Pseudobarbus skeltoni, Chakona & Swartz, 2013

Chakona, Albert & Swartz, Ernst R., 2013, A new redfin species, Pseudobarbus skeltoni (Cyprinidae, Teleostei), from the Cape Floristic Region, South Africa, Zootaxa 3686 (5), pp. 565-577 : 570-576

Pseudobarbus skeltoni sp. nov.

( Figure 3 ; Table 2)

Type material. Holotype: South Africa: Western Cape Province: SAIAB 187243 , mature male, 158.2 mm standard length ( SL), collected on 5 December 2009 by A. Chakona and J. Merron using fyke nets at the pool-riffle interphase at the foot bridge just below gauging weir in the Riviersonderend River upstream of Theewaterskloof Dam, Breede River system, at altitude 327 m (34 o 03' 44.3" S, 19 o 04' 13.8" E).

Paratypes: South Africa: Western Cape Province: SAIAB 187244 , BMNH 2012.9 .17.1, USNM408375 , same data as for holotype ; SAIAB 187241 , collected on 3 December 2009 by A. Chakona, J. Merron and M. Jordaan using seine and hand net in a deep pool (33 o 43' 22" S, 19 o 06' 46.6" E), Krom River, tributary of the Molenaars River, Breede River system ; SAIAB187242 , collected 200 m upstream of locality for SAIAB 187241 , same date and collectors; SAIAB 122627 , collected from the Wit River (Bainskloof), Breede River system by P.H. Skelton and P.H. Greenwood on 14 October 1975 ; SAIAB 187238 collected from the Wit River (Bainskloof), Breede River system by Farquharson. Hologenetype ( GenBank number KF 222578 ) and paragenetype sequences (GenBank numbers KF222577 , KF222579 – KF 222588 ) have been deposited in GenBank following the definitions of Chakrabarty (2010) .

Diagnosis. Pseudobarbus skeltoni sp. nov. reaches the largest size of any Pseudobarbus species and is distinguished from its congeners, except P. burchelli and P. burgi , by possessing two pairs of prominent oral barbels. The new species is distinguished from P. burchelli and P. burgi by the combination of the following features: distinctive terminal mouth in adults (versus subterminal in all size classes of P. burchelli and P. burgi ) ( Fig. 4 ); juveniles, sub-adults and young adults of P. skeltoni have subterminal mouths (as in P. burchelli and P. burgi ), but their lower lips are unretracted and they lack the cartilaginous plate found in both P. burchelli and P. burgi ( Fig. 5 ); a typically longer head relative to standard length (30.5 % vs. 26.8% in burchelli and 25.8% in burgi ); shallow head depth (64.2% of HL compared to 70.1% of HL in P. burchelli and 74.1% of HL in P. burgi ); mature sexually active males develop two clusters of large conical tubercles on the snout, with 22 - 30 tubercles in each cluster (versus 5–10 tubercles per cluster in P. burchelli and 10–15 tubercles per cluster in P. burgi ; multiple rows of large conspicuous tubercles above each nare and antero-dorsal edge of each orbit (vs. single row in congeners) ( Fig. 4 ). Comparisons among the species are given in Table 2.

Description. Proportional measurements and meristics are summarised in Table 2. Body fusiform, dorsal surface slightly convex and ventral surface flattened from anterior base of pelvic fins to tip of lower lip; head relatively large and slightly depressed; head length greater than body depth. Adults have distinct terminal mouths; juveniles, sub-adults and young adults have inferior mouths, but lower lip unretracted and lack cartilaginous plate ( Fig.5C ). Two pairs of simple and well-developed barbels, both pairs are longer than the orbit diameter in adults. Eyes relatively small, located dorsolaterally, closer to tip of snout than to the caudal margin of gill cover, interorbital profile flat, and snout prominent.

*Twelve individuals of P. skeltoni were examined for vertebral counts.

Mature breeding males develop large prominent conical tubercles on the snout and top of the head. Two clusters of tubercles develop on the snout, with 22 to 30 tubercles in each cluster in large individuals. The tubercles measure up to 2.2 mm basal diameter. Multiple rows of tubercles develop above each nare, with 6 to 16 tubercles per cluster. Multiple rows of tubercles also develop above the antero-dorsal edge of each orbit, with up to 12 tubercles per cluster. A cluster of smaller tubercles develop on the posterior end of the head. Numerous minute tubercles form on scales between the nape and the dorsal fin and also on some scales between the dorsal and caudal fins. Tubercles do not form on body scales (apart from the ones mentioned above). Rows of tubercles develop on the primary dorsal ray and pectoral fin rays. Only a few minute tubercles occur in sexually active females (mainly on the snout).

Distal margin of dorsal fin slightly concave with 3 or 4 simple and 7 branched rays, anterior base of dorsal fin is inserted slightly posterior to the origin of pelvic fins. Anterior base of dorsal fin generally closer to base of caudal fin than tip of snout. Fins larger in males than females. Pectoral fin fan-shaped, with 13 or 14 rays, shorter than head length, almost reaching anterior base of pelvic fin in males, but only reaching slightly beyond halfway to anterior base of pelvic fin in females. Posterior edge of pelvic fin gently rounded with 8 rays, shorter than head length, reaching the anus in males and reach within two scale rows from the anus in females. Distal margin of anal fin almost straight, with three simple and five branched rays, origin closer to anterior base of pelvic fin than caudal fin base. Anus situated near the base of the anal fin. Caudal fin is forked, length shorter than head length.

Scales moderately large; lateral line complete, perforating 36–39 scales; 6 or 7 scale rows between lateral line and anterior base of dorsal fin; 5 or 7 scale rows between lateral line and anterior base of pelvic fin; and 4–6 scale rows between lateral line and anterior base of anal fin. Predorsal scales 17–21, smaller than flank scales, embedded in skin, with numerous small tubercles extending from nape to anterior base of dorsal fin. Fifteen to 18 circumpedunculer scales (mode 16 or 17). Naked patch between head and posterior base of pectoral fins, scales between posterior end of pectoral fins and anterior end of pelvic fins reduced and embedded. Scales between nape and anterior base of dorsal fin embedded. Axillary scales of pelvic fin not prominent or elongate.

Thirty-six to 38 vertebrae (mode 37), 19–21 precaudal vertebrae (mode 20), 16–18 caudal vertebrae (mode 17), 12 predorsal vertebrae and 20–22 preanal vertebrae (mode 21). Gill rakers sparse and conspicuous (7–9 on leading arch; observed in two individuals). Three rows of pharyngeal teeth, teeth pattern 2, 3, 5-5, 3, 2 (observed in 2 individuals); teeth are asymmetrical and hooked at their tips.

Colouration. Refer to Figure 3 for general live colouration. In fresh specimens, body olive to golden-tan dorsally and laterally, becoming lighter ventrally. Base of fins yellowish-orange with red patches in some specimens. Juveniles and subadults have well marked spots above lateral line, but spots are absent in adults. Alcohol preserved specimens are dark greyish dorsally and laterally and ash grey or whitish ventrally. Black spots present above lateral line in juveniles and sub-adults, but spots are absent in adults.

Reproduction. Unknown, but spawning possibly occurs around November-December, based on the observation that the males had large tubercles. Development of tubercles in redfins has been found to coincide with reproductive season ( Skelton 1988; Cambray 1994), but this needs to be confirmed for P. skeltoni .

Distribution and habitat. Pseudobarbus skeltoni was collected from only two localities, one in the upper Riviersonderend River (a major tributary of the Breede River) and another in the Krom River (a tributary of the Molenaars River), in the Breede River system ( Fig. 6 ). Some specimens from two lots collected by P.H. Skelton and P.H. Greenwood in 1975 (AMP 3463) and by Farquharson (AMP 1411) from the Wit River (Bainskloof) ( Fig. 6 ) and deposited in the Albany Museum as P. burchelli are morphologically concordant with the new species described herein, and are thus reassigned to P. skeltoni . Pseudobarbus skeltoni has not been collected from the Wit River in recent surveys, suggesting that this population may have possibly been extirpated, because the river section where the specimens were originally collected in 1975 has now been invaded by alien African sharptooth catfish ( Clarias gariepinus (Burchell, 1822)) and North American smallmouth black bass ( Micropterus dolomieu (Lacepède, 1802)) . The Riviersonderend population occurs in sympatry with P. burchelli , as was the case for the Wit River population.

Molecular data revealed a close relationships (sharing of haplotypes) between the Riviersonderend and Krom populations, suggesting that these populations may have been connected historically. It also suggests that P. skeltoni could have occurred in mainstem habitats, or at least used it as conduits for dispersal. Given the comprehensive sampling that was done by the authors across the south-western Cape Floristic Region between December 2008 and December 2009, the Riviersonderend and Krom populations probably represent the remaining natural range of the species. Based on the extensive surveys, genetic results and considering what happened to other threatened fishes of the Breede River system, P. skeltoni probably underwent an initial rapid decline after the introduction of alien fishes (particularly black bass species) due to predation and competition. The alien fishes are now dominant in the mainstem habitats, precluding the possibility for native fish species to breed or even survive as adults in these areas.

In the upper Riviersonderend ( Fig. 7A ), specimens were caught at the pool-riffle interphase using fyke nets, with a single individual caught in gill nets in the deep pools (> 2 m depth). The water was brown and peat stained, moderately flowing at the pool-riffle interphase, and static in the pools. Water temperature was 19.3 o C, pH was 4.5, conductivity was 22 µS/cm and TDS 22 ppm. The river bed is dominated by large boulders and cobbles. The banks along the pool margins were covered with dense stands of Palmiet Prionium serratum ( Fig. 7A ). In the Krom River ( Fig.7B ), the water was clear and the fish were observed in both riffles and pools with moderately flowing water. Samples were collected at 485 m altitude, water temperature was 20.3 o C, pH was 4.4, and conductivity was 11 µS/ cm. The river bed has large boulders, cobbles and gravel and little sand. The riparian zone is dominated by woody vegetation. Submerged rocks were covered with periphyton. The inferior mouth in juveniles and sub-adults suggest that younger age groups of P. skeltoni probably feed by picking prey or algae from rock surfaces, while the distinct terminal mouth in adults suggest a shift towards a more predatory lifestyle and possibly a water column feeder.

Conservation. Ensuring the future survival of P. skeltoni presents a serious challenge to conservation managers due to its restricted distribution and small number of remaining individuals. The upper Riviersonderend population persists above Theewaterskloof Dam, which has large populations of North American black bass species ( Micropterus dolomieu (Lacepède, 1802) , M. punctulatus (Rafinesque, 1819) and M. salmoides (Lacepède, 1802)) , the sharp-tooth catfish ( Clarias gariepinus (Burchell, 1822)) and other alien fishes. The upper Riviersonderend currently has the largest specimens and the highest number of individuals of P. skeltoni (87 adult adults and over 200 subadults were caught in fyke nets in December 2009 and were released alive). The future survival of this population is uncertain, because there is no barrier to prevent upstream migration of non-native fishes from Theewaterskloof Dam. Snorkel counts indicate that the Krom population is represented by less than 120 adult individuals in a stream reach of less than 50 m. Here, the major threat is the presence of rainbow trout, Oncorhynchus mykiss (Walbaum, 1792) , that has invaded beyond popular fly-fishing streams lower down in the catchment.

Etymology. The specific name, skeltoni , is given in honour of Professor Paul H. Skelton in recognition of his lifelong service to taxonomic and systematic research on freshwater fishes in southern Africa, his contribution to the taxonomic revision and systematics of Pseudobarbus and specifically for his mentoring of students on this group of fishes.

Figures

FIGURE 3. Pseudobarbus skeltoni sp. nov. (holotype, SAIAB 187243, 158.2 mm SL): A. illustration by Elain Heemstra of SAIAB; B. lateral view showing live colouration; and C. lateral view showing colouration of alcohol preserved specimen.

FIGURE 4. Lateral view of the heads of (a) P. skeltoni sp. nov. (holotype, SAIAB 187243, 158.2 mm SL), (b) P. burchelli (neotype, SAIAB 126387, 95.9 mm SL) and (c) P. burgi (PG2076, 93.5 mm SL). Note the position of mouth, and a large eye relative to head length in P. burchelli and P.burgi. Also note the clear differences in tubercle expression of mature males of the three species.

FIGURE 5. Ventral view showing the difference in the mouth forms of (A) P. burchelli (specimen from RUSI 058390), (B) P. burgi (specimen from SAIAB 186124) and (C) P. skeltoni sp. nov. subadult (specimen from SAIAB 187241). Both P.burgi and P. burchelli have retracted lips exposing a cartilaginous plate, but lips are unretracted in P. skeltoni and cartilaginous plate is absent (based on seven specimens examined).

FIGURE 6. Map of the Breede River system in the south-western Cape Floristic Region of South Africa showing the distribution of Pseudobarbus skeltoni sp. nov. Individuals of this species were previously collected from the Wit River in 1975, but have not been found in recent surveys. The upper Riviersonderend is the type locality.

FIGURE 7. Pictures showing localities and habitat where specimens of Pseudobarbus skeltoni were collected: (A) upper Riviersonderend (note the foot bridge of hiking trail crossing the river just below the gauging weir), and (B) Krom River.