Table 1.
List of the 28 species of Brachyhypopomus included in this study, with authors and type localities.
Table 2.
List of the nine outgroups species used for the phylogenetic analyses of Brachyhypopomus.
Table 3.
Specimens used in molecular phylogenetic analyses.
Table 4.
Matrix of 60 morphological characters for 28 species of Brachyhypopomus and nine outgroup taxa.
Fig 1.
Parsimony phylogeny for Brachyhypopomus based on morphological data.
Tree is the strict consensus of 127 equally parsimonious trees (60 characters, CI = 0.42, RC = 0.28, RI = 0.68). Numbers below branches denote decay indices. Branch lengths are proportional to reconstructed character state changes. Terminal taxa are 28 species of Brachyhypopomus and 9 outgroup taxa.
Fig 2.
Bayesian inference phylogeny for Brachyhypopomus based on cytb.
Bayesian posterior probabilities shown by nodes (not shown for intraspecific relationships). Branch lengths proportional to substitutions per site. Terminal taxa are 83 individuals representing 26 species of Brachyhypopomus and 16 individuals representing 8 outgroup taxa. Branch lengths for Gymnotus and Sternopygus (represented with dashed lines) are reduced. See Table 3 for list of sequenced specimens. Terminal two letter codes refer to the drainage units described in ‘Geographic and ecological distributions’ (Materials and Methods).
Fig 3.
Bayesian inference phylogeny for Brachyhypopomus based on rag2.
Bayesian posterior probabilities shown by nodes (not shown for intraspecific relationships). Branch lengths proportional to substitutions per site. Terminal taxa are 85 individuals representing 26 species of Brachyhypopomus and 17 individuals representing 9 outgroup taxa. Branch lengths for Gymnotus and Sternopygus (represented with dashed lines) are reduced. See Table 3 for list of sequenced specimens. Terminal two letter codes refer to the drainage units described in ‘Geographic and ecological distributions’ (Materials and Methods).
Fig 4.
Bayesian inference phylogeny for Brachyhypopomus based on cytb and rag2.
Bayesian posterior probabilities shown by nodes (not shown for intraspecific relationships). Branch lengths proportional to substitutions per site. Terminal taxa are 85 individuals representing 26 species of Brachyhypopomus and 18 individuals representing 9 outgroup taxa. Branch lengths for Gymnotus and Sternopygus (represented with dashed lines) are reduced. See Table 3 for list of sequenced specimens. Terminal two letter codes refer to the drainage units described in ‘Geographic and ecological distributions’ (Materials and Methods).
Fig 5.
Bayesian inference phylogeny for Brachyhypopomus based on total evidence (cytb, rag2, and morphology).
Bayesian posterior probabilities shown by nodes (not shown for intraspecific relationships). Branch lengths proportional to substitutions per site. Terminal taxa are 87 individuals representing 28 species of Brachyhypopomus and 18 individuals representing 9 outgroup taxa. Branch lengths show substitutions per site. Branch lengths for Gymnotus and Sternopygus (represented with dashed lines) are reduced. See Table 3 for list of sequenced specimens. Terminal two letter codes refer to the drainage units described in ‘Geographic and ecological distributions’ (Materials and Methods).
Fig 6.
Parsimony phylogeny for Brachyhypopomus based on total evidence (cytb, rag2, and morphology).
Tree shows strict consensus of 3244 equally parsimonious trees each with a length of 4243 steps (CI = 0.36, RCI = 0.29, RI = 0.81). Numbers above branches denote bootstrap proportions; numbers below branches denote decay indices; support values not shown for intraspecific relationships. Branch lengths proportional to reconstructed character state changes. Terminal taxa are 87 individuals representing 28 species of Brachyhypopomus and 18 individuals representing 9 outgroup taxa. Branch lengths for Gymnotus and Sternopygus (represented with dashed lines) are reduced. See Table 3 for list of sequenced specimens. Terminal two letter codes refer to the drainage units described in ‘Geographic and ecological distributions’ (Materials and Methods).
Fig 7.
Species phylogeny for Brachyhypopomus based on Bayesian Inference of cytb, rag2, and morphology, with duplicate individuals for each species removed.
Letters denote well-supported clades and numbers denote poorly-supported clades (see text for details). The label Hypopomidae follows Maldonado-Ocampo et al. (2014). Inset photographs are: B. beebei, B. sp. “belindae”, B. pinnicaudatus, B. bennetti, B. sp. “batesi”, B. bombilla, B. occidentalis, B. brevirostris, and B. janeiroensis. Photographs are uniformly scaled; scale bar = 10 mm.
Fig 8.
Undescribed bone located above the maxilla.
In: (A) Brachyhypopomus beebei, MCP 45450 (WC06.090600), female, 152 mm. (B) Racenisia fimbriipinna, UF 177352, immature, 86 mm TL. Left side, lateral view. Anterior to left.
Fig 9.
Antorbital and infraorbital canal series.
In: (A). Brachyhypopomus brevirostris. MCP 44605 (WC06.010596), male, 374 mm TL. Left side, lateral view. Anterior to left. (B). Brachyhypopomus bennetti. MCP 45346, 1 (WC04.290696), female, 156 mm TL. Left side, lateral view. Anterior to left.
Fig 10.
Macro-photograph of the neurocranium showing the relation between the infraorbital canal aperture and sphenotic spine.
In: (A) Brachyhypopomus brevirostris, MCP 44605 (WC06.010596), male, 374 mm TL. (B) Racenisia fimbriipinna, UF 177352, immature, 86 mm TL. Left side, lateral view. Anterior to left.
Fig 11.
In: (A) Brachyhypopomus diazi, UF 174333 (WC06.210304), paraneotype, immature 132 mm TL. (B) Brachyhypopomus sp. “alberti”, UMSS 07042 (WC43.070707), paratype, 84 mm TL. Left side, lateral view. Anterior to left.
Fig 12.
In: (A). Brachyhypopomus sp. “hendersoni”. MCP 45432 (WC05.130799), male, 161 mm TL. Left side, medial view. Anterior to left. (B). Brachyhypopomus walteri. MCP 44649, (WC04.020698), male, 185 mm TL. Left side, medial view. Anterior to left.
Fig 13.
In: (A) Brachyhypopomus occidentalis USNM 293152, 156 mm TL. (B) Brachyhypopomus sp. “hendersoni” MCP 45432 (WC05.130799), 1, 161 mm TL. Left side, lateral view. Anterior to left.
Fig 14.
Preopercle and associated sensory canals.
In: (A) B. brevirostris. MCP 44605 (WC06.010596), male, 374 mm TL. Left side, lateral view. Anterior to left. Note the incised preopercular sensory canals (borders highlighted with dashed lines). (B). B. beebei. MCP 45421 (WC04.160698), male, 201 mm TL. Left side, lateral view. Anterior to left. Note the preopercular sensory canals are completely independent of preopercle (borders highlighted with solid lines). (C) B. sp. “sullivani”. MCP 45486 (WC02.221299), male, 101 mm TL. Left side, lateral view. Anterior to left. Note only the posterior-most preopercular sensory canal is incised in preopercle (borders highlighted with dashed lines).
Fig 15.
In: (A) B. sp. “hendersoni”. MCP 45432 (WC05.130799), 1, 161 mm TL. Left side, medial view, anterior to left. Note the posterodorsal margin of the dentary is straight and even (Character 51). (B). B. sp. “sullivani”. MCP 45486 (WC02.221299), male, 101 mm TL. Right side (inverted), medial view, anterior to left. Note the posterodorsal margin of the dentary is concave, forming a hook-like process (Character 51).
Fig 16.
Posterior portion of the neurocranium and anterior portion of the Weberian apparatus.
Note the undescribed ossification on the supraoccipital, anterior to the neural complex, in: (A) Brachyhypopomus beebei MCP 45450 (WC06.090600), female, 152 mm; (B) Brachyhypopomus brevirostris MCP 44605 (WC06.010596), male, 374 mm TL (photograph). Left side, lateral view. Anterior to left.
Fig 17.
Ossification of the anterior portion of the palatoquadrate cartilage.
In photographs of cleared and stained specimens of: (A) Brachyhypopomus pinnicaudatus—MCP 45370 (WC03.050497), female, 122 mm (anterior portion of head in lateral view); (B) Brachyhypopomus beebei—MCP 45450 (WC06.090600), female, 152 mm (ventral surface of anterior portion of the neurocranium and associated structures). Note the typical disk-like outline of the ossified element in the palatoquadrate cartilage (PqO) in ventral view–B, and rectangular outline in lateral view–A. Note also, in A, the ascending process of the endopterygoid (End) does not form a contact with the orbitosphenoid (Orb). Abbreviations for bones in the jaw and suspensorium are: Pmx—premaxilla; Mes—mesethmoid; Ven—ventral ethmoid; Max—maxilla; Den—dentary; Ang—anguloarticular; Ret—retroarticular; Qua—quadrate; Sym—symplectic; Met—metapterygoid; Hyo—hyomandibula. Scale bar 1 mm.
Fig 18.
Geographical distributions of Brachyhypopomus and other hypopomids.
Occurrence records are tabulated for five geographical regions and 14 drainage subunits. Crosses show occurrences for each species. The tree topology follows the total evidence BI phylogeny reported in Fig 7. Letters denote well-supported clades and numbers denote poorly-supported clades based on Bayesian Posterior Probabilities (see text and Fig 5). Colored circles at the terminals represent occurrence in one of the five geographical regions (see inset key), or occurrence in both region 1 and 2 (for 3 species), or in region 1 and 5 (for 1 species). Ancestral character states in the internal nodes are optimized by maximum likelihood, with the proportion of the color representing the probability of occurrence in a given region. The ancestral character state for Brachyhypopomus optimizes with high probability to Greater Amazonia. Records for outgroup species are in grey text. The regions listed here are as follows (for details of drainage boundaries see ‘Geographic and ecological distributions‘ in Materials and Methods): Region 1 (Greater Amazonia): (OR) = Orinoco basin; (GU) = Caribbean drainages of the Guianas; (RN) = rio Negro; (UA) = Upper Amazon; (CA) = Central Amazon; (LA) = Lower Amazon; (UM) = Upper Madeira. Region 2 (La Plata–Lagoa dos Patos): (PA) = La Plata drainages; (SE) = Lagoa dos Patos-Mirim system and adjacent coastal drainages of Brazil and Uruguay. Region 3 (Brazilian coastal drainages): (BC) = Atlantic coastal drainages from rio Ribeira de Iguape to rio Paraíba do Sul. Region 4 (São Francisco drainage): (NE) = rio São Francisco. Region 5 = (Trans-Andean drainages): (MA) = Middle America; (PS) = Pacific Slope; (NW) = northwest South America.
Fig 19.
Ecological distributions of Brachyhypopomus and other hypopomids.
(A) Habitat occupancy in terra firme and floodplain systems (letter codes listed below denote occurrences). (B) Occupancy of low (< 30 μScm-1) and high conductivity (> 60μScm-1) systems (crosses denote occurrences). The tree topology follows the total evidence BI phylogeny reported in Fig 7. Letters on the trees denote well-supported clades and numbers denote poorly-supported clades based on Bayesian Posterior Probabilities (see text and Fig 5). Records for outgroup species are in grey text. The circles of the terminals represent exclusive occurrence in either of two habitats (white/black), or eurytopy (grey). Ancestral character states in the internal nodes are optimized by maximum likelihood, with the proportion of white, black, and grey representing the probability of occurrence in a given habitat/conductivity. Letter codes for habitat occupancy: Terra firme systems: S = lowland (< ca. 200 m above sea level) terra firme stream in Paleogene-Neogene formations, Sw = lowland terra firme swamp in Paleogene-Neogene formations; Sh-S = upland (> ca. 200 m above sea level) stream in Proterozoic-Paleozoic shield formations; Pi-S = upland (> ca. 200 m above sea level) stream in Mesozoic-Paleogene Andean or Panamanian piedmont formations. Floodplain systems: W = whitewater, B = blackwater, C = clearwater.
Fig 20.
The occurrence of Brachyhypopomus and other hypopomids in habitats subject to hypoxia.
The tree topology follows the total evidence phylogeny reported in Fig 7. Letters denote well-supported clades and numbers denote poorly-supported clades based on Bayesian Posterior Probabilities (see text and Fig 5). Records for outgroup species are in grey text. Ancestral character states in the internal nodes are optimized by maximum likelihood, with the proportion of white and black representing the probability of occurrence in a given habitat/conductivity environment.? = unknown state in terminal or equivocal ancestral condition.