Marcusenius caudisquamatus, Maake, Pholoshi A., Gon, Ofer & Swartz, Ernst R., 2014

Maake, Pholoshi A., Gon, Ofer & Swartz, Ernst R., 2014, Descriptions of three new species of Marcusenius Gill, 1862 (Teleostei: Mormyridae) from South Africa and Mozambique, Zootaxa 3780 (3), pp. 455-480 : 466-469

Marcusenius caudisquamatus , sp. nov.

Figure 2 c, Table 5

Holotype. SAIAB 88684, 166 mm SL, male, Mhlatuze River, under the bridge on road R102, 28°50'44.2" S, 31°51'59.6" E, South Africa, P. Maake, B. Kramer and B. Mackenzie, 8 November 2009. Hologenetype COI (GenBank number KJ174311 ), Hologenetype cyt b (GenBank number KJ174301 ).

Paratypes. SAIAB 191225, 6: 87.6–175 mm SL, Paragenetypes COI (GenBank KJ174309 and KJ174310 ), Paragenetypes cyt b (GenBank KJ174299 and KJ174300 ), same collection data as holotype, and MRAC, B3-16-P- 3, 113.07 mm SL, Paragenetype COI (GenBank KJ174312 ), paragenetype cyt b (GenBank KJ174302 ), same collection data as holotype. SAIAB 79149, 4: 60.3–114.4 mm SL, KwaMaZulu stream close to where it flows into Goedertrouw Dam, part of the Mhlatuze River system in KwaZulu-Natal, 28°25'30'' S, 31°1'30'' E, J. Engelbrecht and B. Kramer, 12 August 1999. SAIAB 96607, 2: 200–218 mm SL, Nseleni River - Nsezi Lake, 28°44.766' S, 31°58.728' E, O. Weyl, R. Karsing and B. Ellender, 27 May 2010. BMNH 2013 9.4. 74, 215 mm SL, same collection data as previous lot. SAIAB 88692, 2: 242–255 mm SL, Nseleni River, 28°42'57.8" S, 31°59'34.4" E, P. Maake and O. Weyl, 10 November 2009. SAIAB 96543, 2: 192–193 mm SL, Nseleni River - Nature Reserve - Richards Bay, 28°41.965' S, 32°00.077' E, O. Weyl, R. Karsing, R. Jones and B. Ellender, 23 May 2010. USNM 410775, 184 mm SL, same collection data as previous lot. SAIAB 96603, 196 mm SL, Lower Nseleni River opposite Davidson's Farm, 28°41.965' S, 32°00.077' E, O. Weyl, R. Karsing, R. Jones and B. Ellender, 27 May 2010.

Diagnosis. Middle body depth 23.8–25.9% SL; dorsal fin length 15.4–19.2% SL; anal fin length 23.0–24.9% SL; pre-dorsal length 61.7–66.7% SL; pre-anal length 60.7–62.8% SL; caudal peduncle depth 40.4–51.7% of its length (greater than and non-overlapping with M. krameri and M. lucombesi ); dorsal fin rays 20–22; anal fin rays 27–28; scales around caudal peduncle 19; lateral line scales 72–73; anterior GR (4) + (5–6) = (9–10); posterior GR (6) + (9) = 15; conical teeth on upper/lower jaws 5–5.

Description. Measurements and meristic counts are given in Table 5. Head with terminal mouth well in front of eye; mental lobe on lower jaw protruding in front of upper jaw; snout rounded and blunt; pre-anal distance shorter than pre-dorsal distance; distance from origin of anal fin to origin of dorsal fin greater than middle body depth; pre-pelvic distance twice as long as distance between pelvic and anal fin; dorsal and anal fins set well back on the body, dorsal fin situated about three fourths of standard length from snout and opposite anal fin; dorsal fin shorter, originating on vertical at 4th or 5th anal fin ray and ending before anal fin base; distal margin of dorsal and anal fins obliquely orientated, with rays becoming gradually shorter posteriorly; dorsal fin rays 20 (20–22), its anterior rays highest, with distal margin rounded and slightly concave; anal fin rays 27 (27–28); males at sexual maturity have kink in the base of the anal fin distinctly curving inward; anterior anal fin rays of sexually mature males strong and sometimes darker, 3rd–5th rays longer than the first two rays, crescentic and rounded or curved backwards, but are anteriorly sharp or pointy in females and juveniles; pectoral fin rays 10; pectoral fin length 73.4–96.9% of head length; middle body depth 25.1% (23.7–25.9%) of SL; caudal peduncle thicker and stronger, subcylindrical across its entire length, 19.7% (18.6–21.6%) in SL; scales along the caudal peduncle circumference 19 (N = 15); caudal peduncle depth 45.2% (40.4–51.7%) into its length; lateral line scales 70 (72–73); lateral line scales cycloid with reticulate striae; jaws with 5–5 conical teeth; 44 vertebrae (excluding urostyle); anterior GRt = 9 (9–10); on the first gill arch ( Table 6), gill rakers of the anterior row on first arch conicals, thin and more developed (longer and smooth); posterior GRt = 15 on the first gill arch; rakers on the posterior row shorter, thicker and with flattened tips.

Character Holotype N Mean/Median Min Max Std. Dev SL, mm 166 mm 15 157.19 84.64 250 49.23 HL, mm 35.34 mm 15 32.77 19.13 51.3 8.89 Ratio of SL

HL 0.2129 15 0.2114 0.1868 0.2273 0.0117 vBD 0.2404 15 0.2399 0.2305 0.2481 0.0050 mBD 0.2512 15 0.2495 0.2376 0.2588 0.0065 dAD 0.2519 15 0.2556 0.2443 0.2727 0.0081 CPL 0.1966 15 0.2017 0.1864 0.2168 0.0083 PDL 0.6527 15 0.6414 0.6171 0.6668 0.0138 PAL 0.6165 15 0.6186 0.6078 0.6285 0.0052 pD 0.4047 15 0.4056 0.4001 0.4092 0.0028 * MP pA 0.4313 15 0.4339 0.4301 0.4379 0.0026 * MP pPL 0.2301 15 0.2365 0.2301 0.2464 0.0056 pVL 0.4119 15 0.4089 0.3912 0.4234 0.0090 dVA 0.2134 15 0.2127 0.1927 0.2336 0.0110 NB: MP = M. pongolensis , ML = M. lucombesi , MK = M. krameri

Electric organ discharge. Kramer et al. (2007) concluded that the EODs of seven individuals from Mhlatuze River (SAIAB 79149, 4; ZSM 35090, 3) cannot be differentiated from those of M. pongolensis collected from the Incomati and Pongola river systems.

Remarks. Marcusenius caudisquamatus can easily be distinguished by its shorter and thicker caudal peduncle, (40.4–51.37% of peduncle length as opposed to 26.0–40.6% in M. krameri and 28.7–45.66% in M. pongolensis ). The deeper caudal peduncle has more circumpeduncular scales than in other species (19 vs 16 in M. krameri and 16–17 for M. pongolensis , Table 7). The dorsal profile of the head of M. caudisquamatus specimens from the Mhlatuze River has a steep, nearly vertical hump in front of the eye. It is slightly concave behind the eye ( Fig. 2 c) as compared to being convex in M. pongolensis and M. lucombesi ( Figs. 2 a and 2d, respectively), and straight in M. krameri ( Fig. 2 b). The number of GR on the posterior side of the first arch is constant and a reliable character to distinguish this species from both M. pongolensis and M. krameri ( Table 6). Compared to M. krameri and M. pongolensis in which the scales series across their body are easily visible, M. caudisquamatus appears smooth and the scales are almost fully embedded in the skin (observed both in live and preserved specimens). The species is genetically closely related to M. krameri ( Fig. 4 ). Sequence divergences between M. krameri and M. caudisquamatus is 0.9–2.4%, and that is much larger that within the latter species (0.0–0.5%).

Gill arch, anterior series Gill arch, posterior series

Upper limb Lower limb Upper Limb Lower limb

Number of GR 4 5 6 4 5 6 6 7 8 7 8 9 M. caudisquamatus 10* 9* 1 10 10 M. krameri 13* 2 15* 15* 15* M. lucombesi 4* 1 4* 1 5* 5*

M. pongolensis 17 4 21 18 3 18 3 Total Gill Rakers

Gill arch, anterior series Gill arch, posterior series

Number of GR 8 9 10 11 13 14 15 16 17 M. caudisquamatus 9 1 10 * - M. krameri 13* 2 - 15* M. lucombesi 3* 2 5* - M. pongolensis 17 4 18 3 -

Dorsal rays Anal rays Circumpeduncular scales 20 21 22 23 24 25 26 27 28 29 30 31 14 15 16 17 18 19 M. caudisquamatus 6* 8 1 7* 8 - 19* M. krameri 4 26* 3 2 20 11* 3 1 35 * -

M. lucombesi 1 7* 5 4 9* 5* 5 3 -

M. pongolensis 5 15* 25 5 4 33* 13 25* 25 - Coloration. In preservation, specimens are grey to blackish. In 70% alcohol the upper back and belly are covered with a milky-grey mucus layer, especially on the head where opaque ‘mormyrid skin’ (electroreceptor organ area) is present.

Distribution and ecology. Currently this species is known only from the Nseleni and the Mhlatuze rivers in KwaZulu-Natal, South Africa ( Fig. 1 ). The Nseleni River was generally deep with strong currents and all the specimens were collected with gill nets. The Mhlatuze River was wide and shallow, about 1.5 m deep, with a sandy or rocky bottom with dense reed beds on the periphery. Water current was moderate to fast in sections where M. caudisquamatus was collected. The water conductivity was relatively high (503 µS/cm to 707 µS/cm) in summer (November 2009) as compared to the 134 µS/cm recorded by Kramer et al. (2007) in winter (12 August 1999). The water temperature at noon was around 25–26 °C in November 2009, the water was clear and specimens were caught at various depths depending on the sampling station. The Mhlatuze River system has the southernmost record of mormyrids in Africa ( Kramer et al. 2007).

The biology and ecology of the species is unknown, but the characteristic anal fin notch of sexually mature males was observed from about 96.7 mm SL. The largest male was 250 mm and the largest female was 215 mm SL.

Etymology. A combination of the Latin words cauda, for tail, and squama for scale. This name refers to the high number of circumpeduncular scales in this species relative to the other members of the genus in South Africa.

Figures

FIGURE 1. (a) Map of South Africa showing collection sites of Marcusenius pongolensis (black dots), M. krameri (pink diamonds) and M. caudisquamatus (red triangles). Names of major river systems are indicated in italics.

FIGURE 2. Photographs and radiographs of preserved specimens of four species of Marcusenius: (a 1 and a 2) M. pongolensis, SAIAB 188296, 165 mm SL, male, Pongola River; (b 1 and b 2) M. krameri, SAIAB 188295, 116.2 mm, male, holotype, Mogalakwena River; (c 1 and c 2) M. caudisquamatus, SAIAB 88684, 166 mm SL, male, holotype, Mhlatuze River; and (d 1 and d 2) M. lucombesi, SAIAB 73884, 74.0 mm SL, male, holotype, Lucombe River.

FIGURE 4. Maximum likelihood phylogram (lnL = 2537.87) based on mitochondrial cytochrome b sequences from the present study and GenBank using Marcusenius stanleyanus and Campylomormyrus compressirostris as outgroups. Values above the branches are bootstrap support based on 10 0 0 0 parsimony bootstrap replicates. Locality names, and SAIAB and Genbank accession numbers are indicated for each species individually.