Specimens were collected, handled and euthanized in accordance with guidelines published by the Use of Fishes in Research Committee (1987, 2004, 2014). EODs of the Doumé specimen (tag no. JPS-1194) were recorded in a small aquarium filled with water from the collection site, using chloridized-silver wire electrodes connected to an Echo 2 USB analog to digital converter (Echo Audio, Inc.) sampling at 192 kHz/16 bits. We visualized and saved signals using SignalScope virtual oscilloscope software (Faber Acoustical, LLC). We recorded head positivity of the fish in the upward direction and noted water temperature at time of recording. After recording, the fish was euthanized with an overdose of the anesthetic MS222 (tricaine methanesulfonate), tagged, and fixed in 10% formalin. The specimen was subsequently transferred to 70% ethanol and is deposited in the Cornell UniversityMuseum of Vertebrates as CUMV 98155.

In order to investigate the phylogenetic relationships of the three novel mormyrid specimens, we sequenced from each the complete cytochrome b (cyt-b) gene and portions of the 12S and 16S genes from the mitochondrial genome, as well as partial rag2 and the complete S7 introns 1 & 2 and the short intervening exon from the nuclear genome. We added these sequences to an existing 4209 bp alignment of these markers from 38 species belonging to 17 mormyrin genera published in Lavoué et al. (2003), the most complete molecular phylogenetic study of the Mormyrinae to date. In addition to these markers used for phylogenetic analysis, we sequenced a 636 bp fragment of the cytochrome oxidase I (COI) gene, the “barcode” marker from the three specimens for comparison to sequences found in the Barcode of Life Database (Ratnasingham and Hebert 2007).

Because the possible relationship of these new taxa to the genus Hippopotamyrus Pappenheim, 1906 was of special interest, we also sequenced all of these markers from a specimen of Hippopotamyrus castor from Cameroon, the type species of this genus, a species not included in Lavoué et al. (2003). In addition, we added to the expanded matrix cyt-b sequences of a specimen of H. pictus from the White Nile of Ethiopia (complete data for a H. pictus from the Niger basin of West Africa was already included in the dataset) as well as cyt-b sequences retrieved from GenBank for Hippopotamyrus ansorgii (Boulenger, 1905) and H. szaboi Kramer, van der Bank & Wink, 2004, coding missing data as “?” in the matrix.

DNA was extracted from fin clips or epaxial muscle tissue preserved in 95% ethanol using a QIAGEN (QIAGEN, Inc., Valencia, CA) DNeasy kit. Primer sequences for cyt-b, 12S, 16S, rag2, S7 and COI are listed in Table 1. For each marker we carried out PCR in 10 to 30μl reactions with components at the following concentrations: 1x Sigma PCR buffer (Sigma–Aldrich, St. Louis, MO), 0.02 U/μl Sigma JumpStart Taq, 2 mM MgCl₂ (3 mM MgCl₂ for rag2), 0.4 μM of forward and reverse primer, 200 μM of each dNTP and approximately 200 pg/μl template DNA. We used an initial denaturation step of 1 min at 94 °C followed by 35 cycles of 94 °C for 30 s, annealing at 54 °C (59 °C for rag2) for 30 s, and extension at 72 °C for 1.5 min, followed by a final 72 °C extension step for 10 min. We evaluated amplification success on ethidium bromide-stained agarose gels, purified PCR products using Exonuclease I and Shrimp Alkaline Phosphatase and dye-deoxy termination cycle sequencing using ABI Big Dye chemistry followed by Sephadex column purification and data collection on an Applied Biosystems Automated 3730xl analyzer (PE Applied Biosystems, Foster City, CA).

Sequences were edited and combined into contigs for each fragment with Sequencher 4.2 (GeneCodes Corporation, Ann Arbor, MI). Requiring no gaps, alignment of the new coding gene sequences (cyt-b and rag2) to the Lavoué et al. (2003) dataset was trivial; for 12S, 16S and S7 we were able to align these sequences by eye to the previously optimized alignment without need for insertion of additional gaps.

We performed a maximum likelihood phylogenetic analysis of the matrix from Lavoué et al. (2003) with the added taxa in RAxML v.8, implemented on XSEDE (Stamatakis 2014) via the CIPRES Science Gateway web server (Miller et al. 2010) using separate GTRGAMMA evolutionary models for each gene partition and performed a non-parametric bootstrap analysis to estimate support for nodes. Bootstrapping was auto-terminated using the autoMRE criterion. All other settings were left at their default values. The tree was rooted using Myomyrus macrops Boulenger, 1914 as outgroup (Lavoué et al. 2003).

Because the three specimens under study do not obviously belong to any one of the 19 described mormyrin genera, choosing mormyrin species for comparison was not straightforward. We collected counts and measurements from seven sympatric taxa with which the new taxa could conceivably be confused in the field and two extralimital taxa with (in our estimation) the most overall resemblance to the three specimens, both of which are type species of genera not known to occur in Gabon. The sympatric taxa used for comparison are Ivindomyrus marchei (Sauvage, 1879), I. opdenboschi Taverne & Géry, 1975, Marcusenius moorii (Günther, 1867), Stomatorhinus walkeri (Günther, 1867), Paramormyrops kingsleyae (Günther, 1896) and two undescribed species of Paramormyrops Taverne, Thys van den Audenaerde & Heymer, 1977. The extralimital taxa used for comparison are Hippopotamyrus castor (from coastal drainages of Cameroon) and Cyphomyrus psittacus (Boulenger, 1897) from the Congo basin. Most of these species are illustrated in Figure 3.

Figure 3. 

Some mormyrid species used for morphological comparison in this study. A Ivindomyrus marchei female, 135 mm SL, CUMV 98172, tag no. JPS-1233, Sébé River, Ogooué-Lolo, Gabon B Ivindomyrus opdenboschi female, 160 mm SL, CUMV 96829, tag no. JPS-1057, Ivindo River at Loa Loa, Ogooué-Ivindo, Gabon C Boulengeromyrus knoepffleri juvenile 107 mm SL, CUMV 96838, tag no. JPS-1055, Ivindo River at Loa Loa, Ogooué-Ivindo, Gabon D Marcusenius moorii female, 147 mm SL, CUMV 96836, tag no. JPS-1110, Ogooué River at Franceville, Haut-Ogooué, Gabon E Paramormyrops sp. (undescribed species) female, 100 mm SL, CUMV 98119, tag no. JPS-1148, Moumba Creek, Ogooué-Lolo, Gabon F Stomatorhinus walkeri female 95 mm SL, CUMV 80227, tag no. 2814, Diengui Creek, Moyen-Ogooué, Gabon G Hippopotamyrus castor female, 122 mm SL, CUMV 89955, tag no. 6018, Sanaga River at Nachtigal Falls, Centre, Cameroon H Cyphomyrus psittacus female, 109 mm SL, CUMV 96767, tag no. JPS-0438, Congo River at Yangambi, Orientale, D.R. Congo.

We took 29 point-to-point measurements with a digital caliper on each of the three specimens under study for comparison to those taken from 115 individuals of nine species listed above. Twenty-seven of these measurements are defined in Boden et al. (1997): total length (TL), standard length (SL), body depth (BD), caudal peduncle depth (CPD), middle caudal peduncle depth (MCPD), caudal peduncle length (CPL), head length (HL), head depth (HD), head width (HW), snout length (SNL), interorbital width (IOW), eye diameter (ED), postorbital length (POL), internarine distance (DNN), posterior nares-eye distance (DNE), predorsal distance (PDD), preanal distance (PAD), prepelvic distance (PPLD), prepectoral distance (PPCD), dorsal-fin length (DFL), dorsal-fin height (DFH), anal-fin length (AFL), anal-fin height (AFH), pelvic-fin length (PLFL), pectoral-fin length (PCFL), pelvic-to-anal-fin distance (DPLAF), and pectoral-to-anal-fin distance (DPCAF). To these we added two additional measures: head length to the end of the opercle bone (HLBO), and mouth width measured at the inner corner of the front of the mouth (MW).

To facilitate dorsal- and anal-fin ray comparisons among individuals, we report the number of unbranched rays (3) + number of branched rays. (The first two rays are small, the first often only visible in radiographs.)

Specimens were radiographed using a Faxitron Model LX-60 cabinet x-ray system and Kodak Industrex MX125 film. Film images were scanned on a flatbed scanner.

We follow Taverne (1968) in reporting features of the axial skeleton although we exclude the demi-centrum fused to hypurals 3+4 in our vertebral counts. Our nomenclature for hypural bones follows Teugels and Hopkins (1998) and that for intermuscular bones follows Patterson and Johnson (1995).

Sex of the specimens was determined by presence/absence of an “anal-fin notch”: a dorsally directed indentation along the anterior base of the anal fin present in all mature male mormyrids (Pezzanite and Moller 1998).

The EOD waveform of the specimen from Doumé is triphasic and very brief (Fig. 4A). Among mormyrids of Lower Guinea, this EOD is most similar to that recorded from Hippopotamyrus castor from Cameroon (Hopkins et al. 2007), but dissimilar to that of every other species recorded so far in Gabon. Details of the EOD are provided in the species description below, and the recording is available in the archive of the Macaulay Library at the Cornell Lab of Ornithology under accession number ML197475.

Figure 4. 

A Three superimposed electric organ discharge (EOD) waveform recordings of Cryptomyrus ogoouensis holotype CUMV 98155 (Macaulay Library #197475) recorded at 23.2 °C, head positivity upwards, X-axis = 1 millisecond, Y-axis tick marks indicate 20% of EOD peak-to-peak height; P0, P1, P2 mark the positive and negative excursions in the waveform, P0 highlighted by 20× amplification; red dots indicate onset of P0, onset of P1 and offset of P2, respectively B Power spectrum of EOD waveforms in A; red “+” symbols mark peak frequency and frequencies -3 dB below peak at 4300 Hz.

GenBank numbers for the new sequences generated for this study and their appropriate GenSeq codes (Chakrabarty et al. 2013) are given in Table 2. The sequence data (Table 3) confirm that the three specimens are very close relatives while genetically distant from other mormyrins. The cyt-b sequences of the Doumé and Moukalaba specimens differ at 15 sites across the 1140 base pairs, or a p-distance of 1.32%. Between the Doumé specimen and the Mabounié specimen this difference is 17 sites or 1.49%. Between the Moukalaba and the Mabounié specimens the difference is only six sites, or 0.53%. COI genetic distance is smaller between the three specimens: a p-distance of only 0.6% between the Doumé and Moukalaba specimens, 0.8% between the Doumé and the Mabounié specimen, and 0.2% between the Moukalaba and Mabounié specimens. Genetic distances for the 12S and 16S markers are similar to COI, and nuclear rag2 and S7 sequences were identical among all three (Table 3) with the exception of base 564 in the S7 sequence which is heterozygous for C and T in the Doumé specimen (coded as “Y” in the sequence), while homozygous for T in the other two specimens. For all markers, sequences from the two species of Ivindomyrus Taverne & Géry, 1975 and from Boulengeromyrus knoepffleri Taverne & Géry, 1968 are closest to those of the new taxa, with cyt-b p-distances between 6.1 and 6.5% (Table 3).

We note that genetic differentiation observed between the Doumé specimen and the Moukalaba and Mabounié specimens, while small, is within the range seen between some closely related, but distinct mormyrid species. For example cyt-b sequences from the morphologically divergent Paramormyrops gabonensis Taverne, Thys van den Audenaerde & Heymer, 1977 and P. hopkinsi Taverne & Thys van den Audenaerde, 1985 – two species for which there is no evidence of mitochondrial introgression – similarly differ by 1.3%. However, in other cases, populations from different river basins regarded as conspecific can differ by this much. For example, Ivindomyrus marchei from the Ogooué and Ivindo Rivers differ from those in the Nyanga River by just slightly less than 1.3% (Lavoué et al. 2008) as do Congo and Ogooué basin populations of Marcusenius moorii (unpublished data). The degree of genetic differentiation between the Moukalaba River and Mabounié River specimen (0.53%), on the other hand, is squarely within the normal range of intraspecific variation for cyt-b, with the exception of cases where introgressive hybridization is suspected.

Among the more than 300 mormyrid COI sequences in the Barcode of Life Database (BOLD; Ratnasingham and Hebert 2007), the closest match to those from the new taxa is from an Ivindomyrus marchei, with 92% sequence identity. To further explore the relationship of mitochondrial sequence differentiation within and between mormyrid species, we downloaded all of the mormyrid COI sequences archived in BOLD and calculated intraspecific variation in COI for species represented by more than one sequence. Intraspecific distances in the COI dataset are well above the 0.6–0.8% observed between the Doumé specimen and the other two individuals for several nominal species with wide distributions spanning multiple major watersheds, e.g. Brevimyrus niger (Günther, 1866), Gnathonemus petersii (Günther, 1862), and Mormyrops anguilloides (Linnaeus, 1758). However, there are also examples of species pairs with COI sequence differentiation below 1% p-distance within genera Marcusenius Gill, 1862, Campylomormyrus Bleeker, 1874, Cyphomyrus Myers, 1960, and Petrocephalus Marcusen, 1854. While some of these findings could be attributed to identification errors and/or mitochondrial introgression between sympatric species, morphologically distinct species with very low mitochondrial genetic divergence are documented within the riverine species flocks of Campylomormyrus (Feulner et al. 2007) and Paramormyrops (Sullivan et al. 2002, 2004).

Turning to the nuclear markers, the near identity of the nuclear S7 intron sequences among the Mabounié, Doumé and Moukalaba specimens might be thought surprising if these three specimens represent more than one species (less surprising for the identity of the rag2 sequences, as this marker evolves more slowly). However, distinct mormyrid species having identical or near-identical nuclear S7 intron sequences is not unprecedented, as similarly identical S7 intron sequences were found among three species of Campylomormyrus by Feulner et al. (2007).

The aligned dataset and tree described here have been archived on TreeBASE and can be accessed at http://purl.org/phylo/treebase/phylows/study/TB2:S18468?format=html.

In the maximum likelihood tree produced in RAxML (Fig. 5) the three specimens form a monophyletic group, with the Moukalaba River and Mabounié River specimens paired together, sister to the Doumé specimen, as one would predict from the genetic distances. Together, the three specimens appear as the sister lineage to the clade formed by Boulengeromyrus knoepffleri plus the two Ivindomyrus species. All of these together are in turn sister to the clade formed by species of Paramormyrops plus Marcusenius ntemensis (Pellegrin, 1927). Bootstrap support at all of these nodes is 100% with the exception of a 91% value at the node joining the Moukalaba River and Mabounié River specimens. This larger clade that includes taxa either endemic to or most diverse in Lower Guinea is sister to a clade formed by species of genus Stomatorhinus Boulenger, 1898 plus species of genus Pollimyrus Taverne, 1971, with lower bootstrap support. Hippopotamyrus castor from Cameroon, the type species of Hippopotamyrus, clusters far separately from the new taxa from Gabon, sister to two Hippopotamyrus pictus individuals, one from the White Nile River and one from the Niger River. These are sister to a large clade containing species of Marcusenius, Campylomormyrus, Gnathonemus Gill, 1863, Cyphomyrus, Genyomyrus Boulenger, 1898 and a separate pair of species from southern Africa classified as Hippopotamyrus: H. ansorgii and H. szaboi. While incidental to the description of the new taxa treated here, this result implies the “Hippopotamyrus ansorgii complex” of southern Africa treated by Kramer et al. (2004) and Kramer and Swartz (2010) does not belong in Hippopotamyrus Pappenheim and requires classification in a different genus.

Figure 5. 

Maximum likelihood phylogenetic tree calculated in RAxML for 46 mormyrin specimens based on DNA sequences from mitochondrial cytochrome b, 12S, 16S and nuclear rag2 and S7 intron, rooted with Myomyrus macrops. Bootstrap support values at nodes. Some clades collapsed for simplicity. Position of new taxa (magenta) within the “Lower Guinea Clade” (green) and distance from Hippopotamyrus castor, type species of genus Hippopotamyrus (red), highlighted.

To explore the relative contribution of mitochondrial and nuclear data to the phylogenetic result, we ran two additional analyses, one with only mitochondrial data included for the three specimens and the other with only nuclear data included for them. In each case, excluded bases were recoded as missing data, while the full dataset was used for all other specimens. In the nuclear data-only analysis, we used a single OTU for the three specimens since rag2 and S7 sequences were identical among them. In both experimental analyses, the lineage of the three new taxa was resolved as sister to Boulengeromyrus plus Ivindomyrus with a 100% bootstrap proportion, indicating strong signal contributing to this result from both the nuclear and mitochondrial data partitions.

Consistent with the genetic distances reported above, measures and counts of the Moukalaba and Mabounié specimens are more similar to each other than they are to those of the Doumé specimen. In particular, dorsal- and anal-fin bases are short in the former two, with low fin-ray counts, 20/25 and 21/24 total rays respectively, compared to 24/30 in the Doumé specimen, in which the anal-fin origin is situated much further in advance of the dorsal-fin origin. Predorsal distance is nearly equal to preanal distance in the Moukalaba and Mabounié specimens, while it is markedly shorter in the Doumé specimen due to its longer anal fin. Additionally, the caudal peduncle is much deeper and the caudal-fin lobes are shorter in the Moukalaba and Mabounié specimens than in the Doumé specimen. We have never observed a range of four dorsal-fin rays, six anal-fin rays, and differences of 25 percent in anal-fin length relative to standard length among individuals of a single mormyrid species.

While our initial hypothesis that the new taxa are close relatives of Hippopotamyrus castor from Cameroon was ruled out by the molecular result, substantial morphological differences are also obvious between the new taxa and H. castor. The new taxa have fewer scales (44–45 lateral line scales/12 around the caudal peduncle vs. 72–81/16 in H. castor), fewer total vertebrae (40–43 vs. 47), fewer dorsal-fin rays (20–24 vs. 31–33), and a longer anal-fin base (dorsal 78–88% of anal-fin base vs. nearly equal), among other significant differences. Also, while H. castor has elongate paired incisor teeth in the lower jaw (from which its specific epithet derives), their morphology is quite different from the lower teeth of Cryptomyrus. In the three specimens under study, the central incisors are flattened and spatulate. The neighboring pair of teeth are smaller, but similarly spatulate and close or appressed to the inner pair which assume a sort of trowel shape, pointing outward (Fig. 6). In H. castor, the central incisors are cylindrical and club-like and the lateral teeth are greatly reduced (Fig. 6).

Figure 6. 

Mouth and dentary teeth in A Cryptomyrus ogoouensis holotype CUMV 98155 B Cryptomyrus ona holotype MNHN 2003-0425 C Cryptomyrus ona non-type CUMV 9864; dentary teeth in D Boulengeromyrus knoepffleriCUMV 81643 tag no. 2254 E Ivindomyrus marcheiCUMV 96827 tag no. JPS-1043 F Ivindomyrus opdenboschiCUMV 89324 tag no. 5654 G Hippopotamyrus castorCUMV 89955 tag no. 6033 H Paramormyrops sp. “SN4” CUMV 81322. Scale bars = 1 millimeter.

With its moderately swollen chin and small mouth, the head of Cyphomyrus psittacus appears somewhat similar to those of the new taxa and like them C. psittacus has a diffuse band of dark pigment between the origin of the dorsal and anal fins, but the new taxa differ from C. psittacus in being less deep bodied (BD 21–24% SL vs. 30–34%), having a shallower head (HD close to 79% HL vs. 63–75%), a smaller eye (ED 19–20% HL vs. 24–26%), a dorsal fin somewhat shorter than the anal fin (vs. far longer than the anal fin) and fewer midlateral scales (44–45 vs. 54–56). Unlike the spatulate dentary teeth of the new taxa, dentary teeth in species of Cyphomyrus examined are very small, feebly notched, and embedded in the gums.

The three specimens also differ morphologically from individuals of the two species of Ivindomyrus to which the phylogenetic study shows they are closely related. Compared to Ivindomyrus, all three specimens have shorter, deeper caudal peduncles (CPD 31–35% CPL vs. 21–28%), greater interorbital width compared to head length (IOW 33–35% HL vs. 21–30%), a higher dorsal-fin to anal-fin length ratio (DFL 78–83% AFL vs. 65–76%), and considerably fewer midlateral scales (44–45 vs. 53–58). Like the dentary teeth in the new taxa, those in Ivindomyrus are spatulate, however the central pair of incisors are not noticeably elongate along their inner margins as they are in the new taxa (Fig. 6).

Compared to species of Paramormyrops the new taxa are generally deeper bodied (BD 21–25% SL vs. 15–21%) with deeper heads (HD close to 79% HL vs. 63–75%), larger eyes (ED 19% HL vs. 9–17%) and shorter postorbital head length (POL 59–60% HL vs. 61–70%) in addition to having far fewer midlateral scales (44–45 vs. 56–71). Dentary teeth in all Paramormyrops are bicuspid, not spatulate (Fig. 6).

Compared to Marcusenius moorii, a species with a similarly low midlateral line scale count, the new taxa most notably lack a chin swelling that protrudes forward beyond the tip of the snout, have a much more subdued and diffuse area of pigment between the dorsal-fin and anal-fin origins, a larger eye (ED 19–20% HL vs. 15–18%) and unicuspid (vs. bicuspid or notched) dentary teeth.

Two of the three new taxa are of larger size than the largest Stomatorhinus walkeri examined and all three have a smaller head relative to standard length (HL 22–23% SL vs. 28–31%) a larger eye (ED 19% HL vs. 9–13%), a modest chin swelling (vs. none), and the posterior nares are remote from the rictus of the mouth (vs. in close proximity).

Based on the molecular phylogenetic analysis, we conclude that these three morphologically distinctive specimens represent an undescribed lineage of Mormyrinae sister to the clade formed by Boulengeromyrus plus Ivindomyrus. This phylogenetic finding in combination with the morphological differences that exist between these new taxa and Boulengeromyrus and Ivindomyrus precludes placing them in either genus. Creation of an enlarged Boulengeromyrus Taverne & Géry via synonymy in which to subsume all of these taxa is ruled out by the significant morphological differences between them and lack of known morphological synapomorphies. Hence a new genus is required.

We recognize two species: one for the Doumé specimen and the other comprising the Moukalaba and Mabounié specimens, reflecting the greater morphological and genetic similarity of these latter two and their notable morphological differences from the Doumé specimen, discussed above. We choose the Moukalaba River specimen as type for the second species and treat the Mabounié River specimen as a non-type. We describe these new taxa below.

Lower Guinea is defined as the Atlantic drainages of Africa from the Cross River of Cameroon in the north to the Chiloango River of Cabinda/D.R. Congo in the south (Roberts 1975). This artificial key is modified after Hopkins et al. (2007) in which some character states are illustrated. Distribution of each genus within Lower Guinea is indicated in parentheses.

Cyphomyrus psittacus: D.R. Congo, Orientale, Wagenia Falls, CUMV 97543, 4, 88–133 mm SL, tag nos. JPS-0537, JPS-0538, 2 untagged.

Hippopotamyrus castor: Cameroon, Centre, Sanaga R. at Nachtigal Falls, CUMV 89955, 6, 104–136 mm SL, tag nos. 6018, 6019, 6020, 6037, 6038, 6039.

Ivindomyrus marchei: Gabon, Moyen-Ogooué, Ogooué R. at Lambaréné, CUMV 80252, 2, 113 & 115 mm SL, tag nos. 2940, 2941; CUMV 80244, 1, 115 mm SL, tag no. 2908; Ngounié, Loétsi R., CUMV 95128, 3, 105–114 mm SL, tag nos. 6665, 6673, 6674; Ogooué-Ivindo, Ivindo R. at Loa Loa, CUMV 96827, 4, 90–133 mm SL, tag nos. JPS-1101, JPS-1101, JPS-1038, JPS-1043, JPS-1086; Ogooué-Lolo, Sébé R., CUMV uncat., 1, 105 mm SL, field no. TNC14-061; CUMV 89325, 1, 95 mm SL, tag no. 5668.

Ivindomyrus opdenboschi: Gabon, Ogooué-Ivindo, Ivindo R. at Loa Loa, CUMV 83107, 2, 89 & 97 mm SL, tag nos. 4800, 4847; CUMV 89324, 4, 101–114 mm SL, tag no. 5599, 5653, 5654, 5682.

Marcusenius moorii: Gabon, Haut-Ogooué, Ogooué R., CUMV 80460, 1, 123 mm SL, tag no. 3453; Okoloville, CUMV 81623, 1, 124 mm SL, no tag; Moyen-Ogooué, Ogooué R. off point of Lambaréné Island, CUMV 83100, 4, 111–125 mm SL, tag nos. 4950, 4951, 4954, 4972; Ngounié, Louétsi R., CUMV 80346, 1, 126 mm SL, tag no. 3107; CUMV 84517, 1, 122 mm SL, tag no. 2693; Ogooué-Ivindo, Ivindo R., CUMV 89372, 2, 100 & 115 mm SL, tag nos. 5556, 5576; CUMV 96830, 1, 107 mm SL, tag no. JPS-1020; CUMV 81658, 1, 112 mm SL, tag no. 2082; Ogooué R. at Lopé, CUMV 83084, 2, 85 & 85 mm SL, tag nos. 4760, 4761.

Paramormyrops kingsleyae: Gabon, Moyen-Ogooué, Lambaréné, Mikouma Creek, CUMV 80232, 8, 86–108 mm SL, tag nos. 2846–2851, 2854, A3; Ngounié, Biroundou Creek, CUMV 80342, 2, 94 & 121 mm SL, tag nos. 3214, 3215; Bambomo Creek, CUMV 80343, 2, 102 & 109 mm SL, tag nos. 3121, 3128; Woleu-Ntem, Bikagala Creek, CUMV 80348, 2, 102 & 118 mm SL, tag nos. 3209, 3210; CUMV 80251, 7, 85–118 mm SL, tag nos. 3123, 3125, 3129, 3192, 3196, 3197, 3199; Bakanda Creek CUMV 80527, 3, 93–120 mm SL, tag nos. 3221, 3223, 3224.

Paramormyrops sp. (undescribed species): Gabon, Haut-Ogooué, Okoloville, CUMV 80474, 3, 84–101 mm SL; CUMV 80812, 8, 73–99 mm SL; Creek at Bibassa, MRAC 9450 P 0074, 1, 95 mm SL; Ogooué-Ivindo, Balé Creek, CUMV 83110, 1, 90 mm SL; Ivindo R. at Loa Loa rapids, CU89380, 3, 87–115 mm SL.

Paramormyrops sp. “SN4”: Gabon, Haut-Ogooué, Ogooué R. at Franceville, CUMV 80458, 1, 127 mm SL, tag no. 3396; CUMV 80463, 2, 126 & 126 mm SL, tag nos. 3466, 3477; CUMV 80901, 2, 118 & 125 mm SL, tag nos. AG, AH; Mpassa R., CUMV 80501, 3, 112–121 mm SL, tag nos. 3747, 3750, 3755; CUMV 80507, 2, 127, 128 mm SL, tag nos. 3707, 3710; Ogooué-Lolo, Doumé, CUMV 96811, 2, 108 & 117 mm SL, tag nos. JPS-1117, JPS-1119; Woleu-Ntem, Okano R. at Na, CUMV 83059, 3, 86–106 mm SL, tag nos. 4680, 4681, 4684; CUMV 83078, 2, 85 & 86 mm SL, tag nos. 4728, 4729; CUMV 83085, 2, 85 & 87 mm SL, tag nos. 4723, 4725; CUMV 83155, 1, 103 mm SL, tag no. 5.

Stomatorhinus walkeri: Gabon, Ogooué-Lolo, Ogooué R., Haut-Ogooué, Ogooué R., CUMV 80477, 2, 86 & 89 mm SL, tag nos. 3605. 3608; CUMV 81631, 1, 79 mm SL, no tag. Ogooué-Lolo, Doumé, MNHN A-0894 (holotype of Petrocephalus affinis Sauvage), 1, 82 mm SL. Moyen-Ogooué, Ogooué R., BMNH 1867.5.3.15, .16 (syntypes of Mormyrus walkeri Günther), 2, 85, 86 mm SL; CUMV 80237, 2, tag nos. 2877, 2880; CUMV 80248, 1, 82 m mSL, tag no. 2912; CUMV 80255, 1, 75 mm SL, tag no. 2922; Ogooué-Ivindo, Ogooué R. at Lopé, CUMV 83071, 1, 80 mm SL, tag no. 4770.