Introduction
The Sawtooth barracuda, Sphyraena putnamae Jordan & Seale (Family Sphyraenidae), is widely distributed along the Mozambican coast (Fischer et al. Reference Fischer, Sousa, Silva, de Freitas, Poutiers, Schneider, Borges, Féral and Massinga1990) and is well known for hosting sexually adult trematodes (Bray and Justine Reference Bray and J-L2011), including Bucephalidae Poche, 1907. Bucephalidae are cosmopolitan Platyhelminthes comprising about 380 parasitic species in marine and freshwater fish, representing 6.1% of the world’s fauna (Cribb et al. Reference Cribb, Bott, Bray, McNamara, Miller, Nolan and Cutmore2014). The family is comprised of nine subfamilies, the most speciose being Bucephalinae Poche, 1907 with eight genera (Montes et al. Reference Montes, Vercellini, Ostoich, Shimabukuro, Cavallo, Cardarella and Martorelli2023). Within the subfamily, 89% of the 250 species are represented by three genera: Bucephalus Baer, 1827 (77 species), Prosorhynchoides Dollfus, 1929 (78 species), and Rhipidocotyle Diesing, 1858 (67 species) (Corner et al. Reference Corner, Cribb and Cutmore2020).
The trematode genus Rhipidocotyle includes relatively small cosmopolitan parasites which, in their definitive host, inhabit the intestines of a wide range of marine and freshwater piscivorous fishes (Overstreet and Curran Reference Overstreet, Curran, Gibson, Jones and Bray2002; Bott et al. Reference Bott, Healy and Cribb2005; Cribb et al. Reference Cribb, Bott, Bray, McNamara, Miller, Nolan and Cutmore2014). Currently, the genus contains 14 freshwater and 53 marine species, which have been described or recorded from Europe, Asia, Africa, Australia, and the Americas (Ahyong et al. Reference Ahyong, Boyko, Bailly, Bernot, Bieler, Brandao, Daly, De Grave, Gofas, Hernandez, Hughes, Neubauer, Paulay, Boydens, Decock, Dekeyzer, Goharimanesh, Vandepitte, Vanhoorne and Zullini2024).
Adult members of Rhipidocotyle are distinguished from all other typical bucephalids principally by rhynchus morphology that consists of a simple sucker, usually adorned with polygonal cap-like expansion ‘hood’ protruding in a rim (Gibson Reference Gibson1996; Curran and Overstreet Reference Curran and Overstreet2009). The rhynchus assumes various formats, sometimes with small papillae that are more or less extensible (Manter Reference Manter1940). Morphological delimitation of the genus includes a slightly bent Pars prostatica, oblique testes, and a pre-testicular ovary (Curran and Overstreet Reference Curran and Overstreet2009).
The systematics of the subfamily Bucephalinae have been based on broad and general morphological characteristics, leading to several highly speciose genera composed of species that are dissimilar ecologically and phylogenetically (Corner et al. Reference Corner, Cribb and Cutmore2020). Recently, some systematic studies have employed molecular techniques alongside morphological descriptions to differentiate species, further improving databases available for comparison (Atopkin et al. Reference Atopkin, Shedko, Rozhkovan, Nguyen and Besprozvannykh2022).
Members of the three largest Bucephalinae genera (Bucephalus, Rhipidocotyle, and Prosorhynchoides) have been reported from sphyraenid fishes worldwide (Nagaty Reference Nagaty1937; Reimer Reference Reimer1985; Bray and Justine Reference Bray and J-L2011). Nine Rhipidocotyle spp. are known from African marine waters: R. eckmanni Nagaty, Reference Nagaty1937; R. heptatheleta Stunkard, Reference Stunkard1974, and R. khalili from the Red Sea; R. ernsti Reimer, Reference Reimer1985; R. lamberti Reimer, Reference Reimer1985 and R. tonimahnkei Reimer, Reference Reimer1985 from Mozambique; R. paruchini Gavrilyuk-Tkachuk, Reference Gavrilyuk-Tkachuk1979 from Cape Agulhas, Indian Ocean; R. senegalensis Fischthal & Thomas, Reference Fischthal and Thomas1972 from Senegal; and R. ghanensis Fischthal & Thomas, Reference Fischthal and Thomas1968 from Ghana (Nagaty Reference Nagaty1937; Fischthal and Thomas Reference Fischthal and Thomas1968, Reference Fischthal and Thomas1972; Stunkard Reference Stunkard1974; Gavrilyuk-Tkachuk Reference Gavrilyuk-Tkachuk1979; Reimer Reference Reimer1985).
Rhipidocotyle bartolli Bray & Justine, Reference Bray and J-L2011 and R. khalili are the only Rhipidocotyle species currently reported from sawtooth barracuda, S. putnamae, worldwide (Bray & Justine, Reference Bray and J-L2011). In our study, adult trematodes from the intestine of S. putnamae off the western shore of Inhaca Island in Maputo Bay, Mozambique, were studied using an integrative taxonomic approach involving morphological data and DNA characterisation. Based on morphological and molecular traits, the specimens were found to represent two new taxa and placed in the genus Rhipidocotyle.
Material and methods
Fish sampling
Sawtooth barracuda were collected by fishermen at the western Nolwe sand bank offshore of Inhaca Island, Maputo Bay (26°05′43.77″S; 32°50′51.12″E; Figure 1). Thirty-five (n = 35) fish ranging between 25 and 60 cm in length were examined. Fish were dissected, and the intestines opened for collection of trematode parasites.
Parasite collection
Detected trematode parasites were isolated from the intestine of naturally infected sawtooth barracuda. Specimens were washed in saline solution (0.85%) and fixed for morphological examination or genetic characterisation. Specimens for morphological examination by light microscopy (LM) were fixed under pressure between a microscope slide and coverslip in formalin–acetic acid–alcohol (FAA) and then preserved with 70% ethanol. For scanning electron microscopy (SEM), unflattened specimens were fixed in 70% ethanol, with those for DNA analyses fixed in 96% ethanol.
Morphological and morphometrical study
For LM, adult trematodes were stained with acetocarmine, followed by a distaining with 0.5% hydrochloric acid in 70% ethanol solution; gradually dehydrated with serial passage through ethanol at 70, 80, 90, 95%, and subsequent repeated exchanges of anhydrous 100% ethanol; cleared in beechwood creosote; and mounted in Canada balsam (Amato et al. Reference Amato, Boeger, Amato and de Janeiro1991; Eiras et al. Reference Eiras, Takemoto and Pavanelli2006) to produce whole-mount preparations. Photomicrographs and morphometric parameters were obtained with an Olympus BX53 compound microscope and Olympus Soft Imaging Solutions (Olympus, Münster, Germany), with photomicrographs used to compile digital illustrations using Corel DRAW® Graphics Suite X6 software (Corel Corporation, Ottawa, Canada). All measurements pertain to whole mounted specimens and are given in micrometres (two significant figures), unless otherwise stated, as the range followed by the mean in parentheses. The standard deviation is also given alongside the mean in parentheses in Table 2. In the present study, we have utilised the visual key to the marine Rhipidocotyle spp. first developed by Bray and Palm (Reference Bray and Palm2009) for metrical criteria and developed a similar key for comparison. The main criteria included in the visual key are body length and width, egg length, rhynchal length, pre-vitelline distance, pre-uterine distance, pre-mouth distance, post-testicular distance, and cirrus-sac reach, all as a percentage of the body-length.
Scanning electron microscope (SEM)
Unflattened trematodes fixed in 70% ethanol were prepared for SEM analysis by dehydration through a graded ethanol series (70 to 95%) and repeated exchanges of anhydrous 100% ethanol, followed by a graded series of hexamethyldisilazane (40 to 100%) (Merck, Darmstadt, Germany) (Nation Reference Nation1983; Dos Santos et al. Reference Dos Santos, Jansen van Vuuren and Avenant-Oldewage2015). Specimens were dried overnight in a Sanpla dry keeper desiccator cabinet (Sanplatec, Osaka, Japan) and coated with gold using an Emscope SC500 sputter coater (Quorum Technologies, Newhaven, UK). A Vega 3 LMH scanning electron microscope (Tescan, Brno, Czech Republic) was used to study the specimens at 5–6 kV.
Additionally, some specimens fixed in 96% ethanol were bisected, and the anterior was prepared for SEM. This was done by rehydrating the specimens, placing them in DESS (Yoder et al. Reference Yoder, De Ley, King, Mundo-Ocampo, Mann, Blaxter, Poiras and De Ley2006) solution overnight, transferring to distilled water for an hour, transferring to steaming hot (≈ 70 °C) neutral buffered formalin (10% NBF) and allowing to cool to room temperature (10 min), rinsing in 70% ethanol, and then processing as above. The posteriors of these specimens were used for genetic characterisation.
DNA extraction, polymerase chain reaction (PCR) and sequencing
A small piece of tissue was removed from the posterior of 10 specimens fixed in 96% ethanol and genomic DNA extracted using an E.Z.N.A.® Tissue DNA kit (Omega Bio-tek, Inc., GA, USA). The anterior of the specimens was either prepared for LM or SEM study as described above. Genomic DNA was also extracted from three specimens following SEM analyses by removing a piece of dried and gold coated tissue to confirm their identity, again from the posterior. Two ribosomal DNA regions were targeted with partial 28S rDNA amplified using LSU5 (5’ - TAG GTC GAC CCG CTG AAY TTA AGC A - 3’) (Littlewood Reference Littlewood1994) and 1500R (5’ - GCT ATC CTG AGG GAA ACT TCG – 3’) (Snyder and Tkach Reference Snyder and Tkach2001), and entire internal transcribed spacer (ITS) rDNA amplified using 81_f (5’ – GTA ACA AGG TTT CCG TAG GTG AA – 3’) (Gustinelli et al. Reference Gustinelli, Caffara, Florio, Otachi, Wathuta and Fioravanti2010) and ITS2:2 (5’ – CCT GGT TAG TTT CTT TTC CTC CGC – 3’) (Cribb et al. Reference Cribb, Anderson, Adlard and Bray1998). Internal primer 5.8S-2 (5’ – GTC GAT GAA GAG CGC AGC – 3’) (Králová-Hromadová et al. Reference Králová-Hromadová, Scholz, Shinn, Cunningham, Wootten, Hanzelová and Sommerville2003) was also used to specifically target the second internal transcribed spacer (ITS2) rDNA if the entire ITS rDNA amplification failed. PCR reactions (30μl) were run using the following conditions: 5 min at 95°C followed by 35 cycles of 30 seconds at 95°C, 30 seconds at 52°C, 2 minutes at 72°C, and a final extension of 5 minutes at 72°C. Amplification was verified on a 1% agarose gel impregnated with SafeViewTM Classic (Applied Biological Materials Inc., Richmond, Canada) and visualised using a SmartDocTM 2.0 gel visualisation and smartphone imaging system (Accuris instruments, Edison, NJ, USA). Amplicons were sequenced in both directions following Avenant-Oldewage et al. (Reference Avenant-Oldewage, Le Roux, Mashego and Jansen Van Vuuren2014).
Obtained sequence data were aligned and edited, if necessary, and reads were merged and compared using Geneious Prime version 2023.2.1 (http://www.genious.com). Generated data were aligned to closely related taxa following Basic Local Alignment Search Tool (BLAST) (Altschul et al. Reference Altschul, Gish, Miller, Myers and Lipman1990) analyses (see Table 1), with Dollfustrema hefeiense Liu in Zhang et al. 1999 included as outgroup following previous studies (Nolan et al. Reference Nolan, Curran, Miller, Cutmore, Cantacessi and Cribb2015; Hammond et al. Reference Hammond, Cribb and Bott2018, Reference Hammond, Cribb, Nolan and Bott2020; Corner et al. Reference Corner, Cribb and Cutmore2020; Curran et al. Reference Curran, Calhoun, Tkach, Warren and Bullard2022). Only data covering at least 80% of the alignment were retained, and identical sequences for conspecific data were removed. Data with doubtful identification were also excluded. The ITS rDNA alignment was trimmed to only ITS2 (at internal primer 5.8S-2) to better accommodate available data from other studies. Data were aligned using MAFFT (Katoh et al. Reference Katoh, Misawa, Kuma and Miyata2002; Katoh and Standley Reference Katoh and Standley2013) via the EMBL-EBI portal and refined manually in MEGA7 (Kumar et al. Reference Kumar, Stecher and Tamura2016). Genetic distances (uncorrected p-distance) and base pair differences were calculated using MEGA7. Evolutionary histories were reconstructed using both maximum likelihood (ML) and Bayesian inference (BI) approaches. The Tamura-Nei nucleotide substitution model (Tamura and Ne, Reference Tamura and Nei1993) was selected for both ML and BI analyses using the model selection tool in MEGA7, with discrete Gamma distribution (five categories) included. Topologies were supported by 1,000 bootstrap replicates (Felsenstein, Reference Felsenstein1985) for ML and 10 million Markov chain Monte Carlo (MCMC) generations were for BI analyses. A single topology per marker analysed is presented based only on BI analyses due to the similarity between ML and BI approaches and the superior BI nodal support. Nodes annotated with support in format BI/ML, with nodes of inconclusive support (less than 50% bootstrap support or 0.5 posterior probability) not annotated or marked with “-”. All obtained sequence data were deposited to GenBank.
a described from the same host species.
b most similar to R. siphonyaka n. sp.
c most similar to R. nolwe n. sp.
d percentage relative to total body length of specimen.
Results
Class Trematoda Rudolphi, 1808
Subclass Digenea Carus, 1863
Superfamily Bucephaloidea Poche, 1907
Family Bucephalidae Poche, 1907
Subfamily Bucephalinae Poche, 1907
Genus Rhipidocotyle Diesing, 1858
Rhipidocotyle siphonyaka n. sp
Description (Figures 2, 3, 6 and Table 2)
Based on 14 whole-mount and 10 SEM preparations of mature specimens. Body 2000–3082 (2580) μm long, 172–311 (253) μm wide, tapering anteriorly, rounded posteriorly, widest at level of ovary (Figure 2). Tegument with scales (as mulitpointed spines, 29 points), retractable, with retracted scales appearing pit-like (Figures 3B; 6). Rhynchus oval 75.3–243 (120) long, 61.7–189 (98.6) wide, with six small tentacles (Figure 3A). Mouth sinistral opening ventral at pre-uterine region, on anterior half of body, always preceded by protuberance (Figures 2A; 3A, C). Mouth contains inner layers with expanded structures (Figure 3C). Pharynx elliptical, 47.1–160 (78.9) long, 29.9–118 (65.9) wide, positioned in anterior half of body. Oesophagus very short. Caecum tube-like, 235–1310 (601) long, 40–261 (122) wide, extending anteriorly from pharynx, in pre-uterine region, then recurving posteriorly to level of ovary; anterior extremity tapered, posterior extremity rounded. Gonads in post-equatorial region (Figure 2A).
Testes two, sub-spherical in ventral view, smooth, tandem, post-equatorial, partially overlapped, both larger than ovary; anterior testis 98.3–277 (135) long, 85.5–180 (115) wide; posterior testis 92.5–287 (140) long, 82.9–293 (133) wide. Cirrus-sac 401–1200 (649) long, 48.8–167 (92.6) wide, tube-like, sinistral, parallel-sided reaching posterior testis. It encloses oval seminal vesicle in its proximal part. Pars prostatica uniform, straight, 255–1030 (427) long, 44.3–78.3 (60.7) wide; ejaculatory duct large in diameter opens on genital lobe; genital lobe bi-lobed inside genital atrium, 200–376 (272) in diameter; genital atrium large (Figure 2C). Genital pore sub-terminal opens ventrally at very short distance from posterior extremity of body (Figures 2A, C; 3A).
Ovary oval round, smooth, equatorial, pre-testicular, medial in between two vitelline fields, 98–287 (154) long, 79.4–269 (139) wide. Oviduct descends from posterior part of ovary. Mehlis’ gland well developed, large, immediately posterior to ovary surrounding oviduct. Vitelline duct connects to vitelline follicles in sinistral field (Figure 2B). Vitellarium composed by two lateral fields of vitelline follicles on each side, commences at level of posterior margin of posterior testis, proceeding longitudinally up to level anteriorly to ovary; one vitelline follicle field slightly longer than the other; vitelline follicles vary from oval to irregular in shape, dextral vitelline follicles numbering 10–13, sinistral ones 11–15; each vitelline follicle measuring 35.3–81 (57.9) long, 23.7–60.7 (38.2) wide. Laurer’s canal not observed, likely obscured by uterus. Genital pore elliptical, sub terminal at short distance from posterior extremity (Figure 3A); tegument around genital pore corrugated with transversal ridges. Uterus extends from genital atrium, fills most available space between genital pore and posterior testis, passing medially between vitelline fields reaching anterior 40% of body, distinctly anterior to vitelline fields, but posterior to pharynx (midway between the anterior vitelline field and the pharynx) (Figure 2A). Eggs oval, very numerous, operculate, appear golden yellow in colour, 17.3–21 (19.3) long, 13–17 (14.5) wide (Figure 3D). Metraterm not observed, likely obscured by uterus.
Excretory pore terminal (Figure 3A). Excretory vesicle saccular.
Taxonomic summary
Type-host: sawtooth barracuda, Sphyraena putnamae Jordan & Seale (Carangaria: Sphyraenidae)
Type locality: Maputo Bay, western shore of Inhaca Island
Site of infection: Intestines
Infection parameters: Prevalence 42.86% (8 of 35 fish infected); Intensity 1–2.
Specimens deposited: Holotype – ovigerous adult specimen deposited at the Iziko South African Museum, Cape Town, South Africa (SAMC-A096876); paratypes: four specimens deposited at the Iziko South African Museum, Cape Town, South Africa (SAMC-A096877 - 80); four paratype (NHM 2024.9.23.1 - 4) and one hologenophore (NHM 2024.9.23.5) specimens deposited at the Natural History Museum, London, UK.
Representative DNA sequences: 28S rDNA – PQ453504-PQ453510; ITS rDNA – PQ453491-PQ453499.
Zoobank: urn:lsid:zoobank.org:act:31149D12-1DD6-4CC9-9656-488E1C4CC47F
Etymology: The species ‘siphonyaka’ refers to the name of Inhaca’s King (Carlos Sipho Nhaca), whose family nominated ‘KaNyaka’ to an Island with western and southern coasts facing Maputo Bay, where the fish host was collected. It is to honour his commitment to environmental conservation.
Remarks
The current material was first compared with marine Rhipidocotyle spp. described from Sphyraena spp. and thereafter to those morphometrically similar according to the updated visual key for the metrical criteria (See Supplementary Table S1).
Among representatives of Rhipidocotyle, six species are acknowledged from Sphyraena spp. – namely, R. khalili; Rhipidocotyle longleyi Manter, Reference Manter1934; Rhipidocotyle longicirrus (Nagaty, Reference Nagaty1937); Rhipidocotyle barracudae Manter, Reference Manter1940; Rhipidocotyle sphyraenae Yamaguti, Reference Yamaguti1959; and R. bartolli, which are opposed to the present new species by having a caecum and a mouth opening in the post-vitelline field or almost midway through the uterus versus a tube-like caecum extending to pre-uterine region and the mouth opening at the pre-uterine field (Nagaty Reference Nagaty1937; Manter Reference Manter1940; Yamaguti Reference Yamaguti1959; Bray and Justine Reference Bray and J-L2011).
Specifically, R. longleyi and R. khalili have a longer body length, shorter pre-uterine distance, and larger egg size versus shorter body length, longer pre-uterine distance, and smaller egg size in R. siphonyaka n. sp. (Manter Reference Manter1934; Nagaty Reference Nagaty1937). However, R. barracudae, R. longicirrus, R. bartolli, and R. sphyraenae have a shorter body length, larger width, and shorter pre-vitelline distance, whereas R. siphonyaka n. sp. has a longer body length, shorter width, and longer pre-vitelline distance (Manter Reference Manter1940; Yamaguti Reference Yamaguti1959; Bray and Justine Reference Bray and J-L2011).
The new species, R. siphonyaka n. sp., resembles Rhipidocotyle fluminensis Vicente & dos Santos, Reference Vicente and dos Santos1973 from Rio de Janeiro State, Brazil and Rhipidocotyle pseudorhombi Nahhas, Sey & Nakahara, Reference Nahhas, Sey and Nakahara2006 from the Arabian Gulf morphometrically. However, R. fluminensis and R. pseudorhombi are distinct from R. siphonyaka n. sp. in having a shorter pre-vitelline distance, and the mouth and pharynx positioned at the level of vitelline field (Vicente and dos Santos Reference Vicente and dos Santos1973; Nahhas et al. Reference Nahhas, Sey and Nakahara2006). Morphologically, R. pseudorhombi is more similar to R. siphonyaka n. sp. with the uterus reaching the pre-vitelline field and a caecum which is tubular. However, the mouth opens at the pre-uterine field in the new species, whereas it opens almost midway through the uterus and vitellaria in R. pseudorhombi (Nahhas et al. Reference Nahhas, Sey and Nakahara2006).
American Rhipidocotyle species also differ from R. siphonyaka n. sp. by having a shorter body length and a sac-like intestinal caecum (Chandler Reference Chandler1935, Reference Chandler1941; McFarlane Reference McFarlane1935; Linton Reference Linton1940; Hopkins Reference Hopkins1954). The majority of Asian species differ from R. siphonyaka n. sp. by having a shorter body length, uterus and digestive organs positioned at post-vitelline field, and a shorter pre-vitelline distance, except for R. pseudorhombi discussed upward (Chauhan Reference Chauhan1943; Yamaguti Reference Yamaguti1959; Wang Reference Wang1985; Nahhas et al. Reference Nahhas, Sey and Nakahara2006; Bray and Palm Reference Bray and Palm2009; Madhavi and Bray Reference Madhavi and Bray2018).
The intestinal caecum, pharynx, mouth, and ovary of the European Rhipidocotyle are arranged in the post-vitelline field (Dimitrov et al. Reference Dimitrov, Kostadinova and Gibson1996; Bartoli et al. Reference Bartoli, Bray and Gibson2006; Bray and Justine Reference Bray and J-L2011), therefore differing from R. siphonyaka n. sp., which are placed in the pre-vitelline field. The only Australian species, R. labroidei, resembles Asian species in having a uterus and caecum in the post-vitelline field. However, R. laroidei differs from R. siphonyaka n. sp. by a shorter body length and pre-vitelline distance, and a longer cirrus-sac reach and rhynchal length (Jones et al. Reference Jones, Grutter and Cribb2003).
African Rhipidocotyle – namely, R. eckmanni, R. ernsti, heptatheleta, R. lamberti, R. paruchini, R. senegalensis, R. tonimahnkei, and R. ghanensis – are also distinct from R. siphonyaka n. sp. by possessing a short saccular caecum, and the internal organs are arranged in the post-vitelline field versus a tube-like caecum and organs arranged in the pre-vitelline field (Nagaty Reference Nagaty1937; Fischthal and Thomas Reference Fischthal and Thomas1968; Stunkard Reference Stunkard1974; Reimer Reference Reimer1985).
Rhipidocotyle nolwe n. sp
Description (Figures 4, 5, 6 and Table 2)
Based on 26 whole-mount and 8 SEM preparations of mature specimens. Body elongate 1550–2740 (2030) μm long, 237–386 (281) μm wide, tapering towards anterior, rounded posteriorly, widest at level of ovary (Figure 4A). Tegument with scales (as multipointed spines, 29 points), retractable; retracted scales resembling pits (openings) in most specimens (Figure 6). Rhynchus a muscular sucker, oval bearing six (6) tentacles (Figure 5A). Mouth medial, equatorial, ventral, almost midway of uterus surrounded by pharynx (Figure 5D). Pharynx muscular, elliptical, equatorial at pre-vitelline region, mostly medial, occasionally overlapping ovary, 47.4–122 (79.7) long, 54.4–146 (83.6). Caecum tube-like, 312–725 (486) long, 59.5–74.6 (67.9), extends anteriorly from the pharynx, posteriorly reaching pre-ovarian region; tapered in anterior part, posterior extremity of caecum rounded. Gonads in posterior half of body (Figure 4A).
Testes two, oval in dorso-ventral view, smooth, oblique, post-equatorial, partially overlapped to spaced in some specimens 0–29 (4.15) long; both testes larger than ovary; anterior testis 98.8–316 (172) long, 82.2–249 (153) wide, posterior testis, 109–221 (144) long, 85.2–214 (136) wide (Figure 4A). Cirrus-sac elongate 423–687 (538) long, 68.9–192 (96.2) wide, sinistrally in hindbody, almost uniform in diameter throughout entire length, parallel-sided, reaching posterior margin of posterior testis; it encloses ellipsoidal seminal vesicle in its proximal part measuring 69.3–258 (115) long, 45.6–117 (69.8) wide. Pars prostatica elliptical, uniform, straight. Ejaculatory duct large in diameter opens on genital lobe inside genital atrium (Figure 4C). Genital lobe large, 171–342 (286) in diameter. Genital pore subterminal opens ventrally into genital atrium (Figure 4C).
Ovary oval, located at beginning of dextral vitellaria, smooth outline, post-equatorial, pre-testicular, generally smaller than testes, 91.6–252 (141) long, 86.1–198 (113) wide. Oviduct uniform, descending from postero-dextral side of ovary. Laurer’s canal not visible, probably obscured by uterus. Mehlis’ gland developed, far posterior from ovary, overlapping posterior testis. Vitelline duct joins latero-dextral side of Mehlis’ gland, then forming widened part prior to connecting dextral vitelline follicles (Figure 4B). Vitellarium containing two lateral fields of vitelline follicles clustered on each side, commences at level of posterior testes running anteriorly up to ovarian level; one field slightly longer than other; vitelline follicles varying from oval to irregular in shape, each measuring 41.1–67 (55) long, 27–63.2 (40.4) wide; dextral vitelline follicles numbering 12–13, sinistral ones 12–16; transverse vitelline duct between ovary and anterior testis connects vitelline follicles from each side. Uterus filled with eggs extends from beyond genital lobe at 44.5–250 (136) from posterior extremity; it fills most available space between genital pore and posterior testis, passing medially between vitelline fields reaching almost anterior third 55% of body, distinctly beyond anterior extent of pharynx (Figure 4A). Eggs numerous, operculate, appear golden yellow in colour, 18.6–21.9 (20.1) long, 11.8–16.8 (14.1) wide (Figures 4A; 5C). Metraterm not observed, probably obscured by uterus.
Excretory pore subterminal; excretory vesicle saccular.
Taxonomic summary
Type-host: sawtooth barracuda, Sphyraena putnamae Jordan & Seale (Carangaria: Sphyraenidae)
Type locality: Maputo Bay, western shore of Inhaca Island
Site of infection: Intestines
Infection parameters: Prevalence 48.57% (17 of 35 fish infected); Intensity 1–5.
Specimens deposited: Holotype – ovigerous adult specimen deposited at the Iziko South African Museum, Cape Town, South Africa (SAMC-A096881); Paratypes: five specimens deposited at the Iziko South African Museum, Cape Town, South Africa (SAMC-A096882 - 6); six paratype (NHM 2024.9.23.6 - 11) and one hologenophore (NHM 2024.9.23.12) specimens deposited at the Natural History Museum, London, UK.
Representative DNA sequences: 28S rDNA – PQ453500-PQ453503; ITS rDNA – PQ453487-PQ453490.
Zoobank: urn:lsid:zoobank.org:act:06ED6D67-A681-4DCF-A882-E4F52DA5FA8B
Etymology: The species ‘nolwe’ refers to the name of the Nolwe sandy bank along the western and southern coasts facing Inhaca Island, where the fish host was collected.
Remarks
Rhipidocotyle nolwe n. sp. resembles R. siphonyaka n. sp. in having the digestive structures arranged in the pre-vitelline field. However, R. siphonyaka n. sp. is distinguished from R. nolwe n. sp. by a mouth and pharynx positioned in the pre-uterine field, a longer body length and cirrus-sac reach, and shorter pre-vitelline distance and post-testicular distance according to the visual key for the metrical criteria (See Supplementary Table S2).
Species of Rhipidocotyle known from Sphyraena spp. – namely, R. longicirrus, R. barracudae, R. sphyraenae and R. bartolli – have a shorter body length versus longer body length in R. nolwe n. sp. (Nagaty Reference Nagaty1937; Manter Reference Manter1940; Yamaguti Reference Yamaguti1959; Bray and Justine Reference Bray and J-L2011). Two species, Rhipidocotyle longleyi and R. khalili, are distinct from R. nolwe n. sp. by their longer body length (Manter Reference Manter1934; Nagaty Reference Nagaty1937). Rhipidocotyle khalili and R. bartolli share the same host species with new species but are distinct from R. nolwe n. sp. by the larger egg size.
American Rhiphidocotyle are distinct from R. nolwe n. sp. by their shorter pre-vitelline distance and placement of the ovary and intestinal caeca in the post-vitelline field. Furthermore, the pharynx and mouth opening are placed posteriorly to vitellaria (Chandler Reference Chandler1935, Reference Chandler1941; Manter Reference Manter1934, Reference Manter1940; McFarlane 1936; Hopkins Reference Hopkins1954; Vicente and Santos Reference Vicente and dos Santos1973). The pre-vitelline distance is also shorter in the Asian species, therefore differing from the new species (Chauhan Reference Chauhan1943; Velasquez Reference Velasquez1959; Yamaguti Reference Yamaguti1959; Nahhas et al. Reference Nahhas, Sey and Nakahara2006).
Among European Rhipidocotyle, R. nicolli Bartoli, Bray & Gibson, Reference Bartoli, Bray and Gibson2006 is the only species having a uterus extending to the pre-vitelline field as in R. nolwe n. sp., but the former differs in possessing a longer body length (Bartoli et al. Reference Bartoli, Bray and Gibson2006). Rhipidocotyle galeata (Rudolphi, Reference Rudolphi1819), Rhipidocotyle genovi Dimitrov, Kostadinova & Gibson, Reference Dimitrov, Kostadinova and Gibson1996, Rhipidocotyle minima (Wagener, 1852), Rhipidocotyle triglae (van Beneden, 1870), and Rhipidocotyle viperae (van Beneden, 1870) are distinct from R. nolwe n. sp. by possessing a shorter body length, larger egg size, and a short sac-like intestinal caecum which is positioned in the post-vitelline field (Nagaty Reference Nagaty1937; Dimitrov et al. Reference Dimitrov, Kostadinova and Gibson1996; Bartoli et al. Reference Bartoli, Bray and Gibson2006).
African Rhipidocotyle – namely, R. eckmanni, R. ernsti, R. heptatheleta, R. lamberti, R. paruchini, R. senegalensis, R. tonimahnkei, R. Khalili, and R. ghanensis – differ from the R. nolwe n. sp. by the larger size of the eggs (Nagaty Reference Nagaty1937; Fischthal and Thomas Reference Fischthal and Thomas1968, Reference Fischthal and Thomas1972; Stunkard Reference Stunkard1974; Gavrilyuk-Tkachuk Reference Gavrilyuk-Tkachuk1979; Reimer Reference Reimer1985).
Specifically, R. eckmanni from the Red Sea is distinct from R. nolwe n. sp. by its shorter body length, shorter post-testicular distance, and longer pre-mouth distance. Additionally, it has a saccular caecum placed in the post-vitelline field at the level of the reproductive organs (Nagaty Reference Nagaty1937). Rhipidocotyle ghanensis from Ghana is distinct from R. nolwe n. sp. by its larger rhynchal length, longer pre-mouth distance, shorter pre-vitelline distance, and the placement of the intestinal caecum in the post-vitelline field (Fischthal and Thomas Reference Fischthal and Thomas1968). Rhipidocotyle senegalensis differs from R. nolwe n. sp. by its shorter body length and width, longer rhynchal length and pre-uterine distance, and shorter pre-vitelline and post-testicular distance (Fischthal and Thomas Reference Fischthal and Thomas1972).
Mozambican Rhipidocotyle, R. tonimahnkei and R. ernsti have shorter pre-vitelline and pre-uterine distances versus longer pre-vitelline and pre-uterine distances in R. nolwe n. sp. Besides R. tonimahnkei having a uterus extending to a pre-vitelline field similar to that of R. nolwe n. sp., it is distinct by its digestive and female reproductive structures placed in the post-vitelline field (Reimer Reference Reimer1985). Rhipidocotyle ernsti is also distinct from R. nolwe n. sp. by its fusiform body, larger width, longer cirrus-sac reach, and shorter pre-vitelline and pre-uterine distances (Reimer Reference Reimer1985).
Molecular analysis
For 28S rDNA, 11 sequences were generated from the 13 specimens included: four for R. nolwe n. sp. (1249–1263bp) and seven for R. siphonyaka n. sp. (1255–1263bp). Each species was represented by a single haplotype with no intraspecific variation. The 28S rDNA alignment included 37 sequences and was 1391bp long with 982bp conserved, 360bp variable, and 211bp parsimony informative sites. Using included 28S rDNA data, intraspecific distances of up to 0.17 % (2bp) were observed, and an interspecific range of 0.64–8.98% (8–116bp) was calculated. The haplotypes for R. nolwe n. sp. and R. siphonyaka n. sp. differed by 1.27–1.28% (16bp), falling within the interspecific range and supporting their distinctness. Both species were also separated from other taxa by 3.95–11.24% (49–141bp) and 3.71–11.9% (46–149bp) for R. nolwe n. sp. and R. siphonyaka n. sp., respectively (See Supplementary Table S3).
For ITS rDNA, sequences were generated from all 13 specimens included: four for R. nolwe n. sp. (1195–1222bp) and nine for R. siphonyaka n. sp. (601–1227bp). Each species was represented by a single ITS rDNA haplotype with no intraspecific variation. The alignment (trimmed to primer 5.8S-2) with other ITS rDNA data included 35 sequences and was 827 long with 331bp conserved, 399bp variable, and 292bp parsimony informative sites.
Using included ITS rDNA data, intraspecific distances of up to 0.47% (3bp) were observed, and an interspecific range of 1.98–26.05% (11–158bp) was calculated. The ITS rDNA haplotypes for R. nolwe n. sp. and R. siphonyaka n. sp. differed by 4.55–4.76% (27bp), falling within the adjusted interspecific range and supporting their distinctness. Both species were also separated from other taxa by 15.7–26.19% (81–143bp) and 14.51–25.05% (75–137bp) for R. nolwe n. sp. and R. siphonyaka n. sp., respectively (See Supplementary Table S4).
Both 28S and ITS rDNA topologies were similar, with R. nolwe n. sp. and R. siphonyaka n. sp. grouping as closely related sister taxa in a well-supported clade, distinct from all other bucephalids. From the included data, Rhipidocotyle angusticollis Chandler, Reference Chandler1941, together with Rhipidocotyle cf. angusticollis, grouped basally followed by Rhipidocotyle lepisostei Hopkins, Reference Hopkins1954. The remainder of the included data grouped into three major clades, the first with the new species from the present study (Clade 1), the second (Clade 2) with P. megacirrus, P. caecorum, and R. galeata, and the third with the remainder of the included data (Clade 3). The relation of these major clades varied between markers, with Clade 2 sister to Clade 1 based on ITS rDNA, whereas Clade 1 was sister to Clade 3 based on 28S rDNA. The position of the unidentified Prosorhynchoides sp. (TW-2019; LC498575-6) by Shirakashi et al. (Reference Shirakashi, Waki and Ogawa2020) also varied between markers, grouping with the study species (Clade 1) based on 28S rDNA and with Clade 2 based on ITS rDNA.
Discussion
Taxonomy
Within sphyraenids, the great barracuda S. barracuda (Edwards) is the only host species harbouring three species of Rhipidocotyle (see Bray and Justine Reference Bray and J-L2011), with the obtuse barracuda Sphyraena obtusata Cuvier infected with four Aenigmatrema species (Corner et al. Reference Corner, Cribb and Cutmore2020). The sawtooth barracuda has a wide distribution essentially spanning the entirety of the Atlantic, Pacific, and Indian Oceans, also including the Caribbean and Red Sea (Bray and Justine Reference Bray and J-L2011; Bogorodsky et al. Reference Bogorodsky, Alpermann, Mal and Gabr2014; Gottfried et al. Reference Gottfried, Samonds, Ostrowski, Andrianavalona and Ramihangihajason2017). Compared to the widely reported distribution of the host, it appears that it has not been well sampled for trematode fauna, particularly in African waters, and so, conceivably, R. siphonyaka n. sp. and R. nolwe n. sp. might be supported across all or most of its definitive host’s range.
Two morphological characteristics of the material studied here warrant further discussion: the six (6) small tentacles and the tegument with scales. Large tentacles associated with a rhynchus are characteristic of three genera within Bucephalidae – namely, Bucephalus with seven, Alcicornis MacCallum, 1917 with seven to 21, and Aenigmatrema Corner, Cribb & Cutmore, Reference Corner, Cribb and Cutmore2020 with 11 tentacles (Overstreet and Curran Reference Overstreet, Curran, Gibson, Jones and Bray2002). Specifically, the observed tentacles, resembling papillae, are less developed and fewer in number compared to those present in Aenigmatrema, Alcicornis, and Bucephalus. The presence of small tentacles (papillae-like) corroborates with Manter (Reference Manter1940), who stated that the ‘hood’ surmounting the sucker of Rhipidocotyle assumes various forms, and it may bear papillae, which are sometimes more or less extensible, similar to the tentacles of Bucephalus, to which they are probably homologous. Yamaguti (Reference Yamaguti1959) described and illustrated a row of seven double papillae on the rhynchus of R. sphyraenae; Moreover, Nagaty (Reference Nagaty1937) also noticed the presence of papillae in R. khalili; Velasquez (Reference Velasquez1959) for R. eggletoni and R. laruei; Bartoli et al. (Reference Bartoli, Bray and Gibson2006) for R. nicolli; and Nahhas et al. (Reference Nahhas, Sey and Nakahara2006) for R. pseudorhombi. Some reports such as that of Rudolphi (Reference Rudolphi1819) and Stossich (Reference Stossich1887) on R. galeata in the northern Atlantic and Mediterranean Oceans also noticed a rhynchus baring six or seven papillae. However, papillae in Rhipidocotyle species seem to lack any clear pattern in terms of number, size, and arrangement and are sometimes referred to as lobes or protuberances.
Nagaty (Reference Nagaty1937) referred to the presence of scales in the family Bucephalidae. Several reports on tegument with scales in Rhipidocotyle were solely based on light microscopy examination – namely, Yamaguti (Reference Yamaguti1959) for R. sphyraenae; Bartoli et al. (Reference Bartoli, Bray and Gibson2006) for R. minima; and Velasquez (Reference Velasquez1959) for R. laruei. They were first observed with SEM in R. angusticollis by Shalaby and Hassanine et al. (Reference Shalaby and Hassanine1996). The scales on the tegument of Rhipidocotyle species are undefined when utilising light microscopy analysis but are well visible as transverse rows and overlapping each other when using SEM examination (Shalaby and Hassanine et al. Reference Shalaby and Hassanine1996). In the present study, both retracted and extruded states of the scales were observed in specimens of the same taxa. Based on SEM images obtained from the present material, it is apparent that the scales are retracted into the tegument during contraction of the body; therefore, small pits are noticeable when the scales are completely retracted (as seen in the sequence in Figure 6).
Intermediate and definitive hosts
The species-rich nature of bucephalid fauna is strongly connected to the bivalves acting as the first intermediate host, infected fishes as the second intermediate host, and primarily a piscivorous fish as the definitive host, where the metacercaria is ingested along with the second intermediate host through feeding (Muñoz et al. Reference Muñoz, Valdivia and Lopez2014). There are a few notable exceptions to the latter (i.e., apogonids (Bott and Cribb Reference Bott and Cribb2005), cleaner wrasse (Jones et al. Reference Jones, Grutter and Cribb2004), and fang blennies (Roberts-Thomson and Bott Reference Roberts-Thomson and Bott2007)). Members of Bucephalidae infect a range of first intermediate hosts as broad as that of their range of definitive hosts. They have previously been reported from at least 18 bivalve families, the greatest number of host families for any family of digenean (Cribb et al. Reference Cribb, Bray and Littlewood2001).
The life cycles of Rhipidocotyle spp. in naturally infected marine bivalves have been reported for R. transversale Chandler, Reference Chandler1935 and R. lintoni Hopkins, Reference Hopkins1954 in Lyonsia hyalina (Conrad) (Lyionsiidae) (Stunkard Reference Stunkard1976). A study on marine Rhipidocotyle in Australia reported infections by bucephalids in Isognomonidae, Spondylidae, Tellinidae, and Ostreidae bivalves. The bivalves of the family Ostreidae (genus Saccostrea) and Tellinidae (genus Exotica) were heavily infected (Bott et al. Reference Bott, Healy and Cribb2005).
Members of the genus Saccostrea are most diverse and abundant in the banks of the west, south, and north coasts of the Inhaca Island (Paula et al. Reference Paula, Pinto, Guambe, Monteiro, Gove and Guerreiro1998; Mafambissa et al. Reference Mafambissa, Gimo, Andrade and Macia2022, Reference Mafambissa, Lindegarth and Macia2024). Additionally, according to Marcogliese (Reference Marcogliese2023), Rhipidocotyle spp. also use anodontid mussels (Anodontia: Family Lucinidae) as their first intermediate host. For our study area, the extent of a large diversity of bivalves of either group Saccostrea and Anodontia (Paula et al. Reference Paula, Pinto, Guambe, Monteiro, Gove and Guerreiro1998; Branch et al. Reference Branch, Griffiths, Branch and Beckley2000) could support the larval stages of Rhipidocotyle, enabling them to thrive.
The life cycles of Rhiphidocotyle spp. have been mostly studied in freshwater bivalves of the genera Eurynia Rafinesque and Lampsilis Rafinesque for R. papillosa Chauhan, Reference Chauhan1943 and R. septapapillata Krull, 1934 (Woodhead Reference Woodhead1929; Kniskern Reference Kniskern1952); Anodonta for cercariae of R. fennica Gibson, Taskinen & Valtonen, Reference Gibson, Taskinen and Valtonen1992 and R. campanula (Dujardin, 1845) (Gibson et al. Reference Gibson, Taskinen and Valtonen1992; Taskinen and Valtonen Reference Taskinen and Valtonen1995; Taskinen et al. Reference Taskinen, Mäkelä and Valtonen1997); and Nitia Pallary and Vulsella Röding for R. campanula (Baturo Reference Baturo1977; Fol and Abdel-Gaber Reference Fol and Abdel-Gaber2018). All freshwater bivalves infected by Rhipidocotyle belonged to Unionidae, the most species-rich bivalve family, widely distributed across Europe, Asia, North America, and Africa (Lopes-Lima et al. Reference Lopes-Lima, Froufe, Tu Do, Ghamizi, Mock, Kebapçı, Klishko, Kovitvadhi, Kovitvadhi, Paulo, Pfeiffer, Raley, Riccardi, Sereflisan, Sousa, Teixeira, Varandas, Wu, Zanatta, Zieritz and Bogan2017).
Regarding the definitive hosts, Rhipidocotyle spp. are commonly described or/and reported from teleost of Scombridae (n=10); Carangidae (n=8); Sphyraenidae (n=6); Triglidae (n=3); Sciaenidae (n=3); Psettodidae (n=2); and Hexagrammidae, Belonidae, Acropomatidae, Atherinopsidae, Lotidae, Trachinidae, Clupeidae, Serranidae, Sillaginidae, Cynoglossidae, Ariidae, Trachichthyidae, Antennariidae, Chanidae, Labridae, Gempylidae, and Paralichthyidae infected with one (1) species each (see Ahyong et al. Reference Ahyong, Boyko, Bailly, Bernot, Bieler, Brandao, Daly, De Grave, Gofas, Hernandez, Hughes, Neubauer, Paulay, Boydens, Decock, Dekeyzer, Goharimanesh, Vandepitte, Vanhoorne and Zullini2024).
However, in Africa, species of Rhipidocotyle have been described from seven (7) teleost families Carangidae (n=3), Scombridae (n=1), Trachichthyidae (n=1), Sciaenidae (n=1), Antennariidae (1), Psettodidae (n=1), and Chanidae (n=1), arranged in six orders (Nagaty Reference Nagaty1937; Reimer Reference Reimer1985; Fischthal and Thomas Reference Fischthal and Thomas1968, Reference Fischthal and Thomas1972). Only R. khalili has been reported more than twice in a very narrow range of hosts such as Sphyraena japonica Bloch & Schneider, S. obtusata, and S. putnamae (Yamaguti 1953; Madhavi Reference Madhavi1974; Reimer Reference Reimer1985; Bray and Justine Reference Bray and J-L2011). This denotes that Rhipidocotyle spp. in Africa appear to be host-specific, at least to the level of host family.
Molecular taxonomy
The genetic data generated here support the morphological distinctness of R. siphonyaka n. sp. and R. nolwe n. sp. from one another and from other bucephalid taxa, while also illustrating their relatedness. The haplotypes from both species formed well-supported clades, distinct from other taxa and supported by genetic distances while grouping as sister taxa in well-supported clades in all analyses. The presence of sister taxa in the same host species is not uncommon in bucephalids (see Bott et al. Reference Bott, Mille and Cribb2013; Corner et al. Reference Corner, Cribb and Cutmore2020). For example, the Bucephalinae species Prosorhynchoides galaktionovi Hammond, Cribb, Nolan & Bott, Reference Hammond, Cribb, Nolan and Bott2020 and Prosorhynchoides kohnae Hammond, Cribb, Nolan & Bott, Reference Hammond, Cribb, Nolan and Bott2020 were both described from the same host, Tylosurus crocodilus (Péron & Lesueur), but are morphologically and genetically distinct taxa. Similarly, Prosorhynchoides moretonensis Hammond, Cribb & Bott, Reference Hammond, Cribb and Bott2018 and Prosorhynchoides waeschenbachae Hammond, Cribb & Bott, Reference Hammond, Cribb and Bott2018 infects Tylosurus gavialoides (Castelnau), and four Aenigmatrema (Aenigmatrema grandiovum Corner, Cribb & Cutmore, Reference Corner, Cribb and Cutmore2020, Aenigmatrema inopinatum Corner, Cribb & Cutmore, Reference Corner, Cribb and Cutmore2020, Aenigmatrema undecimtentaculatum Corner, Cribb & Cutmore, Reference Corner, Cribb and Cutmore2020, and an unidentified Aenigmatrema sp.) infect S. obtusata. In all three of these cases, the taxa from the same host cluster together as sister taxa (Clade 3). Interestingly, Aenigmatrema from S. obtusata are distant from the present material, which is from a congeneric host. Corner et al. (Reference Corner, Cribb and Cutmore2020) previously proposed that bucephalids adopted sphyraenid fishes as definitive hosts on two separate occasions based on their results. Following this, the present results suggest a third such adoption.
The relation of the new species to other included taxa needs further investigation as this was not constant between the gene regions used. The topologies presented here mimic those of previous studies including similar bucephalid taxa, with Bucephalinae appearing to be polyphyletic (Nolan et al. Reference Nolan, Curran, Miller, Cutmore, Cantacessi and Cribb2015; Hammond et al. Reference Hammond, Cribb and Bott2018; Shirakashi et al. Reference Shirakashi, Waki and Ogawa2020). As such, R. siphonyaka n. sp. and R. nolwe n. sp. do not group with other Rhipidocotyle in a monophyletic clade, but rather independently. The grouping of the unidentified Prosorhynchoides sp. (TW-2019; LC498576) by Shirakashi et al. (Reference Shirakashi, Waki and Ogawa2020) was also not consistent, grouping with the study taxa (Clade 1) in the 28S rDNA topology and basal in Clade 2 based on ITS rDNA. The support for the 28S rDNA grouping was only well supported by BI analyses, with the ITS rDNA groupings less well-supported; thus, the 28S rDNA topology is likely more reliable. Unfortunately, the sequence of Prosorhynchoides sp. by Shirakashi et al. (Reference Shirakashi, Waki and Ogawa2020) was generated using metacercaria; thus, the generic identification may be incorrect. But due to the polyphyletic nature of this group, most generic designations therein may likely change in the future.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/S0022149X24000476.
Acknowledgements
The Marine Biology Research Station in Inhaca is acknowledged for providing laboratory equipment used to collect the parasites. Spectrum Analytical Facility at the Faculty of Science, University of Johannesburg is also acknowledged for providing access to infrastructure for scanning electron micrographs.
Financial support
We thank the Oppenheimer Memorial Trust for a postdoctoral research fellowship to QMDS, the NRF and UJ FRC for funding to AO, and the Ministry of Science and Technology, Higher Education Professional Training of Mozambique for funding DNA analyses.
Competing interest
None.
Ethical standards
Fish sampling was carried out as per a permit issued by the National Administration for the Conservation Areas in Mozambique (ANAC) under number 01/10/2023. The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional guides on the care and use of laboratory animals.
Author contributions
JCD: conceptualisation, sampling, morphological (LM and SEM) analysis, and writing (original draft); QMD: genetic analysis, genetic data curation, SEM analysis, review, and editing; AAO: funding acquisition, review, editing, and approval of final draft.