https://orcid.org/0000-0001-6678-5080
Figures
Abstract
The suborder Gobioidei is among the most diverse groups of vertebrates, comprising about 2310 species. In the fossil record gobioids date back to the early Eocene (c. 50 m.y. ago), and a considerable increase in numbers of described species is evident since the middle Miocene (c. 16 m.y. ago). About 40 skeleton-based gobioid species and > 100 otolith-based species have been described until to date. However, assignment of a fossil gobioid species to specific families has often remained tentative, even if well preserved complete specimens are available. The reasons are that synapomorphies that can be recognized in a fossil skeleton are rare (or absent) and that no phylogenetic framework applicable to gobioid fossils exists. Here we aim to overcome this problem by developing a phylogenetic total evidence framework that is suitable to place a fossil skeleton-based gobioid at family level. Using both literature and newly collected data we assembled a morphological character matrix (48 characters) for 29 extant species, representing all extant gobioid families, and ten fossil gobioid species, and we compiled a multi-gene concatenated alignment (supermatrix; 6271 bp) of published molecular sequence data for the extant species. Bayesian and Maximum Parsimony analyses revealed that our selection of extant species was sufficient to achieve a molecular ‘backbone’ that fully conforms to previous molecular work. Our data revealed that inclusion of all fossil species simultaneously produced very poorly resolved trees, even for some extant taxa. In contrast, addition of a single fossil species to the total evidence data set of the extant species provided new insight in its possible placement at family level, especially in a Bayesian framework. Five out of the ten fossil species were recovered in the same family as had been suggested in previous works based on comparative morphology. The remaining five fossil species had hitherto been left as family incertae sedis. Now, based on our phylogenetic framework, new and mostly well supported hypotheses to which clades they could belong can be presented. We conclude that the total evidence framework presented here will be beneficial for all future work dealing with the phylogenetic placement of a fossil skeleton-based gobioid and thus will help to improve our understanding of the evolutionary history of these fascinating fishes. Moreover, our data highlight that increased sampling of fossil taxa in a total-evidence context is not universally beneficial, as might be expected, but strongly depends on the study group and peculiarities of the morphological data.
Citation: Gierl C, Dohrmann M, Keith P, Humphreys W, Esmaeili HR, Vukić J, et al. (2022) An integrative phylogenetic approach for inferring relationships of fossil gobioids (Teleostei: Gobiiformes). PLoS ONE 17(7): e0271121. https://doi.org/10.1371/journal.pone.0271121
Editor: Juan Marcos Mirande, Fundacion Miguel Lillo, ARGENTINA
Received: September 15, 2021; Accepted: June 23, 2022; Published: July 8, 2022
Copyright: © 2022 Gierl et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files. The data underlying the results presented in the study are available from https://figshare.com/s/ed10b9a5ac382f856a20.
Funding: BR received funding from the Deutsche Forschungsgemeinschaft (grant number RE-1113/20). RS received funding from the Ministry of Culture of the Czech Republic (grant number DKRVO 2019-2023/6.III.d National Museum, 00023272). https://www.dfg.de/ https://www.mkcr.cz/ The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The suborder Gobioidei (Percomorpha: Gobiiformes) is among the most diverse groups of fishes, encompassing about 320 genera and about 2310 species [1, 2]. Gobioids live in marine, brackish, and freshwater habitats, and form an important component of reef faunas [3–5]. They show a wide range of specializations, including the ability to survive on land for long periods of time (mudskippers; e.g. [6]), being diadromous and alternating freshwater and sea water during their lifecycle [7], and forming symbioses with certain shrimp species [8, 9].
Gobioidei belongs to Gobiiformes (sensu Thacker et al. [10]), together with Kurtidae, Apogonidae, and Trichonotidae, and consists of nine families: Rhyacichthyidae, Odontobutidae, Milyeringidae, Eleotridae, Butidae, Thalasseleotrididae, Gobiidae, and Oxudercidae (= formerly Gobionellidae [1]), and the extinct †Pirskeniidae [11] (Fig 1). Phylogenetic relationships within extant Gobioidei are generally well resolved based on molecular data [10, 12, 13]. Morphologically, the extant Gobioidei can be divided into two major groups based on the number of branchiostegal rays (bones supporting the gill membrane [14, 15]). Possession of five branchiostegal rays is considered to be the derived state within Gobioidei, supporting the clade Gobiidae + Oxudercidae (denoted 5brG here; see Fig 1), which is also recovered on the basis of molecular data [10, 12, 13]. In contrast, Gobioidei with six branchiostegal rays (6brG) form a paraphyletic group in molecular phylogenies [12, 16], indicating that this condition is plesiomorphic within Gobioidei. The only gobioid that has seven branchiostegal rays, of which the last ray is expanded as is typical for Gobioidei (see [17]), is the genus †Pirskenius (denoted 7brG here), and phylogenetic analyses placed it within the paraphyletic 6brG [11]; see Fig 1. The closest living relatives of Gobioidei, the Apogonidae, Kurtidae, and Trichonotidae, all have seven branchiostegals and no expansion of the last branchiostegal ray [1, 17–19].
brG = branchiostegal-rayed Gobioidei.
Previous gobioid classifications date back to more than 100 years ago. Regan [20] had already recognized that two main groups within Gobioidei, which he termed Eleotridae and Gobiidae, can be distinguished based on the configuration of the pelvic fins (separate in Eleotridae vs. united in Gobiidae), the shape of the palatine (L- vs. T-shaped), the endopterygoid (present vs. mostly absent), and the composition of the shoulder girdle (presence vs. absence of the dorsal coracoid [= scapula]). Regan’s Eleotridae conforms to the paraphyletic 6brG, and his Gobiidae are reflected in the 5brG. Since then, several efforts have been made to classify groups within Gobioidei using morphological characters (see [21] for a comprehensive compilation). Akihito [22] and Hoese [23] noted that the loss of the anteriormost branchiostegal ray appears to be characteristic for the most advanced groups in Gobioidei (= 5brG). Miller [24] erected the family Rhyacichthyidae and reclassified Gobioidei into Rhyacichthyidae and Gobiidae, with the latter comprising seven subfamilies, Eleotrinae and Xenisthminae (now Eleotridae), Gobionellinae and Tridentigerinae (now included in Oxudercidae), Gobiinae and Kraemeriinae (now Gobiidae), and †Pirskeniinae (referring to †Pirskeniidae Obrhelová, 1961) [10, 12, 25]. Harrison [26] tried to resolve gobioid interrelationships with the help of features of the palatopterygoquadrate complex and indicated some putative relationships among six taxa of the 5brG. Pezold [27] recognized a distinct head pore configuration as an autapomorphy of the subfamily Gobiinae (now in Gobiidae). Hoese and Gill [28] recognized the family Odontobutidae and used 16 characters to propose interrelationships among Rhyacichthyidae, Odontobutidae, and Gobiidae; their Gobiidae included the Butinae, Eleotridinae and Gobiinae, each of which is now assigned family rank [12, 29]. Other authors used morphological data for phylogenetic studies of subgroups of various families, as Larson [30] has done for the Mugilogobius group of the Gobiidae, and Murdy [31] for the Oxudercinae. Most of the more recent works that include morphological characters have focused on smaller groups within Gobioidei (e.g. [29, 32–36]).
The known fossil record of Gobioidei is relatively modest in comparison to their recent diversity and comprises about 40 species that are known from skeletal material. In older studies, these species have been assigned to genera or families by comparing meristic characters, such as counts of vertebrae and fin rays (e.g. [37–40]). More recent authors have also considered the systematic value of certain osteological features (e.g. shape of the palatine, first dorsal fin-pterygiophore formula, number of branchiostegal rays, configuration of the palatine-ectopterygoid complex, presence of the entopterygoid), which enabled a tentative systematic placement of some fossil gobioids in relation to extant taxa [41–47].
However, determining the systematic context of fossil gobioids remains a challenge. Firstly, fossil gobioids apparently include several extinct lineages (e.g. [42, 43, 45, 48–51]). Secondly, a phylogenetic analysis of extant gobioids based on morphological characters is hampered by the rarity of synapomorphic characters [28], and becomes even more difficult in fossil gobioids, which usually preserve only skeletal traits and otoliths [11]. Thirdly, all comparative approaches suffer from a limited knowledge of the range of skeletal characters of extant gobioids, which is also due to the sheer number of species (currently 2310, see above) and the inaccessibility of some rare taxa [52, 53]. This explains why, with the exception of †Pirskeniidae [11], no phylogenetic analyses have yet been conducted to position fossil gobioids within the phylogeny of extant taxa.
The objective of this study was to place ten selected fossil gobioid species within an integrative ("total evidence") phylogenetic framework including published molecular sequence data from extant species in combination with morphological data for both fossil and extant species. The choice of fossils was largely based on our previous works (see section Material and Methods). The overall idea was that placement of fossil gobioids in the context of a rigorous phylogenetic analysis would greatly help to improve our understanding of the evolutionary history of these fascinating fishes.
Institutional abbreviations
BMNH, former acronym for British Museum of Natural History (now NHMUK); FMNH, Field Museum of Natural History, Chicago, Illinois, USA; IRSNB, Royal Institute of Natural Sciences, Brussel, Belgium; LACM, Los Angeles County Museum of Natural History, Los Angeles, California, USA; MNHN, Muséum national d’Histoire naturelle, Paris, France; MRAC, Musée royal de l’Afrique centrale, Tervuren, Belgium; NHMUK, Natural History Museum in London, United Kingdom; NMP, National Museum, Prague, Czech Republic; SMNS, State Museum for Natural History, Stuttgart, Germany; WAM, Western Australian Museum, Welshpool, Australia; ZM-CBSU, Zoological Museum Collection of the Biology Department at Shiraz University, Iran; ZSM, Bavarian State Collection of Zoology, Munich, Germany.
Material and methods
Ethic statement
All specimens originate from museum collections (see Table 1) and no specimens were sacrificed for this work.
Birdsong et al. [60], Hoese and Gill [28] and Gill and Mooi [15] are used for morphology of many or all species. Colour codes for molecular data: Agorreta et al. [12] = green; Near et al. [63] = blue; Thacker et al. [10] = red; for not color-coded cells see entries in Genbank via given accession numbers.
Compilation of the taxon set
The ingroup species of the Gobioidei used in this study comprise 29 extant species; the apogonid Sphaeramia nematoptera is used as outgroup (Table 1). The Odontobutidae, Milyeringidae, Butidae, and Thalasseleotridae are each represented by two species, the Rhyacichthyidae and the Eleotridae each by three, the Oxudercidae by five, and the Gobiidae by ten species. Criteria for species selection were the availability of ‘total-evidence data’ including molecular as well as morphological data of the skeleton and otolith data. Morphological data of four species were coded based solely on published information: the odontobutid Odontobutis obscurus (data from [44, 54, 55]), the milyeringid Typhleotris madagascariensis (data from [56]), the thalasseleotridid Thalasseleotris iota (data from [57–59]), and the oxudercid Eucyclogobius newberryi (data from [60, 61]); no otolith data were available for T. madagascariensis. This selection of extant species, although modest in terms of species numbers of Oxudercidae and Gobiidae, was sufficient to achieve a molecular ‘backbone’ as its phylogenetic analysis produced a well-resolved tree that fully agrees with published hypotheses (see Results).
Ten fossil gobioid species were added to the extant taxon set (Table 1). We selected those that we had examined in previous works [11, 42–44, 46, 62], and added also the oldest known putative gobioid so far, †Carlomonnius quasigobius (Table 1). A short overview of all fossil species is provided in S1 Appendix.
Study and compilation of morphological characters
Literature data, when available, were used to compile phylogenetically informative morphological characters for the extant species (see Table 1 for details). Additionally, X-ray images were produced for the extant species with a Faxitron Ultrafocus facility at the ZSM and examined to determine numbers of vertebrae, fin elements and pterygiophores, and configuration of the caudal skeleton. After radiography, otoliths were extracted from the same specimens and prepared for scanning electron microscopy (SEM) imaging (using a HITACHI SU 5000 Schottky FE-SEM at the Department of Earth- and Environmental Sciences, LMU Munich). SEM images served as the basis for the identification of the otolith characters (Fig 2), which are used here for the first time within a phylogenetic matrix. They include (i) the overall otolith shape (six character states: trapezoid/triangular; long rectangular; high rectangular; quadratic; rounded; longish-ovate), (ii) the posterodorsal projection (two states: present; absent), (iii) the sulcus shape (three states: perciform-like, shoe sole, shoe sole/specialized), and (iv) the sulcus shape and position (three states: shoe sole + centred, shoe sole + shifted anteriorly, not shoe sole). Plesiomorphic otolith character states were defined according to the condition seen in the otolith of the outgroup (see S1 Table for details). The otoliths of the included extant gobioid species are shown in Fig 3, the otolith of the outgroup species Sphaeramia nematoptera is depicted in Fig 2. For the fossil species, skeletal and otolith characters were largely compiled from previous works, but some additional characters could also be added (see Results).
Images depict left otoliths (sagittae) in medial view. Scale bars: 0.5 mm.
Images depict left otoliths (sagittae) in medial view, except for Lesueurigobius sanzi, Aphia minuta, Asterropteryx semipunctata, Dormitator maculatus and Odontobutis obscurus, which represent right sagittae that were mirrored for better comparison. Otoliths were not available for the milyeringid Typhleotris madagascariensis. For sources of otoliths see Table 1. Scale bars: 0.5 mm.
A total of 48 morphological characters were assembled. Thirty-eight characters concern bony structures of the skeleton, four characters relate to cartilage, membrane, or tendon configurations, four refer to otolith morphology, two concern the presence and type of ctenii on the scales, and one is a morphometric character (see S1 Table Part B for all characters). Character states were determined according to literature data and our morphological investigations (based on X-rays, SEM images of otoliths) (see S1 Table Part A for details).
All taxa and characters were assembled in Mesquite 3.61 [87]. We used presence/absence coding (1/0) or up to ten states, depending on the character (see S1 Table).
Preparation of the molecular data matrix
Molecular sequence data for extant species were assembled based on previously published sequences of five markers: rDNA (12S rRNA, tRNA-Val, 16S rRNA), cytb, rag1, zic1, and sreb2. The sequences were mainly from the study of Agorreta et al. [12], supplemented by some data from Near et al. [63] and Thacker et al. [10]. Molecular data were downloaded from GenBank, aligned in AliView 1.26 [88] with MUSCLE [89], followed by manual adjustment and exclusion of ambiguous regions where necessary. Individual gene alignments were then concatenated into a supermatrix (6271 bp) with SeaView 5.0.4 [90]. For details and GenBank accession numbers see Table 1. All data matrices are available on figshare (https://figshare.com/s/ed10b9a5ac382f856a20).
Phylogenetic analyses
We conducted phylogenetic analyses for the extant species based on the morphological character matrix, the molecular supermatrix, and based on the combined molecular and morphological (total evidence) datasets. Adding all ten fossils to the total evidence character set resulted in a collapse of the molecular backbone (see Discussion for possible reasons). Therefore, we used a step-wise approach: (i) a single fossil species was added, (ii) two fossil species of the same genus were added, (iii) four fossil species were added. We used the Bayesian Markov Chain Monte Carlo (MCMC) approach implemented in MrBayes versions 3.2.6 and 3.2.7a [91], as well as implied-weights maximum parsimony (IW-MP, [92]) implemented in TNT 1.5 [93] to infer phylogenies.
In MrBayes, we assigned the Mkv+G model [94, 95] to the morphological data, and separate GTR+G models [95, 96] to each molecular partition. For each analysis, we ran 2 x 4 MCMC chains in parallel for 5 x 106 generations. We used the "sump" command in MrBayes as well as Tracer 1.7.1 [97] to check for convergence and discarded the first 10% of samples of each analysis as burn-in before summarizing the remaining samples in 50% majority-rule consensus (MRC) trees including posterior probability (PP) values for all clades.
In TNT, we employed new-technology searches (with sectorial search, ratchet, drift, and tree fusing enabled; init. addseqs = 100; find min. length = 10) and a concavity constant of K = 12. A 50% majority-rule consensus tree was calculated if more than one most parsimonious tree was found. For assessing clade support in the IW-MP analysis, we used standard bootstrap resampling [98] (500 replicates; new technology search; init. addseq = 10; find min. length = 5). Phylogenetic trees were visualized in FigTree 1.4.4 [99]; tree files are available on figshare (https://figshare.com/s/ed10b9a5ac382f856a20).
Results
Skeleton data
The character state of each character of each species (extant and fossil) is provided in the S1 Table; unknown character states were coded with a question mark.
Two characters could be coded for the first time for †Eleogobis brevis based on the specimens NHMUK PV OR 42779 and 42780, deposited in the Natural History Museum in London: the presence of a single anal fin pterygiophore inserting before the haemal spine of the first caudal vertebra (AP = 1), and the position of the penultimate branchiostegal on the ceratohyal. Re-inspection of the type specimens of †Gobius jarosi revealed that also in this species the penultimate branchiostegal is located on the ceratohyal. Finally, a count of three to four anal fin pterygiophores (AP = 3–4) could be determined for †Lepidocottus aries based on re-inspection of the specimens used in Gierl et al. [44].
Otolith data
Fig 3 depicts the otoliths of the extant species included in this study. Short descriptions of the otoliths are given below.
Family Gobiidae.
Amblygobius phalaena (Valenciennes, 1837).–High-rectangular, margins crenate, dorsal rim strongly lobed, pronounced posterodorsal projection present. Sulcus centered, shoe-sole shape with slender cauda and wide ostium with dorsal lobe.
Aphia minuta (Risso, 1810).–Rounded, margins smooth, short projection slightly below level of ostium tip. Sulcus shifted anteriorly, rather shoe-sole shape, cauda strongly reduced, ostium with dorsal lobe and pointed tip.
Asterropteryx semipunctata Rüppell, 1830.–Rectangular, higher than long. Dorsal, anterior and ventral margins slightly crenulate, small posterodorsal projection. Sulcus centered, shoe-sole shape, with rounded cauda and ostium. See Schwarzhans et al. [50] for additional details of the same otolith.
Cryptocentrus cinctus (Herre, 1936).–High-rectangular, margins smooth, with a projection in the middle of the dorsal margin and a median constriction on anterior and posterior rims. Sulcus centered, shoe-sole shape with small, rounded cauda and a broad ostium with ostial lobe.
Discordipinna griessingeri Hoese and Fourmanoir, 1978.–Long-rectangular, smooth margins with median constrictions anterior and posterior defining four projections. Sulcus centered, shortened, oval specialized, surrounded by a bulging crista.
Glossogobius giuris (Hamilton, 1822).–Long-rectangular, slightly crenate margins, posterodorsal projection and a sub-median projection on the posterior rim. Sulcus centered, shoe-sole shape, cauda elongated, ostium with strong dorsal lobe.
Gobius niger Linnaeus, 1758.–Long-rectangular, dorsal margin slightly undulated, prominent posterodorsal and posteroventral projections, small anteroventral projection. Sulcus centered, shoe-sole shape, elongated with rounded cauda and ostium with dorsal lobe.
Lesueurigobius sanzi (de Buen 1918).–Quadratic with rounded dorsal margin, dorsal and anterior margins faintly lobed, other margins smooth, submedian constriction on posterior margin. Posterodorsal projection short and rounded, weak anteroventral projection. Sulcus centered, shoe-sole shape, rounded cauda, ostium with pointed tip. See Schwarzhans et al. [50] for additional details of the same otolith.
Ptereleotris evides (Jordan and Hubbs, 1925).–High-rectangular, margins sinuate; small posteroventral and more pronounced anteroventral projection. Sulcus centered, shoe-sole shape, with rounded cauda and ostium.
Tigrigobius multifasciatus (Steindachner, 1876).–Quadratic, margins undulated, cusp on dorsal rim. Sulcus shifted anteriorly, shoe-sole shape, cauda rounded, ostium with pronounced elongate tip.
Family Oxudercidae.
Awaous flavus (Valenciennes, 1837).–High-rectangular, dorsal and posterior margins crenate. Sulcus centered, shoe-sole shape. Cauda and ostium rounded, ostium with marked dorsal lobe.
Chlamydogobius eremius (Zietz, 1896).–Quadratic, margins slightly undulated, small posterodorsal projection. Sulcus centered, shoe-sole shape, with small, rounded cauda and broad ostium with marked dorsal lobe.
Eucyclogobius newberryi (Girard, 1856).–Quadratic with slightly curved dorsal margin, small posterodorsal projection, sulcus centered, shoe-sole shape, with well-developed cauda and ostium.
Gobioides broussonnetii Lacepède, 1800.–Rounded, margins irregularly crenate, small posterodorsal cusp. Sulcus centered, shoe-sole shape, rounded cauda, large ostium with rounded tip.
Pomatoschistus flavescens (Fabricius, 1779).–Quadratic, margins smooth. Sulcus centered, shoe-sole shape, weakly developed cauda, rounded ostium.
Family Thalasseleotrididae.
Grahamichthys sp.–Rounded, all margins smooth, no projections. Sulcus centered, shoe-sole shape, cauda and ostium rounded.
Thalasseleotris iota Hoese and Roberts, 2005.–Long-rectangular, margins smooth, with prominent praeventral projection and slightly rounded anterodorsal expansion. Sulcus centered, shoe-sole shape, cauda and ostium slender. See Schwarzhans [58] for additional details of the same otolith.
Family Butidae.
Kribia cf. nana (Boulenger, 1901).–Long-rectangular, all margins relatively smooth, small posterodorsal projection, posteroventral expansion (bulge). Sulcus centered, shoe-sole shape, slightly shifted anteriorly. Rounded cauda about half the length of the ostium.
Oxyeleotris marmorata (Bleeker, 1852).–Long-rectangular, anterior and posterior margins distinctly incised in the middle, pointed posterodorsal and rounded anterodorsal projections. Sulcus centered, slender-to-shoe-sole shape, slightly shifted anteriorly.
Family Eleotridae.
Dormitator maculatus (Bloch, 1792).–Long-rectangular, dorsally rounded. Anterior and ventral margins finely crenulated, other margins smooth. Sulcus centered, shoe-sole shape; cauda almost as large as ostium and posteriorly pointed.
Hypseleotris compressa (Krefft, 1864).–High-rectangular. All margins smooth, anterior and posterior margins with sub-median constriction at the level of the sulcus. Sulcus centered, shoe-sole shape, rounded cauda. Ostium with angled dorsal lobe.
Tateurndina ocellicauda Nichols, 1955.–High-rectangular, all margins relatively smooth except for dorsal margin, which is serrate. Sulcus centered, rounded-to-shoe-sole shape, cauda and ostium only weakly differentiated.
Family Milyeringidae.
Milyeringa veritas Whitley, 1945.–Trapezoid and smooth with posteroventral projection and dorsally pointing anterodorsal projection. Sulcus centered, rounded-to-shoe-sole shape, surrounded by a crista, cauda and ostium not clearly differentiated.
Family Odontobutidae.
Odontobutis obscurus (Temminck and Schlegel, 1845).–Trapezoid, dorsal margin with tip in its posterior half, anterior margin with spine-like dorsal projection, ventral margin crenulated and with spine-like posterior projection, posterior margin deeply sinuate. Sulcus centered, supra-median, relatively small, perciform-like. Cauda slender, posteriorly bent, ostium rounded and distant from anterior rim.
Perccottus glenii Dybowski, 1877.–Trapezoid with sinuate dorsal and ventral margins, posterior margin with median notch, anterior margin with sub-median notch. Sulcus centered, rounded-to-shoe-sole shape, cauda slightly shorter than ostium.
Family Rhyacichthyidae.
Protogobius attiti Watson and Pöllabauer, 1998.–Trapezoid/triangular, rounded with short median projection on posterior margin, dorsal margin smooth, ventral margin lobed. Sulcus shifted anteriorly, perciform-like with slender, relatively short cauda and broad ostium; ostium opened to anterior rim.
Rhyacichthys aspro (Valenciennes, 1837).–Trapezoid/triangular with short median projection on posterior and dorsal margins, dorsal margin slightly undulated with prominent anterodorsal bulge, ventral margin crenate. Sulcus shifted anteriorly, perciform-like with slender, relatively short cauda and broad ostium; ostium anteriorly closed.
Rhyacichthys guilberti Dingerkus and Séret, 1992.–Trapezoid/triangular, ventral margin lobed, dorsal margin smooth. Sulcus shifted anteriorly, perciform-like with slender, relatively short cauda and broad ostium; ostium opened to anterior rim.
Phylogenetic analyses
Phylogenetic relationships of the extant species.
Molecular data.–The Bayesian phylogeny based on molecular data (Fig 4A) recovers all families as monophyletic. Notably, although we used a restricted number of species, the tree is completely congruent with the trees published by Agorreta et al. [12] and Thacker et al. [10]: Rhyacichthyidae and Odontobutidae are sister groups, and together they are sister to the rest of the gobioid families. Milyeringidae, Eleotridae, Butidae, and Thalasseleotrididae are successive sister groups to the 5brG clade, which is composed of well-supported Oxudercidae and Gobiidae.
A Tree based on published DNA data of the 29 gobioid species used in this study (average standard deviation of split frequencies between two independent runs [ASDSF] = 0.003776; for sources of molecular data see Table 1). B Tree based on morphological characters of the extant species only (ASDSF = 0.002446). Scale bar, average number of substitutions per site respectively character changes per character.
Maximum Parsimony analysis produced a single most parsimonious tree (S1 Fig in S1 File), which is similar to the Bayesian tree, but some nodes have poor bootstrap support (BS). The tree is topologically identical to the Bayesian tree on family-level, only within Gobiidae there are some minor differences concerning the positions of Amblygobius and Asterropteryx; BS for several nodes is rather low compared to the support in the Bayesian tree.
Morphological data.–In the Bayesian phylogeny restricted to the extant species, the 5brG clade (Gobiidae + Oxudercidae) is recovered with maximum support, but the internal structure of the clade is completely unresolved (Fig 4B). The thalasseleotridid species G. radiatus and Th. iota group together and are sister to 5brG. The Thalasseleotrididae + 5brG clade is maximally supported, consistent with molecular phylogenetic results. Also, similar to the established molecular phylogeny, Eleotridae, Butidae, and Milyeringidae are closely related to that clade (1.00). However, only Eleotridae and Milyeringidae are recovered as monophyletic (0.91, 0.82), whereas Butidae are paraphyletic (Fig 4B). Monophyly of Odontobutidae is also not resolved, and in contrast to the molecular phylogeny the odontobutid species are closer to the above assemblage (1.00) instead of being sister to Rhyacichthyidae, which are recovered paraphyletic (Fig 4B).
The Maximum Parsimony analysis recovered one most parsimonious tree (S2 Fig in S1 File), which is overall similar to the Bayesian tree, but many nodes have very poor bootstrap support (BS). Within the 5brG clade (BS = 95%) the reciprocal monophyly of Gobiidae and Oxudercidae is not recovered. Thalasseleotrididae is monophyletic and highly supported as sister to 5brG (99%). As in the Bayesian tree, among the remaining families Eleotridae (71%) and Milyeringidae (52%) are supported as monophyletic.
Total-evidence approach.–The Bayesian phylogeny inferred from the total-evidence dataset (= combined molecular and morphological data) including only extant species is topologically identical to the molecular phylogeny but shows maximum support for Rhyacichthyidae and Rhyacichthyidae + Odontobutidae (1.00 and 1.00 vs. 0.62 and 0.64, respectively) (S3 Fig in S1 File).
The Maximum Parsimony analysis recovered one most parsimonious tree (S4 Fig in S1 File), which is similar to the Bayesian tree. Most nodes have good bootstrap support, except for some clades within 5brG that are weakly supported. Concerning the topology, the only difference is that the positions of the oxudercid genera Eucyclogobius, Pomatoschistus and Chlamydogobius are resolved while they form a polytomy in the Bayesian tree (see S3 Fig vs. S4 Fig in S1 File).
Phylogenies including extant and all ten fossil species.
Morphological data.–In the Bayesian phylogeny, the 5brG clade (Gobiidae + Oxudercidae)–including the two fossil Gobius spp.–is recovered only with moderate support (0.70), and internal relationships are again unresolved (Fig 5A). The rest of the tree is topologically similar to the one based on the morphology of the extant species only (Fig 4B), Thalasseleotrididae are now recovered with slightly higher support (0.80 vs. 0.77). The two †Eleogobius spp. are placed between Thalasseleotrididae and 5brG (0.60), but monophyly of the genus is not resolved. The †Pirskenius and †Paralates spp. form a weakly supported clade (0.64) that is sister to the above assemblage (0.73), but only †Pirskenius (i.e., †Pirskeniidae) is resolved as monophyletic (0.87). The remaining two fossil species, †Lepidocottus aries and †Carlomonnius quasigobius, are clearly more related to Thalasseleotrididae + 5brG than to Rhyacichthyidae and Odontobutidae (0.84), but their exact placements are not resolved.
A Tree based on only the morphological data set (ASDSF = 0.006790). B Tree based on the total evidence data set (ASDSF = 0.010856). Scale bars, average number of substitutions per site respectively character changes per character.
The Maximum Parsimony analysis using the same data set recovered six most parsimonious trees. The resulting 50% majority rule consensus tree (S5 Fig in S1 File) is overall consistent with the one based on extant species only (S2 Fig in S1 File) but is very poorly resolved on a deeper level; most nodes have BS < 50%. The two fossil Gobius spp. are included in the 5brG in a clade with Gobius niger, Discordipinna and Lesueurigobius. The two †Eleogobius spp. are sister to 5brG but monophyly of the genus remains unresolved. †Carlomonnius quasigobius is placed in a clade containing Eleotridae and Kribia nana (Butidae). †Pirskenius is monophyletic (64%) but its position, as well as that of the remaining three fossil species, is not further resolved.
Total-evidence approach.–The Bayesian phylogenetic analysis of the total-evidence dataset including all 29 extant and the ten fossil species produced a consensus tree with a largely collapsed backbone (Fig 5B) compared to the tree based on only the morphological data set of the same taxa (Fig 5A). It contains numerous polytomies and shows poor support for many deeper nodes. The 5brG clade (Gobiidae + Oxudercidae) forms a polytomy with the Thalasseleotrididae, the Butidae, the clade of †Paralates + †Pirskenius and three further fossil species (†Eleogobius brevis, †E. gaudanti, †“Gobius” francofurtanus). Nevertheless, several groups can be recovered within this polytomy. Oxudercidae, which were not resolved in the phylogeny based only on morphology (Fig 5A), are now recovered with high support (0.93), as well as some clades within Gobiidae that correspond to the molecular phylogeny, i.e. the Aphia (0.89) and Glossogobius (0.72) clades. †Gobius jarosi is weakly (0.55) placed as sister to the extant G. niger. The clade of †Pirskenius + †Paralates (see above) is recovered with slightly weaker support (0.59 vs. 0.64), while support for monophyly of †Pirskeniidae is slightly increased (0.93 vs. 0.87). †Lepidocottus aries and †Carlomonnius quasigobius are resolved as members of Butidae, albeit with low support (0.61).
The consensus tree of the two most parsimonious trees of the same data set (S6 Fig in S1 File) is even less well resolved, with very weak BS for many nodes. The 5brG clade is recovered, with reciprocally monophyletic Oxudercidae and Gobiidae. In contrast to the Bayesian tree, †“Gobius” francofurtanus is sister to Gobius niger (< 50%) (vs. not resolved in the Bayesian tree), whereas †Gobius jarosi is sister to Lesueurigobius (also < 50%) (vs. sister to G. niger). †Eleogobius is monophyletic and sister to the 5brG clade, but again BS for this is negligible. †Pirskenius is monophyletic (65%) but its position, as well as that of the remaining four fossil species, is not further resolved.
Total-evidence phylogenies including one to four fossil species.
In the following analyses, a single fossil was added to the data set of the extant species and the trees were inferred based on the morphological and molecular data (total-evidence approach).
†Carlomonnius quasigobius.–The Bayesian phylogeny is topologically identical to the molecular and the total evidence phylogenies based on extant species only; †Carlomonnius quasigobius is placed within Butidae (Fig 6A). Support values are high (>0.90) for most clades, but not for Butidae (0.74 vs. 1.00 in the molecular tree). The single Maximum Parsimony tree retained recovered all recent families, like the Bayesian tree (S7 Fig in S1 File). However, in this tree †C. quasigobius is placed as sister to Thalasseleotrididae + 5brG, albeit with very low support (< 50%).
A †Carlomonnius quasigobius was added to the extant species (ASDSF = 0.003939). B †Lepidocottus aries was added to the extant species (ASDSF = 0.000854). Scale bars, average number of substitutions per site respectively character changes per character.
†Lepidocottus aries.–The Bayesian phylogeny (Fig 6B) and the single Maximum Parsimony tree retained (S8 Fig in S1 File) reveal the same topology and almost the same support values as described above for the trees including †C. quasigobius. †Lepidocottus aries is placed within Butidae in both the Bayesian and the Maximum Parsimony tree, with high support values in the former (0.95), but very low support in the latter (< 50%).
†Gobius jarosi and †“Gobius” francofurtanus.–In the Bayesian tree (Fig 7A), topologies and support values are similar as described above. In comparison to the Bayesian total-evidence phylogeny based on extant taxa only (S3 Fig in S1 File), somewhat decreased support values concern two internal gobiid clades: the Aphia-lineage (0.95 vs. 1.00) and the clade containing Glossogobius and two members of the Cryptocentrus-lineage (0.90 vs. 1.00). The two fossil species are recovered as successive sister groups to G. niger with moderate support (0.77) (Fig 5D). In the consensus tree of two Maximum Parsimony trees (S9 Fig in S1 File), †“Gobius” francofurtanus is sister to Gobius niger while the position of †G. jarosi is unresolved in a clade with Lesueurigobius sanzi, Tigrigobius multifasciatus, Asterropteryx semipunctata, Amblygobius phalaena, and Aphia minuta (S9 Fig in S1 File). When only one of the two fossil species is included, a sister-relation with G. niger is apparent in each case, in both the Bayesian and the Maximum Parsimony tree (S10–S13 Figs in S1 File).
A †“Gobius” francofurtanus and †G. jarosi were added to the extant species (ASDSF = 0.014107). B Both †Eleogobius species were added to the extant species (ASDSF = 0.004114). Scale bars, average number of substitutions per site respectively character changes per character.
†Eleogobius brevis and †E. gaudanti.–The Bayesian phylogeny inferred from the total-evidence dataset including all extant species and either †E. brevis or †E. gaudanti (S14, S15 Figs in S1 File) is topologically identical to the molecular phylogeny, and almost no decrease of support values is seen. †Eleogobius brevis is recovered as sister to Gobius niger with good support (0.87), while †E. gaudanti shows a well-supported (0.87) sister relation to the Thalasseleotrididae. When both species of †Eleogobius are added, they are recovered in a polytomy with the Thalasseleotrididae (0.88) (Fig 7B), and there is slightly decreased support for the 5brG clade (0.90 vs. 1.00), and also for some of the gobiid clades (e.g. Tigrigobius + G. niger: 0.91 vs. 1.00) compared to the total evidence tree using only the extant species (S3 Fig in S1 File).
In the Maximum Parsimony trees, when only one of the †Eleogobius spp. is included, its position matches that recovered in the Bayesian analyses, albeit with poor support (S16, S17 Figs in S1 File). When both species are included, they form a clade and are sister to the Thalasseleotrididae (< 50%, S18 Fig in S1 File).
†Pirskenius diatomaceus and †P. radoni.–In the Bayesian tree, topologies are the same as described above; the two species of †Pirskenius are recovered as sister to Thalasseleotrididae (Fig 8A). However, in comparison to the Bayesian total-evidence phylogeny based on extant taxa only, decreased support values occur for the 5brG clade (0.89 vs. 1.00), the Thalasseleotrididae (0.65 vs. 1.00), and for the gobiid clade of Tigrigobius + Gobius niger (0.90 vs. 1.00). Support for the clade Thalasseleotrididae plus one of the †Pirskenius species is similar when only †P. radoni is added (0.69) (S19 Fig in S1 File), but higher when only †P. diatomaceus is involved (0.91) (S20 Fig in S1 File).
A †Pirskenius diatomaceus and †P. radoni were added to the extant species (ASDSF = 0.017673). B †Paralates bleicheri and †Pa. chapelcorneri were added to the extant species (ASDSF = 0.006995). Scale bars, average number of substitutions per site respectively character changes per character.
The single Maximum Parsimony tree including both species of †Pirskenius recovers all recent families as monophyletic (S21 Fig in S1 File) and the genus is recovered as sister to Thalasseleotrididae + 5brG; its monophyly is supported with 74% BS. Adding solely †P. radoni or †P. diatomaceus results in the same topology and similar support values (S22, S23 Figs in S1 File) as seen in the tree including both species.
†Paralates bleicheri and †Pa. chapelcorneri.–Inclusion of both species of †Paralates does not recover a relationship of these two fossil taxa with any of the extant clades in the Bayesian phylogeny, rather they form a polytomy with the Odontobutidae + Rhyacichthyidae clade and the clade containing all other families (Fig 8B). The Bayesian phylogeny that includes only †Pa. bleicheri resolves this species as sister to Odontobutidae (S24 Fig in S1 File), albeit with relatively low support (0.60). In comparison with the total-evidence phylogeny including only extant taxa, support for Odontobutidae decreased (0.61 vs. 1.00), while the high support for all other deeper nodes was not affected. The Bayesian phylogeny that contains only †Pa. chapelcorneri recovers this species as sister to the clade containing the 5brG, Thalasseleotrididae, Butidae and Eleotridae, but with very low support (0.52) (S25 Fig in S1 File; note also decreased support for some backbone nodes in this tree).
In the Maximum Parsimony analyses, inclusion of both species could not resolve their phylogenetic position, as in the Bayesian tree, but here the resolution of the backbone of the tree is even more severely reduced (S26 Fig in S1 File). The Maximum Parsimony results for †Pa. bleicheri (S27 Fig in S1 File) match those of the Bayesian analyses, whereas when only †Pa. chapelcorneri is included, it is placed within Thalasseleotrididae (< 50%, S28 Fig in S1 File).
†Pirskenius spp. and †Paralates spp.–When all four †Paralates and †Pirskenius species were included in the Bayesian analysis, they were recovered in a †Paralates + †Pirskenius clade (0.74), which was resolved as sister to Thalasseleotrididae with moderate support (0.79) (S29 Fig in S1 File). In the Maximum Parsimony tree †Pirskenius is recovered monophyletic (64%) but its position and the positions of the two †Paralates species within a clade together with Butidae, Thalasseleotrididae and 5brG are not resolved; overall, this tree shows very poor resolution at its backbone (S30 Fig in S1 File).
Discussion
In this study, we have assembled for the first time a dataset comprising molecular and morphological data for Gobioidei that encompasses both extant and fossil species. This approach was necessary as a phylogenetic analysis of the extant species based solely on their morphological characters could only resolve those clades for which morphological apomorphies are known, i.e. 5brG, Thalasseleotrididae, Thalasseleotrididae+5brG, and Eleotridae [see 15, 28], while Butidae, Odontobutidae and Rhyacichthyidae each were recovered as paraphyletic. The overall objective was to investigate whether a fossil gobioid species can be confidently placed at family level in the tree of the extant Gobioidei using a Bayesian or Maximum Parsimony total-evidence phylogenetic approach. The results reveal mostly well supported placement at family level when a single fossil species is added to the total evidence data set of the extant species, especially in the Bayesian setting.
Five of the fossil species used here had previously been assigned at family level based on a comparative approach: †Lepidocottus aries had been placed within Butidae [44], †“Gobius” francofurtanus and †Gobius jarosi had been proposed as members of Gobiidae [46, 100], †Pirskenius spp. had been placed in its own family †Pirskeniidae [49] and a sister group relation of †Pirskeniidae to Thalasseleotrididae + 5brG has been suggested [11]. Each of those fossil taxa have been recovered in corresponding positions in the present study (Figs 6B, 7A and 8A). This implies that using comparative morphology has been a very appropriate method to classify those fossils. The family assignment of the remaining five fossil species analyzed here (†Carlomonnius quasigobius, †Eleogobius spp., †Paralates spp.) had been left as incertae sedis in previous work because they possess a mosaic set of characters that is not known among extant gobioids [42, 43, 64].
†Carlomonnius quasigobius originates from the Eocene of Monte Bolca in northern Italy [64], from the lower to middle Eocene (Ypresian to Lutetian, c. 50–40 Ma, see [101]). It is placed in Butidae in our study (Fig 6A), and it shares with some Butidae (especially with Kribia) a very small size, but any comparative approach would not have assigned †C. quasigobius to this family because it has only five branchiostegal rays (vs. six in Butidae) and a continuous dorsal fin (vs. divided). The feature that seems to have placed †C. quasigobius within the Butidae and close to Kribia is the number of 11 branched and segmented caudal fin rays, which is uncommon among other Gobioidei. Nevertheless, given that †C. quasigobius is the oldest gobioid species currently known [64], an assignment to the Gobiidae (with which it shares the number of five branchiostegal rays) seems unlikely and its classification within the Butidae appears to be more plausible. It would expand the fossil record of Butidae from the early Oligocene (c. 30 Ma, [102]) to the early-middle Eocene (c. 50–40 Ma). However, an additional possibility is that †C. quasigobius is a member of an extinct gobioid family or a “stem gobioid”showing a mixture of derived characters (five branchiostegal rays, dorsal postcleithrum absent, 11 (7+6) segmented and branched caudal-fin rays, four pelvic-fin rays) and plesiomorphic ones (dorsal fin continuous, 24 (10+14) vertebrae, autogenous haemal spine of the second preural centrum, first two abdominal centra shortened, first dorsal-fin pterygiophore inserting in the second interneural space) [64].
In case of †Eleogobius and †Paralates, the resulting phylogenies indicate that these genera are not monophyletic and their species may not even belong to the same family (see Figs 7B and 8B, S14–S18 Figs and S24–S28 Figs in S1 File). The two species of †Eleogobius have been reported from the lower and middle Miocene (c. 17–14 Ma) of Central Europe, specifically southern Germany [42], Austria ([103], as Gobius), Switzerland [104, 105], and Croatia [106]. They have been interpreted as belonging to the same genus because they share a T-shaped palatine, absence of an endopterygoid and presence of six branchiostegal rays, and their otoliths are superficially similar, but show differences to recognize the two species [42]. Of those characters, the T-shaped palatine and absence of an endopterygoid can be considered as apomorphic [23, 28], and would support assignment to Gobiidae, which is proposed, based on our phylogenetic results, for †E. brevis (Fig 7A). In contrast, the phylogenetic position of †E. gaudanti near Thalasseleotrididae (Fig 7B) is difficult to understand as no potential synapomorphies are known that can be recognized in a fossil. However, it is more or less consistent with the hypothesis of Gierl and Reichenbacher [42] that †Eleogobius is somewhat “in-between” the 5brG clade and the 6brG gobioids. Furthermore, Bradić-Milinović et al. [41] have recognized a difference in the arrangement of the branchiostegals in the two Eleogobius species, which appears to support the possibility that †Eleogobius is not monophyletic.
In the case of †Paralates, a family assignment had not previously been proposed. †Paralates bleicheri has only been found in lower Oligocene deposits of the southern Upper Rhinegraben [40]. An assignment of †P. bleicheri to the Odontobutidae, as indicated in our analysis, receives little support based on its skeletal traits as all shared characters with the Odontobutidae represent plesiomorphic character states (e.g. seven spines in the first dorsal fin, 8–9 rays in the second dorsal fin, presence of postmaxillary process) (see S1 Table). Finds of fossil skeletons of †P. bleicheri with otoliths preserved in situ would be necessary to reinforce this hypothesis.
†Paralates chapelcorneri originates from the upper Eocene “Chapelcorner Fish bed” of southern England (Isle of Wight) [38, 43]. No otoliths have yet been reported from the “Chapelcorner Fish bed”. The family assignment of this species remains a topic of future research based on new material of this species. Moreover, a possible relationship between †Paralates spp. and †Pirskenius spp. was indicated in the Bayesian and Maximum Parsimony trees (S25, S26 Figs in S1 File), which is possibly due to their specific combination of plesiomorphic (e.g. seven spines in the first dorsal fin, nine anal-fin rays) and apomorphic traits (e.g. 12 abdominal vertebrae, presence of interneural gap).
Our study also yields some new insights from a methodological point of view. Adding morphological data from only extant species to the molecular dataset had practically no influence on the tree topology and support values (S3 Fig in S1 File). Likewise, the inclusion of a single fossil or of two congeneric species did not change the tree topology, only sometimes some support values decreased (Figs 6–8). However, when all fossils were included in the total evidence phylogenetic framework, the resolution of relationships between families and most fossil taxa dramatically collapsed (Fig 5B). Notably, the morphological phylogeny including the extant and all fossil species was less severely collapsed in the backbone of the tree as the 5brG clade and the Thalasseleotrididae were resolved (Fig 5A). It seems that in the case of the total evidence phylogeny the fossil taxa added a high level of conflicting phylogenetic signals, which could not be overcome by the molecular data despite the latter having orders of magnitude more characters and harbouring strong signal for resolving gobioid phylogeny. A possible explanation is that the fossil taxa do not only add morphological information, but also a lot of question marks to the matrix, because, depending on their preservation, some morphological traits cannot be determined. An additional (or alternative) explanation is that many extinct gobioid clades and families, each with a unique character combination, existed in the past [11, 41, 45, 49]. These cannot be ‘forced’ in the tree of extant species and eventually may also be responsible for the collapse of the molecular backbone of the extant families. This highlights that increased sampling of fossil taxa in a total-evidence context is not universally beneficial, as might be expected, but strongly depends on the study group and peculiarities of the morphological data.
Conclusion
We have presented a total-evidence dataset comprising molecular and morphological data of 29 extant gobioid species representing all families. Bayesian and Maximum Parsimony analyses revealed that this dataset is sufficient to achieve a molecular ‘backbone’ that fully conforms to previous molecular work. The new dataset can be used to analyze the family assignment of fossil skeletal-based gobioid species using Bayesian and Maximum Parsimony total-evidence phylogenetic approaches, which has not been possible before.
Our phylogenetic analyses confirmed the family assignment of those fossil gobioid species for which such a placement had been proposed in previous works. It is thus evident that comparative morphology remains an appropriate method to classify some gobioid fossils. However, our phylogenetic analyses could also suggest relationships of fossil gobioid species for cases where the comparative approach did not yield conclusive results. An example is †Carlomonnius quasigobius, which is the oldest gobioid fossil to date and our phylogeny suggests that it could be a possible member of the Butidae, which would expand the known age of fossil butids from the early Oligocene (30 Ma) to the early-middle Eocene (40–50 Ma). Although such positioning of †C. quasigobius remains somewhat speculative for now, it can give hints to look at certain fossil species from a different and new perspective.
We think that the total evidence framework presented here will be beneficial for all future work dealing with the phylogenetic placement of fossil gobioids and thus will help to improve our understanding of the evolutionary history of these fascinating fishes.
Supporting information
S1 Appendix. Short description of the 10 fossil specimens included in this study.
https://doi.org/10.1371/journal.pone.0271121.s001
(DOCX)
S1 Table. Part A.
Distribution of character states of the characters 1–54 among the examined 29 extant and 10 fossil gobioid species. Part B. List of characters, their states and literatur sources. Colour indicates how the character state was determined;? indicates that character state is not known.
https://doi.org/10.1371/journal.pone.0271121.s002
(XLSX)
S1 File. Caption for S1–S30 Figs.
Maximum Parsimony trees with bootstrap values and 50% majority-rule consensus (MRC) Bayesian trees with posterior probabilities based on the different data sets used in this study.
https://doi.org/10.1371/journal.pone.0271121.s003
(DOCX)
Acknowledgments
For providing technical assistance and access to specimens from the SNSB-ZSM collection we thank D. Neumann and U. Schliewen, respectively, (both SNSB-ZSM, Munich, Germany), with special thanks to the latter for help in the acquisition of specimens and insightful discussions. We are grateful to E. Bernard (NHMUK, London, UK), S. Merker (SMNS, Stuttgart, Germany), M. Parrent (MRAC, Tervuren, Belgium), and T. Přikryl (Charles University, Prague, Czech Republic), who all helped to study the specimens kept in the collections of their institutions. Sincere thanks go to W. Schwarzhans (Hamburg, Germany) for providing photographs of the otoliths of Lesueurigobius sanzi, Asterropteryx semipunctata, Eucyclogobius newberry, Thalasseleotris iota and Dormitator maculatus (shown in Fig 3) and D. Nolf and K. Hoedemakers (both IRSNB, Brussels, Belgium) for supplying the otolith SEM image of Odontobutis obscurus (shown in Fig 3). We thank G. Wörheide (LMU Munich, Germany) for providing access to computational resources. Finally we thank the reviewers and the Academic Editor of PLOS One for their constructive and valuable remarks, which greatly helped to improve the manuscript.
References
- 1.
Nelson JS, Grande TC, Wilson MVH. Fishes of the World, Fifth Edition. Hoboken, New Jersey: John Wiley & Sons, inc.; 2016. 752 p.
- 2. Parenti P. A checklist of the gobioid fishes of the world (Percomorpha: Gobiiformes). Iran J Ichthyol. 2021;8:1–480.
- 3. Brandl SJ, Goatley CHR, Bellwood DR, Tornabene L. The hidden half: ecology and evolution of cryptobenthic fishes on coral reefs. Biological Reviews. 2018;93(4):1846–73. Epub May 7. WOS:000446427600008. pmid:29736999
- 4.
Patzner RA, Van Tassell JL, Kovačić M, Kapoor BG, editors. The biology of gobies. 1 ed. Enfield, New Hampshire: Science Publishers Inc.; 2011.
- 5. Tornabene L, Robertson DR, Baldwin CC. Varicus lacerta, a new species of goby (Teleostei, Gobiidae, Gobiosomatini, Nes subgroup) from a mesophotic reef in the southern Caribbean. ZooKeys. 2016;596:143–56. WOS:000378055700010; PubMed Central PMCID: PMC4926659. pmid:27408581
- 6.
Jaafar Z, Murdy EO, editors. Fishes Out of Water: Biology and Ecology of Mudskippers. 1st ed. Raton Boca: Taylor & Francis; 2017.
- 7. Keith P. Biology and ecology of amphidromous Gobiidae of the Indo-Pacific and the Caribbean regions. J Fish Biol. 2003;63(4):831–47. Epub Sep 26. WOS:000186047000001.
- 8. Karplus I, Szlep R, M T. Goby-shrimp partner specificity. I. Distribution in the northern Red Sea and partner specificity. J Exp Mar Biol Ecol. 1981;51(1):1–19. Epub Mar 31, 2003.
- 9.
Karplus I, Thompson AR. The partnership between gobiid fishes and burrowing alpheid shrimps. In: Patzner RA, Van Tassell JL, Kovačić M, Kapoor BG, editors. The biology of gobies. Enfield, NH: Science Publishers Inc.; 2011. p. 559–607.
- 10. Thacker CE, Satoh TP, Katayama E, Harrington RC, Eytan RI, Near TJ. Molecular phylogeny of Percomorpha resolves Trichonotus as the sister lineage to Gobioidei (Teleostei: Gobiiformes) and confirms the polyphyly of Trachinoidei. Molecular phylogenetics and evolution. 2015;93:172–9. Epub Aug. MEDLINE:26265255; PubMed Central PMCID: PMC26265255. pmid:26265255
- 11. Reichenbacher B, Přikryl T, Cerwenka AF, Keith P, Gierl C, Dohrmann M. Freshwater gobies 30 million years ago: New insights into character evolution and phylogenetic relationships of †Pirskeniidae (Gobioidei, Teleostei). PloS one. 2020;15(8):e0237366. Epub Aug 24. WOS:000565550400035. pmid:32834000
- 12. Agorreta A, San Mauro D, Schliewen U, Van Tassell JL, Kovačić M, Zardoya R, et al. Molecular phylogenetics of Gobioidei and phylogenetic placement of European gobies. Molecular phylogenetics and evolution. 2013;69(3):619–33. MEDLINE:23911892. pmid:23911892
- 13. McCraney WT, Thacker CE, Alfaro ME. Supermatrix phylogeny resolves goby lineages and reveals unstable root of Gobiaria. Molecular phylogenetics and evolution. 2020;151:106862. Epub May 28. pmid:32473335
- 14. Akihito, Iwata A, Kobayashi T, Ikeo K, Imanishi T, Ono H, et al. Evolutionary aspects of gobioid fishes based upon a phylogenetic analysis of mitochondrial cytochrome b genes. Gene. 2000;259(1–2):5–15. ISI:000166338300003. pmid:11163956
- 15. Gill AC, Mooi RD. Thalasseleotrididae, new family of marine gobioid fishes from New Zealand and temperate Australia, with a revised definition of its sister taxon, the Gobiidae (Teleostei: Acanthomorpha). Zootaxa. 2012;3266(1):41–52.
- 16. Thacker CE. Biogeography of goby lineages (Gobiiformes: Gobioidei): origin, invasions and extinction throughout the Cenozoic. J Biogeogr. 2015;42(9):1615–25. Epub Jun 13. WOS:000359376600004.
- 17. McAllister DE. Evolution of the branchiostegals and classification of teleostome fishes. Bull natn Mus Can (Biol Ser). 1968;77:1–239.
- 18.
Carpenter KE, Niem VH. The living marine resources of the Western Central Pacific. Volume 5. Bony fishes part 3 (Menidae to Pomacentridae). Rome: Food and Agriculture Organization of the United Nations (FAO); 2001. 2791–3380 p.
- 19. Fraser TH. Comparative osteology of the shallow water cardinal fishes (Perciformes: Apogonidae) with reference to the systematics and evolution of the family. Ichthyological Bulletin of the J L B Smith Institute of Ichthyology. 1972;34:1–105.
- 20. Regan CT. The osteology and classification of the gobioid fishes. The Annals and Magazine of Natural History [Eighth Series]. 1911;8(48):729–33.
- 21.
Van Tassell JL, Tornabene L, Taylor MS. A history of gobioid morphological systematics. In: Patzner RA, Van Tassell JL, Kovačić M, Kapoor BG, editors. The biology of gobies. Enfield, NH: Science Publishers Inc.; 2011. p. 3–22.
- 22. Akihito. A systematic examination of the gobiid fishes based on the Mesopterygoid, Postcleithra, Branchiostegals, pelvic fins, Scapula, and Suborbital. Jpn J Ichthyol. 1969;16(3):93–104. Epub Jun 28, 2010. BCI:BCI197051058231.
- 23.
Hoese DF. Gobioidei: relationships. In: Moser HG, Richards WJ, Cohen DM, Fahay MP, Kendall J A. W.,Richardson SL, editors. Ontogeny and systematics of fishes. Gainesville, Florida: American Society of Ichthyologists and Herpetologists; 1984. p. 588–91.
- 24. Miller PJ. The osteology and adaptive features of Rhyacichthys aspro (Teleostei: Gobioidei) and the classification of gobioid fishes. Journal of Zoology. 1973;171(3):397–434.
- 25. Thacker CE, Hardman MA. Molecular phylogeny of basal gobioid fishes: Rhyacichthyidae, Odontobutidae, Xenisthmidae, Eleotridae (Teleostei: Perciformes: Gobioidei). Molecular phylogenetics and evolution. 2005;37(3):858–71. Epub 2005/06/25. pmid:15975831.
- 26. Harrison IJ. Specialization of the gobioid palatopterygoquadrate complex and its relevance to gobioid systematics. Journal of Natural History. 1989;23(2):325–53.
- 27. Pezold F. Evidence for a monophyletic Gobiinae. Copeia. 1993;1993(3):634–43. BCI:BCI199396109145.
- 28. Hoese DF, Gill AC. Phylogenetic relationships of eleotridid fishes (Perciformes, Gobioidei). B Mar Sci. 1993;52(1):415–40. ISI:A1993KX23000015.
- 29. Thacker CE. Phylogeny of Gobioidei and placement within Acanthomorpha, with a new classification and investigation of diversification and character evolution. Copeia. 2009;2009(1):93–104. ISI:000263748000013.
- 30. Larson HK. A revision of the gobiid fish genus Mugilogobius (Teleostei: Gobioidei), and its systematic placement. Rec West Aust Mus. 2001;62:1–233.
- 31. Murdy EO. A taxonomic revision and cladistic analysis of the oxudercine gobies (Gobiidae: Oxudercinae). Rec West Aust Mus Suppl. 1989;11:1–93. Epub Jun 16, 2009. BCI:BCI199089035472.
- 32. Harold AS, Winterbottom R, Munday PL, Chapman RW. Phylogenetic relationships of Indo-Pacific coral gobies of the genus Gobiodon (Teleostei: Gobiidae), based on morphological and molecular data. B Mar Sci. 2008;82(1):119–36. WOS:000252842600008.
- 33. Lord C, Bellec L, Dettaï A, Bonillo C, Keith P. Does your lip stick? Evolutionary aspects of the mouth morphology of the Indo-Pacific clinging goby of the Sicyopterus genus (Teleostei: Gobioidei: Sicydiinae) based on mitogenome phylogeny. Journal of Zoological Systematics and Evolutionary Research. 2019;57(4):910–25. Epub May 30.
- 34. Pezold F. Phylogenetic analysis of the genus Gobionellus (Teleostei: Gobiidae). Copeia. 2004;2004(2):260–80. WOS:000221336700006.
- 35. Tornabene L, Deis B, Erdmann MV. Evaluating the phylogenetic position of the goby genus Kelloggella (Teleostei: Gobiidae), with notes on osteology of the genus and description of a new species from Niue in the South Central Pacific Ocean. Zool J Linn Soc. 2018;183(1):143–62. Epub Dec 16, 2017. WOS:000432303500006.
- 36. Tornabene L, Greenfield DW, Erdmann MV. A review of the Eviota zebrina complex, with descriptions of four new species (Teleostei, Gobiidae). ZooKeys. 2021;1057:149–84. pmid:34552371
- 37. Gaudant J. Sur les conditions de gisement de l’ichthyofaune oligocène d’Aix-en-Provence (Bouches-du-Rhône): Essai de définition d’un modèle paléoécologique et paléogéographique. Géobios. 1978;11(3):393–7.
- 38. Gaudant J, Quayle WJ. New palaeontological studies on the Chapelcorner fish bed (Upper Eocene, Isle of Wight). Bull Br Mus nat Hist (Geol). 1988;44(1):15–39.
- 39. Gaudant J. Nouvelles observations sur les poissons oligocènes de Monteviale (Vicenza—Italie). Mem Sci Geol. 1978;32:1–9.
- 40. Gaudant J. Sur la présence de Gobiidae (Poissons téléostéens) dans l’Oligocène inférieur de Rouffach (Haut-Rhin). Sci Géol Bull. 1979;32(3):131–7.
- 41. Bradić-Milinović K, Ahnelt H, Rundić L, Schwarzhans W. The lost freshwater goby fish fauna (Teleostei, Gobiidae) from the early Miocene of Klinci (Serbia). Swiss Journal of Palaeontology. 2019;138(2):285–315. Epub June 1.
- 42. Gierl C, Reichenbacher B. A new fossil genus of Gobiiformes from the Miocene characterized by a mosaic set of characters. Copeia. 2015;103(4):792–805. Epub Nov 12. WOS:000366679700006.
- 43. Gierl C, Reichenbacher B. Revision of so-called Pomatoschistus (Gobiiformes, Teleostei) from the late Eocene and early Oligocene. Palaeontologia Electronica. 2017;20.2.33A:1–17. WOS:000405188500014.
- 44. Gierl C, Reichenbacher B, Gaudant J, Erpenbeck D, Pharisat A. An extraordinary gobioid fish fossil from southern France. PloS one. 2013;8(5):e64117. Epub May 15. pmid:23691158; PubMed Central PMCID: PMC3655028.
- 45. Reichenbacher B, Bannikov AF. Diversity of gobioid fishes in the late middle Miocene of northern Moldova, Eastern Paratethys–part I: an extinct clade of Lesueurigobius look-alikes. PalZ. 2022;96:67–112. Epub Aug 26, 2021.
- 46. Reichenbacher B, Gregorová R, Holcová K, Šanda R, Vukić J, Přikryl T. Discovery of the oldest Gobius (Teleostei, Gobiiformes) from a marine ecosystem of Early Miocene age. Journal of Systematic Palaeontology. 2018;16(6):493–513. Epub May 2, 2017. WOS:000424810000002.
- 47. Schwarzhans W, Ahnelt H, Carnevale G, Japundžić S, Bradić K, Bratishko A. Otoliths in situ from Sarmatian (Middle Miocene) fishes of the Paratethys. Part III: tales from the cradle of the Ponto-Caspian gobies. Swiss Journal of Palaeontology. 2017;136(1):45–92. Epub Nov 10, 2016.
- 48. Gaudant J. Mise en évidence des plus anciens Gobioidei (Poissons téléostéens) connus dans le Lutétien inférieur marin de Catalogne (Espagne). Comptes rendus de l’Académie des sciences Série II, Sciences de la terre et des planètes. 1996;322(1):71–6. ISI:A1996TT86800009.
- 49. Obrhelová N. Vergleichende Osteologie der tertiären Süsswasserfische Böhmens (Gobioidei). Sbornik Paläont. 1961;26:103–92.
- 50. Schwarzhans W, Brzobohatý R, Radwańska U. Goby otoliths from the Badenian (middle Miocene) of the Central Paratethys from the Czech Republic, Slovakia and Poland: A baseline for the evolution of the European Gobiidae (Gobiiformes; Teleostei). Bollettino della Società Paleontologica Italiana. 2020;59(2):125–73. WOS:000562503100004.
- 51. Schwarzhans W, Agiadi K, Carnevale G. Late Miocene-Early Pliocene evolution of Mediterranean gobies and their environmental and biogeographic significance. Riv It Paleont Strat. 2020;126(3):657–724. WOS:000584561400003.
- 52. Keith P. Threatened fishes of the world: Rhyacichthys guilberti Dingerkus & Séret, 1992 (Rhyacichthyidae). Environ Biol Fish. 2002;63(1):40–. WOS:000173466700006.
- 53. Lord C, Keith P. Threatened fishes of the world: Protogobius attiti Watson & Pöllabauer, 1998 (Rhyacichthyidae). Environ Biol Fish. 2006;77(1):101–2. WOS:000239741100010.
- 54.
Akihito. Some morphological characters considered to be important in gobiid phylogeny. In: Matsuura K, editor. Indo-Pacific Fish Biology: Proceedings of the Second International Conference on Indo-Pacific Fishes. Tokyo: Ichthyological Society of Japan; 1986. p. 629–39.
- 55. Iwata A, Jeon S-R, Mizuno N, Choi K-C. A revision of the eleotrid goby genus Odontobutis in Japan, Korea and China. Jpn J Ichthyol. 1985;31(4):373–88. Epub Feb 23, 2011. WOS:A1985ADJ0200005.
- 56. Sparks JS, Chakrabarty P. Revision of the endemic Malagasy Cavefish genus Typhleotris (Teleostei: Gobiiformes: Milyeringidae) with discussion of its phylogenetic placement and description of a new species. American Museum Novitates. 2012;3764:1–28. WOS:000312236300001.
- 57. Hoese DF, Larson HK. New Australian fishes. Part 11. A new genus and species of eleotridid (Gobioidei) from Southern Australia with a discussion of relationships. Memoirs of the Museum of Victoria. 1987;48(1):43–50. BCI:BCI198987014125.
- 58.
Schwarzhans W. Reconstruction of the Fossil Marine Bony Fish Fauna (Teleostei) from the Eocene to Pleistocene of New Zealand by Means of Otoliths: With Studies of Recent Congroid, Morid and Trachinoid Otoliths. Milano: Museo Civico di Storia Naturale di Milano; 2019. 390 p.
- 59. Hoese DF, Roberts CD. A new species of the eleotrid genus Thalasseleotris (Teleostei: Gobioidei) from New Zealand coastal waters. J Roy Soc New Zeal. 2005;35(4):417–31. ISI:000234492300004.
- 60. Birdsong RS, Murdy EO, Pezold FL. A study of the vertebral column and median fin osteology in gobioid fishes with comments on gobioid relationships. B Mar Sci. 1988;42(2):174–214. ISI:A1988N503000003.
- 61. Kindermann G, Miljković N, Ahnelt H, Stevenson DE. The osteology of Eucyclogobius newberryi and Quietula guaymasiae (Teleostei: Gobiidae), two closely related Gobionellines from the East Pacific. Annalen des Naturhistorischen Museums in Wien Serie B Botanik und Zoologie. 2007;108:13–56. BCI:BCI200700465922.
- 62.
Gierl C. Articulated gobioid skeletons from the Frankfurt-Formation (Lower Miocene). München: Ludwig-Maximilians-Universität München; 2012.
- 63. Near TJ, Dornburg A, Eytan RI, Keck BP, Smith WL, Kuhn KL, et al. Phylogeny and tempo of diversification in the superradiation of spiny-rayed fishes. Proc Natl Acad Sci U S A. 2013;110(31):12738–43. pmid:23858462; PubMed Central PMCID: PMC3732986.
- 64. Bannikov AF, Carnevale G. †Carlomonnius quasigobius gen. et sp. nov.: the first gobioid fish from the Eocene of Monte Bolca, Italy. Bulletin of Geosciences. 2016;91(1):13–22. Epub Dec 2, 2015. WOS:000372546600002.
- 65. Gaudant J. Présence du genre Lepidocottus Sauvage, 1875 (Teleostei, Gobioidei) dans l’Oligocène inférieur des environs de Céreste (Alpes-de-Haute-Provence, France). Geodiversitas. 2015;37(2):229–35. WOS:000357401300004.
- 66. Reichenbacher B, Gaudant J. On Prolebias meyeri (Agassiz) (Teleostei, Cyprinodontiformes) from the Oligo-Miocene of the Upper Rhinegraben area, with the establishment of a new genus and a new species. Eclogae Geologicae Helvetiae. 2003;96(3):509–20. WOS:000189178600014.
- 67. Přikryl T. A new species of the sleeper goby (Gobioidei, Eleotridae) from the České Středohoří Mountains (Czech Republic, Oligocene) and analysis of the validity of the family Pirskeniidae. Paläontol Z. 2014;88(2):187–96. Epub Jun 28, 2013.
- 68. La Mesa M, Arneri E, Caputo V, Iglesias M. The transparent goby, Aphia minuta: review of biology and fisheries of a paedomorphic European fish. Rev Fish Biol Fisheries. 2005;15(1–2):89–109. WOS:000233725400006.
- 69. Rojo AL. Osteología del chanquete, Aphya minuta (Risso, 1810) (Pisces: Gobiidae). Bol Inst Esp Oceanogr. 1985;2(1):165–79. BCI:BCI198681089950.
- 70. Van Tassell JL, Miller PJ, Brito A. A revision of Vanneaugobius (Teleostei: Gobiidae), with description of a new species. Journal of Natural History. 1988;22(2):545–67. Epub Feb 17, 2007.
- 71. Hoese DF, Fourmanoir P. Discordipinna griessingeri, a new genus and species of gobiid fish from tropical Indo-West Pacific. Jpn J Ichthyol. 1978;25(1):19–24. ISI:A1978FG43000003.
- 72. Esmaeili HR, Baghbani S, Zareian H, Shahryari F. Scale morphology of tank goby Glossogobius giuris (Hamilton-Buchanan, 1822) (Perciformes: Gobiidae) using scanning electron microscope. Journal of Biological Sciences. 2009;9(8):899–903.
- 73. McKay SI, Miller PJ. The affinities of European sand gobies (Teleostei: Gobiidae). Journal of Natural History. 1997;31(10):1457–82. ISI:A1997XZ44500001.
- 74.
Miller PJ. Gobiidae. In: Whitehead PJP, Bauchot M-L, Hureau J-C, Nielsen J, Tortonese E, editors. Fishes of the north-eastern Atlantic and the Mediterranean (FNAM). III. Paris: UNESCO; 1986. p. 1019–85.
- 75. Randall JE, Hoese DF. Revision of the Indo-Pacific dartfishes, genus Ptereleotris (Perciformes: Gobioidei). Indo-Pacific Fishes. 1985;7:1–36.
- 76. Miller PJ. Affinities, origin and adaptive features of the Australian Desert Goby Chlamydogobius eremius (Zietz, 1896) (Teleostei: Gobiidae). Journal of Natural History. 1987;21(3):687–705. Epub Feb 17, 2007. WOS:A1987H259500009.
- 77. Mestermann K, Zander CD. Vergleichende osteologische Untersuchungen an Pomatoschistus-Arten (Gobioidei, Pisces). Zool Jb Anat. 1984;111:501–42.
- 78. Whitley GP. New fishes from Australia and New Zealand. Proceedings of the Royal Zoological Society of New South Wales. 1956;for the Year 1954–55:34–8.
- 79. McDowall RM. Descriptive and taxonomic notes on Grahamichthys radiatus (Valenciennes), Eleotridae. Trans Roy Soc NZ, Zool. 1965;7(2):51–6.
- 80. Schwarzhans W. Reconstruction of the fossil marine bony fish fauna (Teleostei) from the Eocene to Pleistocene of New Zealand by means of otoliths. Memorie della Società Italiana di Scienze Naturali e del Museo di Storia Naturale di Milano. 2019;46:1–326.
- 81.
Systematics Wongrat P., comparative anatomy, and phylogeny of eleotrine gobies (Teleostei: Gobioidei): University of Bristol; 1977.
- 82. Larson HK, Foster R, Humphreys WF, Stevens MI. A new species of the blind cave gudgeon Milyeringa (Pisces: Gobioidei, Eleotridae) from Barrow Island, Western Australia, with a redescription of M. veritas Whitley. Zootaxa. 2013;3616(2):135–50. WOS:000315021500003; PubMed Central PMCID: PMC24758799. pmid:24758799
- 83. Iwata A, Sakai H. Odontobutis hikimius n. sp.: A new freshwater goby from Japan, with a key to species of the genus. Copeia. 2002;2002(1):104–10. BCI:BCI200200185878.
- 84. Shibukawa K, Iwata A, Viravong S. Terateleotris, a new gobioid fish genus from the Laos (Teleostei, Perciformes), with comments on its relationships. Bulletin of the National Science Museum Series A (Zoology). 2001;27(4):229–57. BCI:BCI200200292381.
- 85. Bergman LMR. The cephalic lateralis system of cardinalfishes (Perciformes: Apogonidae) and its application to the taxonomy and systematics of the family: University of Hawaii at Manoa; 2004.
- 86. McAllister DE. Mandibular pore pattern in the sculpin family Cottidae. Nat Mus Can Bull. 1968;223:58–69. BCI:BCI19684900097393.
- 87. Maddison WP, Maddison DR. Mesquite: a modular system for evolutionary analysis. Version 3.61 https://www.mesquiteproject.org. 3.61 ed2019.
- 88. Larsson A. AliView: a fast and lightweight alignment viewer and editor for large datasets. Bioinformatics. 2014;30(22):3276–8. Epub Aug 5. WOS:000344774600022; PubMed Central PMCID: PMC4221126. pmid:25095880
- 89. Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research. 2004;32(5):1792–7. WOS:000220487200025; PubMed Central PMCID: PMC390337. pmid:15034147
- 90. Gouy M, Guindon S, Gascuel O. SeaView version 4: A multiplatform graphical user interface for sequence alignment and phylogenetic tree building. Molecular biology and evolution. 2010;27(2):221–4. WOS:000273704400003. pmid:19854763
- 91. Ronquist F, Teslenko M, van der Mark P, Ayres DL, Darling A, Höhna S, et al. MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Systematic Biology. 2012;61(3):539–42. WOS:000303336200013; PubMed Central PMCID: PMC3329765. pmid:22357727
- 92. Goloboff PA, Farris JS, Nixon KC. TNT, a free program for phylogenetic analysis. Cladistics. 2008;24(5):774–86. WOS:000259611200009.
- 93. Goloboff PA, Catalano SA. TNT version 1.5, including a full implementation of phylogenetic morphometrics. Cladistics. 2016;32(3):221–38. Epub Apr 25. WOS:000376269000001. pmid:34727670
- 94. Lewis PO. A likelihood approach to estimating phylogeny from discrete morphological character data. Systematic Biology. 2001;50(6):913–25. pmid:12116640
- 95. Yang Z. Maximum likelihood phylogenetic estimation from DNA sequences with variable rates over sites: Approximate methods. Journal of Molecular Evolution. 1994;39(3):306–14. Epub Sep 1. pmid:7932792.
- 96. Lanave C, Preparata G, Saccone C, Serio G. A new method for calculating evolutionary substitution rates. Journal of Molecular Evolution. 1984;20(1):86–93. WOS:A1984SN04000011. pmid:6429346
- 97. Rambaut A, Drummond AJ, Xie D, Baele G, Suchard MA. Posterior summarization in Bayesian phylogenetics using Tracer 1.7. Systematic Biology. 2018;67(5):901–4. Epub 2018/05/03. WOS:000443580600012; PubMed Central PMCID: PMC6101584. pmid:29718447
- 98. Felsenstein J. Confidence Limits on Phylogenies: An Approach Using the Bootstrap. Evolution. 1985;39(4):783–91. WOS:A1985APJ8100007. pmid:28561359
- 99.
Rambaut A. FigTree. Tree Figure Drawing Tool version 1.4.4. http://tree.bio.ed.ac.uk/software/figtree/. 1.4.4 ed. Edinburgh: University of Edinburgh; 2018.
- 100. Weiler W. Die Fischfauna des Tertiärs im oberrheinischen Graben, des Mainzer Beckens, des unteren Maintals und der Wetterau, unter besonderer Berücksichtigung des Untermiozäns. Abhandlungen der Senckenbergischen Naturforschenden Gesellschaft. 1963;504:1–75.
- 101. Marramà G, Bannikov AF, Tyler JC, Zorzin R, Carnevale G. Controlled excavations in the Pesciara and Monte Postale sites provide new insights about the palaeoecology and taphonomy of the fish assemblages of the Eocene Bolca Konservat-Lagerstätte, Italy. Palaeogeography, Palaeoclimatology, Palaeoecology. 2016;454:228–45.
- 102. Pandolfi L, Carnevale G, Costeur L, Del Favero L, Fornasiero M, Ghezzo E, et al. Reassessing the earliest Oligocene vertebrate assemblage of Monteviale (Vicenza, Italy). Journal of Systematic Palaeontology. 2017;15(2):83–127. Epub Mar 16, 2016.
- 103. Brzobohatý R, Gaudant J. Gobius brevis (Agassiz, 1839), a gobiid fish with otoliths in situ (Pisces, Teleostei) in the Karpatian (Lower Miocene) of the Vienna Basin. Annalen des Naturhistorischen Museums in Wien—Serie A (Mineralogie und Petrographie, Geologie und Paläontologie, Archäozoologie, Anthropologie und Prähistorie). 2009;111:245–55. BCI:BCI200900349722.
- 104. Jost J, Kälin D, Börner S, Vasilyan D, Lawver D, Reichenbacher B. Vertebrate microfossils from the Upper Freshwater Molasse in the Swiss Molasse Basin: Implications for the evolution of the North Alpine Foreland Basin during the Miocene Climate Optimum. Palaeogeography, Palaeoclimatology, Palaeoecology. 2015;426:22–33. WOS:000353603000003.
- 105. Jost J, Kälin D, Schulz-Mirbach T, Reichenbacher B. Late Early Miocene lake deposits near Mauensee, central Switzerland: fish fauna (otoliths, teeth), accompanying biota and palaeoecology. Eclogae Geologicae Helvetiae. 2006;99(3):309–26. Epub Jan 12, 2007.
- 106. Mandic O, Hajek-Tadesse V, Bakrač K, Reichenbacher B, Grizelj A, Miknić M. Multiproxy reconstruction of the middle Miocene Požega palaeolake in the Southern Pannonian Basin (NE Croatia) prior to the Badenian transgression of the Central Paratethys Sea. Palaeogeography, Palaeoclimatology, Palaeoecology. 2019;516:203–19. Epub Dec 7, 2018. WOS:000456903000018.