Mudskippers are currently included in the subfamily Oxudercinae (Teleostei, Acanthopterygii, Percomorpha, Gobiiformes, Gobioidei, Gobiidae; Nelson, 2006; Thacker et al., 2015), whose diagnostic characters were defined by Hoese (1984).
Murdy (1989) proposed a complete cladistic revision of this group. This revision was very welcome to mudskipper taxonomy, which has been intricate and chaotic ever since the 17th century.

In particular, Murdy described 13 morphospecies of Periophthalmus. Five new species of the genus Periophthalmus were subsequently described (P. magnuspinnatus Lee et al., 1995, P. spilotus Murdy & Takita, 1999; P. walailakae Darumas & Tantichodok, 2002; P. darwini Larson & Takita, 2004; P. takita Jafaar & Larson 2008), while P. novaeguineaensis Eggert was redescribed (Jafaar & Larson 2008), and P. variabilis Eggert restored (Jafaar et al. 2009). One species of Boleophthalmus was also described after Murdy (1989): Boleophthalmus poti Polgar et al., 2013. Therefore, presently oxudercines include 41 morphospecies and 10 genera:


Apocryptes Valenciennes, 1837* (1)

Apocryptodon Bleeker, 1874 (2)

Boleophthalmus Valenciennes, 1837* (6)

Oxuderces Eydoux & Souleyet, 1850 (2)

Parapocryptes (Valenciennes, 1837)* (2)

Periophthalmus Bloch & Schneider, 1801 (18)

Periophthalmodon Bleeker, 1874 (3)

Pseudapocryptes Bleeker, 1874 (2)

Scartelaos Swainson, 1839 (4)

Zappa Murdy, 1989 (1)


*In: Cuvier & Valenciennes, 1837


The monophyly of Oxudercinae has been recently questioned by several molecular studies (Agorreta & Rüber, 2012; Agorreta et al., 2013; Thacker, 2003; Tornabene et al., 2013). In particular, oxudercine gobies appear paraphyletic relative to amblyopines (subfamily Amblyopinae, or worm-eel gobies, which appear as polyphyletic).

Periophthalmus also appears as the sister group of all other oxudercine and amblyopine genera so far examined.
Consistently, a molecular systematic and phylogeographic study (Polgar et al., 2014) showed that this genus has a long evolutionary history (>30 million years) in coastal wetland ecosystems.

The vernacular term 'mudskippers' implies the presence of an amphibious lifestyle. For this reason, 'mudskippers' should be used to indicate those oxudercine species that "are fully terrestrial for some portion of the daily cycle" (Murdy, 1989). Such species are included in the genera Boleophthalmus, Periophthalmodon, Periophthalmus, and Scartelaos. Based on the observations by Polgar et al. (2010) on Zappa confluentus, this latter species should be included in this definition.

Cladogram of the genera of Oxudercinae (Maximum Parsimony tree), based on morphological, osteological, and eco-ethological characters; the gobionelline goby Evorthodus lyricus was used as outgroup (Murdy, 1989). The original dataset was analysed in PAUP v.4 (Swofford, 2002). With permission from the author

Evolutionary history of air-breathing vertebrates.
Relative abundance and diversity of each group is roughly proportional to the band width.
Red bars show the cited groups according to Nelson, 2006.
Redrawn from Graham (1997), with permission from Elsevier

At a higher taxonomic level, oxudercine gobies are percomorphs (Percomorpha; Betancur-R et al., 2014; Near et al., 2013; Thacker et al., 2015), actinopterygian fishes (Actinopterygii, or 'ray-finned fishes') that have reached high degrees of specialization to an aquatic lifestyle (Nelson, 2006).

Also for this reason, the most peculiar characteristic of mudskippers is the high degree of specialization to amphibious life relative to all living and extinct aquatic vertebrates (Clack, 2002; Graham, 1997; Graham et al., 2007).

A recent molecular study (Thacker et al., 2015) proposed the genus Trichonotus (Thichonotidae, or sand-divers), as the sister group of gobies sensu lato (Gobioidei), within the order Gobiiformes. Trichonotus species are marine benthonic fishes living in shallow, sandy bottom habitats. This and other molecular studies (e.g., Near et al., 2013; Smith & Craig, 2007; Smith & Wheeler, 2004) are inconsistent with a previous morphological study (Winterbottom, 1993), which proposed Hoplichthyidae (ghost-flatheads, a group of marine benthonic scorpaeniforms) as the sister group of gobioids.

The oldest gobiid fossil is approximately 34 million years old (Miller, 1974). However, gobioids invaded shallow marine coastal areas at least since the Early Eocene, about 52 millions of years ago (Bajpai & Kapur, 2004; Nolf & Stringer, 2003). Different studies on molecular clocks estimated the origin of Gobioidei at 70-170 million years ago (e.g., Chakrabarti et al., 2012; Near et al., 2013).

No oxudercine fossils are known at present.
The highest species richness and degree of endemism occur in the Indo-Malayan divisional association, while only one species (Periophthalmus barbarus) is present along the coasts of Western Africa (Murdy,1989). The present geographic distribution and the pseudo-congruent pattern of the distribution of some species, between Eastern Africa and Indonesia (Polgar et al., 2014) suggest that the group originated in the Easten Tethis, and then reached the present biodiversity hotspot in Southeast Asia, due to the biogeographic mechanims knwon as 'Go East', together with several other shallow marine animal groups (Renema et al., 2008).

During the Miocene (23.5-5.3 millions of years ago), the distribution of these species was probably restrained within the sub-tropical belt, due to a generalized global cooling. About 9-16 millions of years ago the 'Tethydian gateways' between the Atlantic and the Indian Ocean were completely closed by repeated salinity crises of the newly born Mediterranean Sea, caused by the collision of the African and Eurasian continents (Harzauser et al., 2007). Like all other Atlantic and Indo-Pacific fish species, the West-African ancestor of P. barbarus was probably definitely isolated from Indo-Pacific populations.

The reconstruction of the palaeocurrents' pattern during all the Eocene and Oligocene periods (53.0-23.5 millions of years ago) suggests that the ancestor of P. barbarus could have reached the African continent and the eastern coasts of North and South America in one or more steps if suitable environments had been available (Stille et al., 1996; Ellison et al., 1999). At present though, no oxudercine is known to exist or have existed in the Americas.

Present distribution of oxudercine species and limits of the divisional associations (red lines) based on maximal endemism. Numbers indicate the total number of oxudercine species. Modified from Murdy (1989), with permission from the author.
Green lines indicates the distribution of mangrove forests.
Redrawn from UNEP-WCMC, 2005: Coral Reefs and Mangroves of the World

The present distribution suggests a very close linkage of all oxudercine species to mangrove ecosystems and tropical tidal mudflats.
It is possible that this association established in a relatively early phase of their evolutionary history.
Palaeontological data show that mangrove ecosystems had been widespread in the Tethys Sea since the Paleocene-Eocene (60-35 millions of years ago: Ellison et al., 1999), approximately when the first gobioid fishes appear in the fossil record.

Distribution of modern and extinct mangrove genera based on the fossil record, superimposed on coeval positions of continental land masses.
Redrawn from Ellison et al., 1999, with permission from the first author and from the editor