Mudskippers (Literature Review)

david clayton

Mudskippers (Gobiidae: Oxudercinae) which are known to burrow and reside in the intertidal mudflats or mangrove swamps of the Indo-West Pacific region. Two species of Mudskippers were recorded from Hathab coast namely Boleophthalmus dussumeiri, and Scartelaos histophorus. Within the study area, the highest population density of B. dussumeiri were recorded which is followed by S. histophorus. Populations of mudskippers were highest during winter due to more thickness of mud layer, which reduced during the summer and monsoon with reduction of mud layer. Moreover, here mudskipper distribution varies from upper intertidal zone to lower intertidal zone. The parameters such as mud layer depth, sediment composition, pH, Temperature and salinity to have the significant relationship with the mudskipper distribution.

This paper describes an effort to study the diversity and behaviour of mudskippers occurring in the RAMSAR site Tanjung Piai, Pontian, Johor. Ten species from four genera of mudskippers were reported here. They are all members of the family Gobiidae and subfamily Oxudercinae. The most species rich genus was Periophthalmus (five species) followed by Baleophthalmus and Oxuderces (two species each). The fourth genus Periophthalmodon was represented by one species. Indication is that Malaysia may harbour 19 species representing 12 genera. The giant mudskipper, Periophthalmodon schlosseri was found at all level and stages of the intertidal zones. Being the largest in size of approximately 20cm, Periophthalmodon schloserii also shows interesting courting and mating behaviour and tree climbing. Baleophthalmus boddarti and Scartelaos histophorus are two other common species. This paper illustrates the different species at Tanjung Piai with photographs and endeavours to make notes of their peculiar behaviour. The paper discusses the diversity of the species comparing its occurrence, distribution and abundance. Special notes are made on the importance of this fish group; including discussion on its potential to be developed, promoted and marketed as a new tourism product especially of the mangrove ecosystem.

Mudskippers (Gobiidae: Oxudercinae) are air-breathing gobies, which are widely distributed throughout the West African coast and the Indo-Pacific region. They are closely linked to mangrove and adjacent soft bottom peri-tidal ecosystems. Some species are amongst the best adapted fishes to an amphibious lifestyle. All mudskippers are benthic burrowers in anoxic sediments, and since tidal mudflats are efficient sediment traps, and sinks for nutrients and other chemical compounds, they are constantly in contact with several types of pollutants produced by industrial, agricultural and domestic activities. Due to their natural abundance, considerable resistance to highly polluted conditions, and their benthic habits, mudskippers are frequently used in aquatic ecotoxicological studies. For the same reasons, mudskippers also frequently occur in urbanised or semi-natural coastal areas. Since several species are widely consumed throughout their whole geographical range, these same characteristics also facilitate their aquaculture in several countries, such as Bangladesh, Thailand, Philippines, China, Taiwan and Japan. Even when not directly used, mudskippers are often abundant and are important prey items for many intertidal transient species (marine visitors), and several species of shorebirds. Therefore, there is potential for bioaccumulation of toxicants wherever mudskippers and pollution co-occur. This chapter reviews the ecotoxicology of mudskippers, and their potential for use as biomonitors to better manage coastal swamp ecosystems. The diverse sympatric assemblages of mudskipper species allow for spatially differentiated ecotoxicological investigations along the whole intertidal zone, since adults are often territorial and/or sedentary, and show species-specific patterns of habitat differentiation. A case study is also proposed where this approach could be adopted to address potential health-risk issues in a local population who are regular consumers of mudskippers

Oceanogr. Mar. Biol. Annu. Rev., 1993, 31, 507-577 A. D. Ansell, R. N. Gibson and Margaret Barnes, Editors UCL Press MUDSKIPPERS DAVID A. CLAYTON Biology Department, Sultan Qaboos University, PO Box 36 Al Khod Postal Code 123, Muscat, Sultanate of Oman ABSTRACT The oxudercine gobies commonly known as mudskippers provide a rich source for comparative studies in adaptation to the littoral habitat. In spite of the paucity of information on reproduction, culturing techniques have improved in recent years and the developmental stages of the Chinese and Japanese species are now known. Morphological studies of mature fish havecentred on skeletal characteristics as an adjunct to taxonomy and on vision, respiration and excretion. Mudskippers are euryhaline and ammoniotelic with sophisticated biochemical pathways for ammonia detoxification operating to different extents in the various genera. Free amino acids play a central role in bothexcretion and osmoregulation. The relative importance of the respiratory surfaces of these facultative air-breathing fishes varies in air and water and also between genera. The surfaces include the gills and modified buccal, pharyngeal and opercular epithelia as well as limited, well vascularised areas of the skin. The respiratory rate and energy consumption are reduced during hypoxia, but the full extent of metabolic changes in aestivating or hibernating fishes has yet to be fully explored. Mudskippers usuallyinhabit tidal mudflats and mangroves, but can be found on sandy and rocky shores. Of the common genera, Boleophthalmus and Scartelaos are considered to be more aquatic than Periophthalmus and Periophthalmodon, but complex patterns of zonation indicate that more detailed ecological data are required on this topic as well as on the biotic com ponents and parasites. From simple tidal migrations to burrow construction and complex mud-walled polygonal mosaics, mudskippers exhibit a wide range of territorial behaviour which is matched by their diverse courtship and agonistic displays. INTRODUCTION The literature on the biology of intertidal fishes is extensive and has been reviewed by Gibson (1969, 1982, 1986). Despite the fact that observations on African, Australian and South American species are for the most part sadly lacking (Gibson, 1982), the best known examples of resident intertidal fishes are probably the predominantly Indo-Pacific mudskippers. In general, mudskippers can be defined as those fishes usually found moving about on the mudflats of creeks, estuaries and coastal waters when they are exposed at low tide (Qureshi & Bano, 1971). From the frontispiece of Higson's (1889) "A Naturalist in North Celebes" to Polunin's (1972) "Who says fish can't climb trees?" and the similar pictorial essay on "A fish out of water" (Fukuda, 1985), these perversely photogenic gobies, with their rich behavioural repertoire have caught the popular imagination (Sowerby, 1923; van Dijk, 1959; Whitley, 1960; MacNae, 1968a,c; Kaden, 1978) and warranted mention in museum (Inger, 1952; Whitley, 1968—a reprint 508 DAVID A. CLAYTON of the 1960 article) and zoo (Jes, 1972) bulletins. In some areas mudskippers are of economic importance, because they are eaten in China (Schreitmuller, 1914), Taiwan (Liao et al, 1973), India (Das, 1934; Hora, 1935a; Siddiqi, 1974; Qureshi & Bano, 1971) and provide an alternative fishery during the monsoon in Bombay (Mutsaddi & Bal, 1973). They are extensively cultured in Taiwan (Chen, 1976) and to some Malaysians the raw flesh of these fishes has aphrodisiac properties (Johnstone, 1903). Among Europeans, however, the mudskippers have remainedmerely interesting, if unusual, aquariumfish (Reichelt, 1910; Schreitmuller, 1911; Rodewald, 1913; Simpson, 1955; Klausewitz, 1968; Kluge, 1971; Von Anthouard & Mignot, 1973; Kaden, 1978; Hunziker, 1985; Murdy, 1986; Schneider, 1990a,b). It is doubtful if there is any other group of fishes in which so much general interest is based on so little scientific knowledge. There are four main reasons for this discrepancy. First, much of the popular writing is based on only a few aspects of the fishes' biology. Parenthetically, some of this information was incorrect as in the case of the fishes' respiratory adaptations, the misunderstanding of which began with Higson (1889) and only ended with Sponder & Lauder (1981). Secondly, the scientific information is not easily available, because it is written in a number of languages and spread over a wide variety of often inaccessible sources. As a consequence, relevant liter ature is often overlooked. For example, Yadav et al (1990) omitted the pertinent study of respiratory organs by Singh & Munshi (1969) and similarly, Lele & Kulkarni (1938, 1939) are not mentioned in the osteological study of Lee (1990). Thirdly, the very facet of their behaviour that attracted the original interest, their extreme manoeuvrability in an environment difficult of access for man, meant that they are hard to catch. Individuals caught by blowpipe (Higson, 1889), weighted hooks or a 0 •22 air rifle (Clayton & Vaughan, 1986) are of little use for further study. Fish caught by traps or nets (Burhanuddin & Martosewojo, 1979) or lassos (Khoo, 1966) are obviously preferable, but these are very in efficient techniques that often create unwanted habitat disruption. Consequently, in accounts of the ecology of tropical shores on which mudskippers are found, good qualitative data for other macrofaunal taxa are presented but only the presence or absence of mudskippers is reported (Berry, 1972; Day, 1974; Sasekumar, 1974; Frith etai, 1976; Dorjes, 1978; Branch & Grindley, 1979; Nateewathana & Tantichodok, 1984). Even where the apparent redistribution of mudskippers is discussed (Lipschitz et al, 1975) no quantitative data are presented. Finally, the taxonomy of the group of fishes known as mudskippers has been in a state of confusion. Whilst this is partially a reflection of the problematic systematics of gobies in general (Miller, 1973) it is also exacerbated by problems of specific identification, as illustrated by Brillet's (1969a, 1975) study of the behaviour of Periophthalmus sobrinus and Al-Kadhomiy & Hughes' (1988) study of the comparative morphology of gill structure of Boleophthalmus boddarti. Wherever possible in this review correct specific designations based on Murdy (1989) will be provided. Previous reviews (Gibson, 1969, 1982, 1986) including mudskippers concen trated mainly on recent literature, but this synthesis attempts to include most of the literature irrespective of its age. This is partly because the literature is more limited, but also because some of the earlier work has been misinterpreted or ignored, despite continued relevance for present studies. MUDSKIPPERS 509 TAXONOMY Two brief examples of the taxonomic confusion will suffice to illustrate the problems. Periophthalmus vulgaris was distinguished by Eggert (1935) and has at least eight synonyms, the most common being P. koelreuteri (Pallas) and P. argentilineatus (Cuvier & Valenciennes). Eggert separated P. vulgaris from these on the form of the first dorsal fin which, although variable, did not have the convex, fan-like or irregular shape of these species, and also by differences in colour, pelvic fin and tooth form. Whitley (1931) separated P. vulgaris from P. koelreuteri, but wrongly used it for Euchoristopus kalolo (= Periophthalmus kalolo Lesson), which is unquestionably synonymous with P. koelreuteri (Milward, 1974). Whitley (1953, 1960, 1968) and MacNae (1968a,b) repeated the error. Furthermore, whilst Eggert (1935) recognised four geographically separated subspecies of P. vulgaris, the diagnosticcharacteristicsprovided were inadequate (Herre, 1941). The considerable intraspecific variation in fin spines and rays (Milward, 1974) clearly contributed to this confusion as it has in the identity of other mudskippers. Boleophthalmus dussumieri (Valenciennes) and B. dentatus (Valenciennes) were originally described (Cuvier & Valenciennes, 1837) as differing in denti tion and dorsal fin structure, the latter having stronger canines and united first and second dorsal fins. These two species were later distinguished from a third species, B. chamiri Holly (Holly, 1929) on the basis of the possession of a greater number of teeth in the upper jaw. Tooth number is an age specific variable, however, and the size and fusion of the first and second dorsal fins a sex specific one (Berg, 1949). The correct designation, based on page priority in Cuvier & Valenciennes (1837) is B. dussumieri (Berg, 1949). Nevertheless, and despite Khalaf s (1961) caution, both names persist in more current litera ture (Al Nasiri & Hoda, 1975; Hoda, 1980; 1985). Furthermore, the relation ship between this species and B. boddarti (commonly misspelt as boddaerti) remained problematical because the original descriptions were inadequate, and, as in other cases, the type specimens were no longer extant. Before nomenclatural confusion of the sort illustrated above could be resolved, intraspecific variation in diagnostic characters needed to be assessed using extended size range collections from all locations inhabited by each species. The specific determinations for most, if not all mudskippers, was based entirely on external features many of which have been shown to be of limited value and the more recent osteological criteria used in gobioid classification (Miller, 1973) needed to be incorporated in any such attempt. Interrelationships between mudskipper taxa have recently been reassessed using such criteria (Birdsong et al, 1988; Murdy, 1989) and have resulted in a complete revision of their taxonomy. The mudskipper genera are now all placed in the Family Oxudercinae. The Tribe Oxudercini includes the genera Parapocryptes (2 species), Apocryptodon (2) and Oxuderces (1) and the Tribe Periophthalmini the genera Apocryptes (1), Pseudapocryptes (2), Zappa (1), Scartelaos (4), Boleophthalmus (5), Periophthalmodon (3) and Periophthalmus (12) (Murdy, 1989). This revision of the specific synonyms has reduced the nominal 103 species to a respectable 34, and will benefit future researchers merely wishing to identify their experimental subjects. The revision also further emphasises the necessity for voucher specimens. The usual data on collection should be supplied and the latitude and longitude should be included in the 510 DAVID A. CLAYTON locality information. In any publication authors should provide the museum's catalogue number for the specimen(s). Current programmes of research will be able to rectify their subjects' identity, but continued care will be needed in comparisons made with other work. While some reassignments can be made easily others will be more difficult. For example Periophthalmus barbarus is the only periophthalmid identified in west Africa and Periophthalmus modestus is theonly one inJapan. Conversely, Periophthalmus vulgaris and Periophthalmus koelreuteri are no longer valid species names, and erroneously havebeen applied to a number of extant species. MORPHOLOGY CHROMOSOMES As an adjunct to the more usual morphological classificatory system for the analysis of gobioid fishes, some karyotypic chromosomal analyses of mud skippers have been undertaken, some of which are tabulated in Nishikawa etal (1974). For Apocryptodon madurensis and Pseudapocryptes borneensis [ = Periophthalmus lanceolatusl] Verma (1968) records 48 as the diploid number of chromosomes. There is agreement that the diploid number is 46 in P. cantonensis [ = P. modestus] (Nogusa, 1957, 1960; Arai & Kobayasi, 1973; Nishikawa et al, 1974). In Boleophthalmuspectinirostris (Nogusa, 1957, 1960; Nishikawa et al, 1974), B. dussumieri (Krishnaja & Rege, 1980; Verma, 1968), and B. boddarti (Subrahmanyam, 1969; Verma, 1968; Manna & Prasad, 1974) the number is also 46, but the detailed descriptions of the Indian material differ considerably. There is further confusion in that Manna& Prasad's (1974) identification of B. glaucus mustbe incorrect becausethis species [ = Scartelaos cantoris] (Murdy, 1989) is only known from the Andaman Islands. The most likely appellation is Boleophthalmus boddarti. Subrahmanyam (1969) reported that the chromosomes of B. boddarti were all metacentric with a large hetero- morphic pair which could be the sex chromosomes. Verma (1968) failed to find any similar elements and Manna & Prasad (1974) recorded only six metacentric pairs, the remainder being submetacentric, acrocentric or telocentric. Only acrocentric chromosomes were found in B. dussumieri (Krishnaja & Rege, 1980) and in B. pectinirostris and Periophthalmus modestus there were both acrocentric and telocentric chromosomes (Nishikawa et al, 1974). Nevertheless, the chromosome structure of all three species would, as suggested by Krishnaja & Rege (1980) for Boleophthalmus dussumieri, provide good material for the detection of chromosomal aberrations in mutagenic studies. NERVOUS AND SENSORY SYSTEMS Brain The gross structure of the mudskippers' nervous system has largely been ignored. The dorsal structures of the brain are illustrated for Periophthalmus chrysospilos (Harms, 1935), P. barbarus [ = Periophthalmodon schlosseri] (Lim, 1967), Periophthalmus koelreuteri, Pseudapocrytes lanceolatus (Datta & Das, 1980), Boleophthalmus boddarti (Lim, 1967; Datta & Das, 1980), Boleophthalmus sp. and Apocryptes sp. (Mookerjeeet al, 1950). The illustrations MUDSKIPPERS 511 of Datta & Das (1980) are diagrammatic and while those of Lim (1967) are anatomically correct, they reveal little more than that the brain of the carni vorous Periophthalmus is larger than that of the herbivorous Boleophthalmus. As part of their study on rhythmic changes in neuroendocrine function of B. dentatus, Patel & Desai (1976) show micrographical sections of the hypo thalamic hypophyseal region of the brain. Mechanoreceptors The mechanoreceptors of the lateral line system of mudskippers are limited to epidermal organs (neuromasts) as there are no lateral line canals (Afzelius, 1956). Working with small Periophthalmus barbarus [as P. koelreuteri], he detailed the location and nervous innervation of the neuromasts of the head. While most are embedded deeply in the epithelium, the two rear organs are situated on elevations posterior to the eyes (Afzelius, 1956). He thought they were innervated by the ramus ophthalmicus superficialis, but is is now accepted that lateral line organs are innervated by the acoustic (VIII) nerve (Parker & Haswell, 1962) and it is likely that the nerve Afzelius observed was a section of the anterior lateral line nerve that joined the trigeminal (V). The greatest concentration of neuromasts are along the lower jaw, and as part of a study of their function in these amphibious fishes, it would be interesting to compare this distribution with that in Boleophthalmus or Scartelaos (see p. 514). The struc ture of these organs as revealed by SEM would also prove instructive. Without being able to determine the location of the sound receptor, Diesselhorst (1938) showed that Periophthalmus koelreuteri [?] detected air-borne sounds in the range 258-651 Hz. The sounds used were a conditioned stimulus for food and elicited responses after 45 training presentations. However, because of the variable response, which in some cases was indicatedonly by eye movement. Diesselhorst did not perform any discrimination tests. As a corollary to this, the rapid escape response of many fishes is mediated by the Mauthner reflex in which vibrational stimulation of the acoustic nerve initiates a strong muscular response via the Mauthner cells (Diamond, 1971). The response is presumed to assist in the avoidance of aerial predators, particularly diving or plunging ones. While the role of visual stimuli is acknowledged (Diamond, 1971), it has not been investigated and the study of the mudskippers' startle response is likely to be doubly rewarding in the absence of vibrational cues. Eyes The dorsally protruding eyes of mudskippers are unique in a number of ways and while some of these were described by Graham (1971), it was without reference to the original literature, which in any case is rather confusing (Munk, 1970). The generic designations, Boleophthalmus and Periophthalmus, reflect the great mobility of the eyes of mudskippers, the musculature of which is described for P. koelreuteri from Ceylon [ = P. kalolo or argentilineatus] by Karsten (1923), Periophthalmodon schlosseri and Boleophthalmus or Scartelaos [as Boleophthalmus sp. Val.] by Oliva & Skorepa (1970a,b respectively). The eye muscles of mudskippers are similar to those of other fishes (Karsten, 1923), but because the length is greater than the diameter of the eyeball, they are longer than those of deep sea fishes. Mudskippers also have clear anterior and posterior points of muscle attachment (myodomes) which are more specialised in the 512 DAVID A. CLAYTON carnivorous Periophthalmodon than the herbivorous Boleophthalmus (Oliva & Skorepa, 1970a, b). The eyes move independently of each other and also can be retracted into the head where they are covered by a ventrally placed lid-like skin fold (Karsten, 1923; Munk, 1970). This dermal cup (Graham, 1971) serves as a water reservoir for moistening the eye surface and is a characteristic used to distinguish Scartelaos, Boleophthalmus, Periophthalmus and Periophthalmodon from the other taxa within the Oxudercinae (Murdy, 1989). The cornea is divided into distal and proximal lamellae which together correspond to the cornea of other vertebrates (Karsten, 1923). The outer lamella is presumed to provide additional protection from desiccation and mechanical damage (Graham, 1971). The internal anatomy of the eye has been discussed for Periophthalmodon schlosseri (Volz, 1905a, b,c; Baumeister, 1913), Periophthalmusargentilineatus (Karsten, 1923; Harms, 1914, 1929; Munk, 1970), P. kalolo [as P. koel reuteri] (Hess, 1912, 1913; Munk, 1970), P. chrysospilos (Yew & Wu, 1979), P. modestus [as P. cantonensis] (Yamamoto, 1931), Boleophthalmus boddarti (Volz, 1905a; Baumeister, 1913; Munk, 1970) and B. pectinirostris (Franz, 1910). Despite the absence of the falciform process and the retractor lentis muscle (campanila Halleri) (Volz, 1905a,c) and contrary to Baumeister (1913), the arrangement of lens muscles is essentially the same as in other fishes and enable accommodation to take place (Karsten, 1923). What form accommodation actually takes is debatable, for whereas Hess (1912, 1913) found that the un accommodated eye of Periophthalmus argentilineatus in air focuses light from infinity on (emmetropic) or slightly behind (hypermetropic) the retina, Yamamoto (1931) found the eye of P. modestus to be hypermetropic; slightly in air and strongly so in water. In either case, the accommodatory range is probably insufficient to restore emmetric vision in water and the Periophthalmus eye must be considered to be adapted for aerial vision. It is likely that this generalisation also applies to the other amphibious genera (Munk, 1970). The strongly curved cornea and, in comparison with other fish, the flattened lens of Periophthalmus and Boleophthalmus also contribute to aerial adaptation (Graham, 1971). These genera only differ in the shape of the pupillary aperture (horizontal slit and heartshaped respectively) and the curvature of the retinal surface relative to the anatomical axis of the eye (Munk, 1970). Periophthalmus has a 'ramp' retina which is tilted away from the lens such that the distance from the centre of the lens to the retina increases continuously in the vertical direction (Karsten, 1923; Munk, 1970). Besides accommodation, this is another method that allows objects at different distances from the eye to be in focus. With a 'ramp' retina objects on the ground could be in focus on the dorsal part of the retina at the same time that a distant object located higher in the visual field was in focus on the ventral retina. The density of visual cells in the dorsal part of the retina is greater than that of the ventral part and in general the Periophthalmus retina is better developed than that of Boleophthalmus (Munk, 1970). Without specifying the mechanism, Munk (1970) stated that the eyes are light adapted and while Graham (1971) stated that they contain high concentrations of pigment which gives protection from light and glare, Karsten (1923) found a strongly reflective argentea fibrosa, an adaptation for collecting light, in Periophthalmus. The retina of Boleophthalmus pectinirostris has no retinomotor response (Zhang & He, 1989). There is no change from a dark- MUDSKIPPERS 513 adapted state when it is exposed to a variety of different ambient light intensities and Zhang & He (1989) suggest that the pupillary response is entirely respon sible for adaptation to different light conditions in this species. Contrary to the statements of Harms (1929) and Graham (1971), rods and cones are found in both the dorsal and ventral area of the retina. There are differences in their distribution, however, because Yew & Wu (1979) noted a decrease in both rods and single cones and an increase in double cones from the anterior to the centre of the retina of P. chrysospilos. Increased length of the outer segments of the visual cells also means that objects at various distances would be equally in or out of focus (Munz, 1971). Munk (1970) demonstrated that both Boleophthalmus {boddarti) and Perioph thalmus (argentilineatus and kalolo) have horizontal band-shaped areae which may meet this criterion. The areae extend right across the retina slightly dorsal to the centreof the fundus where the retina and the visual cell layers are at their thickest. In comparison with the rest of the retina, the photoreceptors (rods, cones and double cones) of the areae are longer and thinner and their density is greatly increased. Possibly to avoid reducing the sensitivity of the areae, the optic nerve fibres do not all converge on the optic disc as in other fishes, but are spread out over a larger part of the retinal surface. The high density of visual cells in the areae increases the eyes' resolving power and lowers the threshold of movement perception, especially in the horizontal plane at right angles to the anatomical axis of the eye. Bothfeatures willassist in prey {Periophthalmus) and predator detection {Boleophthalmus, Periophthalmus', Munk, 1970). SKELETON Detailed skeletal characteristics of mudskippers have been investigated by Lele & Kulkarni (1938, 1939), Birdsong etal (1988), Murdy (1989) andLee (1990). Eggert (1929a) and Harris (1961) concentrated on the appendicular skeleton and muscular adaptations for locomotion on land. Pectoral fin musculature, for example, in the form of the adbuctor superficialis muscle being divided into two sections, is a terrestrial adaptation. It is found in Periophthalmus, Perioph thalmodon and Boloephthalmus, but not Scartelaos (Eggert, 1929b; Harris, 1961; Murdy, 1989). Additionally, fin and girdle modifications of Periophthal modon schlosseri [as Pseudapocryptes schlosseri] and Periophthalmus kalolo [as P. koelreuteri] are discussed by Mehta et al (1986, 1987), but like the accounts of Venkateswarlu (1969) and Chatterjee & Siddiqi (1957) add little to the biology of theseanimals. The details of the axial skeleton haveimportance for taxonomy, the spinous dorsal fin pterygiophore formula being of particular importance (Birdsong et al, 1988). From figures of the axial skeleton of P. barbarus (Lele & Kulkarni, 1939) and P. cantonensis (Lee, 1990) the formulae (Murdy, 1989) indicate that these two species should properly be designated as P. kalolo andP. modestus respectively (Murdy, 1989). In comparison withother gobies, the pre-orbital shortening of mudskipper skulls is evident (Lele & Kulkarni, 1938; Murdy, 1989; Lee, 1990), as is the reduction in bones (Lele & Kulkarni, 1938; Afzelius, 1956). Morphologically, little attention has focused on the pharyngeal plates. The marked differences in dentition of these plates relate to different feeding strategies (Milward, 1974) but they may also have taxonomic value. 514 DAVID A. CLAYTON SKIN The skin as a respiratory surface is discussed in the section on respiration (p. 541), but Bai & Kalyani (1960a,b,c) briefly give some detail of the skin of B. boddarti. In their work on scale morphogenesis and regeneration Bai & Kalyani (1960a, b,c) report only that the concentration of Vitamin C (ascorbic acid) in the brain (14.6 mg-lOOg"1) of B. boddarti was considerably higher than that of kidney (8.73), liver (2.09), skin (2.19), muscle (0.51) or blood (0.29). No explanation of the function of Vitamin C was presented although they subsequently reported on imino acid concentrations (Bai & Kalyani, 1961) for which ascorbic acid is a necessary substrate in the conversion of proline to hydroxyproline. These compounds stabilise collagen and hydroxyproline is a chemical marker for its presence. The concentration of the imino acids were higher (16.45 g-lOOg"1) in scales than in skin ((13.91 g-100"1). High brain levels of ascorbic acid are typical of marine fish, but only comparisons between the different genera will show whether the distribution of Vitamin C in the remaining tissues, especially the skin, is of significance to the mudskippers' amphibious mode oflife. The glucose, fructose, galactose and xylose content of the muscles of B. pectinirostris (Yamazoe, 1970) and the fatty acids of thebody and liver of B. boddarti (Misra et al, 1983) have also been measured. ALIMENTARY TRACT Alimentary tract morphology has been histologically described for B. boddarti, Periophthalmodon schlosseri [as Periopthalmus barbarus] (Lim, 1967), Peri ophthalmus vulgaris [ = P. argentilineatus] (Lim, 1971; Milward, 1974), Periophthalmodon schlosseri (Lim, 1971), P. freycineti [as P. schlosseri] and Scartelaos histophorus (Milward, 1974). Miyazaki & Nakamura (1980) describe the muscles in the region of the tongue of Boleophthalmus pectinirostris and Mehta et al (1990) compare the gross morphology of the head and position of the mouth of a number of oxudercine species and attempt to relate this to their method of foraging. The mouth is situated terminally on the lower part of the blunt snout and has fleshy lips which are presumably well supplied with sensory cells, especially in the herbivorous species. The jaw dentition is specialised such that carnivores have conical pointed teeth which may be of unequal size due to replacement (Milward, 1974). More spatulate, horizontally oriented teeth, as in Boleoph thalmus, are more typical of herbivores. Boleophthalmus, Scartelaos and Oxuderces possess a pair of large symphysial teeth (Lim, 1967; Milward, 1974; Murdy, 1989) which are probably used in agonistic situations (Lim, 1967). Pharyngeal plates also reflect diet because canine-like teeth, as illustrated for Periophthalmus barbarus [as P. koelreuteri] (Sponder & Lauder, 1981) are found in carnivores and smaller, more spatulate ones in Boleophthalmus (Lim, 1967; Clayton, unpubl. data). Periophthalmodon andPeriophthalmus haveshort tuberculated gill rakers (Lim, 1971), while those of Boleophthalmus (Lim, 1967) and Scartelaos (Milward, 1974) are much longer and more flexible. The relative gut length has beendiscussed in relation to diet (p. 553). Only Milward (1974) notes any internal divisions of the alimentary tract, noting the presence of a pyloric sphincter in periophthalmid species. Along its entirelength, the tract of all mudskippers studied shows much folding. It is greatest in the stomach and MUDSKIPPERS 515 rectum, but Lim (1967) showed the stomach of Periophthalmon schlosseri without any folding and P. argentilineatus has none in its rectum (Lim, 1971). The only report dealing with enzyme activity is that of Dhage & Mohamed (1977) in which the amylase activity of P. koelreuteri [IP. kalolo] decreased antero posterior^ along the tract. KIDNEY Periophthalmus is one of the few teleost genera to retain in maturity functional nephrons in the pronephros or head kidney, the gross and fine structure of which has been investigated by Safer et al (1982) and Safer & El-Sayed (1986). In P. waltoni [as P. koelreuteri] the paired renal corpuscles in the body of the head kidney and the single nephron of the anterior lobes are structurally indistin guishable from other kidney tubules. Haemopoietic tissue forms a supportive matrix throughout, with that of the anterior lobes being predominantly erythropoeitic. Ultrastructurally, the nephronic tubuleconsists of two proximal segments, a distal segment and a collecting duct. In the proximal segments ciliated cells assist in filtrate movement. Histochemical evidence that macromolecule and active transport mechanisms are present (El-Sayed & Safer, 1985) is supported by the anatomical evidence. Wandering cells that may be phagocytic are found here and in the proximal tubules of the body kidney (El-Sayed & Safer, 1986). The distal tubule is usually absent in marine teleosts and variably present in those which are euryhaline. That of P. waltoni has fewer mitochondria and less elaborate folding than euryhaline fishes adapted to sea water. While there was little enzyme activity in the distal tubule (El-Sayed & Safer, 1985), prior to sacrifice the fish were not adapted to freshwater and the distal tubule was otherwise structurally designed for secretory activity. Furthermore, the distal segment also contained many microbodies, the function of which is not yet known (Safer & El-Sayed, 1985). Also in the head kidney is the homologue of the adrenal gland. In Perioph thalmus viridis [?] the suprarenal or chromaffin (medullary) cells are inter spersed with the interrenal (cortical) cells that are located mainly around the posterior cardinal vein. Only adrenaline could be detected (Banerji, 1973). REPRODUCTION, DEVELOPMENT AND GROWTH REPRODUCTION The most detailed study of the reproductive cycle in mudskippers is that of Boleophthalmus dussumieri from Bombay (19°N) (Mutsaddi & Bai, 1970). The period of active maturation (gonadal stages IV—VI) lasts from February to May in males and March to June in females and the fish spawn once a year over the period July to September. Further north, in Korangi Creek (25°N) however, B. dussumieri (Hoda, 1986a) and B. dentatus [ = B. dussumieri] (Hoda & Akhtar, 1985) spawn twice a year, first in April to May and then in July to September. Working at an intermediate latitude (Jodia coast, 22 °N), Soni & George (1986) comment that B. dentatus spawns once a year in January to February. Lati tudinal differences are clearly inadequate to reconcile the disparate observations and further work is necessary. The fish in the population sampled from Korangi (Hoda, 1986a) were smaller than those from Bombay (Mutsaddi & Bai, 1970) and reached (50%) maturity at about 70 mm. Fish from Bombay attained 516 DAVID A. CLAYTON maturity at 96-110mm. Similarly, the fecundity of Korangi specimens was lower (970-4113) than that of the females from Bombay (1028-7197). Fecun dity of 15-50 g B. pectinirostris was much higher at 10 000-23 000 (Zhang et al., 1989) and in China (Fujian coast, 24°N) this species spawns only once a year (Xie & Zhang, 1990). Direct comparison between the Indian and Chinese data is difficult because different measures (length, weight) of size were used. As weight relationships are unreliable (see p. 517), both measures should be given in future. In the Ganges estuary at Alampur in west Bengal (22°25'N: 86°40'E), Acentrogobius, Boleophthalmus Periophthalmodon and Periophthalmus spawn between May or June and September or October. This is the monsoon period when pH and salinity are reduced and turbidity, temperature and the standing crop of plankton increases (Sircar & Har, 1975). Harms' (1935) supposition of viviparity in Boleophthalmus and Periophthalmus from Batavia (Jakarta) needs verification, for in other species eggs are laid in a burrow. The adhesion of eggs to the walls and roof of the egg chamber is achieved by filamentous attachment threads in P. chrysospilos (Harms, 1929), P. kalolo (Magnus, 1972), P. cantonensis [ = P. modestus] (Kimura, 1958), Boleophthalmus boddarti (Jones, 1937) and B. pectinirostris (Chen & Ting, 1984; Liao et al, 1973; Zhang et al, 1987, 1989; Zhang & Zhang, 1988; Hong et al, 1988). According to Brillet (1976) the eggs of Periophthalmus sobrinus [ = P. argentilineatus] are without threads, but as the eggs were only collected from burrows, it is possible that they were not noticed. Similarly, Asano (1936) omitted to mention their presence despite presenting a figure of pre-hatch eggs on the wall of the egg chamber of Periophthalmus sp. [ = P. modestus], a species which clearly possess them (Kimura, 1958). DEVELOPMENT Periophthalmids appear to require a longer period to hatch than do boleophthalmids. Hatching times for P. modestus are reported as 104 h (Kimura, 1958) and 170 h (Kobayashi et al, 1971) after fertilisation. The hatching period for Periophthalmus sobrinus [ = P. argentilineatus. ] as calculated from data in Brillet (1976) is 115 h. Liao et al (1973) found that the eggs of Boleophthalmus pectinirostris hatched between 66 h and 86 h after fertilisation, data that are supported by Chen & Ting's (1984) value of 66 h and by the value of 88 h given by Zhang et al (1989). Shorter hatching times were due to higher rearing temperatures, which may also have contributed to larval mortality. The research of the last three groups of authors and of Chen (1976, 1982) and Qiu (1989) has been directed towards the artificial propagation of larval fish for commercial farming. Previously, Pearse (1932) using animals ready to spawn, had little success in fertilisation of eggs and none in hatching eggs. Kimura (1958) was unsuccessful in inducing ovulation in Periophthalmus modestus by the use of frog cerebral hypophysis extract. Greater successwas achievedin Boleophthalmus pectinirostris where injections of carp pituitary and synchorin (Liao et al, 1973), human chorionic gonadotrophin (HCG; Chen & Ting, 1984) and HCG and luteinising hormone releasing hormone and pimozide (Hong & Wang, 1989) and combinations thereof (Zhang et al, 1989) produced varying degrees of success. In all cases milt was provided from surgically removed testes. Only Zhang et al. (1987, 1989) systematically controlled egg rearing condi tions and reported that the optimum temperature and salinity for hatching were MUDSKIPPERS 517 28°C and 15-25%o respectively. Outside these values hatch rates dropped dramatically. The failure of eggs to hatch at higher salinities is undoubtedly a contributing factor in the disparity between the potential and the realised niche of these species. In the Red Sea Periophthalmus is absent from the mangroves of the Sinai, an apparently suitable habitat (Fishelson, 1971; Por et al, 1977), but where salinity can reach 40-70%o (Por, 1974). Larval stages successfully hatched from eggs are illustrated by Kimura (1958), Kobayashi etal (1971), Zhangetal. (1987), Hong et al (1988), Zhang & Zhang (1988), Zhang et al (1989) and Yhang & Zhang (1990). Both Kobayashi et al (1972) and Zhang et al (1987, 1989) were successful in rearing larval fish beyond the fifth day and present feeding regimes for the developing larvae. In both cases, the diet was similar, but diverged as the periophthalmid was offered and began to take meat between 40-50 days after hatching when the 14-15mm larvae became substratum bound and amphibious (Kobayashi et al, 1972). In Boleophthalmus pectinirostris the equivalent age and total length for the young to become benthic was 42 days and 20 mm (Zhang et al, 1989). Field observations support these data on the size at which mudskippers become amphibious. The smallest amphibious Periophthalmus argentilineatus found in the mangrove are 11-13 mm (Brillet, 1976). Seasonal collection of 15-30 mm Boleophthalmus pectinirostris fry from the mudflats was the only way to obtain stocks for commercial rearing until successful egg culture was achieved (Chen, 1976). The fry were placed in specially prepared ponds of 0.1-1.0 hectares at densities of between 30-50 000 fry-hectare"1. Beforethe fish reach a size suitable for sale the ponds required careful pre- and post-stocking management over 1-2 years (Chen, 1976). As fish growth is greatest above 28°C (Chen, 1982) tropical pond culturing could reduce this management period (Macintosh, 1982). GROWTH The only data on longevity of mudskippers is for Boleophthalmus dussumieri from the west coast of the Indian continent where, from a study of otolith rings of B. dentatus [ = B. dussumieri], Soni & George (1986) showed that 130 mm total length fish were 2 years old. Both Hoda (1986b, 1987) and Soni & George (1986) present length/weightrelationships expressed in the form of the equation W= aLb, but the analyses are not comparable because Soni & George used total length (TL) and Hoda used standard length (SL). Furthermore, the equation presented in Soni & George (1986) is clearly incorrect. The fish in the sampled population ranged from 20-133 mm TL and 0.1-10.5 g, yet the equation gives the weight of an 80 mm fish as 256 g. Hoda also gave equations for the relationship between SL and TL but these also seem unreliable because the slope and intercept values in the equations for the combined male and female data sets are at variance with those of the separate (male and female) data sets. This is most obvious in the equation for B. dentatus where similar intercept and slope values for male (6.21, 0.734 respectively) and female (6.184, 0.730) change considerably in the combined data set (6.957, 0.743). Additionally, separate analyses for B. dussumieri and B. dentatus were presented (Hoda, 1986b, 1987) and the latter name is a junior synonym of the former (Murdy, 1989). More information on length/weight relationships would be of value to other areas of investigation, for size related changes in structure or function complicate 518 DAVID A. CLAYTON comparative studies (see sections on Osmoregulation, p. 524, and Respiration, p. 532). The data on the range of standard length of the specimens Murdy (1989) examined in his taxonomic revision of the Oxudercinae provide some guidelines for the limits of comparisons between species (see also p. 540). EXCRETION As amphibious fishes, the mudskippers' excretory mechanisms are particularly interesting, but as they are only just being revealed, the full potential of this aspect of the fishes' physiology in relation to their degree of amphibiousness has yet to be fully explored. Basic information on the rates of excretion under comparable conditions is required to complement the available and sometimes contradictory data (Table I). Allowing for differences in experimental condi tions, there is still considerable variation in the rates of excretion of ammonia and urea. Although retaining similar percent levels of urea excretion, the same authorities working on the same species show several-fold changes in the absolute levels of ammonia and urea excretion {B. pectinirostris and Periophthalmus modestus: Morii et al., 1978, 1979 in Table I). The ammonia: urea ratio {% urea Table I) shows no correlation with the fishes presumed water dependence {Scartelaos > Boleophthalmus > Periophthalmus). Unfortunately, measurements were made at a variety of salinities which has a marked influence on this ratio as shown in P. modestus by Iwata et al (1981) (Table I). Additionally, more studies need to be done on Scartelaos and Pseudapocryptes, especially as the latter has a response to the drying out of its habitat that is similar to lungfishes (Hora, 1935a, b). Furthermore, because there are several closely related genera within a single subfamily, better comparative material can be obtained for Pseudapocryptes than for any lungfish. The basic issue in excretion has centred on differences in nitrogen metabolism between the mudskippers' aquatic or surface phases of activity. In marine teleosts there are a number of waste nitrogen compounds and in mudskippers over 96% is in the form of ammonia and urea (Morii et al., 1978). More than three- quarters of waste nitrogen is excreted by the gills, a pathway presumably not available to a mudskipper out of water (Gregory, 1977). Setting the experimental paradigm, Gordon et al. (1969) compared the 24 h rate of ammonia and urea excretion between starved Periophthalmus sobrinus [ = argentilineatus or kalolo] which had been confined in sea water for 24 h with that of similar fish which had spent the first 12 h out of water before being returned to sea water. They found no differences in these 24 h rates. At this point the fish exposed to air had only been back in sea water for 12 h, but when the excretion rate of both compounds was calculated at the end of 24 h in sea water, it was found to be double that of the 24 h rate for fish maintained continuously in sea water. This suggested to Gordon et al (1969) that nitrogenous waste, accumulated in the body during aerial exposure, was released on return to water. Further calcu lation demonstrated that while the rate of ammonia production only doubled when the fish was out of water, there was a 3.5 fold increase in the rate of urea production (Gordon et al., 1969; Gordon, 1970). This shift of the am monia: urea ratio in the direction of urea was also shown for P. modestus [as P. cantonensis] and Boleophthalmus pectinirostris (Morii et al, 1978). Gregory (1977) showed that in sea water, Periophthalmus novaeguineaensis [as P. expeditionium], P. gracilis and Scartelaos histophorus all had comparable Table I Ammonia and urea excretion in mudskippers. For ease of comparison all data are converted to /ig-g-24 h"1. aData from 24 h graphical presentation. Species Ammonia-N (Mg-g-24h_l) Urea-N % Urea (Mg-g-24h_l) Fish size Temperature (°C) Fed (F) or starved (g) Scartelaos histophorus 156 37 Boleophthalmus pectinirostris 34 104 19 1-5 4 10 33-46 11 9 36-45 (S) Room % Reference sea water s 100 Gregory, 1977 20+2 s 25 28 + 2 s 25 Morii et al., 1978 Morii et al., 1979a Periophthalmus gracilis 184 31 14 1.0 Room s 100 Gregory, 1977 P. novaeguineaensis 259 129 33 1-5 Room s 100 Gregory, 1977 P. sobrinus 165 242 59 0.5-15.0 26.5 + 3.5 s 100 3.0-15.0 26.5±3.5 s 40 Gordon et al., 1965 Gordon et al., 1969 Gordon et al., 1965 P. modestus 2 a o GO m 259 40 2117 4973 55 0.5-5.0 21+2 s 100 2150 8467 66 0.5-5.0 21+2 F 100 106 11 9 4.0-6.3 s 25 440 60 2 3.8-6.5 222 55 20 1.0-3.0 s 80 s Air 13 167 40 19 1.0-3.0 101 62 38 1.0-3.0 20±2 29±2 20±1 20±1 20±1 23 28 55 1.0-3.0 20+1 s 25 s 20 s 50 Gordon et al., 1978 Gordon et al., 1978 Morii et al., 1978 Morii et al., 1981a Iwata et al., 1981 Iwata et al., 1981 Iwata et al., 1981 Iwata et al., 1981 on 520 DAVID A. CLAYTON levels of ammonia excretion to that of Periophthalmus sobrinus, but consider ably lower levels of urea excretion (Gordon et al., 1965). Urea excretion, expressed as a percentage of total excretion in these three species, is therefore considerably lower than that of P. sobrinus (Gregory, 1977; see also Table I). No air exposure trials were done, but Gregory (1977) found that the fish were deficient in enzymes necessary for the production of urea by the ornithine-urea cycle. He suggested that, alternatively, the urea could have been produced by purine catabolism because the enzymes necessary for the breakdown of uric acid are present in the periophthalmids. Neither this suggestion (seealso Chew & Ip, 1987) nor the supposition by Gordon et al (1968) that P. sobrinus exhibits a shift towards ureotelism while out of water have found support. When nitrogen excretion is monitored over 4-5 days in both P. modestus and Boleophthalmus pectinirostris the general responseis that urea levels remain fairly constant while ammonia initially declines and thereafter increases (Morii et al., 1978). This pattern does not depend on whether the fish are in or out of water although the absolute levels of excretion in water are several times greater. A compensatory increase in the rate of nitrogenous excretion when fish return to water (Gordon et al, 1968) has been partially confirmed in more detailed studies. Ignoring the high excretory levels reported for Periophthalmus modestus in comparison with other authors (Table I), Gordon et al. (1978) kept fish out of water for 16.5 h and found that the rate of urea-N excretion showed a three-fold increase above control (in sea water) rates for the first 4 h after return to sea water and a two-fold increase in ammonia-N excretion for the first 5 h. Working with a 6-h sampling programme Morii et al. (1979) monitored the excretion on return to water of P. modestus and Boleophthalmus pectinirostris after 12, 24, 36, or 48 h out of water. For the shortest period out of water (12 h), for example there was an approximate 1.5 fold increase in ammonia-N excretion above control rates in both Periophthalmus modestus and Boleophthalmus pectinirostris during the first 6 h back in water. Contrary to Gordon et al. (1978), no change in urea rates were found. As the period out of water increased there was an increase in the rates of excretion of both products during this initial period (Morii et al., 1979). This trend was most pronounced for ammonia in B. pectinirostris, but in the absenceof statistical analysis the significance of any of these trends is not known. Morii et al. (1978) also attempted to show the partitioning of urea and ammonia excretion between the gills, skin and cloaca. When Periophthalmus modestus was out of water, there was no great difference between concentrations excreted through the skin and cloaca. In anuric (cloaca closed) P. modestus and Boleophthalmus pectinirostris in water the nitrogen concentration excreted by the skin (body behind head) was numerically higher in urea than ammonia. The reverse was found in the gills (actually head and gills). Verification of these data and conclusions are required, if only because the data show a 10-fold increase of urea production by anuric fish when compared with that of normal fish in water. In the absence of any change in the rate of urea excretion on return to water, Morii et al. (1978) suggest that the ammonia accumulated during the period out of water was not converted to urea. The data on blood concentration are con sistent with this interpretation, although again statistical confirmation would be beneficial. In Periophthalmus modestus urea concentration was lower in fish out of water than in fish in water, and decreased with longer periods out of water. MUDSKIPPERS 521 For Boleophthalmus pectinirostris the data show an opposite trend, but in both species, fish out of water exhibit a lower ratio of urea-N/ammonia-N than those in water (Morii et al, 1979). These results were replicated by Morii (1979) and extended for 5 days instead of two and show a broadly similar distribution and change with time of the levels of urea-N and ammonia-N in the liver, muscle, skin and gill tissues. After 3 days out of water, ammonia levels increased dramatically in muscle, skin, and especially gill tissue, of B. pectinirostris. It is likely that these data represent pathological changes because fish in this group died (Morii, 1979). Similarly, Periophthalmus modestus that died in sea water are also included in the data sets. Certainly Boleophthalmus pectinirostris is likely to have a much lower tolerance to excretory product level than Perio phthalmus modestus as both urea and ammonia concentrations in the latter are several times greater than in the former species (Morii, 1979; Morii et al, (1979). The absence of any accumulation of urea in any of the tissues of either fish can, however, be taken as confirmation that ammonia is not converted to urea during the period out of water (Morii, 1979). Certainly in P. modestus kept out of water the distribution of urea was more uniform among the tissues (muscle 34 jag-g-day-1; liver 39 /xg-g-day-1; 5-day data, Morii, 1979; muscle 65 /xg-g-day"1; liver 29 /xg-g-day-1; 7-day data, Iwata et al., 1981) than ammonia where the concentration was greatest in muscle (142, 186 ^g-g-day-1) and lowest in liver (52, 59 /xg-g-day-1, Morii, 1979; Iwata et al., 1981 respectively). As the liver accounts for less than 2% and the muscle for 48% of the body weight in P. modestus (Iwata, 1988), muscle is likely to be the major storage site of ammonia during the period out of water (Iwata et al, 1981). In the absence of ureotelism, the detoxification of nitrogenous waste appears to be achieved through the metabolism of non-essential amino acids which are more usually thought to be associated with osmoregulation (Iwata et al, 1981). In comparison with fish kept in 20% sea water, and paralleling the change in ammonia concentration, there was a significant doubling of total free amino acids (FAA) in the muscle of P. modestus out of water. Markedly increased levels of the non-essential amino acids alanine, aspartate and glutamate and especially taurine were found, such that the FAA comprised about 70% of the total increment of nitrogen (above that of fish in water); ammonia and urea made up only 21% and 4% respectively (Iwata et al, 1981). The enzyme glutamate dehydrogenase (GDH) also plays a central role in removing ammonia from the tissues of fish subjected to conditions of water shortage including being out of water (Iwata & Kakuta, 1983). This enzyme catalyses reductive animation of a-keto-glutarate and Iwata et al. (1981) and Iwata & Kakuta (1983) measured its activity in the reductive amination (or glutarate forming) direction in the skin, gills, muscle and liver of P. modestus. In fish kept out of water the activity of muscle GDH was four times higher than in 20% sea water. Liver GDH responded similarly, but was about five times higher inactivity than that of muscle (Iwata etal., 1981; Iwata & Kakuta, 1983). The kinetic constant Km (substrate concentration at which the reaction rate is half maximum) of GDH for NH4+ in muscle is lower than that of liver, skin and gill GDH, permitting a rapid response of the enzyme to tissue ammonia concentration. GDH activity in the liver was also greater than in the muscle (and gills) of Boleophthalmus boddarti and Periophthalmodon schlosseri (Chew & Ip, 1987). Enzyme activity in the reductive amination direction was 16 {Boleophthalmus 522 DAVID A. CLAYTON boddarti) and 20 {Periophthalmodon schlosseri) times greater than in the oxi dative deamination (or ammonia forming) direction indicating that liver GDH is not involved in ammoniagenesis, but is likely to be central to the regulation of amino acid pool in the whole body (Chew & Ip, 1987). The metabolic pathways for the production of ammonia from other FAA is also discussed by Chew & Ip (1987). Chew & Ip (1990) confirm that liver GDH activity is higher than that of muscle in Boleophthalmus boddarti and also showed that GDH activity was greater, but of a similar pattern in Periophthalmus chrysospilos. For Boleophthalmus boddarti, however, the enzyme activity in the amination direction was 32 times greater than in the deamination direction: double that reported previously (Chew & Ip, 1987). It is difficult to relate the doubling of the amination/deamination ratio to salinity since Chew & Ip (1990) found no difference within liver or muscle GDH activity between fish exposed for 3 days to either 10% or 80% sea water. In the earlier experiment, the fish were main tained in 50% sea water for 1 day prior to enzyme analysis (Chew & Ip, 1987). This is possibly why the earlier paper was not mentioned in the later one. Aspartate transaminase (or glutamate-oxaloacetate transaminase, GOT) and, to a much lesser extent, alanine transaminase (or glutamate pyruvate trans aminase, GPT) activity was also much greater in the liver than in the gills or muscle of B. boddarti and Periophthalmodon schlosseri (Chew & Ip, 1987). While the activity of GOT was also higher than that of GPT in Periophthalmus modestus, their activity was higher in the muscle rather than the liver (Iwata, 1988). The relative activity of these enzymes suggest that aspartate (= FAA aspartic acid) rather than alanine is the major substrate for ammoniagenesis in the muscle and liver mitochondria of Boleophthalmus boddarti (Chew & Ip, 1987) and Periophthalmus modestus (Iwata, 1988) and the liver of Periophthal modon schlosseri (Chew & Ip, 1987). In support of this suggestion, the aspartate content was significantly less than that of alanine in the muscle of Boleoph thalmusboddartiand Periophthalmodonschlosseri (Siau & Ip, 1987). Under the catalytic action of mitochondrial GOT, aspartate and a-ketoglutarate form oxaloacetate and glutamate, the latter being deaminated by GDH to yield aketoglutarate and ammonia. In muscle of P. schlosseri the FAA glutamine is the most effective substrate for ammoniagenesis, but the low glutaminase activity (which could catalyse ammonia and glutamate formation) and other data suggested to Chew & Ip (1987) that an alternative pathway for ammoniagenesis involving an initial transamination to a-ketoglutaramate which is hydrolysed to a-ketoglutarate. A more parsimonious explanation of the low glutaminase activity found by Chew & Ip (1987) may be that, contrary to expectations, the muscles were actually subjected to ammonia loading or stress. By comparing Periophthalmus modestus out of water with those subjected to an increased ammonia load by being kept in NH4C1 solution, Iwata (1988) further confirmed the role of FAA in ammonia detoxification. The increases in the qualitatively similar amino acids (glutamate, alanine, glutamine and glycine) under both conditions strongly suggest that their production was triggered by ammonia rather than dehydration stress although the impairment of gas exchange may be the reason for the greater increase in alanine in the ammonia-stressed fish (Iwata, 1988). In contrast to the situation in Boleophthalmus boddarti and Periophthalmodon schlosseri (Chew & Ip, 1987), glutamine synthetase in the muscles of Periophthalmus modestus was actively involved in the removal of ammonia by catalysing its use in the synthesis of glutamine from glutarate (Iwata, MUDSKIPPERS 523 1988). Chew & Ip (1987) maintained their fish in 50% sea water but, as Iwata et al (1981) demonstrated, salinity changes have consequences for levels of ammonia and urea excretion. This raises the general observation that fish rearing conditions are extremely importantin determining the adaptiveresponseof enzyme systems to physiological change. It is not clear why most studies (Table I) have used fish adapted to 20% sea water unless it is to maximise ammonia and urea production between the adapted fish and those kept out of water. Following the original experimental paradigm, biochemical investigation of fish returning to water after being emersed would be most instructive. Whatever further details arise from such studies, including the role of taurine (Iwata etal, 1981; Iwata, 1988), it is clear that mudskippers have a greater tolerance than other water-breathing fish including gobies. This ammonia tolerance can be accountedfor by their ability to synthesise non-essential amino acids (Iwata, 1984, 1988). P. modestus was kept in very different conditions (Iwata, 1988) to the other species (Chew & Ip, 1987) and it is possible that glutamate synthetase also becomes active in species of Boleophthalmus, Scartelaos and Periophthalmodon under conditions of extreme ammonia stress. Further comparisons that could be made include the assay of brain enzyme activity. In Periophthalmus modestus there is a powerful GDH-glutamine synthetase ammonia detoxification system (Iwata, 1988). It would be instructive to know if this is a general mechanism or specific to those species which have a high tissue tolerance for ammonia. SALINITY TOLERANCE AND OSMOREGULATION SALINITY TOLERANCE Mudskippers can generally be regarded as euryhaline and percent mortality is the simplest and crudest measure of mudskippers' response to changes in salinity. Typically, groups of fish have been placed in water of salinity ranging from 0—100% sea water for a number of days during which deaths were monitored (Table II). Leaving aside ethical issues, the execution of even this simple procedure leaves much to be desired and experimental procedures need to be standardised and carefully described before good comparative data can be obtained. The timing of analytical procedures used to investigate any physio logical mechanisms of adaptation to changing environmental conditions is important, if only in determining the sequence of such mechanisms. Vague experimental protocols (Bhan & Mansuri, 1978a; 5, 10 or 15 days acclima tion?), experimental design (Gordon et al, 1965; adaptation period varied between 2 and several days), or independent non-comparable measures of response (Mansuri & Bhan, 1978; Chew & Ip, 1990; Table II) at best simply confuse the analysis. At worst, and given the reported mortality of fish in different salinities as described below, it may mean that pathological changes are being observed. Mansuri & Bhan (1978) report high mortality of Periophthalmus dipes in 50 and 70% sea water, but little in 0% and 100% sea water. Conversely at 0% and 100% salinities, P. chrysospilos and Boleophthalmus boddarti exhibited their highest mortality but survived well (only 8% mortality) in 50% and 80% sea water (Chew & Ip, 1990). While Mansuri etal (1982) showed that B. dussumieri [as B. dentatus] had a better survival rate at all salinities than Periophthalmus dipes, mortality in both species was greatest in 50% and 70% Table II Experimental protocols used by various authors for osmoregulation studies in mudskippers Author Species Fish Group Temp. Fed (F) Immersed Holding Adaptation Sea water Salinity Experi Independent size size or (I) or dilutions of sea mental variables starved Air/water (% sea (%) water duration (S) (A/W) (%o) (days) (°C) (g) Gordon et al., 1965 Gordon et al., 1978 Mansuri & Bhan, 1978 P. sobrinus 3—15 P. modestus 2—3 P. dipes 20-24 Mansuri et al., 1982 P. dipes 2-7 100,80,60,40,20,0 23-30 S 100 6 22-24 S 100 100 27-28 F 100 90,70,50,30,10,0 — 100 90,70,50,30,20,0 20 — 14 and % Mortality 6 Final % survival 15 Daily % survival Time at 50% and 32 50 100% mortality, 22-25 blood serum 25 Mortality at 3 and 7 days, 5—12 25 B. boddarti water) 20 B. dentatus P. chrysospilos (days) 12-16 25-30 Chew & Ip, 1990 condition time 100,80,50,10,0 50 - 34 7 P. vulgaris 0.1—3.4 — F — 50 2 100,40,14 35 1 Water intake, Na+ fluxes Lee et al., 1987 P. chrysospilos 6—12 - - - 50 3 100,80,50,30,0 34 1 Na+, K+,volume regulation Bhan & Mansuri, 1978a, b P. dipes Lee & Ip, 1987 P. chrysospilos 20-24 6-12 - - F S - - 5 100,90,70,50,30,0 A/W 50 3 100,50,30 10 — 3 36 Tissue water and mineral content T3, T4 cAMP activity B. boddarti Siau & Ip, 1987 P. schlosseri - - A/W 50 A/W 25 100,50,30 - Enzyme activity - B. boddarti Fenwick & Lam, 1988a, b P. schlosseri 78-112 S > tissue water and 10-22 proteins Dall & Milward, 1969 > 24-26 6 25, Air - 7 Na+, K+, Ca++ levels n r > H O MUDSKIPPERS 525 sea water. Additionally both species had the greater survival times in winter (12-16°C) than in the summer (22-25°C). This apparent confirmation of the earlier data for Periophthalmus (Mansuri & Bhan, 1978) must be offset against the large discrepancy between the two data sets. Mansuri & Bhan (1978) found that some fish survived for 15 days in both salinities (50% sea water, 90% mortality; 70% sea water, 70% mortality) whereas in later experiments (Mansuri et al, 1982) all fish died within 2 (summer) or 4 (winter) days. If, as the data on summer and winter survival suggests, temperature is an important variable, then the discrepancy is even more surprising because the fish survived longer at high temperatures (27-28°C; Mansuri & Bhan, 1978). The specific identity of P. dipes is also unclear. P. dipes is a synonym of P. argentilineatus, but at 12—16 cm and 20-24 g, the specimens seem to be too large to be this species. The difference between the results obtained by Chew & Ip (1990) and Mansuri & Bhan (1978) could be related to the experimenters' choice of holding condi tions for the fish prior to the salinity change. P. chrysospilos and Boleoph thalmus boddarti were able to move between air and 50% sea water (= 17%o salinity; Chew & Ip, 1990), while Periophthalmus dipes was maintained in 100% sea water of unknown salinity (Mansuri & Bhan, 1978). As P. dipes can tolerate salinity changes in the range of 24%o—47%o salinity (Bhan & Mansuri, 1978a), the acclimation salinity is of some importance. Where P. sobrinus [ = kalolo or argentilineatus] all died within 3 days of being transferred from 100% sea water (= 34%o salinity), if they were first acclimated in 20% sea water for 6 days, there was no mortality (Gordon et al, 1965). Especially in groups, mortality among totally submerged mudskippers may be due to asphyxiation rather than salinity stress (cf. Hora, 1935b). OSMOREGULATION The body water content of mudskippers is maintained at constant levels over a wide range of salinities. Only P. chrysospilos in air and deionised water showed any significant change (decrease and increase respectively) in water content after 3 days of exposure to a variety of salinities (Lee et al., 1987), supporting data on P. sobrinus where, however, the significant increase of muscle water content in fresh water over the 100% sea water condition was reported only from moribund fish (Gordon et al, 1965). Nevertheless, P. sobrinus [ = kalolo or argentilineatus] showed a 10-15% weight increase following transfer to all salinities, which Gordon et al, (1965) attributed to the immersion of the fish. For fish directly transferred to fresh water, the increase was 20% and the fish died within 3 days, but if previously adapted to 20% sea water, there was no change in weight (Gordon et al, 1965). In air in direct sunlight, P. sobrinus lost 12% body weight before dying after 50 min, whereas in shade and high humidity the body weight loss could be doubled and the fish survived for 24 h (Gordon et al, 1969). In P. dipes [?= P. argentilineatus] the percent water content of muscle, gill, liver, heart and kidney remained stable at all salinities including fresh water, whereas in comparison with the levels in 100% se? water, sodium, potassium, and to a lesser extent, calcium, but not phosphorus, were reduced in these tissues (Bhan & Mansuri, 1978a). The total protein and fat content of red and white muscle, kidney and gills was reduced, however, especially at lower (< 30%) 526 DAVID A. CLAYTON salinities (Mansuri & Bhan, 1978). In most salinities white muscle used glycogen rather than fat as the main energy source (Bhan & Mansuri, 1978b) and succinic dehydrogenase activity was lower than in other tissues (Bhan & Mansuri, 1978c). Gills of fish adapted to 10% sea water and fresh water exhibited the highest level of succinic dehydrogenase activity (Bhan & Mansuri, 1978c). The regulation of plasma osmolarity (and NaCl concentration) was nearly perfect in P. sobrinus kept for up to 6 days in 20—100% sea water (Gordon et al, 1965). Only dying fish that had been directly transferred to fresh water showed a significant dilution of their blood. On the basis that when previously adapted to 20% sea water, a single fish in fresh water maintained osmolarity in the normal range, Gordon et al (1965) concluded that the species was probably a perfect osmoregulator. No other mudskipper approaches this level of perfor mance. After 3 days in 100% sea water plasma osmolarity of P. chrysospilos reached 440 mmol-kg"1 a significant increase above the 324 mmol-kg"1 control level offish free to enter 50% sea water (Lee etal, 1987). No significant changes were found for fish kept in 30, 50 or 80% sea water (Lee et al., 1987), but after 7 days in 80% sea water the 315 mosmolal value reported by Chew & Ip (1990) for P. chrysospilos kept under similar control conditions repres ented a significant increase above the 299 mosmohal control level. Furthermore, the plasma osmolarity of 260 mosmolal of fish in 10% sea water indicated a significant dilution of the blood. Chew and Ip (1990) did not test their fish in 100% sea water but additionally showed that plasma osmolarity of Boleophthalmus boddarti (219 mosmolal) was significantly less than that of Periophthalmus chrysospilos and also decreased (to 252 mosmolal) in 10% sea water after 7 days. However, as the decrease in 10% sea water was greater in P. chrysospilos, and Boleophthalmus boddarti was able to maintain its plasma osmolarity (275 mosmolal) in 80% sea water, it is likely that extracellular aniso-osmotic regulatory mechanisms are more efficient in B. boddarti (Chew & Ip, 1990), supporting the suggestion by Mansuri et al. (1982) that Boleophthalmus was a better regulator than Periophthalmus. In comparison with the condition in 100% sea water, after a day Boleoph thalmus boddarti showed increased plasma levels of sodium at all salinities except for fresh water where there was a decrease (Mansuri et al., 1982). For potassium and calcium there was no similar simple relationship between salinity and plasma concentration. Plasma potassium increased greatly in 50% and 70% sea water (the salinities in which most fish died) and calcium only decreased in fresh water (Mansuri et al, 1982). Whether the increased plasma sodium is accounted for by the reduced intracellular (tissue) levels (Bhan & Mansuri, 1978a) is a moot point, but there was little correlation between intra- and extracellular levels of potassium and calcium (Bhan & Mansuri, 1978a; Mansuri et al., 1982). If only because of the absence of the units of measurement, the data of Mansuri et al., (1982) must be questionable. Truly comparative data with a similar control condition are not available, but in Periophthalmus chrysospilos after 1 day plasma sodium levels were similar (80—82 mM) at salinities above (100%, 80% sea water) and below (30% sea water) the freely accessible 50% sea water control condition, where the plasma concentration was significantly higher at 92mM (Lee et al., 1987). After 3 days for sodium and 5 days for potassium, however, a more stable state was reached in which plasma concen trations were higher than those of the control condition, such that after 5 days sodium values were between 134-151 mM and potassium between 14-16.8 mM MUDSKIPPERS 527 (values calculated from Figure in Lee et al., 1987). These data need reconciling with the osmolarity data presented in the same paper. In comparison with control fish, a 3-day restriction to 50% sea water produced no change in osmolarity (Lee et al, 1987; Chew & Ip, 1990) while 3 days in 100% sea water significantly increased it (Lee et al, 1987). Plasma levels of sodium and potassium did not reflect this lack of change and were increased to similar levels above control in both conditions. The removal (in 50% sea water) or addition (in 100% sea water) of other osmotically active substances must be implicated. The presence of NPS (Ninhydrin Positive Substances) including FAA in tissues is clearly related to the regulation of intracellular fluid content, but they may also be involved in plasma osmotic regulation, although Lee et al, (1987) found no significant changes in NPS levels of the liver and muscles of the fish after 3 days. Fish in higher salinities for 7 days, however, showed higher levels of NPS in both tissues. There was a similar direct relationship between NPS and salinity after 7 days in P. modestus (Iwata et al, 1981). Chew & Ip (1990) confirm this 7-day finding and additionally showed that the decrease in plasma osmolarity of P. chrysospilos was accompanied by a decrease in the level of NPS in muscle compared with control fish. Furthermore, GDH activity in the aminating direction was significantly greater in fish maintained in 80% sea water than in fish in 10% sea water for 7 days. The differences between short term (1 day) and long term (7 days) plasma levels of sodium and potassium were related to changing membrane permeability such that both sodium influx (k) and efflux (k) rate constants increased with increasing salinity; the latter linearly (Lee et al, 1987). The efflux rate constant of P. argentilineatus [as P. vulgaris] also showed a linear relationship to salinity (Dall & Milward, 1969), but the slope was much shallower, possibly because the fish were ligated, excluding urinary excretion. There was an efflux of 15—20% from the body and fins, the remainder was from the branchial region. While individual rates were variable , drinking of P. argentilineatus was inversely related to salinity such that the rate in 100% sea water (35 %o) was tripled in 14% sea water (5%o; Dall & Milward, 1969). In comparison with a cardinal fish, the mudskipper had similar efflux rates but a reversed drinking regime and Dall & Milward (1969) concluded that the wide salinity tolerance coupled with behavioural adaptations were adequate to account for salt regu lation in Periophthalmus. For example, adaptation to low salinities could be achieved by remaining out of water for long periods so that hydration could be offset by evaporation (Dall & Milward, 1969). Studies of osmoregulation and excretion frequently used animals that were not fed prior to sacrifice while in captivity, and care is required in interpreting results from starved fish. The depletion of muscle glycogen reported in P. dipes [? P. argentilineatus] as a result of exposure to different levels of salinity (Bhan & Mansuri, 1978c) was also reported for Boleophthalmus boddarti as a response to starvation (Lim & Ip, 1989). In Bhan & Mansuri's (1978c) experiments with Periophthalmus dipes the fish were fed while in captivity (Table II) but this is the exception. If only because of the variety and inter-relationships between the biochemical pathways concerned, the effects of fasting need to be assessed independently of those of osmoregulation and excretion. In P. chrysospilos the short term (1 day), regulation of plasma sodium and potassium, as evidenced by the increased Na+ and K+ efflux rates with in creasing salinity, was matched by increased Na+K+ATPase activity in the 528 DAVID A. CLAYTON gills. This active mechanism that requires energy to regulate ionic control was further reflected in the increased oxygen consumption rate at higher salinities (Lee etal, 1987). Plasma osmolarity increases in mudskippers out of water, presumably because of dehydration, but varied experimental protocols again leave much to be desired. Gordon et al. (1978) subjected P. modestus to rapid (20% body weight change, h"1) and slow (2%h_1) dehydration in air at 25-28°C. After the fish had lost 20% of their body weight in the latter condition there was an increase of 40% and 60% in plasma osmotic and sodium concentrations respectively. The fish dehydrated so rapidly that they became moribund and the experiments lasted only 23 h, but the 24 h air exposure of P. chrysospilos did not result in any weight loss and only a 16% increase in plasma osmolarity (Lee et al, 1987). Despite the difference between species, the (unspecified) conditions were appar ently far less severe in the latter experiment. By constantly moistening filter paper, Fenwick & Lam (1988a) were able to maintain Periophthalmodon schlosseri in air at 25 ± 1°C for 7 days and body weight and plasma sodium and potassium concentration remained at control levels. Only calcium levels in creased significantly and measurement of influx and efflux rates showed a net uptake of calcium by P. schlosseri in water, whereas in air the efflux rate was effectively zero (Fenwick & Lam, 1988b). This suggests that the gills are the primary site of calcium efflux and that P. schlosseri is able to perform considerable cutaneous calcium transport. Being able to prevent the increase in calcium plasma concentration of fish maintained in air, it appears that calcitonin only has a regulatory effect under conditions of hypercalcemia and is otherwise not involved in osmoregulation (Fenwick & Lam, 1988a). HORMONAL CONTROL OF OSMOREGULATION Using histological evidence of the synthesis and utilisation of a variety of neuro endocrine glands in Boleophthalmus dentatus, Patel & Desai (1976) correlated gland activity with environmental conditions at capture. They made monthly collections of fish at low tide from July to October; half were sacrificed immedi ately and the rest after being kept on the shore until they had been immersed for an hour by the flood tide. Examination of the hypothalamic neurohypo physeal complex, the caudal neurosecretory system, the corpuscles of Stannius, inter-renal cells and thyroid gland showed that their activities were correlated with the water temperature and salinity at sacrifice. Water pH was also measured and found to vary in the same way as salinity, but was otherwise ignored. The changes in neuroendocrine activity were dependent on the relative salinity and temperature of the tides and are best interpreted in terms of the ebb and flood tides of October against those of the remaining months. In comparison with flood tides, the ebb tides of October had a marginally higher salinity (28 %o, +0.6%o) and a lower temperature (26°C, -1.7°C) and all neuroendocrine glands, except the thyroid, were in a phase of synthesis. In the October flood tides this situation was reversed, such that products were being utilised while the thyroid, having been secretory, entered its phase of synthesis. For the July to September period the relationships between the activity of the neuroendocrine glands and environmental conditions were maintained, except that it was the flood tides that exhibited marginally higher salinity (median 36.9%o, range, 25.6-37.2%o: + 1.6%o, 0.1-2.1%o) and lower temperature (27°C, 27-28°C: MUDSKIPPERS 529 -2°C, 0.7-2°C). Patel & Desai (1976) relate the neuroendocrine response only to salinity, but in view of the wide salinity and temperature ranges that these fish can tolerate, the level and speed of the response for such small environmental changes needs to be verified. Nevertheless, some evidence for thyroidal secretion at higher salinities and prolactin secretion from the eta cells of the rostral pars distalis (Patel & Desai, 1976) at lower salinities, is provided by Lee & Ip (1987) who showed that plasma thyroxine (T4), 3,5,3'-triido-L-thyroxine (T3), pro lactin and cyclic adenosine 3',5'-monophosphate (cAMP) are implicated in control of osmoregulation in Periophthalmus chrysospilos and Boleophthalmus boddarti. Furthermore, in comparison with fish in 15% and 100% sea water, the nuclear and cell sizes of prolactin—but not growth-hormone-secreting, cells of the pituitary of P. chrysospilos were considerably enlarged (Ogasawara et al., 1991). Prolactin acts to reduce salt loss and in P. chrysospilos plasma concentration increased significantly in 30% and 0% sea water and in fish out of water. Whatever the mechanism of salt reduction induced by prolactin, it is difficult to account for the significantly reduced plasma concentration when the fish was submerged in 50% sea water. This was the same concentration as the control condition, and was not matched by an equivalent reduction in the other 'higher' salinity (100% sea water). In Boleophthalmus boddarti plasma prolactin concentration only increased above control values in 100% sea water, but the mechanism of prolactin action in hyperosmotic regulation is unclear. Relative to the control condition, the level of T4 in Periophthalmus chrysospilos was higher in air and, irrespective of salinity, was reduced while the fish were submerged. T4 helps the fish to cope with terrestrial stress and is otherwise not involved in osmoregulation and the intermediate position of the T4 concen tration in control fish is to be expected. In Boleophthalmus boddarti, T4 was only elevated in fish out of water and how terrestrial adaptation is mediated by T4 in either fish is unknown. T3 appeared to be unaffected by changing salinity in Periophthalmus chrysospilos while in Boleophthalmus boddarti it was higher in 100% sea water and reduced in 30% sea water and in air suggesting that it is of importance in osmoregulation in the latter but not the former species (Lee & Ip, 1987). Consistent with the function of plasma cAMP in stimulating chloride secretion are the findings, in fully submerged fish, that levels increase with increasing salinity in Periophthalmus chrysospilos (30, 50 100% sea water) and in Boleophthalmus boddarti in 100% sea water. The absence of any difference between control cAMP levels where the fish is free to move into and out of 50% sea water and those of B. boddarti submerged in 30% and 50% sea water is explained by the suggestion that these conditions approximate to that of the natural habitat where the fish spends most of its time swimming along the water's edge half submerged (Lee & Ip, 1987). This explanation was also offered for the absence of any change of plasma T4 concentrations in B. boddarti at all sea water concentrations, including 100%. RESPIRATION AND RESPIRATORY SURFACES HISTORICAL ASPECTS The investigation of the mudskippers' respiratory capabilities was initiated by Higson (1889). Based on the observation that mudskippers always kept their tails 530 DAVID A. CLAYTON in water he believed that these fish used caudal respiration. Crude experimental support for this supposition was provided by Haddon (1889). He noted that blood circulation to the tail in Periophthalmus was strong and showed that a fish which had its tail painted with gold size died while control fish did not (Haddon, 1889). These finding were dismissed (Rauther, 1910; Harms, 1929; Hora, 1933; Das, 1935), but did draw attention to the role of cutaneous respiration in these fishes. Vascularisation of the skin epithelium necessary for cutaneous respiration was demonstrated (Rauther, 1910; Harms, 1929; Schottle, 1932) but, contrary to Graham's (1976) assertion, no mudskipper has any specialisation of the caudal region for respiratory purposes. The general question of the partitioning of respiratory activity between the gills and accessory respiratory organs, however, continued to interest researchers. Das's (1933, 1934) work is an example of the methodology adopted in early investigations of air breathing fishes. Like the airbreathers, Pseudapocryptes lanceolatus comes to the surface to swallow air and Das attempted to drown them by keeping them submerged and denying them access to the surface (Das, 1934). Whilst these mudskippers survived considerably longer than other species such as Clarias batrachus and Anabas testudineus, they still died. This was probably due to deoxygenation of the water (Hora, 1935b) although in some conditions, osmotic stress may be responsible (Ogasawara etal., 1991). Nevertheless, the idea persisted that these fish were obligate airbreathers and Willem & Boelaert (1937) used Rauther's (1910) finding of a vascularised diverticulum in the buccopharyngeal cavity of Periophthalmus to suggest that mudskippers utilise both the gills for aquatic and the buccopharyngeal cavity for aerial respiration. Both modes of respiration were considered to operate simultaneously irrespective of whether the fishes were in or out of water. Thus, respiration in immersed P. papilio [ = P. barbarus] involved the presence of a supply of air from the surface. Willem & Boelaert (1937) argued that the lamellar arrangement of Arch I and its clear separation from the second arch supported this suggestion. Furthermore, eye retraction, which was often accompanied by slight bulging of the opercular region, was thought to be a mechanism whereby the the two media, water and air, could be mixed. Eye retraction depressed the ceiling of the buccal cavity compressing the buccopharyngeal air and moving the interface between the air and the water meniscus. The suggested presence of air in immersed fish has not been pursued, but the alternative suggestion, that water remained in the buccopharyngeal space during terrestrial excursions (Stebbins & Kalk, 1961), was seriously considered. They noted that P. sobrinus gulped air on land and, following Willem & Boelaert (1937), discussed ways in which the air and water could be mixed to provide an adequate oxygen supply to the gills. While Gordon et al. (1968) disputed that P. sobrinus held water during terrestrial excursions, MacNae (1968a, b) believed that mudskippers did not need any accessory respiratory organs beyond their gills and dismissed the earlier (Rauther, 1910; Schottle, 1932) histological evidence. Later P. modestus [as P. cantonensis] was reported as maintaining buccal water on land (Gordon et al., 1978), but the cineradiographic study of Sponder & Lauder (1981) showed conclusively that it was not necessary during terrestrial excursions in P. barbarus [as P. koelreuteri]. What the exact role, if any, retained water plays in respiration or feeding (Sponder & Lauder, 1981) remains a moot point but in common with other marine fishes, mudskippers are Table III Comparison of oxygen consumption in ml 02-kg'h_1 of mudskippers immersed in sea water and emersed in air. Only rounded up means are presented, the original references should be consulted for exact data. Bold entries indicate significant differences between oxygen consumption in the two respiratory media. aData calculated from graph in Tamura et al. (1976) as their tabulated values are more likely to be for 02 consumption at 25 °C Oxygen consumption Species Size (g) Temperature (°C) P. modestus 0.5-1.5 20 P. modestus 4.0-8.0 20 Water Air Reference 85 106 167 97 2 Periophthalmus P. P. P. P. P. argentilineatus or kalolo argentilineatus novaeguineaensis chrysospilos chrysospilos Periophthalmodon P. freycineti 1.1-2.5 24 84 94 2.2-9.2 20 65 63 4.6-6.7 25 103 90 Gordon et al., 1978 Tamura et al., 1976a Gordon et al., 1969 Milward, 1974 Milward, 1974 20-25 29 88 48 Natarajan & Rajulu, 1983 6-12 25 378 306 11.6-51.4 20 53 48 Milward, 1974 33-46 20 72 47 5.7-9.7 20 57 50 Tamura et al., 1976 Milward, 1974 Lee et al., 1987 Scartelaos S. histophorus S. histophorus C a C/2 5 *TD *3 00 532 DAVID A. CLAYTON now considered as facultative air breathers, respiring aerially with gills, modified buccal or pharyngeal epithelium and a vascularised skin (Graham, 1976, see also p. 536). Despite their obvious suitability, however, the contribution of studies of mud skippers to an understanding of respiratory adaptation of marine air-breathing fishes has been minimal, accounting for less than 15% of species mentioned in reviews of the topic (Graham, 1976; Bridges, 1988). The respiratory rates of mudskippers confined to either the aquatic or aerial respiratory medium (Table III) shows that they also conform to the generalisation that respiration rates of marine species in water are equal to or greater than the aerial rate (Bridges, 1988). This relationship also holds for Periophthalmus, Boleophthalmus and Scartelaos freely able to select aerial or aquatic respiration (Tamura et al., 1976; Niva et al., 1979, Natarajan & Rajulu, 1983). RESPIRATION IN AIR AND WATER Oxygen consumption in water and in air increases with increasing temperature. For Periophthalmus modestus [as P. cantonensis] (10—30°C; Gordon et al., 1978; 10-15°C, Tamura et al., 1976), P. argentilineatus [as P. vulgaris] and Periophthalmodon freycineti [as Periophthalmodon schlosseri] (20—35°C; Milward, 1974) and Scartelaos histophorus [as Boleophthalmus chinensis] (10—15°C; Tamura et al., 1976) there are no significant differences between the aerial and aquatic rates. By contrast, however, the aerial rate is significantly less than the aquatic rate for Periophthalmus modestus and Scartelaos histophorus in the temperature range 20—35°C (Tamura et al., 1976) and for S. histophorus in the range 15—35 °C where the disparity increased with increasing temperature (Milward, 1974). Gordon etal. (1978) could offer no reasonable explanation for the differences between their result and that of Tamura et al. (1976): the state of thermal acclimation including a seasonal one and the possibility of population differences seem unlikely. In Periophthalmus chrysospilos however, smaller fish have a greater dependence on cutaneous rather than branchial respiration (Low et al., 1990) and the small (0.5-1.5g) P. modestus used by Gordon et al. (1978) could have been aerially more efficient than the larger (4—8 g) fish used by Tamura et al. (1976). The physiological findings for Australian Scartelaos histophorus match ecological reality in that the fish is usually confined to surface pools at low tide (Milward, 1974). A similar explanation could be offered for Japanese S. histophorus, except that when the relative role of the gill and skin are considered, it appears that the gill of Scartelaos has a greater ability for oxygen uptake than that of the more terrestrial Periophthalmus (Tamura et al., 1976). Except for P. chrysospilos, which showed no change (Lee et al, 1987), when mudskippers could choose between respiring in air or water, the total oxygen consumption increased over that recorded in either medium separately. For P. modestus and Scartelaos histophorus both the aquatic and aerial rates were reduced when compared with the single medium values (Tamura et al., 1976). In Periophthalmus koelreuteri, Natarajan & Rajulu (1983) found no difference in the values while for Boleophthalmus boddarti there was an increase in the aquatic rate when the animal was free to choose. The aerial rate in air alone was not calculated (Niva et al., 1979). Without some method of comparing levels under the different conditions, the relative contribution of activity to the MUDSKIPPERS 533 differences observed in oxygen consumption is difficult to determine. As fish move between the respiratory media, oxygen consumption may well be expected to be higher and whereas fish become quiescent under water it is not clear if basal rates were measured in all cases. That oxygen consumption can be extremely variable is demonstrated by the routine metabolic rates in sea water of Periophthalmus sobrinus (Gordon et al, 1978). During the 5-h period of measurement, the fish would remain quiescent, but at irregular intervals would swim for variable periods. The routine metabolic rate varied between 160—430 ml 02*kg'h_1 and for small fish the maximum rate was five times larger than the resting metabolic rate. In air the resting oxygen consumption rate of P. barbarus could be more than tripled during activity (Hillman & Withers, 1987). The bimodal oxygen uptake reported under free choice conditions (Tamura et al., 1976; Niva et al., 1979; Natarajan & Rajulu, 1983) is simply the percent values of total oxygen obtained from the two media. It is invariably greater from water than air and is not to be confused with either cutaneous and branchial respiration or the bimodal respiration of air-breathing freshwater fish where most oxygen is taken aerially and most carbon dioxide released aquatically (Graham, 1976). Indeed, no work on mudskippers has attempted to partition gas exchange in this way and only Tamura et al. (1976), by preventing gill venti lation, have strictly compared the relative contribution of cutaneous and branchial respiration. In reporting on respiratory quotients, only Schottle (1932) provides any information on carbon dioxide release. In air the mean oxygen consumption of P. argentilineatus [as P. vulgaris for a single small specimen and P. dipus for two larger ones] was 151.4 ml 02.kg-l.rT1 and the mean carbon dioxide output 109.5 ml COrkg~1.h~1 with a corresponding RQ of 0.78. For a single Periophthalmodon schlosseri the values were 123.5, 90.23 and 0.72 respectively. When the body behind the gills and the head are separated by a thin rubber septum, oxygen consumption anterior and posterior to the septum can be measured independently. Thus in water Periophthalmus argentilineatus and Scartelaos histophorus obtain about 20—25% and 10—17% respectively from respiratory surfaces behind the head (Milward, 1974). In Periophthalmus sobrinus [ = argentilineatus or kalolo] the equivalent value was slightly less than half, but by assuming that the head represented about a quarter of the total body surface, Teal & Carey (1967) suggest that 60% of oxygen is exchanged through the skin. Unfortunately, their tabulated results cannot be meaningfully interpreted and do not substantiate their conclusions. Total cutaneous respiration in P. modestus and Scartelaos histophorus accounted for 48% ± 12% and 36% ± 18% in water and 76% ± 10% and 43% ± 11% respectively in air (Tamura et al., 1976). In air, gill oxygen uptake was much reduced being only 31% of the 'standard' for uptake in water in Periophthalmus and 58% of that in Scartelaos; aerial skin respiration remained much higher at 96% and 78% respectively. Given that oxygen uptake in both species in air was reduced to about 60% of that in water, it appears that the reduction in Periophthalmus was largely due to a reduction in gill uptake. For Scartelaos, although the gill reduction was the greater, oxygen uptake was also reduced via the skin. Partitioning studies are necessarily restrictive of fish movement, but the experimental conditions in the free choice situation also introduce limitations that have implications for the partitioning of oxygen uptake between water and air. By only providing a small 5-cm diameter semi-circular hole in the centre of the float separating the aquatic and aerial parts of their respirometer, Niva 534 DAVID A. CLAYTON et al. (1978) will have increased aquatic activity and limited aerial respiration to gulping air. Furthermore, cutaneous respiration under such conditions would be reduced. Similarly, in the respirometer used by Tamura et al. (1976), the opportunity for aerial cutaneous respiration would be less than that available to fish in more natural conditions. If only because it is the only known example of diving brachycardia in fishes, the reduction in heart rate of Periophthalmodon freycineti [as P. australis] (Garey, 1962) should be re-examined. The presumed confirmation (Gordon et al, 1969; Graham, 1976) by Bandurski et al. (1968) was no more than the normal brachycardiac response of fish to hypoxia. Indeed, it was an incidental observation made during a preliminary experiment where the heart rate of P. freycineti [as P. australis] in an aerated bag was noted to change from 95-110 beats-min-1 to 65 beats-min-1 after substitution of a nitrogen gas phase (Bandurski et al, 1968). Working with Periophthalmus sobrinus [ = P. argentilineatus or kalolo], Gordon et al.. (1969) were unable to demonstrate any similar brachycardia and the fish actually showed an initial, but slight increase in heart beat frequency when transferred between the two media, irrespective of the direction of transfer. The only finding in common with Garey (1962) was that the heart rate decreased after 3-4min in either water or air. As a result of electrode insertion, however, the fish were dying during recording (Gordon etal., 1969) and even this similarity is suspect. The whole problem will undoubtedly benefit from the improved methodologies now available. RESPONSES TO HYPOXIA Garey (1962) suggested that diving brachycardia in mudskippers was an adaptation to hypoxia in their burrows and reported that the mud in one burrow was almost free of oxygen. In the water in burrows of P. modestus the dissolved oxygen concentration was certainly low, ranging between 0.2 and 0.7 ml 02.1-1 and the inertness of P. modestus in closed jars of sea water when the oxygen level fell to between 0.7 and 1.0 ml 02.1~' lent further support to the idea that mudskippers had a limited tolerance to hypoxia (Gordon et al., 1978). In a similar experimental situation, Niva et al. (1979) found that Boleophthalmus boddarti died when the dissolved oxygen in the water fell to 2.3 to 1.8 ml 02.1_I. However, mudskippers spend long periods in their burrow at high tide and at low tide in inclement weather (Clayton & Vaughan, 1988; Lim & Ip, 1989) and burrows contain an adequate supply of oxygen for eggs to hatch (Kobayashi et al., 1971). Replication of the 'closed jar' experiment with Periophthalmus chrysospilos also produced inert fish when the oxygen level in the sea water fell after 2 h to 0.75 ml 02.1_1 (Chew et al., 1990). Further experiments showed that this response was not, as Gordon et al. (1978) con cluded, due to a limited tolerance for hypoxia. Fish were exposed to varying degrees of environmental hypoxia by substituting part of the oxygen by an equivalent amount of nitrogen. The more usual response of fish to hypoxia is to increase the respiratory rate (Bridges, 1988), but in P. chrysospilos the rate at 1.0 ml 02.1~' of dissolved oxygen was approximately a quarter of that of fish under normoxia and the respiratory rate decreased sharply within 5 min of exposure with the result that at 0.8 ml 02.1_I, P. chrysospilos was able to survive for at least 6 h (Chew et al., 1990). Rather, it is possible that the 'closed jar' fish succumbed to the increased carbon dioxide level and concomitant pH MUDSKIPPERS 535 increase in the sea water (Chew et al, 1990). The possibility that fish built up an oxygen debt during hypoxic exposure was expressed by Teal & Carey (1967) who reported a doubling of respiration in the period following return to normoxic conditions. As the anaerobic end product of carbohydrate metabolism, evidence of high lactate (lactic acid) levels would provide support for this suggestion, but as Teal & Carey (1967) did not measure it, it was left to Bandurski et al. (1968) to demonstrate that in Periophthalmodon freycineti anoxic conditions more than doubled brain and tripled muscle resting values of lactate. During hypoxia, blood lactate levels in Periophthalmus chrysospilos showed a 6-fold increase (control value 1.23 ^mol-ml"1), but the respiratory rate was only slightly increased during recovery in normoxic conditions. This low level of oxygen debt repayment suggested that there must have been a depressionof glycolytic activity and energy consumption during hypoxia (Chew et al., 1990). This was confirmed by Ip et al. (1991a) who compared a variety of compounds including ATP, glycogen and lactate in P. chrysospilos exposed to normoxic (environmental) hypoxia and severe exercise (functional hypoxia). Experimental hypoxia produced no increase in lactate above the normoxic control condition and only when exercised did anaerobic metabolism occur, with ATP and glycogen decreasing and lactate increasing in muscle tissue. This mudskipper's response to hypoxia thus involves slowing of the respiratory rate and a reduced energy consumption. It does not involve glycolysis and previous reports of lactate accumulation (Bandurski etal, 1968; Gordon etal., 1969) are likely to be due to functional hypoxia. After 6 h hypoxic exposure, however, lactate did accumulate in the gills of P. chrysospilos (Ip et al., 1990) and Periophthalmodon schlosseri (Ip & Low, 1990). Lactate was not found to accumulate in branchial tissue of Boleophthalmus boddarti, suggesting that this fish has biochemical adaptations for dealing with hypoxia that are different from those of Periophthalmus and Periophthalmodon (Ip et al., 1990). While lactate accumulation indicates anaerobiosis, neither Periophthalmus chrysospilos nor Boleophthalmus boddarti are anaerobic animals because there was no evidence of the metabolic products of glycolysis in such animals: no branchial succinate or proprionate (Ip et al., 1990) or ethanol (Chew et al., 1990) was found. As suggested earlier for studies of nitrogenous excretion, similar studies on Pseudapocryptes wouldbe informative. If the observation that Periophthalmodon schlosseri [as Pseudapocryptes schlosseri] is also capable of aestivation (becoming torpid for periods during the hottest part of the year; Hora, 1933) can be confirmed, a strong comparative study could be undertaken, especially if the torpor induced during the cold of winter (hibernation response) of Peri ophthalmus (Kobayashi et al., 1971) and Scartelaos [as Boleophthalmus] (Tamura et al, 1976) are included (see also p. 545). OXYGEN TRANSPORT Virtually no work on oxygen transport in the blood has been undertaken in mudskippers. Venkateswarlu (1966) reports that the total amount of iron in the blood of Periophthalmus schlosseri [ = Periophthalmodon schlosseri] and Boleophthalmus boddarti are approximately the same (27—45 mg-100 ml"1 and 25-42 mg-100 ml"1 blood respectively) and greater than that of the fully aquatic goby Glossogobius giuris. Without any indication of the number of individuals or their weight the data are of minimal value as are those on oxygen 536 DAVID A. CLAYTON consumption at an unspecified temperature. For two specimens of Perioph thalmus sp., Pradhan (1961) reported that the amount of haemoglobin (16.5 and 15 g-100 ml"1 blood for 21.2 and 61.5 g fish respectively) was greater than that of a variety of freshwater fishes of equivalent size. Vivekanandan & Pandian (1979) report on the leukocyte (15.2 x 103-mm"3), erythrocyte (3.032 X106mm-3) count and haemoglobin concentration (14.8 ml-100 ml"1 blood) in Boleophthalmus boddarti. The haemoglobin and red blood corpuscle (RBC) counts were significantly correlated. The highest values were found in males, the lowest in juveniles and intermediate levels in females. The same measures were repeated over 48 h for males with free access to water and air (control) and for those restricted to either the aquatic or the aerial medium. The water and air values of the percentage of haemoglobin and RBC counts were reduced in comparison with the control. It should be noted, however, that the fish exposed to air died within 51 h. All of these studies would benefit from more controlled repetition. Similar values for the blood parameters of B. boddarti were reported by Manickam & Natarajan (1985). Variables measured included erythrocyte count, haemoglobin concentration, haemocrit, mean cell volume, mean cell haemoglobin, mean haemoglobin concentration, oxygen capacity and standard bicarbonates, all of which were higher in B. boddarti than in Periophthalmodon schlosseri [as Pseudapocryptes schlosseri] (Manickam & Natarajan, 1985). The authors also report, but without statistical analysis, that (for both species) there were clear-cut sex differences with male values exceeding those of females. Additional data on oxygen dissociation curves and its associated Bohr effect would be instructive. RESPIRATORY SURFACES The main gas exchange surfaces in mudskippers include the gills, bucco pharyngeal and opercular cavity membranes and the skin (Graham, 1976) but the fins (Milward, 1974) and the nasal sac (Rauther, 1910) have also been proposed as suitable structures. Gills Gill structure has been investigated by light (Rauther, 1910; Schottle, 1932; Singh & Munshi, 1969; Milward, 1974; Hughes & Munshi, 1979; Al-Kadhomiy & Hughes, 1988) and electron microscopy (Welsch & Storch, 1976; Hughes & Munshi, 1979; Low et al., 1988; Yadav et al., 1990). The general (Schottle, 1932; Das, 1934) and specific (Niva et al., 1981; Al-Kadhomiy & Hughes, 1988) gill blood circulation and musculature (Schottle, 1932; Willem & Boelaert, 1937; Singh & Munshi, 1969) of some mudskippers has also been described. Morphometric measurements of the gills of a number of species have been taken and are summarised in Table IV. The gill areas of a number of mud skippers were calculated by Schottle (1932) and re-presented by Graham (1976), but are omitted here as only the length and not the weight of the fish were recorded (see also p. 540). As with earlier tabulated summaries (Hughes & AlKadhomiy, 1986; Low et al., 1990), to aid comparison between species, the measurements presented are those calculated for specimens of a weight dictated by Tamura & Moriyama (1976). Theirs was the first major study of mudskipper gill morphometries, but does not include regression slopes that describe gill Table IV Comparison of gill measurements for species of the genera Periophthalmus, Periophthalmodon, Boleophthalmus and Pseudapocryptes Body weight (g) Total filament Total filament Number of Mean bilateral Total number of Gill area number length (mm) secondary secondary secondary (mm2) lamellae on lamellae area lamellae both sides (mm2) Reference Periophthalmus 2 P. modestus 5.3 306 383 47.0 0.036 P. chrysospilos 5.3 242 264 20.35 0.044 P. schlosseri 5.3 244 469 49.8 0.063 P. schlosseri 5.3 307 314 52 0.02 17,977 10,797 660 23,527 23,634 1425 482 Tamura & Moriyama, 1976 Low et al., 1990 Periophthalmodon Boleophthalmus B. pectinirostris 35.2 486 2090 28.4 0.088 B. boddarti 35.2 608 1878 36.6 0.051 B. boddarti 35.2 491 1419 40.9 0.064 B. dussumieri 35.2 3088 23.53 0.054 — 59,356 69,391 58,934 72,672 473 Yadav etal., 1990 Low et al., 1990 3330 Tamura & Moriyama, 1976 3513 3768 Niva et al., 1981 Low et al., 1990 3891 Hughes & Al Kadhomiy, 1986 Pseudapocryptes P. lanceolatus 8.2 518 1222 44.2 0.058 52,650 3302 Yadav & Singh, 1989 c O ^3 g * 538 DAVID A. CLAYTON measurements in relation to fish weight. Consequently Hughes & Al-Kadhomiy (1986) and Low et al. (1990) used their regression equations to produce values for the various gill parameters so they could directly compare their results with those of maximum sized Scartelaos histophorus [as Boleophthalmus chinensis] and Periophthalmus modestus [as P. cantonensis] (53 & 8.8 g respectively) as calculated by Tamura & Moriyama (1976). However, Tamura & Moriyama's data were presumed dimensions, calculated from regressions in body weight for Anabas testudineus and Saccobranchusfossilis. Instead of using such adulterated data, it is more realistic to use the averaged data set values of the gill parameters presented by Tamura & Moriyama (1976). Table IV therefore replicates the comparisons made by Hughes & Al-Kadhomiy (1986) and Low et al. (1990) at the more appropriate (35.2 g & 5.3 g) fish size. The total filament number reported for Tamura & Moriyama (1976) by Low et al. (1990) and in Table IV remains the same because Low et al. (1990) erroneously presented the averaged values (of 35.2 & 5.3 g) and not those of the maximum sized (53 & 8.8 g) fish. Data for small (1—6 cm, 0.75—2.3 g) Periophthalmodon schlosseri from the Andman (Andaman) islands (Yadav et al., 1990) is included in the Table, but it is likely that this is an incorrect designation. P. schlosseri can exceed 20 cm and 100 g and given the small size of the specimens, it is more likely that the fish was Periophthalmodon septemradiatus or Periophthalmus minutus, both of which are from the correct geographical area (Murdy, 1989). Less circumstantial, however, is the evidence of the structure of the gills. In this case, branched filaments seem to be a feature of the gills of Periophthalmodon, being found in P. freycineti (Milward, 1974) and P. schlosseri (Low et al, 1990), but not in the specimens investigated by Yadav et al. (1990). When presented for a hypothetical fish of a given weight, the calculated values for total gill area within boleophthalmids and within periophthalmids are similar (Table IV). The exception is the high value obtained for P. schlosseri which results from the increased filament length and bilateral secondary lamellae area reported for this species (Yadav et al. (1990). Gill morphometric data enabled Hughes & Al-Kadhomiy (1986) to distinguish between genera such that when values for their fish of appropriate weight were compared with that of Tamura & Moriyama (1976), their data on Boleophthalmus dussumieri [as B. boddarti] matched that of Scartelaos histophorus and Boleophthalmus boddarti better than that of Periophthalmus modestus. While they were correct to conclude that their fish was a boleophthalmid and not a periophthalmid, the comparison of gill parameters between species based on values obtained from regression equations should be treated with considerable caution. By way of illustration, Table V provides values of gill areas calculated from regression equations. The Table clearly shows that gill area increases with increasing weight and that generally the gill areas of the boleophthalmids are greater than those of the other genera. This is certainly true among fish of 20 g and greater where, for example, the gill area of a 30-g P. chrysospilos is only 77% that of an equivalent sized Boleophthalmus dussumieri. B. dussumieri also has a gill area of a similar magnitude to other 30-g boleophthalmids. At the mean weight of Periophthalmus chrysospilos (5.3 g), the comparison would lead to the conclusion that Boleophthalmus dussumieri has a gill area closer to that of Periophthalmus chrysospilos and Periophthalmodon schlosseri than the other boleophthalmids. At this size the gill area of Periophthalmus chrysospilos is Table V Gill areas (A) for fishes of different weights (W) calculated from regression equations. Fish weights are given at arbitrary 10-g intervals except that 5.3, 35.2 and 53 g are the averaged and presumed maximum weight individuals from Tamura & Moriyama (1976). aYadav & Singh (1989) measured the gills of 35 fish (1-21 g), but then reduced the data to seven averaged body weights from which the regression equations were calculated. Species Regression Gill equation 5.3 A = 92.654W10496 A = 281.28W°-709 A = 679W04812 B. boddarti Pseudapocryptes lanceolatus A=607.8W0827 Periophthalmus chrysospilos A = 97.6W°'9577 Periophthalmodon schlosseri A=100W09312 area (mm2) calculated for fish of 10 20 Fish used in weight (g;} 30 calculation 35.2 53 N Reference Weight (g) Length 2 (mm) Boleophthalmus dussumieri 533 1039 2150 3290 3891 5979 14 3.6-35.4 40-240 Hughes & Al Khadomiy 1986 B. boddarti 918 1439 2353 3136 3513 4695 9 1-12 55-110 Niva et al., 1981 Low et al., 1990 1515 2056 2870 3487 3768 4588 10 2-35 2376 4081 7240 10124 11555 116209 7 1-21 482 885 1720 2536 2955 4373 9 2-13 473 853 1627 2374 2755 4033 7 3-111 a a 00 _ 75-165 — — Yadav & Singh, 1989a Low et al., 1990 Low et al., 1990 oo 540 DAVID A. CLAYTON 90% that of Boleophthalmus dussumieri. Nevertheless, the results appear to support the generalisation that gills are reduced in more terrestrial species (Graham, 1976). Similarly, in comparing 10-g fish, Low etal. (1990) conclude that the natural preference of B. boddarti for an aquatic environment can be explained by their having more than double the gill area of either Periophthalmus chrysospilos or Periophthalmodon schlosseri. Furthermore, Pseudapocryptes lanceolatus can be considered the most aquatic as the gill area is greater than that of all the other genera (Table V). This conclusion is partially at variance with that of Yadav & Singh (1989). They calculated gill area for fish of a length equivalent to specimens of Periophthalmus vulgaris [ = P. argentilineatus] and Boleophthalmus viridis [ = Scartelaos histophorus] for which Schottle (1932) provided data. On this basis Pseudapocryptes was more aquatic than Perio phthalmus having more than 1.5 times the gill area, but equivalent to Scartelaos with which it had a comparable gill area. These are inappropriate comparisons because species have different length/weight relationships and will have widely different weights at a similar length. As a corollary to this observation, the data in Table V for Periophthalmus chrysospilos and Pseudapocryptes lanceolatus above 20 g have no biological significance because neither of these species grow to such sizes. Similar comparisons could have been presented for other gill variables and would have also shown the size dependent nature of the conclusions drawn. The measurement of gills is a time-consuming operation and so small data sets are to be expected. However, while regression equations are precise descriptions of the relationship between two variables, they can be influenced considerably by one data point, especially in small data sets (Clayton, 1990). As a minimum, data sets need to include the complete size range of the species with measure ments of equal numbers of fish at each size. Niva et al. (1981) only have one fish heavier than 10 g while Low et al. (1990) have only three and Hughes & Al-Kadhomiy (1986) only two that are lighter than 10 g. The measurements of any one of these fish will have a greater influence on the slope of the regression line than that of the other fish. The cause for caution in the interpretation of data derived from regression equations does not imply that size dependence is only a statistical artefact. The growth related changes in the size or form of the respiratory surfaces are of clear physiological relevance. Greater spacing between secondary lamellae is an adaptation to aerial exposure, reducing the collapse of the gill in air (Graham, 1976) and is a characteristic of all mudskippers so far investigated. This is the only parameter for which there is a consistently negative slope value for the regression equation for all species (Niva et al., 1981; Hughes & Al-Kadhomiy, 1986; Low et al., 1990; Yadav et al., 1990) indicating that the frequency of secondary lamellae decreases with increasing body weight. Low et al. (1990) found the decrease was greater for Periophthalmus chrysospilos (slope —0.237) than for Periophthalmodon schlosseri (—0.052) which was similar to the more aquatic Boleophthalmus boddarti (-0.031). Other features of the gill of Periophthalmodon schlosseri, however, exhibited greater terrestrial adaptation. The total filament length and gill area was similar to that of Periophthalmus chrysospilos and reduced in comparison with that of Boleophthalmus boddarti (Low et al., 1990). The branched filaments reduce the risk of collapse and the tissue fusion between secondary lamellae may reduce the risk of dehydration (Schottle, 1932; Low et al., 1988). Other authors report slope values of a similar MUDSKIPPERS 541 magnitude for B. boddarti (-0.083, Niva et al., 1981) and Periophthalmodon schlosseri (-0.048, Yadav et al., 1990), but for Boleophthalmus dussumieri the slope (-0.229, Hughes & Al-Kadhomiy, 1986) is similar to that of Periophthalmus chrysospilos. In other respects Boleophthalmus gills are better adapted to aquatic respiration because the filaments are longer and more of the secondary lamellae are aligned parallel to the respiratory water current (Low et al., 1988). Conversely, the branched filaments of Periophthalmodon (Milward, 1974; Low etal., 1988) and the interlamellar fusion (Schottle, 1932; Low etal., 1988; Yadav et al., 1990) will impede water flow and reduce their gill efficiency in water. Gill area is reduced in species that are more dependent on aerial respiration (Graham, 1976) and is reflected in the low values of the gilhskin area ratio for Periophthalmus (0.27-0.46 for P. argentilineatus [as vulgaris and dipus], P. kalolo [as koelreuteri] and P. chrysospilos (Schottle, 1932; Low et al., 1990) and for Periophthalmodon schlosseri (0.22-0.5; Low et al., 1990) when compared with that of Scartelaos histophorus (0.72; [as Boleophthalmus viridis], Schottle, 1932) and B. boddarti (0.67-0.77; Low et al., 1990). Contrary to the generalisation, however, are the lower values (0.48-0.52) reported for B. boddarti (Niva et al., 1981). Additionally, the ratio for Periophthalmus chrysospilos and Periophthalmodon schlosseri increased with increasing fish size (Low et al, 1990), a finding that argues for the greater amphibiousness of smaller individuals except that the largest ratio is still relatively low. Skin The skin of mudskippers undoubtedly has a respiratory function (Rauther, 1910; Schottle, 1932), but the gill:skin area ratio is a crude method of measuring its contribution to total respiration. The capacity for gaseous exchange is variably located and dependent on adequate skin vascularisation. Beginning with the species with most skin vascularisation, Periophthalmus argentilineatus, the blood circulation is greatest on the skin of the upper and front sides of the head and gill covers, reduced in the tail and body and is virtually absent from the ventral surfaces (Schottle, 1932). In the region of scale pockets, blood vessels in finger-like projections of the dermis (corium) rise to the epidermal layer where they branch umbrella-like into several intra-epithelial capillaries running parallel to the surface. Intra-epithelial capillaries also occur in P. modestus [as P. cantonensis] (Maekawa et al., 1968). In Periophthalmodon schlosseri where the epidermis is thicker, the dermal papillae are histologically more complicated, but also have intra-epithelial branches and capillaries so close to the surface that the epidermis can be seen microscopically to bulge out. Scartelaos histophorus [as Boleophthalmus viridis] has the simplest form of skin respiration that consists of blood vessels in flat dermal papillae that lie over the scales. Whereas the epidermis on the head and trunk has several layers of cells, it is reduced to one layer of very flat turgor cells above the scales, which are confined to the caudal region. More highly differentiated papillae, with presumably better respiratory properties, are found on the outer surface of the operculum, on the dorsal surface of the head and in a narrow zone along the upper trunk to the beginning of the second dorsal fin. The opercular button papillae consist of a single capillary loop towards the surface while elsewhere 542 DAVID A. CLAYTON the skin papillae are larger and contain a network of capillaries. In B. boddarti there are large diameter button-shaped papillae on the dorsal surface of the head and to the side of the start of the first dorsal fin. Additionally, the fish has greatly differentiated true respiratory papillae each associated with a single scale. They are distributed over the upper side of the head up to a line joining the ventral point of attachment of the branchiostegal membrane to the corner of the mouth and on the side of the trunk in a wedge-shaped zone pointing caudo-dorsally up to the middle of the second dorsal fin. These papillae are also found in the skin of the basal section of the ventral fins. Where the blood vessels break through the dermis into the epidermis, the efferent and afferent vessels always run close together and parallel to the scale before bending upwards into the papilla (Schottle, 1932). The possibility that the positioning of these vessels may act as a counter-current heat exchanger is worth exploring. The centrallylocated arteriole branches into a large capillary bed which connects to peripheral ring-like venules. As both vessels are pigmented this arrangement is easily visible and has also been found in B. dussumieri [as B. dentatus] by Salih & Al Jaffery (1980) and Al Kadhomiy & Hughes (1988) [as B. boddarti]. Whereas there is a standard methodology for the measurement of gill and skin area, the other accessory respiratory surfaces are less amenable to quantification and, therefore, seem to have been somewhat neglected in the discussion of the physiology of respiration. Only Niva et al. (1981) measured the opercular chamber area, finding that its rate of growth was less than that of either the gills or the skin. Accessory organs There are many anatomical descriptions of the accessory organs (Rauther, 1910; Harms, 1929; Schottle, 1932; Das 1933, 1934; Willem & Boelaert, 1937, Marlier, 1938; Milward, 1974; Singh & Munshi, 1969; Klika & Tesik, 1980; Singh et al., 1989; Singh et al., 1990; Yadav et al., 1990). In all species of mudskipper the buccal and pharyngeal cavities and the opercular space outside the gill arches are large and capable of considerable extension for retention of air. Additionally, the extent of the opercular chamber is increased in one {Boleophthalmus, Pseudapocryptes and Scartelaos) or two directions {Perioph thalmus and Periophthalmodon). Ventro-anteriorly, a diverticulum extends forwards beneath the branchial arches and the floor of the buccal cavity. This ventral extension (Schottle, 1932) is found in all species and has subsequently been called the infra- (Singh & Munshi, 1969) or inferior branchial recess (Yadav et al, 1990). Dorsally a flat pharyngeal side chamber has a wide con nection to the opercular cavity through the dorsally restricted opening of the first gill slit. The opening is restricted because the gill arch is fused to the pharynx wall by a ventro-medial membrane which is vascularised on the opercular side (Schottle, 1932). The dorsal extension of the opercular chamber is also called the pharyngeal diverticulum (Klika & Tesik, 1980) or the suprabranchial chamber (Singh & Munshi, 1969; Yadav et al., 1990). Neither the dorsal extension nor the restricted opening of the gill arch is present in Boleophthalmus or Scartelaos. Information relating to the extent of vascularisation of these accessory organs can be derived from two sources: the descriptions of the histology of the acces sory organs (Rauther, 1910; Harms, 1929; Schottle, 1932, Singh & Munshi, 1969; Milward, 1974; Klika & Tesik, 1980) and that of the blood vessels MUDSKIPPERS 543 supplying them. The latter have been described in Periophthalmus argentilineatus (Schottle, 1932; Singh & Munshi, 1969), Periophthalmodon schlosseri, Scartelaos histophorus (Schottle, 1932), Boleophthalmus boddarti (Schottle, 1932; Niva et al, 1981) and Pseudapocryptes lanceolatus (Das, 1934; Singh etal., 1990). For all species, the pharyngeal and opercular chamber, including the branchiostegal apparatus which greatly contributes to the inflation capability of the opercular chamber, and the ventral extension are well vascularised as are the dorsal extensions in those species that possess them. Contrary to the generalisation made earlier that an increase in area of more heavily vascularised epithelium is accompanied by a reduction of gill area in increasingly amphibious species (Schottle, 1932), the buccal cavity is vascularised to a variable degree. While Scartelaos histophorus and Boleophthalmus boddarti have fewer intra-epithelial blood vessels than the well vascularised buccal epithelia of Periophthalmus chrysospilos and Periophthalmodon schlosseri, Schottle (1932) found almost none in Periophthalmus argentilineatus and P. modestus [as P. cantonensis]. Singh et al. (1989) found that the development of the vascularised respiratory epithelium of the buccal, opercular and suprabranchial chambers was broadly similar in Periophthalmodon schlosseri, Boleophthalmus boddarti and Pseuda pocryptes lanceolatus. The gill material of the embryonic fourth gill arch gives rise to a separate gill mass on either side of the pharynx which later begins to form the respiratory epithelium of all three chambers. In the suprabranchial chamber, gill lamellae from the dorsal ends of the arches also contribute towards the respiratory epithelium (Singh et al., 1989). In P. lanceolatus and Perioph thalmodon schlosseri the gill tissue transferred from the arches, and particularly the first gill arch, forms a series of longitudinal antero-posteriorly oriented ridges on the inner surface of the operculum (Singh et al., 1990). The ridges are more pronounced in Pseudapocryptes lanceolatus. As the ridges were found to be rich in carbonic anhydrase, Singh et al (1990) consider them to be an organ for the release of carbon dioxide that is independent of oxygen uptake. From the large number of published light and electron micrographs of the respiratory surface (see pp. 536, 542), it is clear that the air/blood barrier is only a few /xm thick, and where it has been measured, variability in width is apparent. The range of skin diffusion distances must be presumed to vary from that in the respiratory papillae to that in the non-vascularised sections of the epithelia. The presence of specialised cells including mucus, chloride and acidophil cells (Hughes & Munshi, 1979; Welsch & Storch, 1976) and mitochondria-rich cells (Hughes & Al Kadhomiy, 1986) all increase the diffusion distance. The diffusion distances in the gills of P. argentilineatus are given as 0.9 pan (Welsch & Storch, 1976) and between 1—3.2 /xm (Singh & Munshi, 1969) while for Boleophthalmus boddarti the mean and range are 1.43 /xm and 0.1—1.56 /xm respectively (Hughes & Munshi, 1979; Niva et al., 1981). For B. dussumieri the diffusion distance of the gill membranes ranged between 4-5 /xm and 10—12 /xm depending on the site of measurement (Hughes & Al Kadhomiy, 1986), but even at its thickest, the distances were generally less than that for skin of the trunk (mean 121 /xm range 4-230 /xm) or snout (50, 5-144 /xm ) or the outer (725, 700-1200 /xm) or inner operculum (9, 4-230 /xm ) (Al Kadhomiy & Hughes, 1988), the last values being the only ones available for any accessory respiratory surface. There are profuse networks of capillaries in the dorsal, pelvic, caudal and particularly the pectoral fins of Australian mudskippers (Milward, 1974). 544 DAVID A. CLAYTON Following Harms (1929) and Schottle (1932), Milward (1974) was unable to find any histological support for any respiratory function of the nasal cavity as suggested by Rauther (1910) after he noted the sac-like enlargement of the olfactory chamber in mudskippers. Datta & Das (1980) revealed no respiratory function in their detailed anatomical study of the olfactory apparatus of Periophthalmus koelreuteri [?= P. novemradiatus], Boleophthalmus boddarti and Pseudapocryptes lanceolatus. TEMPERATURE REGULATION The range of temperature to which intertidal amphibious fishes are exposed are among the greatest met by any fish. Within wide tolerance limits, mudskippers are extremely eurythermal using body colour changes (Stebbins & Kalk, 1961), evaporative cooling and behavioural thermoregulation (Tytler & Vaughan, 1983). Generally, body temperatures match those of the mud or surface water from which the fish were caught (Stebbins & Kalk, 1961; Burhanuddin & Martosewojo, 1979; Tytler & Vaughan, 1983) although Gordon et al (1968) found no correlation between substratum and body temperature. Upper tolerance limits have been measured in a variety of ways. In full sunlight at shade tempera ture of 43-44°C (Gordon etal, 1969), Periophthalmus sobrinus died within 50 min when their body temperatures were 33-35°C (Gordon et al., 1968, 1969). In water P. sobrinus stopped breathing at a mean temperature of 41.9°C (n = 5, range 40.4—42.3), but subsequently recovered when placed in water at 31°C (Stebbins & Kalk, 1961). They did not note the time of exposure, but in water at 38-39°C, P. waltoni survived for 9 min before becoming comatose (Al-Naqi, 1977). Survival time decreased with increasing temperature until at 43 °C it was only 48 s. All fish subsequently recovered at lower temperatures (Al-Naqi, 1977). Lower limits are equally vague, P. sobrinus was immobilised at 10°C and recovered at 20°C (Stebbins & Kalk, 1961) and opercular movements, but not respiration, of P. modestus and Scartelaos histophorus ceased at 5°C and began again at 10°C (Tamura et al., 1976). At 4°C Periophthalmus waltoni became comatose after 77 s and at -1.5°C coma ensured after 26 s. Fish were resilient in that after 95 min exposure at 4°C and 0°C, recovery at room temperature took 97 and 215 min respectively; 24 min at —1.5°C required 58 min recovery time (Al-Naqi, 1977). The burrow is an important temperature refuge especially during periods of high surface temperature, when differences can be as high as 11°C (Tytler & Vaughan, 1983). Fish avoid high surface water temperatures (37°C, Stebbins & Kalk, 1961) and under these conditions it would be expected that the duration of terrestrial excursions would decrease and refuging increase, but no con vincing data are available. At the other extreme when surface temperatures are below that of the burrow, Periophthalmus waltoni and Boleophthalmus dussumieri in Kuwait (29 °N) only emerge when the temperature rises to about 10-11 °C, equalling that of the burrow (Tytler & Vaughan, 1983). Other mech anisms of behavioural thermoregulation include basking in the shallower, warmer edges of standing pools in cold weather and following the advancing tide or climbing onto shaded structures that project from the tidal flat in the splash zone when the weather is hot (Tytler & Vaughan, 1983). Rolling onto the side is also assumed to be involved in thermoregulation (Stebbins & Kalk; 1961, MUDSKIPPERS 545 Khoo, 1966; Clayton & Vaughan, 1988) and Boleophthalmus engage in pectoral fin flapping in which the well spread fin is first moved anteriorly and applied to the opercular region and then placed posteriorly against the flank (Clayton & Vaughan, 1988). This action may also serve to keep the skin or fin rays moist, but Stebbins & Kalk (1961) thought that the similar action in Periophthalmus helped to mix the air and water retained in the opercular chambers. The skin and fins could be moistened for both evaporative cooling and cutaneous respiratory purposes, and the former possibility is supported by the observation that B. boddarti roll onto the side at which a jet of air is directed (Ip et al., 1991b). So far only relatively short term responses to environmental temperatures have been discussed, but mudskippers also exhibit long term responses to seasonal fluctuations in temperature. These responses have been little studied, but are more likely to be physiological than behavioural. In tropical regions (23 °N) Pseudapocryptes and Periophthalmodon probably aestivate, retreating into their burrows for extended periods during the hottest part of the year (Hora, 1933, 1935a). Further north in Japan (33 °N), deep burrow temperature in the cold season dropped to 5—5.6°C and Periophthalmus modestus remained in their burrows from November to March, only emerging when the air temperature reached 18°C (Kobayashi et al., 1971). For Scartelaos histophorus air emergence temperature was 22°C and was not reached until May (Tamura et al., 1976). Understanding how these species cope with the low burrow temperature for such long periods would be interesting. ECOLOGY HABITAT AND DISTRIBUTION The distribution of mudskippers is commonly assumed to be based on the presence of a suitable muddy substratum (MacNae, 1968a; Quereshi & Bano, 1971) but they are also found on other substrata. At Laboean (= Labuhan) and Popole Island (=Popau Popole, 6°24'S:105°48'E) along the west coast of Java, Periophthalmus kalolo [as P. harmsi and P. koelreuteri] were found on (sea?) grass surfaces and on sandy-rocky shores (Harms, 1929; Eggert, 1935) and at Perbaoegan (= Perbaungan, 2°03'N:99°58'E) P. chrysospilos "only occurs on the sandy beach despite the presence of a large muddy area not far away" (Eggert, 1935). P. argentilineatus [as P. kalolo] inhabits rock strewn shores in the New Hebrides and Polynesia (Vanuato region, 16°S:168°E) (McCulloch & Ogilby, 1919). Similarly, Gordon etal. (1968) found P. sobrinus [ = P. argentilineatus or kalolo] on cobblestone beaches and on the rocky coastline of Nosy Be (13°24'S:48°17'E) off the west coast of Madagascar. Mangroves and intertidal mud and sand flats were equally populated both here (Gordon et al., 1968) and at Inhaca Island (26°01'S:32°58'E) off the east coast of Africa (Stebbins and Kalk, 1961). P. modestus [as P. cantonensis] was found on sand, gravel and mud substrata in the Gum river in Korea (36°N:126°43'E) (Ryu & Lee, 1979). Despite such observations and the certainty of Eggert's statement, MacNae is probably correct in suggesting that such records were of wandering immature individuals. This was true of mudskippers he collected 546 DAVID A. CLAYTON from a rocky promontory adjacent to mangrove areas (MacNae, 1968b). Distribution according to size was reported by Ryu & Lee (1979) and additional data on the size and maturity of mudskippers taken from non-muddy areas would help resolve the issue. MacNae's explanation may also account for the presence of Boleophthalmus sp. [ = boddarti or dussumieri] on the rocky littoral of Port Okha (22°28'N: 69°05'E) where they were found on the Okha and Balapur reefs but not the ones at Dwarka, Adatra and Hanuman. The mudskippers were in the midlittoral zone of both reefs, but only at Balapur was the environment muddy (Gopalakhrishnan, 1970). Distribution is also assumed to be based on a species' ability to live out of water (Hora, 1935a; MacNae, 1968a, b; Gunderman & Popper, 1984). Species of Apocrytes and Scartelaos are considered to live in very soft mud around mean sea level while Boleophthalmus spp. are found higher on the shore. The distri bution of Periophthalmus and Periophthalmodon spp. are at the level of the highest tides and beyond (Hora, 1935a; MacNae, 1968a; Berry, 1972). Pseuda pocryptes is considered as a semi-aquatic form and in the Ganges Delta is con fined to spring tide pools. Of all the oxudercine genera, however, Pseudapocryptes is best adapted to withstand drought, becoming torpid at the bottom of its 2 m deep burrow when the pools dry up (Hora, 1935a). Thus on the open mudflats of Kuwait (29°20'N:47°50'E) Periophthalmus waltoni [as P. koelreuteri] is found in a zone extending from above mean high high water (MHHW) to below mean low high water (MLHW). This zone overlaps that of Boleophthalmus dussumieri [as B. boddarti] which extends from above MLHW to mean sea level (MSL). Scartelaos tenius [as S. viridis] is found only in areas above mean high low water (MHLW) where standing water remains on the surface at low tide. At the upper limits of its distribution S. tenius overlaps into the area occupied by Boleophthalmus dussumieri (Clayton, 1986). Apocrytes madurensis is also commonly found in this zone (Clayton & Abo Seedo, 1986). As in all reports of mudskipper distributions, however, this zonation is based on qualitative data. A more quantitative assessment of the distribution of Periophthalmus waltoni and Boleophthalmus dussumieri on the same mudflats, used the number of burrows as an indicator of fish density (Al Taher, 1990). The problems of density estimates based on burrow counts are well known (cf. crab density: Warren, 1990) and equally apply to the study of mudskippers. In this study, however, each species' burrow was easily identified and it was known that a foraging P. waltoni commonly occupied a home range of 4-5m2 in which the fish used several burrows (Clayton & Snowdon, unpubl. data). Al Taher (1990) was able to confirm the relative positions of the species on the shore and additionally demonstrated a seasonal variation in distribution. Between May and September the mean number of Periophthalmus waltoni burrows along two transects was 4.15-25m-2 but increased to 8.55-25m~2 between November and March: the increase occurring at MHHW and mean high tide level (MHTL). As fewer burrows were found at MLHW in this latter period, it is probable that there was a seasonal migration of fish up the shore during the cooler months when longer periods of emersion would be less stressful to the fish. On other mudflats, this simple overlapping zonation is confounded in a number of ways. The most obvious of these is the absence of a generic rep resentative from the series. Periophthalmodon is only found to the east of the northeastern coast of India (Murdy, 1989) although P. septemradiatus [as P. MUDSKIPPERS 547 tridecemradiatus] is also reported from Korangi Creek (24°48'N:67°08'E) on the coast of Pakistan (Qureshi & Bano, 1971; Hoda & Akthar, 1983; Hoda, 1985) as well as from Dubla Island (= Dubla shoal, 21°44'N:89°32'E) in Bangladesh (Qureshi & Bano, 1971) Nevertheless, the genus is absent from Kuwait mudflats. The Arabian Gulf also marks the most westward extent of the distribution of Boleophthalmus and Scartelaos and only Periophthalmus {argentilineatus, kalolo, barbarus) are found to the west. Whether the absence of one or more genera from a particular locality means that the remaining genera extend their niche is one area for future study. In the mixed species {Avicennia, Bruguira, Rhizophora) mangrove forests of the Morrumbene (23°45'S:35°20'E) (Day, 1974) and Mngazane (31°42'S: 29°25'E) (Branch & Grindley, 1979) only Periophthalmus argentilineatus [as P. sobrinus] is found. In terms of indicating the zonation of this mudskipper within the estuaries, the data are inconclusive. The only indication that P. argentilineatus may occur extensively, both within mangroves and out on the open mudflats comes from its absence from a list of 11 species (1 annelid, 3 molluscs, 7 crustaceans) which were only found among mangroves (Day, 1974). The extent of the niche of P. argentilineatus [as P. koelreuteri] in the Avicennia mangroves of Melita Bay (15°16'N:39°49'E) in the Red Sea is similarly unclear (Fishelson, 1971). In mangrove forests at Ku Yao Yai (8°N:98°30'E) in southern Thailand Periophthalmus koelreuteri [ = ?P. kalolo] is confined to the mangrove and Scartelaos histophorus [as Periophthalmus viridis] the open mudflat (Natee wathana & Tantichodok, 1984). Similarly in those of An Nam Bor on Phuket Island (7°51'N:98°25/E) where P. argentilineatus [as P. vulgaris] is found with Scartelaos histophorus [as S. viridis], the latter species occupies the open mudflat and the former is found in the Rhizophora and Sonnertia mangrove forest (Frith et al., 1976). At Koh Surin Nua on Surin Island (9°25'N: 98°50'E), Periophthalmus argentilineatus [as P. vulgaris] is also found in the Bruguira and Rhizophora mangrove forests, but Periophthalmus koelreuteri [ = IP. kalolo] is the species present on the mudflat. (Frith, 1977). On a more localised scale, the ecological correlates of the presence or absence of a species would be instructive. Why for example is Boleophthalmus present in the mangrove forests of Ao Nam Bor but absent from Kih Surin Nua? It is unlikely that competitive exclusion by Scartelaos koelreuteri at the latter location is responsible for its absence from the mudflat and other explanations need to be investigated. Similarly, as Periophthalmodon is absent from these locations leaving a vacant niche in the landward edge of the mangrove, why is Periophthalmus argentilineatus only found in Ao Nam Bor (Frith et al., 1976), but not at Koh Surin Nao (Frith, 1977)? Confusion over classification of equivalent mangrove zones is unlikely as both studies were carried out by Frith. Nursall (1981) attributed the disappearance of P. novaeguineaensis [as P. expeditionium] and increased occurrence of Periophthalmodon freycineti [as P. schlosseri] in mangroves near Townsville (19°13'S:146°48'E) between 1972 and 1976 to topographical changes. The former species preferred to inhabit steep banks of the tidal channel. As siltation occurred, these became shallow, sluggish, low-sided creeks that were favoured by the latter species. The mangroves are another confounding variable in the simple picture of mudskipper zonation because they provide a considerably increased range of habitats. Nevertheless, the generalised picture of zonation of mudskippers seems 548 DAVID A. CLAYTON to hold within Malaysian mangroves (Berry, 1972). Here, Scartelaos and Boleophthalmus are found in the mangrove pioneer and foreshore zones and Periophthalmus and Periophthalmodon occupy the mangrove forest and its terrestrial edge. However, while Periophthalmodon is the only fish to be found in the driest zone, which is subjected to tides on only 2—3 days a year, it is also found in the pioneer and foreshore mudflat zones where the residents experience daily tidal immersion. A similar generic association is recorded from Port Swettenham [Pelabohan Kelang, 2°57'N:101°24'E) where Periophthalmodon also occurs with Boleophthalmus in the streams and channels that create greater topographical variation and tidal penetration of the mangrove (Sasekumar, 1974). ASSOCIATIONS In Japan, the estuarine Apocryptodonpunctatus [as A. bleekeri] lives in alpheid shrimp burrows (Dotsu, 1961). In Malaysian mangroves the association of Scartelaos viridis and the crab Macrophthalmus latreilli and Boleophthalmus boddarti with Metaplax crenulatus, Uca coarctata and U. dussumieri was first reported by MacNae (1968a) and subsequently by Berry (1972) and Chapman (1976). The complex faunal zonation of mangrove and mudflat inhabitants and the limited geographical distribution of individual species mean that mudskipper faunal associations are difficult to construct with any degree of accuracy and the absence of Macrophthalmus latreilli and Uca coarctata from faunal lists that include mudskippers (Sasekumar, 1974; Frith et al., 1976; Frith, 1977; Nateewathana & Tantichodok, 1984) advise caution in attempting such generali sations. Additionally, while Metaplax crenulatus is found in association with Boleophthalmus it also occurs with Periophthalmodon (Berry, 1972; Sasekumar, 1974) and Scartelaos (Frith et al., 1976). Nevertheless, in the same way that a specific mangrove fauna is a point for discussion, there may be some valid associations between mudskippers and other faunal elements. Furthermore, since fish distributions are given only in terms of their presence or absence from a particular transect or station, it is not possible to judge fully the level of association. At present the best that can be attempted is to point out some generic rather than specific associations. The latter can be explored during additional detailed survey work. While associations probably involve all macrofaunal groups, including polychaetes and molluscs, existing observations are restricted to the burrowing surface-active crustaceans. The general association of mudskippers and decapod crustaceans, particularly grapsids and ocypodids has long been recognised, but the xanthids Erycarcinus natalensis and E. orientalis, the synalpheid Alpheus sp. (usually A. eurythrosine or A. crassimanus), the pagurids Clibanarius padavensis and Diogenes avarus and the callianasids Thalassina anomola and Upogebia sp. (rarely identified) also need to be included (Table VI). Scartelaos has the fewest associations, but this simply may reflect limited sampling at lower tidal levels. Nevertheless, an investigation of the joint distribution of mudskippers and the crustacean genera {Clibanarius, Alpheus, Uca, Macrophthalmus, Metaplax) that occur with them could provide greater insight into habitat selection and utilisation by the fishes because, in the case of Macrophthalmus and Uca, much more information is available on the ecological correlates of their distribution (cf. Jones, 1984). MUDSKIPPERS 549 Table VI Associations between mudskipper genera and burrowing crustaceans. Species names are only used if they are given by all authors. Data drawn from Gopalakrishnan, 1970; Berry, 1972; Day, 1974; Sasekumar, 1974; Frith etal., 1976; Frith, 1977; Dorjes, 1978; Branch & Grindley, 1979; Nandi & Choudhury, 1983; Nateewathana & Tantichodok, 1984. Periophthalmodon Periophthalmus Thalassina anomala X X X Upogebia sp. X X X Boleophthalmus Scartelaos Callianassidae Paguridae Clibanarius padavensis Diogenes avarus X X X X X X X X Synalphidae Alpheus spp. X X X X X X X X X Xanthidae Eurycarcinus sp. Ocypodidae Uca rosea X U. lactea annulipes X U. vocans X X U. mani X X X U. dussumieri X X X X U. triangularis Tylodiplax tetratylophora Ilyoplax spp. Macrophthalmus spp. Grapsidae Chiromanthes spp. Metaplax spp. Sesarma spp. X X X X X X X X X X X X X X X X X X X X X PREDATORS One common explanation for the amphibious mode of life adopted by mud skippers has been that it has enabled them to exploit new food resources (Inger, 1952) and to escape marine competitors or predators (Pearse, 1929, 1933; Stebbins & Kalk, 1961). At high tide, however, mudskippers are available to such predators. Boleophthalmus tenuis [ = Scartelaos tenuis] are taken by Hydrophis cyanocinctus, blue banded sea snakes (Volsloe, 1939) and the catfish Arius sagar (Sasekumar et al, 1984). Stonefish, Leptosyanceja melanostigma and possibly catsharks, Chiloscyllium griseus are predators of Boleophthalmus boddarti [ = B. dussumieri] (Clayton & Vaughan, 1988). In captivity Perioph thalmodon schlosseri eat Scartelaos viridis (Khoo, 1966), and although this may be a laboratory artefact, the possibility that mudskippers eat other fish is sup ported by field observations of the predation of Periophthalmus barbarus by the same species of Periophthalmodon (Lim, 1971). Furthermore, fish remains in the stomachs of Periophthalmodon schlosseri [ = P. freycineti] were tentatively identified as of periophthalmid origin (Milward, 1974). Table VII Heron, bittern and egret predators of mudskippers in the Sundarban region of the Ganges delta (based on Mukherjee, 197la, b) Predator species Smaller egret No. birds % diet Total examined as fish species fish eaten 95 16 220 Periophthalmus Boleophthalmus N Rank N Rank 1292 1 225 N 67 Egretta intermedia intermedia Eastern large egret 70 80 24 1166 1 109 8 41 Egretta alba modesta Little egret 138 66 25 1073 2 95 86 Egretta garzetta garzetta Purple heron Ardea purpurea manilensis 70 Indian grey heron 76 57 32 963 3 81 76 Other fishes Rank N 2 202 3 183 Family Cyprinodontidae Bagridae 2 106 2 106 Mugilidae Cyprinodontidae 1 212 Bagridae 3 86 Ambassidae Bagridae Anguillidae 1 300 2 198 1 210 3 176 2 19 •3 11 Cyprinidae Bagridae •2 10 Channidae •2 10 Mastocembelidae 18 Bagridae a > < 5 > n r > H 34 24 1303 2 203 Ardea cinerea rectirostris Little green heron 26 29 10 108 1 55 Butorides striatus chloriceps Chestnut bittern 8 19 7 80 1 40 Ixobrychus cinnamomeus Night heron Nycticorax n. nycticorax 78 10 7 70 1 32 147 Bagridae Cyprinodontidae O MUDSKIPPERS 551 In an extensive survey of the stomach contents of the water birds of the Sundarban region of the Ganges delta (Mukherjee 1971a, b), mudskippers were the commonest fishes taken by herons, bitterns and egrets (Table VII). Of the 61 prey species of fishes only the bagrid Mystus guilo, the anguillid Anguilla bengalensis and the cyprinodonts Oryzias melastigmus and Aplocheilus panchax were as numerically significant as Periophthalmus in the diet of these birds. Only Periophthalmus, however, was consumed by all the birds and indeed the three least piscivorous species (little green heron, chestnut bittern and night heron) can almost be considered as mudskipper specialists because mudskippers constituted such a large percentage of the piscivorous portion of their diet. The smaller egret additionally takes Scartelaos histophorus [as Boleophthalmus viridis] (Mukherjee, 1971b). On Singapore Island Halcyon chloris humii, the white-collared kingfisher is largely a bird of the mangrove belt and associated mudflats and takes crabs and Periophthalmus. As an occasional visitor to the intertidal mudflats Alcedo atthis bengalensis, the Indian common kingfisher, occasionally takes small Periophthalmus schlosseri [? = Periophthalmodon schlosseri] (Gibson-Hill, 1948). Mudskippers are also taken by gulls and terns (Clayton & Vaughan, 1986). Especially in mangrove areas, there are a number of reptiles that seem to take crabs and other invertebrates in their prey (Berry, 1972) and which are likely predators of mudskippers. Morton & Morton (1983) suggest that Bennett's water snake, the viper Enhydris bennetti, from Hong Kong may include mud skippers in its diet and in Madagascar the skink Cryptoblepharus boutoni takes small Periophthalmus koelreuteri [ = P. kalolo or argentilineatus], grapsid crabs and insects in its daily foraging forays into the intertidal zone (Fricke, 1970). Boleophthalmus boddarti is second only to the swimming crab as a preferred food of the Ganges estuarine crocodile Crocodylus porosus, at least in captivity (Banerjee et al, 1989). In Malaysia, snakes (Eggert, 1935) monitor lizards and fishing hawks take mudskippers (Johnstone, 1903). On the Solomon Islands, Slooff & Marks (1965) caught female mosquitoes Ades (Geoskusea) longiforceps that were gorged on the blood of Periophthalmus sp. [ = P.kalolo or argentilineatus]. PARASITES Parasitologically, mudskippers are somewhat neglected, but in most investigations reported to date, new species of parasite have been described. At the general level, larval ascarid nematodes infect the guts of Periophthalmodon schlosseri (Khoo, 1966) and cestode, agamofilarial and acanthocephalid cysts have been found in Periophthalmus koelreuteri (Pearse, 1933), Periophthalmodon schlosseri and Boleophthalmus boddarti (Khoo, 1966). Unusually all these three species were found to be infested with Gnathia sp., a parasitic copepod of the gills (Pearse, 1933) but heavy infestation of parasitic copepods in the gills and buccal cavity of Pseudapocryptes lanceolatus were also present (Das, 1934). In Periophthalmus cantonensis [ = P. modestus] stomach spirurioids and duodenal larval cestodes were also noted by Pearse (1932). In Boleophthalmus boddarti and B. viridis [ = Scartelaos histophorus] no external infections on the skin or gills could be found by Choudhury & Nandi (1973). However, the former, but not the latter, was found to contain protozoan parasites. In the gut and gall bladder two species of myxosporidian sporozoans were identified and a third 552 DAVID A. CLAYTON species was recorded for the first time from an estuarine as opposed to a fresh water fish (Choudhury & Nandi, 1973). Pearse (1933) also noted the presence of intestinal flagellates in Boleophthalmus boddarti and Morii & Kasama (1989) have reported on the bacterial flora in the digestive tracts of B. pectinirostris and Periophthalmus cantonensis [ = P. modestus]. Only some of the bacteria found in the mud substratum on which the fish live were found in the digestive tracts, Vibrio or coryneforms being the commonest. It was suggested that osmo regulatory sea water is the main factor controlling the intestinal flora of the fish (Morii & Kasama, 1989). From P. barbarus in west Africa a new species of acanthocephalan parasite (Troncy & Vassiliades, 1974 [as P. papilio] and a hemiurid trematode (Fischthal & Thomas, 1971 [as P. koelreuteri]) were described. In the mesentries of Periophthalmodon schlosseri from Singapore, Khan & Ip (1988) also found and described a new acanthocephalan species. POLLUTION STUDIES Benthic organisms are useful in monitoring heavy metal pollution as they often concentrate potential pollutants in their tissues. In two marine areas surrounded by heavy industrialisation, the Ariake Sea in Japan and Bombay Harbour in India, the tissues of Boleophthalmus pectinirostris (Uchida et al., 1971) and B. boddarti (Patel et al., 1985) respectively, have been assayed for heavy metal levels. The Indian study was more comprehensive, but of the elements common to both studies, iron accumulation was 10-15 times greater than that of zinc. Cadmium appeared not to be accumulated to any great extent by these fishes. In the Indian study mercury was not sampled, but the oversight was corrected by Patel & Chandry (1988) and Mahajan & Srinivasan (1988), the latter authors finding greater accumulation of mercury (2.6 /xg-g"1 dry weight) in B. boddarti in the 4 months following the monsoon than in the 4 months before it (2.36 fxg-g~x). This was also true of sediment values and was accounted for by an input of freshwater containing industrial effluent into the sea. The effects of mercury and fluoride poisoning on mudskippers have been investigated through experimental exposure of the fishes to the pollutants. In B. dussumieri [as B. dentatus] from the west coast of India, mercury had an inhibitory effect on intestinal (Lakshmi et al., 1991) and gill ATPases (Lakshmi et al., 1990). Mercury pollution adjacent to industrial areas was chronic because its concen tration was more than twice that required to kill the fish (Lakshmi et al., 1991). Also from the Gujarat coast of western India, B. dussumieriexposed to sublethal doses of sodium fluoride exhibited cell membrane rupture, degeneration and vacuolarisation of the liver (Shaikh & Hiradhar, 1987) and inhibition of liver and muscle acid and alkaline phosphatase activity (Shaikh & Hiradhar, 1988). B. boddarti from Bombay Harbour (Bangera & Patel, 1984) and Perioph thalmus schlosseri [ = Periophthalmodon schlosseri] {Paid et al., 1975) have been used to monitor levels of radionuclides. Tissue concentrations of the elements recorded in all these studies on pollution monitoring were greater than that found in the water column, but were often considerably less than that found in the sediments. FEEDING The behaviour of Periophthalmus and Periophthalmodon when catching decapod crustaceans consists of a swift tail-powered lunge (Khoo, 1966; Lim, 1971), MUDSKIPPERS 553 after which the fish may or may not return to the water (Stebbins & Kalk, 1961; Gordon et al., 1969; Sponder & Lauder, 1981), the latter issue being of interest to respiratory physiologists (see p. 530). Appetitive behaviour is variable. In Periophthalmus waltoni waiting, stalking and fast approaches are all used, but, successful or not, they avoid returning to the attack site for a long time, possibly because prey in the area do not re-surface immediately the fish has moved away (Clayton & Snowden, unpubl.). Boleophthalmids graze on the surface sediment with distinctive side to side movements of the anterior body and skim off the algae and diatoms that are their predominant food items (Khoo, 1966; Lim, 1967; Clayton & Vaughan, 1988). Feeding occupies two-thirds of the time for which B. dussumieri is active, choosing areas of its territory where the diatom densities are high (Clayton & Wright, 1989). Scartelaids also feed by scraping the top layer of the sediment (Khoo, 1966) but, additionally, may arch slightly into the air before biting into the substratum (pers. obs.). Whether these lunges are directed at prey move ments is not known. Generally, mudskippers of the genera Periophthalmus and Periophthalmodon are considered as carnivores while those of Boleophthalmus are herbivores and Scartelaos are omnivores. There is no information on the other genera although the structure of their dentition, including that of their pharyngeal plates (see Sponder & Lauder (1981) for SEM of plate of P. barbarus), should clarify the issue. While Apocrytodon madurensis is very small (40-50 mm), its pharyngeal plates (and the manner of feeding) are very similar to that of Boleophthalmus dussumieri (Clayton, unpubl. data). Relative gut lengths for the different genera are consistent with the generalisation (Pearse, 1932; Table VIII). Differences in the way in which the ratios were calculated prevent cross-study comparisons, but within studies (Table VIII) the carnivorous species have the shortest, omnivores intermediate and herbivores the longest intestinal length. Two other species are herbivores: Apocryptes bato with a ratio of 2-0 (Hora, 1935a) and Pseudapocrytes lanceolatus [as Apocryptes lanceolatus] for which recorded values are 3-43 (Pearse, 1932) and 2-7 (Hora, 1935a). Table VIII Relative gut lengths for genera of mudskippers Periophthalmus Periophthalmodon Boleophthalmus Scartelaos 0.5-0.6 1.09 2.55 1.266 0.76 0.64 _ 0.39 1.0 0.28-0.45 — 0.43 — 1.45 0.6 — 0.8 — — 0.42 — — — — — — 2.0 0.6 Reference Khoo, 1966 Pearse, 1932 Pearse, 1933 Hora, 1935a Lim, 1967 Lim, 1971 Milward, 1974 The detailed study of the diet, however, reveals that all the fish are omni vorous to some extent and that there are dietary differences between juveniles and adults. Mutsaddi & Bai (1969a) examined the stomach contents of over 1000 specimens of Boleophthalmus boddarti collected from the Bombay coast at monthly intervals over a year. The fish ingest algae, diatoms, polychaetes, nematodes, crustaceans and teleost eggs. Monthly percent occurrence values 554 DAVID A. CLAYTON showed that crustaceans formed the dominant food for juvenile fish (total n for 1 year = 330, 16—95 mm SL, no gonadal development), while for adults (n = 680, 96-185 mm) plant material, particularly diatoms, were dominant. There was also considerable seasonal variation in their diet. Crustacean larvae were common in January and February but were superseded by copepods in March and April. Nematodes were present in the diet mainly in July to October and diatoms were only a significant component from July to December. There was similar variation in the adults' diet; teleost eggs only appearing in July and nematodes predominantly in September to December. These variations in diet probably reflect the seasonal availability of the prey items rather than any selective feeding process. This was demonstrated for diatoms where there was a good correlation between the percent occurrence in the stomach and their availability in the mud (Mutsaddi & Bai, 1969a). Prince Jayaseelan & Krishnamurthy (1980) record that Boleophthalmus sp. [ = dussumieri] from the Pichavaran mangroves (11°29'N:20°49'E) is a herbivore, but later (Krishnamurthy et al., 1984) mentioned that it ingested nematodes and missed a good opportunity to verify if the seasonal variations of nematode species abundance was reflected in the mudskippers' diet. Based on the percentage of empty stomachs, juveniles apparently have a lower feeding intensity than adults (Mutsaddi & Bai, 1969a). This result contrasts with that of Sarker et al. (1980) on 159 Pseudapocryptes dentatus [ = Boleophthalmus dussumieri] from the Shatt Al Arab in the northern Arabian Gulf. They found that food consumption, as measured by weight of stomach contents per gram body weight, was least in the largest fish (160—209 mm total length). The collection was made on a single day (26 November 1976) and no data on seasonal variation are available. This may partially explain why Sarker et al (1980) failed to find any animal material in the fishes' diet. As strict herbivores only benthic diatoms were taken including, in decreasing order of importance in the diet, the bacillariophytes Pleurosigma, Navicula, Nitzschia, and Synedra, the chlorophyte Closteriopsis, and the cyanophyte Oscillatoria (Sarker et al, 1980). Using the same collection offish, Pankow & Hug (1979) provided taxonomic information on the ingested diatoms and identified 114 species, of which Nitzschia (23 spp.) and Navicula (15 spp.) were the dominant genera. Diatoms of the genera Nitzschia, Navicula and Oscillatoria were also predominant in the diet of Boleophthalmus boddarti in Malaysia (Lim, 1967). In addition algae, fungal material and some nematodes were eaten. In contrast to the findings of Mutsaddi & Bai (1969a), Khoo (1966) found that 80% of the diet of 32-41 mm juveniles consisted of diatoms; in 20-40 mm SL juveniles Lim (1967) found the diet to be strictly bacillariophylic. The similarities and differences between the diets of Boleophthalmus boddarti and Scartelaos viridis [ = S. histophorus] are well illustrated by Khoo (1966). In both species the diatoms Oscillatoria sp., Pleurosigma spp., Nitzschia sp. and the cyanobacta Lyngbya sp. are important components of the diet. Both also take harpacticoid copepods, chromadoroid molluscs and nematodes but these form a significant part of the diet only in Scartelaos (Khoo, 1966). The diet of Scartelaos histophorus from Australia (Milward, 1974) is also clearly that of an omnivore. Diatoms and nematodes were the major food items in all sizes offish (19-112 mm SL) while ostracods and harpacticoid copepods may be less important in the larger fish (66-112 mm). Algal filaments, dipteran larvae, kinorhynchs and crustacean nauplii were also included in the diet (Milward, 1974). MUDSKIPPERS 555 There are differences in diet between juveniles and adults. In a study of the combined age and locality differences in the feeding of P. barbarus [now Periophthalmodon schlosseri] (Lim, 1967), those individuals inhabiting a brackish water stream habitat took Formicidae, Coleoptera, dipteran nymphs and molluscs, the last being the only food item in common with those individuals that lived downstream in an estuarine habitat. Crabs of the genera Sesarma and Uca, polychaetes and fish fry completed the diet of these estuarine fish. In the stream habitat, ants were eaten by the juveniles and beetles by the adults; in the estuarine habitat polychaetes were only consumed by juvenile fish (Lim, 1967). The catholic diet of species of Periophthalmus can partially be attributed to their extreme opportunistic approach to prey items. Crabs (61 %), insects (29%), of which over a quarter were ants, small gastropods (9%), nemertine worms, Orthoptera and arachnids were taken by Periophthalmus vulgaris [ = P. argentilineatus] (Lim, 1971). Also from the same general area, Macintosh (1979) identified the crab prey of P. vulgaris [ = P. argentilineatus] as small individuals of Uca rosea, while Frith & Brunenmeister (1980) report that the same species took small U. forcipata. In terms of percentage by volume of stomach contents, shrimp (25%), spiders (12%), copepods (10%), snails (9%) and crabs (8%), were the commonest items taken by Periophthalmus argentilineatus [as P. cantonensis] from India (Pearse, 1932). As in other periophthalmids insect larvae, ostracods and amphipods completed the diet. Foraging P. waltoni in Kuwait take the small ocypodid crabs Tylodiplax indica, Cleistostoma dotilleforme and Ilyoplax stevensi (Al Taher, 1990; as P. koelreuteri) and Uca sp., Alpheus sp., nematodes and polychaetes (El Zaidy et al., 1975; as P. chryso spilos) and along the Gujurat coast of north western India, Siddiqi (1974) reports that the gut content of P. waltoni was mostly small-shelled molluscs. On Madagascar, P. sobrinus [ = P. argentilineatus or kalolo] takes polychaetes {Polydora, Dendronereis), crustaceans {Upogebia sp., Uca spp. including U. chlorophthalmus; Dotilla fenestrata), tanaids, and insects including ephydrid flies and dipteran larvae (Stebbins & Kalk, 1961; MacNae & Kalk, 1962). Secondary sources (e.g. Lim, 1971) quote Khoo (1966) as recording Perio phthalmus chrysospilos from Singapore as a herbivore because only diatoms, filamentous algae and other plant material were found in the stomachs of six individuals. Khoo (1966) qualifies these findings, however, by pointing out that the relative gut length of this species is that of a carnivore and that the other two individuals examined, which were from a different location, contained crus taceans {Ilyoplax sp.), harpacticoid copepods and polychaetes. It was concluded that the fish has the ability temporarily to utilise plant food sources when unable to take animal prey. The reason for the inability to take animal prey, however, remains obscure. In a study of Queensland's mudskippers (Milward, 1974), the diet of three sympatric species of Periophthalmus were compared. Arthropods dominated the diet and included harpacticoid copepods, brachyurans, dipteran larvae and adults. Tanaids, ostracods, collembolans and ants were also eaten and the com monest non-arthropod items were polychaetes and gastropods. Nematodes, foraminiferans and algal filaments were present in insignificant amounts. Copepods and dipteran larvae were mostly taken by the small fish and adult insects and crabs by the large ones. This size difference in diet applied both within and between species such that P. expeditionium [ = P. novaeguineaensis] (22-95 mm SL) was intermediate in size between P. gracilis (18—64 mm) and 556 DAVID A. CLAYTON P. vulgaris [ = P. argentilineatus] (14—104 mm) and resulted in some resource partitioning (Milward, 1974). Species of Periophthalmodon are crab specialists. Only Lim (1971) mentions the presence of ants and small bivalve molluscs as part of the diet of P. schlosseri the major portion being composed of crabs of the genera Metaplax, Ilyoplax, Sesarma and Uca. Similarly, the diet of Periophthalmodon freycineti [as P. schlosseri] consisted predominantly of crabs of the genera Macrophthalmus and Uca with a few ants and fish. The fish were probably smaller periopthalmids (Milward, 1974; see also p. 549). In terms of volumetric composition of identifiable food items, crabs also dominated the diet of Periophthalmodon schlosseri from Klang (3°24'N:101°23'E) (Sasekumar etal., 1984). However, the sipunculid Phascolosoma arcuatum was also common and mangrove plant detritus and gastropods were present in small quantities (Sasekumar et al, 1984). Khoo (1966) only found sesarmid crabs in the diet of Periophthalmodon schlosseri while Macintosh (1979) found only Uca dussumieri and U. rosea. In a series of prey presentations, only 15—25 mm carapace width fiddler crabs were eaten by Periophthalmodon; crabs larger than 30 mm carapace width were attacked and sometimes injured but not ingested. It was suggested this sizedependent predation was one reason for the excess of large males in the fiddler crab population (Macintosh, 1979). BEHAVIOUR GENERAL There are many short descriptive accounts of general behaviour (Petit, 1921; Champeau, 1951; van Dijk, 1959, 1960, 1978; Klauzewitz, 1967; MacNae, 1968c; Vestergaard, 1972; Burhanuddin & Martosewojo, 1979), locomotion (van Dijk, 1960; De & Nandi, 1984), courtship and nesting (Petit, 1922, 1928; Freitas, 1961a, b; Brillet, 1969a; Magnus, 1972) and aggression and territoriality (Mutsaddi & Bai, 1969b; Brillet, 1969b,c; Nursall, 1974; Clayton & Vaughan, 1982). Similarities in the general aspects of the biology and life cycles of small substratum-bound fish (Gibson, 1969, 1982) can be extended to include behaviour. In Madagascar the behaviour of Periophthalmus sobrinus [ = P. argentilineatus] and Lophalticus kirkii, an amphibious blenny that inhabits rocky shores, have many behavioural characteristics in common (Brillet, 1986). There are also important differences, but these might have been less marked had mudskippers from a similar rocky habitat been studied (cf. Gordon et al, 1968;. see p. 545). As well as providing a detailed functional analysis of skeletal and musculature involvement in fin adaptations for terrestrial locomotion, Harris (1961) also describes the behaviour. In water P. barbarus [as P. koelreuteri] employs paddling, fast swimming and skimming whereas on land, they use crutching, skipping and climbing. Apart from climbing, Boleophthalmus dussumieri [as B. boddarti] exhibits similar locomotory behaviour (Clayton & Vaughan, 1988). RHYTHMIC BEHAVIOUR While Gordon et al. (1968) found that Periophthalmus sobrinus [ = argentilineatus or kalolo] was active at all hours, Stebbins & Kalk, (1961) and Magnus (1972) MUDSKIPPERS 557 only reported diurnal activity. The study of activity rhythms tends to support the latter observations. Al Naqi (1977) found that P. waltoni [as P. chrysospilos] was strongly diurnal. Under a 12L:12D regime in the laboratory it showed two activity peaks in the light period, the larger starting at the beginning of the period, the smaller occurring towards the end of it. Under continuous light or dark a similar periodicity of activity was maintained, but at a much lower level in the dark (Eissa et al., 1978). Behavioural rhythms have also been investigated in Boleophthalmus pectinirostris [as B. chinensis] (Ishibashi, 1972; Ishibashi & Nishikawa, 1973) and Periophthalmus modestus [as P. cantonensis] (Nishikawa & Ishibashi, 1973,1975a,b). Using a photocell to record entry and exit from an artificial burrow Ishibashi (1972) demonstrated that, under constant conditions, Boleophthalmus pectinirostris was rhythmically active. Its peaks of activity coincided with the time of low tide on the shore from which it was collected and overall activity was low at times of neap tides and high at spring tides. This 15-day activity cycle was confirmed under conditions of continuous light or darkness by Ishibashi & Nishikawa (1973). They briefly report that the fish have both a circadian and a tidal component to their rhythm, but a more detailed verification is required. The equally brief report on endogenous rhythms in Periophthalmus (Nishikawa & Ishibashi, 1975a) explains some of the false starts in experimental procedure (19-h feeding entrainment and operant con ditioning) leading to their later work on entrainment of activity by feeding (Nishikawa & Ishibashi, 1975b). Under conditions of continuous darkness, the free running activity apparently showed some synchronisation with times of low tide, but the data were only averaged on a 24-h and not a tidal periodicity. Clear synchronisation of activity was achieved with 12-h (for 88 days) and 24-h (for 22 days) feeding regimes, the anticipation of which was evident from prefeeding bursts of activity, particularly under the shorter regime (Nishikawa & Ishibashi, 1975b). BURROWING AND TERRITORIALITY It has long been known that mudskippers build burrows (Petit, 1922, 1928; Harms, 1935; Champeau, 1951), but the common observation offish migrating across the intertidal zone following changing tidal levels (Hora, 1935a; Stebbins & Kalk, 1961; Khoo, 1966; Gordon etal., 1968; MacNae, 1968b) made the idea of territoriality in terms of a permanent burrow occupancy difficult to accept (Gibson, 1969). Thus, in his short study of mudskippers in Australia, Nursall (1981) who cited MacNae (1968a), was of the opinion that none is terri torial. The mudskippers concerned were Periophthalmus argentilineatus [as P. vulgaris] and P. gracilis, Periophthalmus which migrate with the tide, and P. minutus [as Periophthalmus sp.] and Periophthalmodon freycineti [as P. schlosseri], species that were associated more clearly with identifiable burrows. These observations contrast with those showing that Periophthalmus argentilineatus and P. gracilis in the same study area, construct and inhabit burrows (Milward, 1974). Milward was not quoted by Nursall (1981). Kobayashi et al (1971) imply that P. modestus [as P. cantonensis] also migrates with the tide, resting on the edge of the high tide line. However, they spend the cold season in burrows and subsequently use them for spawning. Of the two forms of territories of P. sobrinus [ = P. argentilineatus ] in Madagascar, one involves fish moving between high shore burrows on the bank 558 DAVID A. CLAYTON and low shore feeding territories in the tidal channel bed (Brillet, 1975). More conventional, permanent residency territories occur in areas away from banked channels. Territory size depends on fish density but each usually contains a single burrow system from which the territory is defended (Brillet, 1975), although in P. waltoni there are several burrow systems in a single territory (Clayton & Snowden, unpubl.). Single burrow systems were also found in territories of Boleophthalmus (Mutsaddi & Bai, 1969b; Clayton & Vaughan, 1986, 1988). As well as resident burrow occupants that are territorial, popula tions also contain non-territorial errant fish that may range more widely over the mudflat. In B. dussumieri, territorial individuals live in well-defined, conti guous, polygonal mud-walled territories (Clayton & Vaughan 1982, 1988). If residents are removed, smaller errant individuals occupy vacant territories within one or two tidal cycles (Clayton & Vaughan, 1986). This division also applies to Periophthalmus argentilineatus where errant fish are usually smaller and less aggressive than the territorial ones (Brillet, 1975). In captivity the difference results in the development of a dominance system comprising dominant, territorial burrow occupants and subordinate non-territorial ones (Brillet, 1969c, 1975). Thus fish that follow the tide are likely to be errant individuals, including juveniles and adults unable to establish dominance. The core, but not the geometric centre, of the territory is the burrow. Burrow structure and construction behaviour has been described for P. modestus (Kobashi et al., 1971; Matoba & Dotsu, 1977), P. sobrinus [ = P. argentilineatus or kalolo] (Freitas, 1961b; Brillet, 1969a, 1976), Boleophthalmus dussumieri (Clayton & Vaughan, 1986) and Periophthalmodon schlosseri (Harms, 1929; Verwey, 1930). The generalised structure consists of a lower, near vertical tunnel connected to the surface by two, less steep entrances. The variability of burrow architecture has been illustrated schematically (Brillet, 1969a, 1976; Matoba & Dotsu, 1977), photographically in cut away section (Kobashi et al, 1971) and by casts (Freitas, 1961b; Clayton & Vaughan, 1986); the last authors clearly showing that tunnels of smaller infauna anastomose with those of the fish. The specific architecture is so profoundly affected by local topography and tidal conditions (Brillet, 1969a, 1976) that generalisations of species-specific constructions (MacNae, 1968b) are to be treated with caution. While surface embellishments variably consist of turrets and saucer-like depressions with or without an elevated rim, in some species there are none at all (Milward, 1974). Species identification should not be attempted on burrow structure alone. The situation is exacerbated in situations where burrows may be appropriated by other species. Additionally, fish escape down the nearest burrow at the approach of an observer and may not be the true occupant: this may also account for the reports of paired fish (Freitas, 1961b; MacNae, 1968b). Structural modifications related to breeding are unlikely to be directly related to surface structures (cf. MacNae, 1968b), but consist of diverticulae from the base of the descending tunnel in which the eggs are laid (Kobashi et al., 1971; Brillet, 1976). Burrow depth can exceed 1.5 m (Clayton & Vaughan, 1986) and the general assumption is that they descend to the water table, but Brillet (1976) noted that eggs require periods out of water for successful hatching. However, while the breeding burrows of Periophthalmusmodestus contain an egg chamber (Kobayashi et al., 1971), Matoba & Dotsu (1977) also showed that, irrespective of the diameter of the original, the opening of the breeding burrow is restricted by mud deposition to such an extent that females may have to squeeze through an MUDSKIPPERS 559 opening smaller than themselves. Construction and maintenance is performed by removing mouthfuls of mud from the burrow and depositing them on the surface (Petit, 1928; Brillet, 1969a; Clayton & Vaughan, 1988) where, in periophthalmids, they may be used in turret construction or burrow modification (Brillet, 1969a; Matoba & Dotsu, 1977). Boleophthalmus dussumieri is unique in that the mud gobbets are used to construct mud walls at some distance from their burrow. Neighbouring fish may contribute to building the same wall (Clayton & Vaughan, 1986). Wall construction behaviour is density dependent and serves to reduce aggression in neighbouring territorial fish (Clayton, 1987). Burrows function as a refuge, observation post and nest (Brillet, 1975). Additionally, because mudskippers are poor swimmers, the burrow will be an important refuge from piscivorous predators at high tide (Milward, 1974). HABITAT SELECTION Habitat selection has not been studied in juvenile fishes, despite its relevance to resolving the issue of occupancy of non-muddy shores. The salinity preferences of adult Periophthalmus sobrinus (Gordon et al., 1968) and P. waltoni (Al-Naqi, 1977) were assessed by choice trials in groups of fish pre-adapted to different levels of salinity. Both studies showed that seawater-adapted fish preferred the terrestrial habitat to any dilution of sea water and that 100% sea water was generally the least preferred salinity. In a group choice experiment without preadaption, P. modestus [as P. cantonensis] also preferred land over water of any salinity (Gordon et al, 1985). Fish also preferred water at 30 °C over water at 10, 20 or 35°C and darkness to light. In terms of behavioural selection of microhabitat, these results are difficult to interpret, but in any case are con founded by social groupings and inappropriate analyses. Dominance relations in captive groups would interfere with fish distribution and the preference for the terrestrial habitat could have simply resulted from the fact that it occupied considerably more of the choice chamber than did the water. SOCIAL BEHAVIOUR Social behaviour in mudskippers is similar to that of other substratum bound fishes, involving erection of fins in a variety of combinations and leaping into the air (Gibson, 1982). Courtship has been described in Periophthalmus (Khoo, 1966; Brillet 1969b, 1970, 1980a, 1984; Magnus, 1972) and Boleophthalmus (Khoo, 1966; Lim, 1967; Clayton & Vaughan, 1988). As in many species agonistic elements occur during courtship, but these have been separately described for Periophthalmus (Brillet, 1969c,d, 1970, 1975, 1980b, 1981, 1983) and Boleophthalmus (Lim, 1967; Clayton, 1987; Clayton & Vaughan, 1988). The territorial status of interacting fish determines the nature of the social interaction. Errant or subordinate fish exhibit a marked lack of aggression in interactions with each other and do not resist or display to an attack by a dominant fish. In established territories, dominant fish avoid each other or use a ritualised approach, but directly attack subordinate ones (Brillet, 1969c,d). The direct attack is similar to the approach to a prey and is without fin erection whereas the ritualised approach is slower and less direct with fin displays occurring at closer approach. In prolonged encounters the fish circle each other in an 'O' formation (Brillet, 1969c). In a more quantitative analysis of agonistic behaviour, 16 non-exclusive elements were reported (Brillet, 1980a, 1981). Fish 560 DAVID A. CLAYTON that initiated agonistic encounters always chased their opponents and invariably used tail-assisted bounds in doing so. Dorsal fin erection, posturing by raising the body on pelvic and pectoral fins and physical contact including head-butting and biting were also common. The last two elements were difficult to distin guish, because they were extremely fast, erupting from essentially static display postures. Tail undulation and butting or biting were much reduced in juvenile fish (43—55 mm TL) that initiated encounters when compared with adults (90—100 mm TL). The recipient of such attacks rarely reciprocated, only occasionally erecting its dorsal fins (Brillet, 1980b). The influence of sex and size was assessed by recording victories and defeats among small captive groups (Brillet, 1981). Larger individuals, irrespective of sex, dominated smaller individuals but during encounters between fish of the same size males dominated females. Males initiated and won a greater percentage of encounters than females. Dorsal fin erection constituted the greatest percentage of all displays in small males and least in large males with that of females being intermediate. In contests initiated by large males, dorsal fin erection of the opponent was low irrespective of whether it was female or a large or small male. When responding in a contest initiated by a female or a small male, large males gave a stronger response than females (Brillet, 1981). Size was also an important factor in resolution of intraspecific agonistic encounters in four species of Periophthalmus (Nursall, 1974). The specific identity of these fish was only known later (Nursall, 1981) and in Nursall (1974) P. argentilineatus, P. novaeguineaensis, P. gracilis and P. minutus were des cribed as BB, RS, SBB and RF respectively. While the largest species P. argen tilineatus dominated all the other species, P. novaeguineaensis, the secondranked in size retreated before the smaller P. gracilis and P. minutus, which in turn ignored each other. Species-specific differences in agonistic behaviour are evident because, whereas P. argentilineatus rarely raised the first dorsal fin alone (Brillet, 1980b), the periophthalmids observed by Nursall (1981) showed various combinations of first, second and caudal fin erection in their displays. Nevertheless, and in agreement with Brillet's (1981) observations, the more aggressive species used such displays more frequently. Combinations of first and second dorsal and caudal fin displays were a feature of the agonistic displays of Boleophthalmus dussumieri (Clayton & Vaughan, 1988). While chromatophore body colour changes have been noted in courtship (Magnus, 1972; Matoba & Dotsu, 1977), only Nursall (1981) indicates that colour change is involved in agonistic display. While there are clear species-specific differences in courtship (Brillet, 1984), the general organisation of mudskipper courtship is similar for B. dussumieri (Clayton & Vaughan, 1988), Periophthalmus kalolo [as P. koelreuteri] (Brillet, 1984), P. argentilineatus (Brillet, 1969c) and P. modestus (Matoba & Dotsu, 1977). The last authors provided diagrammatic representations of the malefemale interactions during the complete sequence of the 'nuptial parade' as Brillet (1969b) so aptly describes it. From the vicinity of their burrows, males advertise their presence by leaping into the air with all fins fully spread and, at the approach of a female, give a series of substratum bound displays. These displays, which involve orientation of the body in relation to the female, mouth gaping and buccopharyngeal expansion or contraction and fin display, may be extended, repeated or become aggressive depending on the responsiveness of the female. The female's role is MUDSKIPPERS 561 largely passive requiring only that she remains stationary or follows the male. Furthermore, the fewer dorsal fin erections she gives, the more likely the courtship is to progress. After successful display, which in periophthalmids involves some tail waving, the male leads the female to the burrow. In Boleophthalmus tail waving has become a more pronounced static display termed quivering (Clayton & Vaughan, 1988). The male usually enters the nest first and spawning can take up to 5 h. Many agonistic elements occur during courtship, sometimes leading to its termination especially if interrupted by the male chasing an intruder (Brillet, 1970). Dorsal fin erection is an agonistic act that must be absent from the later stages of successful courtship such as tail beating or quivering (Brillet, 1969b, 1980a; Clayton & Vaughan, 1988). The dorsal fin displays always involve both dorsal fins in P. argentilineatus, whereas in the sympatric species P. kalolo, they may be raised independently as well as together (Brillet, 1984). The absence of first dorsal fin erection when male P. kalolo are close to females is seen as a mechanism that leads to reproductive isolation of the two species. The ability to erect the fins independently, on the other hand, is seen as a mechanism that increases the chances of successful courtship. In P. argentilineatus there is no scope for signalling different levels of aggression (Brillet, 1984). CONCLUSIONS That mudskippers can provide a rich source of comparative studies in adaptation to the littoral habitat should now be apparent. While some suggestions for future study have been mentioned in the relevant sections, there are a number of issues that bear repetition. The problems of specific identification should be diminished by the recent taxonomic revision of the Oxudercinae. If researchers also provide voucher specimens the confusion that previously hindered progress should also be reduced. Now that the artificial propagation of larval fish is possible, the potential for commercial rearing of mudskippers has increased. Successful aquaculture, however, requires that all aspects of reproductive biology are understood and more data on natural reproduction and larval development are needed. The study of larval stages and the factors that determine benthic settlement would also contribute towards understanding adult distribution and habitat selection, two areas where there is currently little information. Indirectly these studies may benefit the investigation of the physiological mechanisms of respiration, excretion and osmoregulation. It is likely that some of the con flicting results in these physiological studies are related to variations in responsiveness and/or the mechanisms of adaptation that are dependent on factors of age or size. More control of these variables is certainly required, as is attention to antecedent experimental conditions. That these are important is borne out by the tolerance to a wide range of salinities exhibited by fish that have free access to water and air when compared with those confined to one medium. This also illustrates that the mudskippers' behavioural control mechanisms are well developed, a feature more typically associated with thermoregulation than with other physiological systems. The full potential of comparative study in these closely related gobies has yet to be explored. Few data on Pseudapocryptes and Apocryptes are available, yet the former have considerable capabilities to resist drought. In northern latitudes the cold winters induce mudskippers to 562 DAVID A. CLAYTON hibernate, while in subtropical and tropical latitudes the summer heat causes them to aestivate. The partitioning of gaseous exchange between the various components of the respiratory system, and the characteristics of blood transport mechanisms need to be elucidated. Mudskippers provide wide opportunities for ecological and behavioural studies. Unfortunately, their pelagic larvae and particularly the difficulty of access to breeding burrows, prevent any realistic measures of reproductive success from being made and exclude any serious investigation of evolutionary issues. Nevertheless, mudskippers are eminently suitable for studies of the proximate causes of territorial, courtship and agonistic behaviour. Studies of resource partitioning and predator-prey relationships are two areas for further study. ACKNOWLEDGEMENTS This review was initially supported by Grant SZ032 from the University of Kuwait, continued at the Plymouth Marine Laboratory and completed at Sultan Qaboos University. REFERENCES Note: References or parts of references enclosed in brackets [ ] are translations of the original text. Afzelius, B. A., 1956. Seitenorgane und Schleimkanalknocken bei Periophthalmus koelreuteri und Gobius minutus. Zeitschrift fur Anatomie und Entwicklungsgeschichte, 119, 470-484. Al-Kadhomiy, N. K. & Hughes, G. M. 1988. Histological study of different regions of the skin and gills in the mudskipper Boleophthalmus boddarti with respect to their respiratory function. Journal of the Marine Biological Association of the United Kingdom, 68, 413-422. Al-Naqi, Z. A., 1977. Ecological and physiological studies on the mud-skipper fish, Periophthalmus chrysospilos, inhabiting the muddy shores of Kuwait. MSc thesis, Kuwait University, Kuwait, 133pp. Al-Nasiri, S. K. & Hoda, S. M. S., 1975. Survey of the fish fauna of Shatt-Al-Arab (from Abu Al-Khasib to Karmat Ali). Bulletin of the Basrah Natural History Museum, 2, 36—46. Al Taher, E. Y., 1990. The impact of the Doha west power station effluent on the upper eulittoral benthos of western Sulaibikhat Bay, Kuwait. MSc Thesis, Kuwait University, Kuwait, 208pp. Arai, R. & Kobayasi, H., 1973. A chromosome study on thirteen species of Japanese gobiid fishes. Japanese Journal of Ichthyology, 20, 1-16. Asano, N., 1936. The egg laying habits of the South Sea mudskipper, Periophthalmus sp. Ecological Review (Sendai), 2, 137—1139. Bai, P. A. & Kalyani, M., 1960a. On the composition of the scales and skin of Boleophthalmus boddarti Pall, and Mugil dussumieri (Val.). Journal of Animal Morphology and Physiology, 7, 71-77. Bai, P. A. & Kalyani, M., 1960b. Studies on ascorbic acid in fish tissues I. Journal of Animal Morphology and Physiology, 7, 162—165. Bai, P. A. & Kalyani, ML, 1960c. Studies on the ascorbic acid content of fish tissues. II. Journal of the Zoological Society of India, 12, 216—219. MUDSKIPPERS 563 Bai, P. A. & Kalyani, M., 1961. The relative content of imino acids in the skin and scales of some teleosts. Journal of Animal Morphology and Physiology, 8, 126-129. Bandurski, R. S., Bradstreet, E. & Scholander, P. F., 1968. Metabolic changes in the mud-skipper during asphyxia or exercise. Comparative Biochemistry and Physiology, 24, 271-274. Banerjee, R., Nandi, N. C. & Raut, S. K., 1989. Food and feeding habits of the estuarine crocodile Crocodylus porosus Schneider in captivity. Acta Biologica Cracoviensia Series Zoologica, 30, 95—98. Banerji, T. K., 1973. A comparative histological investigation on the interrenal gland of some Indian teleosts. Anatomische Anzeiger, 133, 20—32. Bangera, V. S. & Patel, B., 1984. Natural radionuclides in sediment and in arcid clam (Anadara granosa L.) and gobiid mudskipper {Boleophthalmus boddaerti Cuv. & Val.). Indian Journal of Marine Science, 13, 5—9. Baumeister, L., 1913. Uber die Augen der Schlammspringer {Periophthalmus und Boleophthalmus). Zoologische Jahrbucher, Abteilungfur Anatomieund Ontogenie der Tiere, 35, 341-354. Berg, L. S., 1949. [Freshwater fishes of Iran and neighbouring countries]. Studies of the Zoological Institute of the Academy of Science of the USSR Moscow, 8, 753-758. (In Russian). Berry, A. J., 1972. The natural history of west Malaysian mangrove faunas. Malaya Nature Journal, 25, 135-162. Bhan, S. & Mansuri, A. P., 1978a. Adaptations to osmotic stress in the euryhaline teleost Periophthalmus dipes: Tissue water & mineral content. Journal of Marine Science, 7, 134—136. Bhan, S. & Mansuri, A. P., 1978b. Adaptations to osmotic stress in the euryhaline teleost Periophthalmus dipes Bleeker: II. Changes in glycogen marine Indian marine and fat levels of tissues. Geobios, 5, (1), 3—8. Bhan, S. & Mansuri, A. P., 1978c. Adaptations to osmotic stress in the marine eury haline teleost Periophthalmus dipes: III Changes in tissue succinic dehydrogenase enzyme levels. Journal of Animal Morphology and Physiology, 25, 132-138. Birdsong, R. S., Murdy, E. O. & Pezold, F. L., 1988. A study of the vertebral column and median fin osteology in gobioid fishes with comments on gobioid relationships. Bulletin of Marine Science, 42, 174—215. Branch, G. M. & Grindley, J. R., 1979. Ecology of southern African estuaries. Part XI Mngazana: a mangrove estuary in Transkei. South African Journal of Zoology, 14, 3, 150-170. Bridges, C. R., 1988. Respiratory adaptations in intertidal fish. American Zoologist, 28, 79-96. Brillet, C, 1969a. Etude du comportement constructeur des poissons amphibies Periophtalmidae. Terre et Vie, 23, 496—520. Brillet, C, 1969b. Premiere description de la parade nuptiale des poissons amphibies Periopthalmidae. Comptes Rendus Hebdomadaires des Seances de VAcademie des Sciences, Serie D, 269, 1889-1892. Brillet, C, 1969c. Observations sur le comportement agressif des poissons amphibies periophtalmes {Periophtalmus koelreuteri Pallas). Comptes Rendus Hebdomadaires des Seances UAcademie des Sciences, Serie D, 269, 1114—1117. Brillet, C, 1969d. Comportement territorial et hierarchie sociale chez les poissons amphibies du genre Periophtalmus (Teleosteen, Gboioidae). Comptes Rendus Hebdomadaires des Seances de VAcademie des Sciences, Serie D, 269, 1322-1325. Brillet, C, 1970. Relations entre comportement sexual, territoire et agressivitie chez les periophtalmes. Comptes Rendus Hebdomadaires des Seances de VAcademie des Sciences, Serie D, 269, 1507-1510. 564 DAVID A. CLAYTON Brillet, C, 1975. Relations entre territoire et comportement agressif chez Periophthalmus sobrinus Eggert (Pisces, Periophthalmidae) au laboratoire et en milieu naturel. Zeitschrift fur Tierpsychologie, 39, 283-331. Brillet, C, 1976. Structure du terrier, reproduction et comportement des jeunes chez le poisson amphibie Periophthalmus sobrinus Eggert. Terre et la Vie, 30, 465-483. Brillet, C, 1980a. Comportement sexuel du poisson amphibie Periophthalmus sobrinus Eggert. Ses rapports avec le comportement agonistique. Review Ecologie {Terre et Vie), 34, 427-468. Brillet, C, 1980b. Comportement agonistique du poisson amphibie Periophthalmus sobrinus Eggert. Analyse quantitative-1. Frequence des divers elements des repertoire. Biology of Behavior, 5, 297-315. Brillet, C, 1981. Comportement agonistique du poisson amphibie Periophthalmus sobrinus Eggert. Analyse quantitative-II. Influence de la taile et du sexe sur l'establissement des relations de dominance/subordination et sur l'utilisation du repertoire agonistique. Biology of Behavior, 6, 35-57. Brillet, C, 1983. Agonistic behaviour in the amphibious fish Periophthalmus sobrinus Eggert. Biology of Behavior, 8, 49-66. Brillet, C, 1984. Etude comparativede la parade nuptialechez deux especes de poissons amphibies sympatriques (Pisces, Periophthalmidae). Comptes Rendus Hebdomadaires des Seances de VAcademie des Sciences, Serie D, 298, 347-350. Brillet, C, 1986. Notes on the behaviour of the amphibious fish Lophalticus kirkii Gunther (Pisces Salariidae). Comparison with Periophthalmus sobrinus Eggert. Review Ecologie (Terre et Vie), 41, 361-375. Burhanuddin & Martosewojo, S. 1979. [Observations on the natural history of ikan gelodok Periophthalmus koelreuteri in Par Island]. In, Proceedingsof theSeminar on Mangrove Forest Ecosystems, Jakarta, edited by S. Soemodihardjo et al, Indonesia, pp. 86-92. (In Malay with English abstract). Champeau, M. F., 1951. Des moeurs de Periophthalmus koelreuteri gobiide de l'Ouest Africa. Deuxieme Conference International des Africanists de VOuest, Bissau, 1947, 3, 271-277. Chapman, V. J., 1976. Mangrove Vegetation. J. Cramer, Berlin, 448 pp. Chatterjee, T. K. & Siddiqi S. Z., 1957. Periophthalmus weberi Eggert (Pisces: Gobiidae), a rare gobioid fish from Indian waters. Matsya, 1, 82 only. Chen, T. P., 1976. Culture of the mudskipper. In, Aquaculture Practices in Taiwan. Fishery News (Books) Ltd, Farnham, England, 62 pp. Chen, T. P., 1982. Mudskipper culture. Thai Fish Gazette, 35, 321-324. Chen, Y.-D., & Ting, Y.-Y., 1984. [The preliminary report on inducing breeding experiments with mudskipper, Boleophthalmus pectinirostris (L.) by injections of HCG.] Bulletin of the Taiwan Fisheries Research Institute, No. 37, 133-144. (In Chinese with English abstract). Chew, S. F. & Ip, Y. K., 1987. Ammoniagenesis in mudskippers Boleophthalmus boddarti and Periophthalmus schlosseri. Comparative Biochemistry and Physio logy, 87B, 941-948. Chew, S. F. & Ip, Y. K., 1990. Differences in the responses of two mudskippers, Boleophthalmus boddaerti and Periophthalmus chrysospilos to changesin salinity. Journal of Experimental Zoology, 256, 227—231. Chew, S. F., Lim, A. L. L., Low, W. P., Lee, C. G. L., Chan, K. M. & Ip, Y. K., 1990. Can the mudskipper Periophthalmus chrysospilos tolerate acute environ mental hypoxic exposure? Fish Physiology and Biochemistry, 8, 221-227. Choudhury, A. & Nandi, N. C, 1973. Studies on myxosporidian parasites (Protozoa) froman estuarine gobiidfish of westBengal. Proceedings of the Zoological Society of Calcutta, 26, 45-55. Clayton, D. A., 1986. Ecology of mudflats with particular reference to those of the northernArabianGulf. In, Marine Environment andPollution: Proceedings of the Conference on the Marine Environment and Pollution, edited by R. Halwagy MUDSKIPPERS 565 et al, Kuwait University, Kuwait Foundation for Advancement of Science and Environmental Protection Council, Kuwait, pp. 83-96. Clayton, D. A., 1987. Why mudskippers build walls. Behaviour, 102, 185-195. Clayton, D. A., 1990. Crustacean allometric growth: A case for caution. Crustaceana, 58, 270-290. Clayton, D. A. & Abo Seedo, F. 1986. Osteichthyes. In, D. A. Jones, A Field Guide to the Sea Shores ofKuwait and the Arabian Gulf. University of Kuwait, Blandford Press, Kuwait, pp. 167-175. Clayton, D. A. & Vaughan, T. C, 1982. Pentagonal territories of the mudskipper Boleophthalmus boddarti (Pisces: Gobiidae). Copeia, 1982, 233-235. Clayton, D. A. & Vaughan, T. C, 1986. Territorial acquisition in the mudskipper Boleophthalmus boddarti (Pisces: Gobiidae) on the mudflats of Kuwait. Journal of Zoology, London, 209, 501-519. Clayton, D. A. & Vaughan, T. C, 1988. Ethogram of Boleophthalmus boddarti (Teleostei: Gobiidae), a mudskipper found on the mudflats of Kuwait. Journal of the Univer sity of Kuwait, (Science), 15, 115-138. Clayton, D. A. & Wright, J. M., 1989. Mud-walled territories and feeding behaviour of Boleophthalmus boddarti (Pisces: Gobiidae) on the mudflats of Kuwait. Journal of Ethology, 7, 91-95. Cuvier, G. L. G. C. D. & Valenciennes, A., 1837. Histoire Naturelle des Poissons, Vol 12, F. G. Levrault, Paris, 509 pp. Dall, W. & Milward, E., 1969. Water intake, gut absorption and sodium fluxes in amphi bious and aquatic fishes. Comparative Biochemistry and Physiology, 30, 247—260. Das, B. K., 1933. On the bionomics, structure and physiology of respiration in an estuarine air-breathing fish, Pseudapocryptes lanceolatus (Bloch and Schneider), with special reference to a new mode of aerial respiration. Current Science, 1, 389-393. Das, B. K., 1934. The habits and structure of Pseudapocryptes lanceolatus, a fish in the first stages of structural adaption to aerial respiration. Proceedings of the Royal Society, Series B, 115, 422-430. Datta, N. C. & Das. A., 1980. Anatomy of the olfactory apparatus of some Indian gobioids (Pisces: Perciformes). Zoologische Anzeiger, 205, 241-252. Day, J. H., 1974. The ecology of Morrumbene Estuary Mocambique. Transactions of the Royal Society of South Africa, 41, 43-97. De, J. K. & Nandi, N. C, 1984. A note on the locomotory behaviour of the mudskipper Boleophthalmus boddarti. Indian Journal of Fisheries, 31, 407—409. Dhage, K. P. & Mohamed, O. M. M. J., 1977. Study of the amylase activity in mudskipper Periophthalmus koelreuteri (Pallas). Indian Journal of Fisheries, 24, 129-134. Diamond, J., 1971. The Mauthner cell. In, Fish Physiology, Vol. 5, edited by W. S. Hoar & D. J. Randall, Academic Press, London, pp. 265—346. Diesselhorst, G., 1938. Horversuche an Fishen ohne Weberschen Apparat. Zeitschriftfur Vergleichende Physiologie, 25, 748—783. Dorjes, J., 1978. Sedimentologische und faunistiche Untersuchungen an Watten in Taiwan. II. Faunistiche und aktuopalaontologische Studien. Senckenbergiana Maritima, 10, 117-143. Dotsu, Y., 1961. The bionomics and life history of the gobioid fish. Apocryptodon bleekeri (Day). Bulletin of the Faculty of Fisheries, Nagasaki University, 10, 127-131. Eggert, B., 1929a. Bestimmungstabelle und Beschreibung des Arten der Family Periophthalmus. Zeitschrift fur Wissenschaftliche Zoologie, 133, 398—409. Eggert, B., 1929b. Die Gobiidenflosse und ihre Anpassung an das Landleben. Zeitschrift fur Wissenschaftliche Zoologie, 133, 411—440. Eggert, B. 1935. Beitrag zur Systematik, Biologie und geographischen Verbreitung der Periophthalminae. Zoologische Jahrbucher, Abteilung fur Systematik, Okologie und Geographie, 67, 29—114. 566 DAVID A. CLAYTON Eissa, S. M. El-Zaidy, S. & Al-Naqi, Z., 1978. Circadian activity rhythm of mud skipper fish, Periophthalmus chrysospilos (Bleeker) in Kuwait. Bulletin of the Faculty of Science, Cairo University, 48, 213-225. El-Sayed, N. K. & Safer, A. M., 1985. Enzyme activity in the kidney of the mudskipper Periophthalmus koelreuteri (Pallas). Journalof the University ofKuwait, (Science), 12, 245-256. El-Sayed, N. K. & Safer, A. M., 1986. Wandering cells in the nephron of the mud skipper Periophthalmus koelreuteri (Pallas). Cell Biology International Reports, 10, 755-762. El-Zaidy, S., Eissa, S. M., & Al-Naqi, Z., 1975. Ecological studies on the mud-skipper fish Periophthalmus chrysospilos (Bleeker) in Kuwait. Bulletin of the Faculty of Science, Cairo University, 48, 113—134. Fenwick, J. C. & Lam, T. J., 1988a. Effects of calcitonin on plasma calcium and phosphate in the mudskipper, Periophthalmodon schlosseri (Teleostei), in water and during exposure to air. General and Comparative Endocrinology, 70, 224230. Fenwick, J. C. & Lam, T. J., 1988b. Calcium fluxes in the teleost fish Tilapia (Oreochromis) in water and in both water and air in the marble goby {Oxyeleotris) and the mudskipper (Periophthalmodon). Physiological Zoology, 61, 119-125. Fischthal, J. H. & Thomas, J. D., 1971. Some hemiurid trematodes of marine fishes from Ghana. Proceedings of the Helminthological Society of Washington, 38, 181-190. Fishelson, L., 1971. Ecology and distribution of the benthic fauna in the shallow waters of the Red Sea. Marine Biology, 10, 113-133. Franz, V., 1910. Die japanischen Knockenfische der Sammlungen Harberer und Doflein. Abhandlungen Bayerische Akademie der Wissenschaften, Mathematisch-Naturwissenschaftliche Klasse, Supplement, 4, 1—32. Freitas, A. J., 1961a. Alqumas notas sobre a reproducao do Periophthalmus sobrinus. Boletim do Instituto de Investigacao Cientica de Mocambique, 2, 73—76. Freitas, A. J., 1961b. Uma nota sobre a estructura dos ninhos do peixe Periophthalmus sobrinus. Boletimdo Instituto de Investigacao Cienticade Mocambique, 2, 389—390. Fricke, H. W., 1970. Die okologische Spezialisierung der Eidechse Crytoblepharus boutoni cognatus (Boettger) auf das Leben in der Gezeitenzone (Reptilia, Skinkidae). Oecologia (Berlin), 5, 380-391. Frith, D. W., 1977. A preliminary list of macrofauna from a mangrove forest and adjacent biotopes at Surin Island, Western Peninsular Thailand. Phuket Marine Biological Centre, Research Bulletin, 17, 1-14. Frith, D. W., Tantanasiriwong, R. & Bhatia, O., 1976. Zonation and abundance of macrofauna on a mangrove shore, Phuket Island. Phuket Marine Biological Centre, Research Bulletin, 10, 1-37. Frith, D. W. & Brunenmeister, S., 1980. Ecological and population studies of fiddler crabs (Ocypodidae, Genus Uca) on a mangrove shore at Phuket Island, Western peninsular Thailand. Crustaceana, 39, 157-183. Fukuda, T., 1985. A fish out of water. Discover, 6, 56-61. Garey, W. F., 1962. Cardiac responses of fishes in asphyxic environments. Biological Bulletin (Woods Hole, Massachusetts), 122, 362—368. Gibson, R. N., 1969. The biology and behaviour of littoral fish. Oceanography and Marine Biology: an Annual Review, 7, 367—410. Gibson, R. N., 1982. Recent studies on the biology of intertidal fishes. Oceanography and Marine Biology: an Annual Review, 20, 363-414. Gibson, R. N., 1986. Intertidal teleosts: life in a fluctuating environment. In, The Behaviour of Teleost Fishes, edited by T. J. Pitcher, Croom Helm, London, pp. 388-408. Gibson-Hill, C. A., 1948. A note on the food habits of three kingfishers occurring on Singapore Island. Journal of the Bombay Natural History Society, 48, 146-152. MUDSKIPPERS 567 Gopalakrishnan, P., 1970. Some observations on the shore ecology of the Okha coast. Journal of the Marine Biological Association of India, 12, 15-34. Gordon, M. S., 1970. Patterns of nitrogen excretion in amphibious fishes. In, Urea and Kidney, edited by K. Schmidt-Nielson, Excerpta Medica, Amsterdam, pp. 238-242. Gordon, M. S., Boetius, J., Boetius, I., Evans, D. H., McCarthy, R. & Oglesby, L., 1965. Salinity adaptation in the mudskipper fish Periophthalmus sobrinus. Hvalradets Skrifter, 48, 85-93. Gordon, M. S., Boetius, J., Evans, D. H. & Oglesby, L. C, 1968. Additional observa tions on the natural history of the mudskipper, Periophthalmus sobrinus. Copeia, 1968, 853-857. Gordon, M. S., Boetius, I., Evans, D. H., McCarthy, R. & Oglesby, L. C, 1969. Aspects of the physiology of terrestrial life in amphibious fishes. I. The mud skipper, Periophthalmus sobrinus. Journal ofExperimental Biology, 50, 141—149. Gordon, M. S., Gabaldon, D. J. & Yip, A. Y.-W., 1985. Exploratory observations on microhabitat selection within the intertidal zone by the Chinese mudskipper fish Periophthalmus cantonensis. Marine Biology, 85, 209—215. Gordon, M. S., Ng, W. W.-S. & Yip, A. Y.-W., 1978. Aspects of the physiology of terrestrial life in amphibious fishes. III. The Chinese mudskipper Periophthalmus cantonensis. Journal of Experimental Biology, 72, 57—75. Graham, J. B., 1971. Aerial vision in amphibious fishes. Fauna (Rancho Mirage, California), 3, 14-23. Graham, J. B., 1976. Respiratory adaptations of marine air-breathing fishes. In Respiration of Amphibious Vertebrates, edited by G. M. Hughes, Academic Press, London, pp. 165-187. Gregory, R. B., 1977. Synthesis and total excretion of waste nitrogen by fish of the Periophthalmus (mudskipper) and Scartelaos families. Comparative Biochemistry and Physiology, 57A, 33-36. Gundermann, N. & Popper, D. M., 1984. Notes on the Indo-Pacific mangal fishes and on mangrove related fisheries. In, Hydrobiology of the Mangal, edited by F. D. Por & I. Dor, W. Junk, The Hague, pp. 201-206. Haddon, A. C, 1889. Zoological notes from the Torres Straits: Caudal respiration in Periophthalmus. Nature, 39, 285 only. Harms, J. W., 1929. Die Realization von Genen und die consecutive Adaptation. 1: Phasen in der Differenzierung der Anlagenkomplexe und die Frage der Landtierwerdung. Zeitschrift fur wissenschaftliche Zoologie, 133, 211-397. Harms, J. W., 1935. Die Realisation von Genen und die consecutive Adaptation. 4: Mitteilung Experimentell hervorgerufener Medienwechsel: Wasser zu Feuchtluft bwz. zu Trockenlaft bei Gobiiformes (Gobius, Boleophthalmus und Perioph thalmus). Zeitschrift fur wissenschaftliche Zoologie, 146, 417-462. Harms, W., 1914. Uber die Augen der am Grunde der Gewasser lebenden Fische. Zoologische Anzeiger, 44, 35-41. Harris, V. A., 1961. On the locomotion of the mudskipper Periophthalmus koelreuteri (Pallas) (Gobiidae). Proceedings of the Zoological Society of London, 134, 107-135. Herre, A. W. C. T., 1941. A list of fishes from the Andaman Islands. Memoirs of the Indian Museum, 13, 331—403. Hess, C, 1912. Untersuchungen zur vergleichenden Physiologie und Morphologie des Ciliarringes. Zoologische Jahrbiicher Supplement, 15, 155—175. Hess, C, 1913. Gesichtssin. In, Winterstein's Handbuch der Vergleichenden Physio logie, 4, 555-840. Higson, S. J., 1889. A Naturalist in North Celebes. Murray, London. Hillman, S. S. & Withers P. C, 1987. Oxygen consumption during aerial activity in aquatic and amphibious fish. Copeia, 1987, 232—234. Hoda, S. M. S., 1980. A contribution to the gobioid fishes of Pakistan. Proceedings of the First Pakistan Congress of Zoology B. 469-482. 568 DAVID A. CLAYTON Hoda, S. M. S., 1985. Fishes of the coast of Pakistan. Pakistan Agriculture, 7, 38—44. Hoda, S. M. S., 1986a. Maturation and fecundity of the mudskipper Boleophthalmus dussumieri Cuv. & Val. from the Karachi Coast. Mahasagar—Bulletin of the National Institute of Oceanography, 19, 73-78. Hoda, S. M. S., 1986b. On the growth of Boleophthalmus dussumieri C.V. and B. dentatus C.V. (Family Gobiidae) from the Karachi coast of Pakistan. Biologia (Lahore), 32, (Suppl.), 55-62. Hoda, S. M. S., 1987. Relative growths of body parts and length-weight relationships in Boleophthalmus dussumieri and B. dentatus of Karachi Coast. Indian Journal of Fisheries, 34, 120-127. Hoda, S. M. S. & Akhtar, Y., 1983. Mudskippers from mangrove swamps of Karachi coast. In, Mangroves of Pakistan. Proceedings of the National Workshop on Mangroves held at Karachi 8-10 August, 1983. Pakistan Agricultural Research Council, Islamabad, pp. 49-52. Hoda, S. M. S. Akhtar, Y., 1985. Maturation and fecundity of mudskipper Boleophthalmus dentatus in the northern Arabian Sea. Indian Journal of Fisheries, 32, 64—74. Holly, M., 1929. Beitrage zur Kenntnis der Fischfauna Persiens. Zoologische Anzeiger, 85, 183-185. Hong, W.-S., Dai, Q.-G., Zhang, Q.-Y., Zang, J. & Cai, Y.-Y., 1988. [Observations on the early development of the mudskipper Boleophthalmus pectinirostris (Linnaeus).] Tropic Oceanology/Redai Haiyang, 5, (2), 1—8. (In Chinese with English abstract). Hong, W.-S. & Wang, Y.-L., 1989. [Introduction of ovulation in female mudskipper (Boleophthalmus pectinirostris) by HCG analogue ofLHRH & pimozide. ] China Journal of Oceanology & Limnology, 7, 283—287. (In Chinese with English abstract). Hora, S. L., 1933. Respiration in fishes. Journal ofthe Bombay Natural History Society, 36, 538-560. Hora, S. L., 1935a. Ecology and bionomics of the gobioid fishes of the Gangetic delta. International Congress of Zoology, 12, 841—865. Hora, S. L., 1935b. Physiology, bionomics and evolution of the air breathing fishes of India. Transactions of the National Institute of Science, India, 1, 1-16. Hughes, G. M. & Al-Kadhomiy, N. K., 1986. Gill morphometry of the mudskipper Boleophthalmus boddarti. Journal of the Marine Biological Association of the United Kingdom, 66, 671-682. Hughes, G. M. & Munshi, J. D., 1979. Fine structure of the gills of some Indian air-breathing fishes. Journal of Morphology, 160, 169—194. Hunziker, R. E., 1985. Gobies for freshwater and brackish aquaria. Tropical Fish Hobbyist, 34, 10-18. Inger, R. F., 1952. Walking fishes of south eastern Asia travel on land. Chicago Natural History Museum Bulletin, 23, 4—7. Ip, Y. K., Chew, S. F. & Low, W. P., 1991a. Effects of hypoxia on the mudskipper, Periophthalmus chrysospilos (Bleeker, 1853). Journal ofFish Biology, 38, 621—623. Ip, Y. K., Chew, S. F. & Ping, C. T., 1991b. Evaporation and the turning behaviour of the mudskipper Boleophthalmus boddaerti. Zoological Science (Tokyo), 8, 621-623. Ip, Y. K. & Low, W. P., 1990. Lactate production in the gills of the mudskipper Periophthalmodon schlosseri exposed to hypoxia. Journal of Experimental Zoology, 253, 99-101. Ip, Y. K., Low, W. P., Lim, A. L. L. & Chew, S. F., 1990. Changes in lactate content in the gills of the mudskippers Periophthalmus chrysospilos and Boleophthalmus boddaerti in response to environmental hypoxia. Journal of Fish Biology, 36, 481-488. Ishibashi, T., 1972. [Periodical behaviour of mudskippers.] Zoological Magazine (Tokyo), 81 409-410. (In Japanese). MUDSKIPPERS 569 Ishibashi, T. & Nishikawa, M., 1973. [Behavioural rhythms of gobioid fish, Boleoph thalmus chinensis (Osbeck) with special reference on the effect of light condition]. Zoological Magazine (Tokyo), 82, 311 only. (In Japanese). Iwata, K., 1984. A high ammonia tolerance in the mudskipper Periophthalmus can tonensis. Zoological Science, 1, 877 only. Iwata, K., 1988. Nitrogen metabolism in the mudskipper Periophthalmus cantonensis: changes in free amino acids and related compounds in various tissues under conditions of ammonia loading, with special reference to its high ammonia tolerance. Comparative Biochemistry and Physiology, 91A, 499—508. Iwata, K. & Kakuta, I., 1983. A comparison of catalytic properties of glutamate dehydrogenase from liver and muscle between amphibious Periophthalmus can tonensis and water-breathing gobiid fishes Tridentiger osbcurus obscurus. Bulletin of the Japanese Society of Scientific Fisheries, 49, 1903—1908. Iwata, K., Kakuta, I., Ikeda, M., Kimoto, S. & Wada, N., 1981. Nitrogen metabolism in the mudskipper Periophthalmus cantonensis: a role of free amino acids in detoxification of ammonia produced during its terrestrial life. Comparative Bio chemistry and Physiology, 86A, 589—596. Jes, H., 1972. Eine Uferanlage fur Schlammspringer und Winkerkrabben im Kolner Aquarium am Zoo. Zeitschriftefur Kolner Zoo, 15, 111—117. Johnstone, J., 1903. Report on the marine fishes. Fasciculi Malayenses: Zoology, 12, 293-303. Jones, D. A., 1984. Crabs of the mangal ecosystem. In, Hydrobiology of the Mangal, edited by F. D. Por & I. Por, W. Junk, The Hague, pp. 207-209. Jones, S., 1937. Observations on the breeding habits and development of certain brackish water fishes of Adyar, Madras. Proceedings of the Indian Academy of Science, 5, 261-289. Kaden, J., 1978. Schlammspringer sind ganz anders. Aquarien-Magazin, 12, 552—555. Karsten, H., 1923. Das Auge von Periophthalmus koelreuteri. Jenaische Zeitschriftfur Naturwissenschaft, 59, 115-154. Khalaf, K. T., 1961. The Marine and Freshwater Fishes of Iraq. Ar Rabitta Press, Baghdad, 143 pp. Khan, M. M. & Ip, Y. K., 1988. Pallisentis singaporensis (Acanthocephala: Quadrigyridea) from the mudskipper Periophthalmodon schlosseri in Singapore. Journal of Singapore National Academy of Sciences, 17, 24—27. Khoo, K. G., 1966. Studies on the biology of periophthalmid fishes in Singapore. BSc thesis, University of Singapore, 61 pp. Kimura, M., 1958. Early development of the mudskipper Periophthalmus cantonensis (Osbeck). Bulletin of the Japanese Society of Scientific Fisheries, 23, 754-757. Klausewitz, W., 1967. Uber einige Bewegungsweisen der Schlammspringer (Perioph thalmus). Natur und Museum, 97, 6, 211—222. Klausewitz, W., 1968. Some observations on East African mudskippers. Tropical Fish Hobbyist, 16, 52-65. Klika, E. & Tesik, I., 1980. Die Ultrastruktur der Schleimhaut des akzessorischen Luftatmungsorganes des Schlammspringerfisches Periophthalmus barbarus. GiT Fachzeitschrift fur das Laboratorium, 24, 189—196. Kluge, K., 1971. Gologo. Aquarien Den Hag, 42, 108-109. Kobayashi, T., Dotsu, Y. & Takita, T., 1971. [Nest and nesting behaviour of the mud skipper Periophthalmus cantonensis, in Ariake Sound. ] Bulletin of the Facultyof Fisheries, Nagasaki University, 32, 27-39. (In Japanese with English abstract). Kobayashi, T., Dotsu, Y. & Miura, N., 1972. [Egg development and rearing experi ments of the larvae of the mudskipper Periophthalmuscantonensis. ] Contributions from the Fisheries Experimental Station ofNakasagi University, No. 35, 49-62. (In Japanese with English abstract). Krishnaja, A. P. & Rege, M. S., 1980. Some observations on the chromosomes of certain teleosts using a simple method. Indian Journal of Experimental Biology, 18, 268-270. 570 DAVID A. CLAYTON Krishnamurthy, K., Sultan Ali, M. A. & Prince Jeyaseelan, M. A., 1984. Structure and dynamics of the aquatic food web community with special reference to nematodes in mangrove ecosystems. In, Proceedings of the Asian Symposium on Mangrove Environment-Research & Management, edited by E. Soepadmo et al, University of Malaya and UNESCO, Kuala Lumpur, pp. 429-452. Lakshmi, R., Kundu, R. & Mansuri A. S. P., 1990. Toxicity of mercury to mudskipper Boleophthalmus dentatus C. & V. Part I. Changes in the activity of ATPases in gills. Acta Hydrochimica et Hydrobiologica, 18, 605—611. Lakshmi, R., Kundu, R., Thomas, E. & Mansuri, A. S. P., 1991. Mercuric chloride induced inhibition of different ATPases in the intestine of Boleophthalmus dentatus. Ecotoxicology and Environment Safety, 21, 18—24. Lee, C. G. L. & Ip, Y. K., 1987. Environmental effect on plasma thyroxine (T4), 3,5,3'-triodo-L-thyronine (T3), prolactin and cyclic adenosine 3',5'-monophosphate (cAMP) content in the mudskippers Periophthalmus chrysospilos and Boleophthalmus boddarti. Comparative Biochemistry and Physiology, 87A, 1009—1014. Lee, C. G. L., Low, W. P. & Ip, Y. K., 1987. Na+, K+ and volume regulation in the mudskipper, Periophthalmus chrysospilos. Comparative Biochemistry and Physiology, 87A, 439-448. Lee, C.-L., 1990. [Osteological study of the mudhopper, Periophthalmus cantonensis (Perciformes, Gobiidae) from Korea]. Korean Journal ofZoology, 33, 402-410. (In Korean with English abstract). Lele, S. H. & Kulkarni, R. D., 1938. The skeleton of Periophthalmus barbarus (Linn.). I. The skull. Journal of the University of Bombay, New Series, 6, 76—91. Lele, S. H. & Kulkarni, R. D., 1939. The skeleton of Periophthalmus barbarus (Linn.). II. Branchial arches, vertebral column and appendicular skeleton. Journal of the University of Bombay, New Series, 7, 123—134. Liao, I.-C, Chao, N.-H., Tseng, L.-C. & Kuo, S.-C, 1973. [Studies on the artificial propagation of Boleophthalmus chinensis (Osbeck)—I: Observations on embryonic development and early larvae.] J.CR.R. Fish Series, No. 15, 29-42. Lim, A. L. L. & Ip, Y. K., 1989., Effect of fasting on glycogen metabolism and activities of glycolytic and gluconeogenic enzymes in the mudskipper Boleophthalmus boddaerti. Journal of Fish Biology, 34, 349—367. Lim, G. S., 1967. Studies on the feeding biology of Periophthalmus barbarus (Linneus) and Boleophthalmus boddaerti (Pallas) Perciformes—Periophthalmidae. BSc thesis, University of Malaya, 37 pp. Lim, L. S., 1971. The feeding biology of Periophthalmus vulgaris (Eggert) and Peri ophthalmodon schlosseri (Pallas) Perciformes, Periophthalmidae. MSc thesis, University of Malaya, 33 pp. Lipschitz, H. D., Berjak, P., Pammenter, N. W., & Davey, J. E., 1975. A note on the apparent redistribution of crabs in Saco mangrove swamp Inhaca Island, Mocambique. South African Journal of Science, 71, 55-57. Low, W. P., Ip, Y. K. & Lane, D. J. W., 1990. A comparative study of the gill morphometry in the mudskippers Periophthalmus chrysospilos, Boleophthalmus boddaerti and Periophthalmodon schlosseri. Zoological Science (Tokyo), 7, 29-38. Low, W. P., Lane, D. J. W. & Ip, Y. K., 1988. A comparative study of terrestrial adaptations of the gills in three mudskippers—Periophthalmus chrysospilos, Boleophthalmus boddaerti, and Periophthalmodon schlosseri. Biological Bulletin (Woods Hole, Massachusetts), 175, 434-438., Macintosh, D. J., 1979. Predation of fiddler crabs (Uca spp.) in estuarine mangroves. Biotrop Special Publication No. 10, Mangrove and Estuarine Vegetation in South East Asia, edited by P. B. L. Srivastava, Serdang, Malaysia, pp. 101-110. Macintosh, D. J., 1982. Fisheries and aquaculture significance of mangrove swamps with special reference to the Indo-West Pacific region. In, Recent Advances in Aquaculture, edited by J. F. Muir & R. Roberts, Croom Helm, London, pp. 3—36. MUDSKIPPERS 571 MacNae, W., 1968a. Mudskippers. African Wildlife, 22, 240-248. MacNae, W., 1968b. A general account of the fauna and flora of mangrove swamps and forests in the Indo-West-Pacific region. Advances in Marine Biology, 6, 73-270. MacNae, W., 1968c. Mangroves and their fauna. Australian Natural History, 7, 17-21. MacNae, W. & Kalk, M., 1962. The ecology of the mangrove swamps at Inhaca Island, Mocambique. Journal of Animal Ecology, 50, 19-34. Maekawa, K., Fukuda, Y., Okada, Z. & Imamura, K., 1968. [Intraepithelial blood capillaries in the air breathing organs of fish.] Bulletin of the Tokyo Medical College, 26, 793-799. (In Japanese with English abstract). Magnus, D. B. E., 1972. On the ecology of the reproduction behaviour of Periopthalmus kalolo Lesson at the east African coast. Proceedings of the International Congress of Zoology, 17, Theme 6, 1-4. Mahajan, B. A. & Srinivassan, M., 1988. Mercury pollution in the estuarine region around Bombay. Environmental and Technological Letters, 9, 331—336. Manickam, P. & Natarajan, G. M., 1985. Observations on the blood parameters of two air breathing mudskippers of South India. Journal of CurrentBioscience, 2, 19-22. Manna, G. K. & Prasad, R., 1974. Chromosome analysis in three species of fishes belonging to family Gobiidae. Cytologia, 39, 609—628. Mansuri, A. P. & Bhan, S., 1978. Adaptations to the osmotic stress in the marine euryhaline teleost, Periophthalmus dipes IV. Changes in total protein levels of various tissues. Journal of Animal Morphology and Physiology, 25, 139—146. Mansuri, A. P., Bhan, S. & Mathai, T., 1982. Salinity tolerance and survival of mud skippers, Boleophthalmus dentatus and Periophthalmus dipes of Saurashtra coast. Journal of Animal Morphology and Physiology, 29, 142—148. Marlier, G., 1938. Considerations sur les organes accessoires servant a la respiration aerienne chez les teleosteens. Annales de la Societe Royale Zoologique de Belgique, 69, 163-185. Matoba, M. & Dotsu, Y., 1977. Pre-spawning behaviour of the mudskipper Perioph thalmus cantonensis in Ariake Sound. Bulletin of the Faculty of Fisheries, Nagasaki University, No. 43, 23-33. McCulloch, A. R. & Ogilby, J. D., 1919. Some Australian fishes of the family Gobiidae. Records of the Australian Museum, 12, 193—291. Mehta, R., Mehta, H. S. & Devi, K., 1986. Evolutionary modifications in the girdles and fins of the gobiid fish Periophthalmus schlosseri (Pallas). Journal of the Andaman Science Association, 2, 47—50. Mehta, R., Mehta, H. S. & Devi, K. 1987., Adaptational modifications of pelvic fins in some gobiid fishes. Journal of the Andaman Science Association, 3, 152—154. Mehta, R., Mehta, H. S. & Devi, K., 1990. Structure and shape of mouth in some gobioid fishes. Environment and Ecology, 8, 668—671. Miller, P. J., 1973. The osteology and adaptive features of Rhyacichthys aspo (Teleostei: Gobioidei) and the classification of gobioid fishes. Journal of Zoology, London, 111, 397-434. Milward, N. E., 1974. Studies on the taxonomy, ecology and physiology of Queensland mudskippers. PhD thesis, University of Queensland, 276 pp. Misra, S., Dutta, A. K., Dhar, T., Ghosh, A., Choudhury, A. & Dutta, J., 1983. Fatty acids of the mudskipper Boleophthalmusboddaerti. Journal ofthe Science ofFood and Agriculture, 34, 1413-1418. Miyazaki, M. & Nakamura, T., 1980. [On the shape of muscles of the tongue of Boleophthalmus pectinirostris.] Journal Kurume Medical Association, 43, 807811. (In Japanese with English abstract). Mookerjee, H. K., Ganguly, D. N. & Mookherji, P. S., 1950. Study on the structures of the brains of some Indian fishes in relation to their feeding habits. Proceedings of the Zoological Society, Bengal, 3, 119-152. Morii, H., 1979. Changes with time of ammonia and urea concentrations in the blood and tissue of mudskipper fish, Periophthalmus cantonensis and Boleophthalmus 572 DAVID A. CLAYTON pectinirostris kept in water and on land. Comparative Biochemistry and Physiology, 64A, 235-243. Morii, H. & Kasama, K., 1989. [Bacterial flora in the digestive tracts of mudskipper fishes Boleophthalmus pectinirostris and Periophthalmus cantonensis. ] Nippon Suisan Gakkaishi, 55, 733-741. (In Japanese with English abstract). Morii, H., Nishikata, K. & Tamura, O., 1978. Nitrogen excretion of mudskipper fish Periophthalmus cantonensis and Boleophthalmus pectinirostris in water and on land. Comparative Biochemistry and Physiology, 60A, 189-193. Morii, H., Nishikata, K. & Tamura, O., 1981. Ammonia and urea excretion from mudskipper fishes Periophthalmus cantonensis and Boleophthalmus pectinirostris transferred from land to water. Comparative Biochemistry and Physiology, 63A, 23-28. Morton, B. & Morton J., 1983. The Sea Shore Ecology of Hong Kong. Hong Kong University Press, Hong Kong, 350 pp. Mukherjee, A. K., 1971a. Food-habits of water-birds of the Sundarban, 24-Parganas District, West Bengal, India—II. Herons and bitterns. Journal of the Bombay Natural History Society, 68, 37-64. Mukherjee, A. K., 1971b. Food-habits of water-birds of the Sundarban, 24-Parganas District, West Bengal, India—III. Egrets. Journal of the Bombay Natural History Society, 68, 691-716. Munk, O., 1970. On the occurrence and significance of horizontal band-shaped retinal areae in teleosts. Videnskabelige Meddelelserfra Dansk Naturhistorisk Forening i Khobenhavn, 133, 85-120. Munz, F. W., 1971. Vision: Visual pigments. In, Fish Physiology, Vol. 5, edited by W. S. Hoar & D. J. Randall, Academic Press, London, pp. 1-32. Murdy, E. O., 1986. Mudskippers of Malaysia. The lords of the mudflat. Freshwater and Marine Aquarium, November 1986, 20-23. Murdy, E. O., 1989. A taxonomic revision and cladistic analysis of the oxudercine gobies (Gobiidae: Oxudercinae). Records of the Australian Museum, Supplement 11, 1-93. Mutsaddi, K. B. & Bai, D. V., 1969a. Food and (Cuv. & Val.). Journal of the University Mutsaddi, K. B. & Bai, D. V., 1969b. Some Boleophthalmus dussumierei (Cuv. and feeding of Boleophthalmus dussumierei of Bombay, 38, 42—55. observations on habits and habitat of Val.). Journal of the University of Bombay, 39, 33-41. Mutsaddi, K. B. & Bai, D. V., 1970. Maturation and spawning of Boleophthalmus dussumierei. Journal of the University of Bombay, 39, 58-76. Mutsaddi, K. B. & Bai, D. V., 1973. Gobies from the intertidal region of Bombay. Indian Journal of Fisheries, 20, 476-486. Nandi, S. & Choudhury, A., 1983. Quantitative studies on the benthic macrofauna of Sagar Island, intertidal zone, Sunderbans, India. Mahasager—Bulletin of the National Institute of Oceanography, 16, 409—414. Natarajan, G. M. & Rajulu, G. S., 1983. Bimodal oxygen uptake in the amphibious mud-skipper Periophthalmus koelreuteri (Pallas) (Gobiidae: Teleostei). Journalof Animal Morphology and Physiology, 30, 177—180. Nateewathana, A. & Tantichodok, P. 1984. Species composition, density and biomass of macrofauna of a mangrove. In, Proceedings of the Asian Symposium on the Mangrove Environment-Research & Management, edited by E. Soepadmo et al, University of Malaya and UNESCO, Kuala Lumpur, pp. 258-285. Nishikawa, M. & Ishibashi. T., 1973. [Behavioural rhythms of gobioid fish, Boleoph thalmus chinensis (Osbeck) with special reference to the effect of light condition. ] Zoological Magazine (Tokyo), 82, 311 only. (In Japanese). Nishikawa, M., & Ishibashi, T., 1975a. [Endogenous rhythm and time sense in the mud skipper Periophthalmus cantonensis (Osbeck)]. Zoological Magazine (Tokyo), 84, 354 only. (In Japanese). MUDSKIPPERS 573 Nishikawa, M. & Ishibashi, T., 1975b. [Entrainment of the activity rhythm by the cycle of feeding in the mud-skipper Periophthalmus cantonensis]. Zoological Magazine (Tokyo), 84, 184-189. (In Japanese with English abstract). Nishikawa, S., Amaoka, K. & Nakanishi K., 1974. A comparative study of chromosomes of twelve species of gobioid fishes in Japan. Japanese Journal ofIchthyology, 21, 61-71. Niva B., Ojha, J. & Munshi, J. S. D., 1979. Bimodal oxygen uptake in relation to body weight of the amphibious mud-skipper Boleophthalmus boddaerti (Pall.). Indian Journal of Experimental Biology, 17, 752-756. Niva, B., Ojha, J. & Munshi, J. S. D., 1981. Morphometries of the respiratory organs of an estuarine goby. Boleophthalmusboddaerti. Japanese Journal ofIchthyology, 27, 316-326. Nogusa, S., 1957. Chromosome studies in Pisces VII. A comparative study of the chromosomes in six species of the Gobiidae. Japanese Journal of Ichthyology, 6, 141-146. Nogusa, S., 1960. A comparative study of the chromosomes in fishes with particular considerations on taxonomy and evolution. Memoires of Hyogo University, Agriculture, 3, 1—62. Nursall, J. R., 1974. Character displacement and fish behaviour, especially in coral reef communities. American Zoologist, 14, 1099-1118. Nursall, J. R., 1981. Behaviour and habitat affecting the distribution of five species of sympatric mudskippers in Queensland. Bulletin of Marine Science, 31, 730-735. Ogasawara, T., Ip, Y. K., Hasegawa, S., Hagiwara, Y. & Hirano, T., 1991. Changes in prolactin cell activity in the mudskipper Periophthalmus chrysospilos in response to hypotonic environment. Zoological Science (Tokyo), 8, 89-95. Oliva, O. & Skorepa, V., 1970a. The eye muscles in Periophthalmus Bloch et Schneider, 1801 and Chauliodus Bloch et Schneider, 1801 (Osteichthyes). Vestnik Ceskoslovenske Spolecnosti Zoologicke Acta Societatis Zoologicae Bohemoslovacae, 34, 50-57. Oliva, O. & Skorepa, V., 1970b. The eye muscles in Argyropelecus olfersi (Cuvier, 1829) and Boleophthalmus Val., 1837. Vestnik Ceskoslovenske Spolecnosti Zool ogicke Acta Societatis Zoologicae Bohemoslovacae, 34, 234-239. Pankow, H. & Hug, M. F., 1979. Diatoms in the stomach content of Pseudapocryptes dentatus a mudskipper from the Shatt-al-Arab estuary (Iraq). Wissenschaftliche Zeitschschrift der Wilhem-Pieck-Universitat Rostock. Mathematisch-Naturwissenschaftliche Reihe, 6, 547-554. Parker, T. J. & Haswell, W. A., 1962. A Textbook ofZoology. Macmillan, London, 952 pp. Patel, B., Bangera, V. S., Patel, S. & Balani, M. C, 1985. Heavy metals in the Bombay harbour area. Marine Pollution Bulletin, 16, 22—28. Patel, B. & Chandry, J. S., 1988. Mercury in the biotic and abiotic matrices along Bombay coast. Indian Journal of Marine Science, 17, 55-58. Patel, B. & Desai, K., 1976. Tidal rhythms and cyclic changes in the neuroendocrine complex of Boleophthalmus dentatus Cuv. & Val. Annals of Zoology (Agra), 12 69-95. Patel, B., Mulay, C. D. & Ganguly, A. K., 1975. Radioecology of Bombay harbour—a tidal estuary. Estuarine and Coastal Marine Science, 3, 13—42. Pearse, A. S., 1929. Observations on certain littoral and terrestrial animals at Tortugas, Florida, with special reference to migrations from marine to terrestrial habitats. Papers from the Tortugas Laboratory ofthe Carnegie Institute of Washington, No. 391. 205-223. Pearse, A. S., 1932. Observations on the ecology of certain fishes and crustaceans along the bank of the Malta River at Port Canning. Records of the Indian Museum, Calcutta, 34, 289-298. Pearse, A. S., 1933. The gobies at Paknam. Journal of the Siamese Natural History Society, Supplement, 9, 173—178. 574 DAVID A. CLAYTON Petit, M. G., 1921. Observations sur certains poissons des cotes de Madagascar pres- entant une adaptation a la locomotion terrestre. Bulletin du Museum National d'Histoire Naturelie Zoologique, 28, 216—220. Petit, M. G., 1922. Les Periophthalmes, poissons fouisseurs. Bulletin du Museum National d'Histoire Naturelie Zoologique, 27, 404-408. Petit, M. G., 1928. Novelle observations sur la biologie des Periophthalmes. Bulletin du Museum National d'Histoire Naturelle Zoologique, 34, 197-199. Polunin, I., 1972. Who says fish can't climb trees? National Geographic Magazine, 141, 84-92. Por, F. D., 1974. The hard-bottom mangroves of Sinai. Israel Journal of Zoology, 23, 211 only. Por, F. D., Dor, I, & Amir, A., 1977. The mangal of Sinai: Limits of an ecosystem. Helgoldnder wissenschaftliche Meersuntersuchungen, 30, 295—314. Pradhan, S. V., 1961. A study of blood of a few Indian fishes. I. Haemoglobin. Pro ceedings of the Indian Academy of Sciences, Section B, 54, 251-256. Prince Jeyaseelan, M. A. & Krishnamurthy, K., 1980. Role of mangrove forest of Pichavaram as fish nursery. Proceedings of the Indian National Science Academy, Part B, 46, 548-553. Qui, L.-Y., 1989. [Observation on two species of Periophthalmidae along Jiangsu coast]. Oceanologia et Limnologia Sinica, 20, 577-580. (In Chinese with English abstract). Qureshi, M. R. & Bano, A., 1971. Common gobioid fishes of sub-families Periophthalminae and Apocrypteinae. Pakistan Journal of Science, 23, 143-146. Rauther, M., 1910. Die akzessorischen Atmungsorgane der Knochenfische. Ergebnisse der Fortschritte der Zoologie, 2, 517—585. Reichelt, J., 1910. Boleophthalmus pectinirostris L. und Periophthalmus koelreuteri (Pall.) Bl. Blatterfur Aquarien- und Terrarienkunde, 21, 309-311. Rodewald, C, 1913. Boleophthalmus pectinirostris L. (?). Blatterfur Aquarien- und Terrarienkunde, 24, 565—566. Ryu, B. & Lee, J. 1979. [The life form of Periophthalmus cantonensis in the Gum river in summer.] Bulletin of the Korean Fisheries Society, 12, 71—78. (In Korean). Safer, A. M. A. & El-Sayed, N. K., 1986. The fine structure of the nephronic tubule of the mudskipper Periophthalmus koelreuteri (Pallas). Journal of Morphology, 187, 109-121. Safer, A. M. A., Tytler, P. & El-Sayed, N., 1982. The structure of the head kidney in the mudskipper, Periophthalmus koelreuteri (Pallas). JournalofMorphology, 174, 121-131. Salih, M. S. & Al-Jaffery, A., 1980. Morphological study of Boleophthalmus dentatus Cuvier & Valenciennes, 1837 with reference to its feeding habits. I. Habitat, external features and integument. Bulletin of the Biological Research Center, Baghdad., 12, 11-44. Sarker, A. L., Al-Daham, N. K. & Bhatti, M. N., 1980. Food habits of the mudskipper, Pseudapocryptes dentatus (Val.). Journal of Fish Biology, 17, 635—639. Sasekumar, A., 1974. Distribution of macrofauna in a Malayan mangrove shore. Journal of Animal Ecology, 43, 52—69. Sasekumar, A., Ong, T. L. & Thong, K. L., 1984. Predation of mangrove fauna by marine fishes. In, Proceedings of the Asian Symposium on the Mangrove Environment-Research & Management, edited by E. Soepadmo et al, University of Malaya and UNESCO, Kuala Lumpur, pp. 378-384. Schneider, H., 1990a. Schlammspringer in der Mangrove. Aquarium (Minden), 249, 19—23. Schneider, H., 1990b. Das Naturerlebnis: Schlammspringer in der Mangroven. Aquaria (St. Gallen), 35, 287-295. Schottle, E., 1932. Morphologie und Physiologie der Atmung bei wasser-, schlamm- und landlebenen Gobiiformes. Zeitschriftfur Wissenschaftliche Zoologie, 140, 1-114. Schreitmuller, W., 1911. Neues uber Boleophthalmus pectinirostris Linne. Blatter fur Aquarien- und Terrarienkunde, 22, 29—31. MUDSKIPPERS 575 Schreitmuller, W., 1914. Boleophthalmus pectinirostris Linne und Boleophthalmus boddaerti Pallas. Blatter fur Aquarien- und Terrarienkunde, 25, 41-43. Shaikh, Y. A. & Hiradhar, P. K., 1987. Histological observations on changes induced in some tissues of the edible mudskipper Boleophthalmus dussumieri exposed to sublethal concentrations of sodium fluoride. Journal of Animal Morphology and Physiology, 34, 69-75. Shaikh, Y. A. & Hiradhar, P. K., 1988. Alterations in the acid and alkaline phosphatase quantitiesin fluoride-exposed estuarine goby Boleophthalmus dussumieri. Fluoride, 21, 131-136. Siau, H. & Ip, Y. K., 1987. Activities of enzymes associated with phosphoenolypyruvate metabolism in the mudskippers, Boleophthalmus boddaerti and Periophthalmus schlosseri. Comparative Biochemistry and Physiology, 88B, 119—125. Siddiqi, S. A., 1974. Natural distribution and ecology of dark blotched mudskipper, Periophthalmus waltoni Koumans (Pisces: Gobiidae). Journal of the Marine Biological Association of India, 16, 843-844. Simpson, D. A., 1955. Fish out of water. Aquarium Journal, 26, 181-183. Singh, B. N. & Munshi, J. S. D., 1969. On the respiratory organs and mechanics of breathing in Periophthalmus vulgaris (Eggert). Zoologische Anzeiger, 183, 92-110. Singh, B. R., Mishra, A. P., Prasad, M. S. & Singh, I., 1989. Development of neomorphic air-breathing organs in three genera of estuarine gobies. Gegenbaurs Morphologische Jahrbuch, 135, 529-540. Singh, B. R., Yadav, A. N., Prasad, M. S., Mishra, A. P. & Singh, I., 1990. Neomorphic organ for C02 elimination in air-breathing teleosts. European Archives of Biology, 101, 257-268. Sircar, A. K. & Har, S. P., 1975. Ecology of an estuarine breeding ground of some gobioid fishes in west Bengal, India. In, Special Symposium on Marine Sciences, edited by B. Morton & J. R. Lee. Pacific Science Association, Government Printer, Hong Kong, pp. 63-64. Slooff, R. & Marks, E. N., 1965. Mosquitoes (Culicidae) biting a fish (Periophthalmidae). Journal of Medical Entomology, 2, 16 only. Soni, V. C, & George, B., 1986. Age determination and length weight relationship in the mudskipper Boleophthalmus dentatus. Indian Journal of Fisheries, 33, 321—234. Sowerby, A. De C, 1923. The mud skipper. China Journal of Science and Arts, 1, 343-345. Sponder, D. L. & Lauder, G. V., 1981. Terrestrial feeding in the mudskipper Perio phthalmus (Pisces: Teleostei) : A cineradiographic analysis. Journal of Zoology, London, 193, 517-530. Stebbins, R. C. & Kalk, M., 1961. Observations on the natural history of the mud skipper Periophthalmus sobrinus. Copeia, 1961, 18—27. Subrahmanyam, K., 1969. A karyotypic study of the estuarine fish Boleophthalmus boddaerti (Pallas) with calcium treatment. Current Science, 38, 437-439. Tamura, S. O., Morii, H. & Yuzuriha, M., 1976. Respiration of the amphibious fishes Periophthalmus cantonensis and Boleophthalmus chinensis in water and on land. Journal of Experimental Biology, 65, 97-107. Tamura, O. & Moriyama, T., 1976. On the morphological feature of the gill of amphi bious and air breathing fishes. Bulletin of the Faculty of Fisheries, Nagasaki University, 41, 1—8. Teal, J. M. & Carey, F. G., 1967. Skin respiration and oxygen debt in the mudskipper, Periophthalmus sobrinus. Copeia, 1967, 677-679. Troncy, P.-M. & Vassiliades, G., 1974. Acanthocephales parasites de poissons d'Afrique. Bulletin de VInstitut Fondamental d'Afrique Noire, Serie A: Sciences Naturelles, 36, 902-910. Tytler, P. & Vaughan, T., 1983. Thermal ecology of the mudskipper, Periophthalmus koelreuteri (Pallas) and Boleophthalmus boddarti (Pallas) of Kuwait Bay. Journal of Fish Biology, 23, 327-337. 576 DAVID A. CLAYTON Uchida, Y., Enomoto, N. & Miyaguchi, M., 1971. [On the heavy metal contents in marine products from Ariake Sea.] Agricultural Bulletin of Saga University, 32, 45—49. (In Japanese with English abstract). Van Dijk, D. E., 1959. A fish that walks, jumps and burrows. Nucleon, 1, 36—38. Van Dijk, D. E., 1960. Locomotion and attitudes of the mudskipper, Periophthalmus, a semi-terrestrial fish. South African Science, 56, 158—162. Van Dijk, D. E., 1978. Transition from aquatic to semi-terrestrial life in vertebrates: the gobioid mudskipper Periophthalmus, as model. IsraelJournal ofMedical Science, 14, 905-906. Venkateswarlu, T., 1966. A note on the oxygen carrying capacity of the blood in three gobiid fishes. Naturwissenschaften, 12, 3 only. Venkateswarlu, T., 1969. On Periophthalmus schlosseri Bl. Schn. from Vellar estuary, Portonovo, S. India. Science and Culture, 35, 492-493. Verma, G. K., 1968. Studies on the structure and behaviour of chromosomes of certain fresh water and marine gobid fishes. National Academy of Science of India, 37, 178 only. Verwey, J., 1930. Einiges uber die Biologie Ost-Indischer Mangrovekrabben. Treubia, 12, 167-261. Vestergaard, K., 1972. The biology of some mangrove fish. In, Anonymous, The Study Tour to East Africa, 1970, University of Copenhagen, pp. 317-319. Vivekanandan, E. & Pandian, T. J., 1979. Erythrocyte count and haemoglobin concen tration of some tropical fishes. Journal ofMadurai Kamaraj University (Science), 8, 71-75. Volsoe, H., 1939. The sea snakes of the Iranian Gulf and the Gulf of Oman. With a summary of the biology of the sea snakes. In, Danish Investigations in Iran Part 1, edited by K. Jessen & R. Sparch, Einar Munksgaard, Copenhagen, pp. 9-45. Volz, W., 1905a. Uber das Auge von Periophthalmus und Boleophthalmus. Mitteilungen der Naturforschenden Gesellschaft in Bern, 108—111. Volz, W., 1905b. Zur Kenntnis des Auges von Periophthalmus und Boleophthalmus. Zoologische Jahrbiicher, Abteilung fur Anatomie und Ontogenie der Tiere, 22, 331-346. Volz, W., 1905c. Sur l'oeil de Periophthalmus et Boleophthalmus. Archives des Sciences Physiques et Naturelles (Geneve), 4, 579—580. Von Anthouard, M. & Mignot, A., 1973. Schlammspringer. Aquarien Terrarien, 25, 330-333. Warren, J. H., 1990. The use of open burrows to estimate abundances of intertidal estuarine crabs. Australian Journal of Ecology, 15, 277—280. Welsch, U. & Storch, V., 1976. Elektronenmikroskopische Beobachtungen am Kiemenepithel des amphibischen Teleosteers Periophthalmus vulgaris. Zeitschrift fur Mikroskopisch-Anatomische Forschung, 90, 447—457. Whitley, G. P., 1931. New names for Australian fishes. Australian Zoologist, 6, 310—334. Whitley, G. P., 1953. Fishes collected by the Australian museum expedition, 1952. Records of the Australian Museum, 23, 123—132. Whitley, G. P., 1960. The mud-skipper. The Education Gazette, 54, 142-143. Whitley, G. P., 1968. Mud-skippers. The Australian Museum Leaflet, 17, 1-6. Willem, V. & Boelaert, R., 1937. Les manoeuvres respiratoires de Periophthalmus. Bulletin de L'Academie Royale des Sciences des Lettres et des Beaux-Arts de Belgique (Series 5), 23, 942-959. Xie, X.-G. & Zhang, Q.-Y., 1990. [Studies on the gonadal development in female mudskipper, Boleophthalmus pectinirostris]. Journal of Oceanography in Taiwan Straits, 9, 217-221. (In Chinese with English abstract). MUDSKIPPERS 577 Yadav, A. N., Prasad, M. S. & Singh, B. R., 1990. Gross structure of the respiratory organs and dimensions of the gill in the mudskipper, Periophthalmodon schlosseri (Bleeker). Journal of Fish Biology, 37, 383-392. Yadav, A. V. & Singh, B. R., 1989. Gross structure and dimensions of the gill in an air breathing estuarine goby, Pseudapocryptes lanceolatus. Japanese Journal of Ichthyology, 36, 252-259. Yamamoto, S., 1931. Die statische Refraktion des Auge verchiedener Tiere. Journal of Okayama Medical Society, 43, 1407-1425. Yamazoe, Y., 1970. Nutritional studies on fish in Ariake Bay (IV). Free sugars com position of Mutugoro (Boleophthalmus pectinirostris) and Warasubo (Odontamblyops rubicundis). Journal of the Japanese Society of Food and Nutrition, 23, 21-23. (In Japanese with English abstract). Yang, K.-G. & Zhang, Q.-Y., 1990. [Development stages of larval, juvenile and young mudskippers Boleophthalmus pectinirostris.] Journal of Xiamen University, Natural Science, 29, 76-80. (In Chinese with English abstract). Yew, D. T. & Wu, M. F., 1979. Ratio of rods and different cones in fishes of shallow sea. Anatomische Anzeiger, 145, 155-160. Zhang, H.-G. & He, D., 1989. [The comparison of the retino-motor response of Epinephalus fario and Boleophthalmus pectinirostris under various light condi tions.] Journalof Xiamen University, Natural Science, 28, 647-650. (In Chinese with English abstract). Zhang, Q.-Y., Hong, W.-S., Dai, Q.-N., Cai, Y.-Y. & Zhang, J., 1987. Studies on the artificial propagation and larval rearing of the mudskipper (Boleophthalmus pectinirostris L.).] Journal ofXiamen University, Natural Science, 26, 366-373. (In Chinese with English abstract). Zhang, Q.-Y., Hong, J.-L., Dai, Q.-N., Zhang, J., Cai, Y.-Y., & Huang, J.-L., 1989. Studies on induced ovulation, embryonic development and larval rearing of the mudskipper (Boleophthalmus pectinirostris). Aquaculture, 83, 375-385. Zhang, Q.-Y. & Zhang, J., 1988. [On the feeding, growth and survival of larval mud skipper (Boloeophthalmus pectinirostris)]. Journal of Fisheries, China, 12, 203-211. (In Chinese with English abstract).

Log In

Mudskippers (Literature Review)

Mudskippers (Literature Review)

Related Papers

RELATED PAPERS