens of the intertidal mud creeper Cerithideopsilla cingulata (formerly Cerithidea cingulata) showed unusual and somewhat discontinuous growth (abnormally narrow shells with a disproportionately large final whorl) indicative of stressful conditions (Robert Moolenbeek, pers. comm.), stressful intervals and/or possibly pollutants. During the 2000s, the remaining natural areas of the khor continued to suffer disruption due to infrastructure projects, including additional dredging in the lower channel and construction of an electric power and desalination plant beside one embayment of the intertidal zone. Most recently, at some point between August 2008 and May 2010, a corner of the seaward mangrove forest was obliterated for improved road access, a square kilometre of peripheral saltbushes and cyanobacterial mats was filled, and major dredging, wharfage and other infrastructure works were begun in the channel immediately below the mangrove forest. All of these development activities have been associated with smaller scale rearrangement of the original landscape in the vicinity of the mudskipper study site, including improved off-road access, construction of several small earthen boat ramps, and finally extensive perimangal bulldozing (for reasons not immediately apparent).

The latest phase of development, occurring after mid-2008 and continuing to the present, has been paralleled by changes in the prevailing mollusc and other invertebrate fauna of the khor. It is possible that those faunal changes and the presence of mudskippers are related, and that both are related to changes in the physical, chemical and/or hydrological parameters of the site. One possible hydrological change might be an increased flow rate of sea water into and out of the upper channel, possibly facilitating access by certain planktonic larvae (molluscs, mudskippers and others) that were previously excluded. The same kinds of changes could possibly have acted indirectly as well, by selectively excluding species that might have preyed on mudskipper larvae or young.

Changes in patterns of sedimentation and/or substrate change may also have resulted and could be determinative. From general observation, the authors are inclined to think that sedimentation at the P. waltoni site is influenced by adjacent earthworks, which seem to have created a 'pocket' where somewhat finer or slower settling sediment collects, and from which drainage is slightly impeded. Those works may also have involved removal of some of the original sediment cover and/or disruption of its original thin but cohesive veneer of algae/cyanobacteria. Did these changes create (or re-create) a differentiated local habitat conducive to colonisation by incoming mudskipper larvae, that had not existed in previous years? If so, then mudskipper colonisation can be dated from 2005 or later, when the works in question were completed. It is possible, however, that some of the observed changes to the local substrate were engineered by the mudskippers themselves.

On balance, the authors are inclined to favour the hypothesis of recent re-colonisation by P. waltoni due to serendipitous local contingencies, rather than to posit its persistence, unnoticed, for several decades. That hypothesis presumes that re-colonisation was facilitated, inadvertently, by alteration of physical parameters in such a way as to promote larval recruitment and/or the favourable modification of habitat on a local scale. A larger former population ought to have been noticed (even now, the site is within easy view of off-road vehicles, fishermen and picnickers), whereas the current population seems much too small to have persisted for so long. This view entails the paradox that broader environmental destruction in the surrounding area may actually be responsible for the return of the small population of mudskippers that now exists. There is, however, no proof of either alternative at the moment.

Whatever the truth, the re-discovery of P. waltoni is a reminder that there is no substitute for careful field observation. And that injunction is a fitting segue to our final topics ... Were two species of mudskippers present in the UAE in the early 1970s?

John Stewart-Smith, a former UAE resident, aircraft pilot and amateur naturalist, was a co-founder and first Chairman of the Emirates Natural History Group (ENHG) when it was formed in Abu Dhabi in 1977, and he has been honoured as a life member of the ENHG. He read the initial press report of the re-discovery of mudskippers in the UAE and wrote to give a brief account of his own acquaintance with the mudskippers present in the early 1970s. Some of his observations have already been quoted above. He was kind enough to correspond further in response to questions about various details. His description of the site and of the behaviour he witnessed added a provocative detail not seen in our UAE study population: territorial walls. He wrote:

"The mudskippers were in an area that was also populated by small crabs that waved one large claw about when approached. We noticed that each mudskipper seemed to build a 'reserved' area outside its burrow by pushing the wet sand into small walls that enclosed an irregular piece of sand that it treated as its property. The 'owners' reacted fiercely to any mudskipper approaching their boundary, sometimes jumping into the air and grunting."

P. waltoni is not associated in the literature with any reports of territorial walls, but at Sulaibikhat Bay in Kuwait a second mudskipper, Boleophthalmus dussumieri Valenciennes, 1837 (Fig. 29) (discussed in much of the earlier literature from Kuwait as B. boddarti), is present at high densities and is notorious for building mud-walled polygonal territories (Fig. 30) and for its exuberant territorial and mating displays (Clayton 1987, Clayton & Wells 1987, Clayton & Vaughan 1986, Clayton & Vaughan 1988, Clayton & Wright 1989). In Kuwait B. dussumieri is present sympatrically with P. waltoni although it occupies a tidal range that is slightly lower overall. B. dussumieri was present within and adjacent to the P. waltoni study site of Clayton & Snowden (2000).

B. dussumieri is similar in appearance to P. waltoni, and although it differs in colouration and can grow somewhat larger (to ca. 25 cm), it would be easy for a novice or casual observer to overlook those distinctions (especially as colour may vary with breeding status) or to fail to ascribe taxonomic significance to them in the context of what appeared to be a single population. We presented the above information to Stewart-Smith, referring him to additional resources, and asked whether he thought it was possible that B. dussumieri was also present on the UAE mudflats that he visited in the early 1970s. His reply was judicious but clearly affirmative:

"After careful perusal of http://www.mudskipper.it/ [Polgar (2010)] I now believe that both Boleophthalmus dussumieri and Periophthalmus waltoni were present in the same general area. Some of the mudskippers definitely had built retaining walls that held a layer of water outside their burrows, indicating B. dussumeri, although I do not remember any 'grazing' feeding action by them. Some of them engaged in an aggressive 'defensive' reaction to any other fish approaching their boundary walls. The size difference between these species doesn't help me to a firm decision, but my memory leads me to think that some were the smaller P. waltoni. The size and location of their erectile dorsal fin supports this thought. I did notice that there were two different patterns of dorsal fin but assumed these were sexual variations, i.e., larger males and smaller females. It is so easy to get it wrong when one assumes something without any evidence!"

The description of the retaining walls, layers of water and demonstrative territorial defence are all highly consistent with previous descriptions of B. dussumieri at Sulaibikhat Bay in Kuwait (Clayton 1987, Clayton & Wells 1987, Clayton & Vaughan 1986, Clayton & Vaughan 1988, Clayton & Wright 1989), although Stewart-Smith did not recall the characteristic side-to-side head movement of B. dussumieri when grazing on the surface film of diatoms, cyanobacteria and filamentous algae within its territory (Polgar 2010). In the authors' opinion, however, the single most persuasive detail is his recognition of "two different patterns of dorsal fin" (which he attributed at the time to sexual dimorphism), whereas both P. waltoni and B. dussumieri are monomorphic.

Taken as a whole, Stewart-Smith's information strongly suggests that a second and distinctive species of mudskipper, Boleophthalmus dussumieri, was present in the UAE into the modern era, but has been 'lost' before it was ever 'found'.

An East Coast record, too

The original newspaper report of our study population (Todorova 2012) elicited, about six months later, a record from the East Coast of the UAE, so far the only one known. Kevin Budd of the Breeding Centre for Endangered Arabian Wildlife, in Sharjah, related in conversation that in 1997, when he was still newly arrived in the country, he observed several mudskippers at Khor Kalba, the largest mangrove forest on the Gulf of Oman, in the narrow fringe of mudflats on the eastern edge of the mangroves, just inland of the main beachfront ridge. He was confident in distinguishing the animals he saw, presumably P. waltoni, from the smaller and non-amphibious, burrowing intertidal goby species (taxon/taxa unknown) that can also be found in and around UAE mangroves. However, being new to the UAE and being primarily engaged in work with terrestrial mammals, he did not appreciate at the time that this was an exceptional record.

And it does genuinely appear to have been exceptional. Prior to 1997, Khor Kalba had been the subject of at least two conservation surveys and reports (Hornby 1996, Stuart & Stuart 1996) and from mid-1998 through 2006 it was extensively explored on various occasions by Feulner and others in connection with research on Terebralia palustris (Feulner 2000) and other intertidal molluscs (Feulner & Hornby 2006), as well as intertidal crabs. None of those investigations recorded mudskippers, nor were mudskippers suspected.

Budd's East Coast visit in 1997 was in company with Chris and Tilde Stuart, South Africa-based consultants who at the time were engaged in follow-up field work at Khor Kalba. They have recalled and promptly confirmed seeing mudskippers there (Chris Stuart, pers. comm.).

Our own study suggests that, during the recruitment season, juvenile fish can potentially be found in habitats where permanent colonies did not previously exist, and where it is questionable whether they will become established (see "Juveniles and juvenile recruitment", above). That is potentially significant because Budd's Khor Kalba visit was, by his estimate, during the period August to October, which we now know includes or closely follows the recruitment period on the UAE's Arabian Gulf coast.

The Stuarts believe they may have filmed the mudskippers seen at Khor Kalba and they have volunteered, once they return from extended international travel, to make those films (if any) available. Reasonable images should permit a determination whether the fish shown are adults or juveniles, and therefore give a better indication whether a potentially breeding population has previously existed at Khor Kalba. In the meantime, staff of the Sharjah Environment and Protected Areas Authority, which has authority over Khor Kalba, are also keeping a lookout for mudskippers there.

Epilogue

This paper is intended to make a record of the presence and behaviour of Periopthalmus waltoni in the UAE and to try to put our observations in the larger context of knowledge about this species. The known UAE population is too small and vulnerable to support extensive follow-up study. There can be no better justification for our account than to quote, once again, John Stewart-Smith, writing in 2012 of his experience four decades earlier:

"I wish I had taken more careful note of these fish, never thinking that they would become rare."

Acknowledgements

It is a pleasure for the authors to acknowledge the ready cooperation of John Stewart-Smith, Colin Richardson, Kevin Budd and Chris and Tilde Stuart in providing information about prior records of mudskippers in the UAE. Dr. Richard Hornby commented on an early draft of this paper and provided helpful information about a number of UAE intertidal invertebrate taxa. Particular thanks are extended to the concerned local authorities who facilitated our research within a protected area; regrettably, they cannot presently be identified more specifically without risk to the study population.

References

Al-Behbehani, B.E. and Ebrahim, H.M.A. 2010. Environmental Studies on the Mudskippers in the Intertidal Zone of Kuwait Bay. Nature and Science 8(5): 79-89 [Online]. Available at: <http://www.sciencepub.net/nature>. [Authors' note: This paper, published online, is valuable principally for its extensive citation and reiteration of the content of earlier literature relating to P. waltoni and B. dussumieri.]

Askari, S.A.A., Velayatzadeh, M. and Mohammadi, M. 2010. Mercury Concentration in Mudskipper (Periophthalmus waltoni) and Flat Fish (Cynoglossus arel) in Bandar-e-Emam and Bandar Abbas. Journal of Fisheries (Summer 2010) 4(2(14)): 51-56.

ARKive. Jewels of the UAE: Fish: Available at: <http://www.arkive.org/uae/en/fish/latin/>. [Accessed 28/12/12]

Baeck, G.W., Takita, T., and Yoon, Y.H. 2008. Lifestyle of Korean mudskipper Periophthalmus magnuspinnatus with reference to a congeneric species Periophthalmus modestus. Ichthyoliogical Research 55: 43-52. [Cited in Ishimatsu & Gonzalez 2011.]

Beech, M.J., Al Abdessalaam, T.Z., and Hoolihan, J.P. 2005. Marine Fish. In P. Hellyer and S. Aspinall (eds.), The Emirates--A Natural History. Trident Press. pp. 261-282 and species checklist pp. 361-364. Available at: <http://viewer.zmags.co.uk/showmag.php?mid=srt hf&preview=1& x=1#/page0/>

Borges, R., Faria, C., Gil, F. and Goncalves, E.J. 2011. Early Development of Gobies. In R.A. Patzner, J.L. Van Tassell, M. Kova i and B. Kapoor (eds.), The Biology of Gobies. Science Publisher: Enfield. pp. 403-462.

Brillet, C. 1980. Comportement sexuel du poisson amphibie Periophthalmus sobrinus Eggert: ses rapports avec le comportement agonistique. Revue d'Ecologie, Terre et la Vie 34 (3): 427-468. [Cited in Polgar (2010)]

Brillet, C. 1984. Etude comparative de la parade nuptiale chez deux especes de Poissons amphibies sympatriques (Pisces,

Periophthalmidae). C. R. Hebd. Seances Acad. Sc. Serie D, Paris, 198 (12): 347-350. [Cited in Polgar (2010)]

Bu-Olayan, A.H. and Thomas, B.V. 2010. Trace Metals Toxicity and Bioaccumulation in Mudskipper Periophthalmus waltoni Koumans 1941 (Gobiidae: Perciformes). Turk. Jour. Fisheries and Aquatic Sciences 8: 215-218.

Clayton D.A. 1987. Why mudskippers build walls. Behaviour 102: 185-195.

Clayton D.A., 1993. Mudskippers. Oceanogr. Mar. Bull. Annu. Rev. A.D. Ansell, R.N. Gibson and M. Barnes (eds.), UCL Press, 31: 507-577.

Clayton, D.A. and Snowden, R. 2000. Surface activity in the mudskipper Periophthalmus waltoni Koumans 1941 in relation to prey activity and environmental factors. Tropical Zoology 13: 239-249

Clayton, D.A. and Vaughan, T.C. 1986. Territorial acquisition in the mudskipper Boleophthalmus boddarti (Pisces: Gobiidae) on the mudflats of Kuwait. Journ. Zool. (A) 209: 501-519.

Clayton, D.A. and Vaughan, T.C. 1988. Ethogram of Boleophthalmus boddarti (Pallas), a mudskipper found on the mudflats of Kuwait. Journal of the University of Kuwait (Sciences) 15(1): 115-138.

Clayton, D.A. and Wells, K. 1987. Discovering Kuwait's Wildlife. Fahad Al-Marzouk, Kuwait. 253 pp.

Clayton, D.A. and Wright, J.M. 1989. Mud-walled territories and feeding behaviour of Boleophthalmus boddarti (Pisces, Gobiidae), on the mud flats of Kuwait. Journal Ethol. 7(2): 91-95.

Colombini I, Berti R., Nocita A. and Chelazzi I., 1996. Foraging Strategy of the Mudskipper Periophthalmus sobrinus Eggert in a Kenyan mangrove. J. Experimental Marine Biology and Ecology, 197 (2): 219-235. [Cited in Al-Behbehani & Ibrahim (2010).]

Egan, D. 2007. Snakes of Arabia. Motivate Publishing, Dubai. 208 pp.

Encyclopedia of Life. Periophthalmus waltoni Walton's Mudskipper. Available at: <http://eol.org/pages/1157620>. [Accessed 20/02/12]

Feulner, G.R. 2000. The Large Mangrove Mud Creeper Terebralia palustris (Linnaeus 1767) in Non-Mangrove Environments in Southeastern Arabia. Tribulus 10.2: 15-26. Available at: <http://www.enhg.org/Portals/1/trib/V10N2/Tribulus V10N2.pdf>.

Feulner, G.R. and Cunningham, P.L. 2000. The freshwater goby Awaous aeneofuscus in the Wadi Hatta watershed (U.A.E./Oman). Tribulus 10.1: 12-15. Available at: <http://www.enhg.org/Portals/1/trib/V10N1/Tribulus V10N1.pdf>.

Feulner, G.R. and Hornby, R.J. 2006. Intertidal Molluscs in UAE Lagoons. Tribulus 16.2: 17-23. Available at: <http://www.enhg.org/Portals/1/trib/V16N2/Tribulus V16N2.pdf>.

Froese, R., and Pauly, D. (eds.) 2012a. FishBase: All fishes of United Arab Emirates: n = 189 (Incomplete). World Wide Web electronic publication. Available at: <http://www.fishbase.org/Country/CountryChecklis t.php>. [Accessed 28/07/12]

Froese, R., and Pauly, D. (eds.) 2012b. FishBase: Species in Apocryptodon--Fish Identification. World Wide Web electronic publication. Available at: <http://fishbase.sinica.edu.tw/identification/Specie sList.php?genus=Apocryptodon>. [Accessed 09/08/12]

Froese, R., and Pauly, D. (eds.) 2012c. FishBase: References used for Apocryptodon madurensis. World Wide Web electronic publication. Available at: <http://fishbase.sinica.edu.tw/References/Summar yRefList.php?id=7328&genusname=Apocryptodo n&speciesname=madurensis&lang=russian>. [Accessed 09/08/12]

Fishwise. Periophthalmus waltoni--Koumans 1941. Available at: <http://www.fishwise.co.za/>. [Accessed 20/02/12]

George, D. and John, D. 2005. The Marine Environment. In P. Hellyer and S. Aspinall (eds.) The Emirates--A Natural History. Trident Press. pp. 111-121. Available at: <http://viewer.zmags.co.uk/showmag.php?mid=srt hf&preview=1& x=1#/page0/>

Graham J.B. (ed.) 1997. Air-breathing Fishes. Evolution, Diversity and Adaptation. Academic Press, San Diego California. 299 pp.

Hagen Aqualab. 2005. Mudskipper Standard Operating Procedures. April 1, 2003--Revised December 2005. Available at: <http://www.uoguelph.ca/~aqualab/forms/mudskip perSOP2004.pdf>. [Accessed 20/02/12]

Hogarth, P.J. 1999. The Biology of Mangroves. Oxford University Press. 228 pp.

Hornby, R.J. 1996. Khor Kalba Proposed National Park--a Provisional Management Plan. Unpublished report to the Arabian Leopard Trust. November 1996.

Ip, Y.K., Chew, S.F. and Ping, C.T. 1991. Evaporation and the turning behaviour of the mudskipper Boleophthalmus boddarti. Zoological Science, Tokyo 8(3): 621-624. [Cited in Clayton & Snowden (2000).]

Ishimatsu A., Hishida Y., Takita T., Kanda T., Oikawa S., Takeda T. and Khoo K.H. 1998. Mudskippers Store Air in Their Burrows. Nature, 391 : 237-238. [Cited in Polgar (2010)]

Ishimatsu, A. and Gonzalez, T.T. 2011. Mudskippers: Front-Runners in the Modern Invasion of Land. In R.A. Patzner, J.L. Van Tassell, M. Kova i and B. Kapoor (eds.), The Biology of Gobies. Science Publisher: Enfield. pp. 609-638. Ishimatsu A., Yoshida Y., Itoki N., Takeda T., Lee H.J. and Graham J.B. 2007. Mudskippers brood their eggs in air but submerge them for hatching. J. Experimental Biology 210: 3946-3954.

Kim, J.K., Baek, H.-J., Kim, J.-W., Chang, D.-S., and Kim, J.I. 2011. Sexual Maturity and Early Life History of the Mudskipper Scartelaos gigas (Pisces, Gobiidae): Implications for Conservation. Fisheries and Aquatic Sciences 14(4): 403-410. Available at: <http://dx.doi.org/10.5657/FAS.2011.0403>.

Kobayashi T., Dotsu Y. and Miura N. 1972. Egg development and rearing experiments of the larvae of the mud skipper, Periophthalmus cantonensis (in Japanese with English summary). Bulletin of the Faculty of Fisheries, Nagasaki University 33: 49-62. [Cited in Ishimatsu et al. (2007) and Polgar (2010)]

Kooseg, N., Rahmani, A., Kamrani, E., Taherizadeh, M. and Alinia, M. 2011. Investigation and Measurement of heavy Metals, Lead and Cadmium, in Muscle of Periophthalmus waltoni in Persian Gulf (Khamir Port, Dargahan Island and Golshahr). Jour. Life Science and Biomedicine 1(1): 13-17. Available at: <http://jlsb.science-line.com/>.

Lee, H.J. and Graham, J.B. 2002. Their Game is Mud. Natural History Magazine 9/02: 42-47. Available at: <http://naturalhistorymag.com/htmlsite/0902/0902 feature.html>.

Lee, H.J., Martinez, C.A., Hertzberg, K.J., Hamilton, A.L., and Graham, J.B. 2005. Burrow air phase maintenance and respiration by the mudskipper Scartelaos histophorus (Gobiidae: Oxudercinae). Jour. Experimental Biol. 208: 169-177. Available at: <http://jeb.biologists.org/content/208/1/169.full.pdf>.

Llewellyn-Smith, R. 1999. Khor Kalba Nature Reserve: A Review. Unpublished report by the Arabian Leopard Trust to the Sharjah Department of Environment and Protected Areas. November 1999. 14 pp.

Llewellyn-Smith, R. 2011. Fresh water input to Khor Hulaylah--an unusual feature of Ra's al-Khaimah's wetlands. Tribulus 19: 154-157.

Llewellyn-Smith, R., Khuluti, I., Tayseer, Y. and Khalifa, M. 2007. Proposal for a Hulayla Nature Reserve, Ra's al Khaimah. Unpublished report by the Wildlife & Protected Areas Division of the Environmental Protection & Development Agency, Ra's al Khaimah. 43 pp.

Mhaisen, F.T., and Al-Maliki, N.S. 1997. Parasites, diseases and food of the dark-blotched mudskipper Periophthalmus waltoni (Perciformes:Gobiidae) in the Khor Al-Zubair estuary (Iraq). Oceanographic Literature Review (July 1997) 44(7): 738

Murdy, E.O. 1989. A taxonomic revision and cladistic analysis of the oxudercine gobies (Gobiidae: Oxudercinae). Records of the Australian Museum 11: 1-93.

Murdy, E.O. 2011. Systematics of Oxudercinae. In R.A. Patzner, J.L. Van Tassell, M. Kova i and B. Kapoor (eds.), The Biology of Gobies. Science Publisher: Enfield. pp. 99-106.

Polgar, G. (ed.) 2010. The Mudskipper. World Wide Web electronic publication: <http://www.mudskipper.it>. [Accessed June-July 2012].

Polgar, G., Burreson, E.M., Stefani, F., and Kamrani, E. 2009. Leeches on Mudskippers: Host-Parasite Interaction at the Water's Edge. J. Parasitol. 95(4):1021-1025. Available at: <http://www.journalofparasitology.org/doi/abs/10.1 645/GE-1718.1>.

Randall, J.E. 1995. Coastal Fishes of Oman. Crawford House, Bathurst, Australia. 439 pp.

Richardson, C. 1994. United Arab Emirates: Dhayah, Rams, Ghalilah and Hulayla Island (15). In A Directory of Wetlands of he Middle East. Available at: <http://www.wetlands.org/RSIS/WKBASE/MiddleE astDir/UAE.htm>.

Sarafraz, J., Abdoli, A., Hassanzadeh, B., Kamrani, E. and Akbarian M.A. 2011. Determination of the age and growth of the mudskipper Periophthalmus waltoni Koumans, 1955 (Actinopterygii: Perciformes) on the mudflats of Qeshm Island and Bandar Abbas, Iran. Progress in Biological Sciences. Abstract available at: <http://journals.ut.ac.ir/page/download- 9c5Sj2qHxAk.artdl>. [Accessed 20/02/12].

Schultze H.-P. 1999. The Fossil Record of the Intertidal Zone. In M.H. Horn, K.L.M. Martin & M.A. Chotkowski (eds.), Intertidal Fishes: Life in Two Worlds. Academic Press, San Diego California, 399 pp: 373-392.

Sponder, D.L., and Lauder, G.V. 1981. Terrestrial feeding in the Mudskipper Periophthalmus (Pisces: Teleostei): A radiographic analysis. J. Zool. Lond. 193: 517-530.

Stebbins, R.C. and Kalk, M. 1961. Observations on the Natural History of the Mudskipper, Periophthalmus sobrinus. Copeia 1 : 18-27. [Cited in Clayton & Snowden (2000)]

Stuart, C. and Stuart, T. 1996. Proposal for the creation of Khor Kalba Wetland National Park. Unpublished report to the Arabian Leopard Trust. 13 pp.

Swift, S.A. and Bower, A.S. 2003. Formation and Circulation of Dense Water in the Persian/Arabian Gulf. J. Geophys. Res. 108, no. C1, 3004, doi:10.1029/2002JC001360.

Takeda, T., Hayashi, M, Toba, A., Soyano, K., and Ishimatsu, A. 2012. Ecology of the Australian mudskipper Periophthalmus minutus, an amphibious fish inhabiting a mudflat in the highest intertidal zone. Australian Jour. Zool. 59(5): 312-320. Available at: <http://dx.doi.org/10.1071/ZO11059>.

Taylor, J.R., Cook, M.M., Kirkpatrick, A.L., Galleher, S.N., Eme, J., and Bennett, W.A. 2005. Thermal Tactics of Air-Breathing and Non-Air-Breathing Gobiids Inhabiting Mangrove Tidepools on Pulau Hoga, Indonesia. Copeia 2005(4): 885-892.

Todorova, V. 2012. Return of the fish that skipped. The National. 16 March 2012, p. 5. Available at: <www.thenational.ae>.

Tytler, P., and Vaughan, T. 1983. Thermal Ecology of the Mudskippers Periophthalmus koelreuteri (Pallas) and Boleophthalmus boddarti (Pallas), of Kuwait Bay. Journal of Fish Biology 23(3): 327-337. [Cited in Clayton & Snowden (2000)]

UAE Interact. Natural UAE: Dhayah marshes. Available at: <http://www.uaeinteract.com/nature/natu ems/en4 6.asp>. [Accessed 26/02/12 and 28/07/12]

Yang, K.Y.J. 1996. The feeding ecology of the mudskipper Boleophthalmus pectinirostris (Pisces: Periophthalmidae) at the Mai Po marshes nature reserve, Hong Kong. Thesis for the degree of Master of Philosophy. Faculty of Science, Hong Kong University. 193 pp. [Cited in Clayton & Snowden (2000).]

Zajonz, U., Beech, M. and Gill, A.C. 2002. Fishes of Sabkha-Related Habitats. In H.-J. Barth and B. Boer (eds.), Sabkha Ecosystems--Vol I: The Arabian Peninsula and Adjacent Countries. Kluwer Academic Publishers. pp.283-298.

Gary R. Feulner

Dubai, U.A.E.

email: grfeulner@gmail.com

Binish Roobas

Trivandrum, Kerala, India

email: johanruphus@hotmail.com
COPYRIGHT 2013 Emirates Natural History Group
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Feulner, Gary R.; Roobas, Binish
Publication:Tribulus
Date:Jan 1, 2013
Words:20575
Previous Article:Potential alien species threat for the Arabian Collared Kingfisher at its only UAE breeding locality.
Next Article:Report of a short-term reconnaissance study of sand dune temperature profiles.
Topics:

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters