A New Trichodina Species (Peritrichia: Mobilida) from Anuran Tadpole Hosts, Sclerophrys spp. in the Okavango Panhandle, Botswana, with Comments on this Taxon

Acta Protozool. (2019) 58: 141–153 www.ejournals.eu/Acta-Protozoologica doi:10.4467/16890027AP.19.014.11915 ACTA PROTOZOOLOGICA A New Trichodina Species (Peritrichia: Mobilida) from Anuran Tadpole Hosts, Sclerophrys spp. in the Okavango Panhandle, Botswana, with Comments on this Taxon lsid:zoobank.org:pub:D89B6A84-603E-49BF-9D38-746496A448D0 Gerhard P. de Jager1, Linda Basson1, Jacqueline van Marwijk2 1 Department of Zoology and Entomology, University of the Free State, Bloemfontein, South Africa 2 Department of Microbial, Biochemical, and Food Biotechnology, University of the Free State, Bloemfontein, South Africa Abstract. Mobiline taxonomic data is mostly inferred from populations collected in and on hosts associated with aquaculture. Even though these conditions may be conducive for studies relating to the hosts, accurate taxonomic inferences of the symbionts will be problematic. The site for the present study was the Okavango Panhandle region in Northern Botswana, an isolated, natural area with minimal anthropogenic influences. Morphometric and molecular evidence revealed that anuran tadpole trichodinids, up to now reported as Trichodina heterodentata Duncan 1977 and T. hypsilepis Wellborn 1967 from multiple host types, are in fact a new, more host specific species. This study includes comprehensive denticle descriptions of both the anuran hosted trichodinid and the morphologically similar T. hypsilepis restricted to teleost hosts (previously T. heterodentata). Keywords: Trichodina koloti sp. nov., Sclerophrys spp., morphology, 18S ribosomal DNA, southern Africa. Address for correspondence: Linda Basson, Department of Zoology and Entomology, University of the Free State, PO Box 339, Bloemfontein, 9300, South Africa, E-mail: [email protected] INTRODUCTION The protist microcosm is one of the richest groups in regards to their morphological variation, and exceeds that of all the other eukaryotic kingdoms (Sogin and Silberman 1998). They are predominantly unicellular organisms that inhabit a heterogenous array of environments, ranging from free living to parasitic in nature. Within this complex group, the ciliated protists, with their characteristic rows of cilia for locomotory actions, are the most identifiable (Lynn 2017). Trichodinids are members of the family Trichodinidae Raabe, 1959 and together with its two sister families (Urceolariidae Dujardin, 1840 and Trichodinopsidae Kent, 1881) make up the order Mobilida Kahl, 1933 (Oligohymenophorea de Puytorac et al., 1974: Peritrichia Stein, 1859). The largest group within this family (and order) is the genus Trichodina Ehrenberg, 1838 that consists of more than 300 described species to date (Tang et al. 2013), all of them symbiotic on or in a diverse range of hosts. G. P. de Jager et al. 142 Trichodinids are usually associated with teleost fishes, especially as ectosymbionts, but have been described from the gills and skin of anuran tadpoles (Raabe 1950, Lom 1961, Chen 1963, Arthur and Lom 1984, Kazubski 1988 and Kruger et al. 1995), however, a recent wave of anuran symbiont research from South America (Dias et al. 2009, Fernandes et al. 2011, Pala et al. 2018) has illustrated that some species are more abundant than previously thought. During parasitological surveys in the Nxamasere Floodplain in the Okavango Panhandle, Botswana by the Aquatic Ecology group from the University of the Free State, it was noticed that certain tadpoles have a single trichodinid species infestation on their gills and skin. The geographically isolated Okavango River System has hitherto no known introduced aquatic fish and anuran species, implying that all collected hosts and their symbionts are indigenous to this system. The Nxamasere Floodplain located on the eastern banks of the Okavango River in the system’s panhandle in northern Botswana creates seasonal refugia for aquatic life during the dry periods of the year, temporarily trapping fish and tadpoles that in turn become an important food source for the assorted water birds that surround the plain. The Guttural Toad, Sclerophrys gutturalis (Power, 1927), is a large (140 mm in length) pale ground coloured toad with dark patches over its ventral side and a distinct vertebral line along the midline of the back (du Preez and Carruthers 2009) that has a distribution throughout East and sub-Saharan Africa. The tadpoles of S. gutturalis naturally occur in any permanent or semi-permanent pools, usually in the shallow water during the day, moving to the deeper patches at night. Sclerophrys gutturalis tadpoles usually reach metamorphoses after ten weeks, when their front legs break through and their gills are completely resorbed. The Western Olive Toad, Sclerophrys poweri (Hewitt, 1935), on the other hand, is a thickset and robust toad that reaches a maximum size of 100 mm (du Preez and Carruthers 2009) found in southern Angola, northern Namibia, Botswana and central South Africa (Channing et al. 2012). These tadpoles prefer the peripheral edges of temporary shallow pools and according to Channing et al. (2012) the whole development of S. poweri takes about ten weeks from hatching to adulthood. The present study illustrated that the Botswanan trichodinid species from African anuran tadpoles is indeed the same species described as T. heterodentata Duncan, 1977 from anuran tadpoles in other parts of the world. The study also revealed that this anuran hosted trichodinid, although superficially similar to T. hypsilepis Wellborn, 1967 (syn. T. heterodentata Duncan, 1977) from teleost hosts (de Jager and Basson 2019), differ morphologically (specifically in denticle structure) and molecularly (18S SSU rDNA). These distinctions led to the description of a new freshwater trichodinid taxon, specific to amphibian tadpole hosts. MATERIALS AND METHODS Tadpole specimens, Sclerophrys gutturalis and S. poweri were collected during the winter of 2015 and both the winter and summer seasons of 2016, from six isolated standing pools in the Nxamasere Floodplain using shrimp nets (Fig. 1). After collection, hosts were separated into aerated tanks, as per locality. Water used for the aquariums was collected from the collection localities, to keep the pH and conductivity the same as the natural environment. Hosts were euthanised according national ethics regulations. All sampling localities was recorded with a Garmin Geographical Positioning System (GPS) (Table 1). The nuclear apparatus was stained using Mayer’s haematoxylin straining method, as suggested by Wellborn (1967) and Basson et al. (1983). Silver impregnation for morphometric comparison of the aboral denticle structure was adapted from Klein’s (1926) “dry silver” impregnation technique, as described by Lom (1958), Wellborn (1967) and Basson et al. (1983). The prescribed measurements initially proposed by Lom (1958) and later adapted by van As and Basson (1989) were done for all sampled populations. Twenty five suitable adult trichodinids per collected population were measured and photographed using a Zeiss Axiophot compound microscope fitted with an AxioCam ICc 5 digital camera. All morphometric measurements are provided in µm, and represented as minimum to maximum (mean ± standard deviation). In the case of the number of denticles and number of radial pins per denticle the mode was used rather than the mean. Body diameter is determined as the adhesive disc plus the border membrane. Denticles of individual representative trichodinids from every population were re-drawn, analysed and described using the method devised by van As and Basson (1989). PCR amplification of small rRNA gene, sequencing and phylogenetic analyses Material for molecular analysis was collected in the field with a fine glass pipette, fixed in absolute ethanol and kept at 4°C for the duration of the expedition and analysed back at the laboratory on the university campus. Genomic DNA extraction and PCR amplification was performed using the REDExtract-N-Amp Tissue PCR Kit according to the manufacturer’s instructions. The 18S SSU rRNA gene region was PCR amplified, using the following primer set: ERIB 1 EukA forward primer (ACC TGG TTG ATC CTG CCA G); ERIB 10 EukA reverse primer (CTT CCG CAG GTT CAC CTA CGG) as per Medlin et al. (1988). Amplification cycling parameters were as follows; initial denaturation (94°C) for 5 minutes followed by 35 cycles each of denaturation (94°C) for 40 seconds, annealing (56°C) for one minute and extension (72°C) for 1 minute 30 seconds. Final extension occurred at 72°C for 10 minutes. PCR New Anuran Trichodinid from Botswana 143 Table 1. Longitude and latitude of collection pools in the Nxamasere Plain. Collection sites NX5 and NX6 share the same coordinates, as these are two separate pools during the drier latter months of the year, but are connected during the earlier months of the year. Standing pool (Population) LONGITUDE LATITUDE Nxamasere 1 (NX1) S 18° 35, 770’ E 22° 01, 551’ Nxamasere 2 (NX2) S 18° 36, 007’ E 22° 01, 349’ Nxamasere 3 (NX3) S 18° 35, 396’ E 22° 00, 766’ Nxamasere 4 (NX4) S 18° 35, 247’ E 22° 00, 198’ Nxamasere 5 (NX5) S 18° 34, 984’ E 22° 00, 035’ Nxamasere 6 (NX6) products were sub-cloned into the pSMART® vector (Lucigen) and sequenced using the BigDye® Terminator (v 3.1) Cycle Sequencing Kit (Applied Biosystems®) as per manufacturer’s instructions, using SR1 and SL2 primers. Data was collected using a 3130xl Genetic Analyser (Hitachi). Phylogenetic analysis The extracted sequences were imported into Geneious v.7.1.3 (Biomatters, Auckland, New Zealand), where the cloned sequences were De Novo assembled into a consensus sequence which in turn was multiple aligned, using Geneious Alignment default settings (93% similarity cost matrix) with 18S SSU rRNA mobiline sequences acquired from the GenBank/NCBI database (Clark et al. 2016) (Table 2). Members from the Urceolariidae family (U. korschelti, U. parakorschelti and U. urechi) were chosen as outgroup. Sequences from the above mentioned alignment were used for constructing phylogenetic trees, using Maximum-Likelihood (ML) (Saitou and Nei 1987) and Bayesian Inference (BI) methods. The optimal evolutionary model for both Maximum Likelihood and Bayesian Inference was the GTR+I+G model. Bayesian Inference (BI) completed using MrBayes 3.1 (Ronquist and Huelsenback 2003) with 1,000,000 generations, sampling every 100 generations. The evolutionary history was inferred by using the Maximum Likelihood method based on the General Time Reversible model (Nei and Kumar 2000), where the bootstrap test (1,000 replicates) was used for the percentage of replicate trees in which the associated taxa clustered together (Felsenstein 1985). The tree with the highest log likelihood (–11,634.29) is shown. The percentage of trees in which the associated taxa clustered together is shown next to the branches. Initial tree(s) for the heuristic search were obtained automatically by applying Neighbour-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood (MCL) approach, and then selecting the topology with superior log likelihood value. A discrete Gamma distribution was used to model evolutionary rate differences among sites (5 categories (+G, parameter = 0.1668)). The rate variation model allowed for some sites to be evolutionarily invariable ([+I], 36.34% sites). The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. The analysis involved 40 nucleotide sequences. There were a total of 2,251 positions in the final dataset. Evolutionary analyses were conducted in MEGA X (Kumar et al. 2018). Evolutionary distances of representative consensus sequences from different localities and hosts were computed using the Kimura 2-parameter method (Kimura 1980) in MEGA X (Kumar et al. 2018) and are in the units of the number of base substitutions per site (Table 3). RESULTS Type species: Trichodina koloti sp. nov. Type locality and type host: Sclerophrys gutteralis (Power, 1923), Nxamasere Floodplain, Okavango panhandle, Botswana (S 18° 35, 770’ E 22° 01, 551’) Additional locality and host: Sclerophrys powerii (Hewitt, 1926), Nxamasere Floodplain, Okavango panhandle, Botswana (S 18° 34, 984’ E 22° 00, 035’) Type-specimens: Holotype, slide 2016/12/15-01 and paratype, slides 2015/07/09-01 and 2015/07/12-09 18S nucleotide accession number: MT214940 Etymology: koloti (noun); [Tswana, a language spoken in southern Africa and the Okavango Delta region] – meaning tadpole Morphology and morphometrics Adult trichodinid specimens collected from tadpoles had a C-shaped macronucleus with a mean external diameter of 31.1–57.4 µm (47.0±9.5), thickness of 4.5–11.3 µm (7.2±2.5) and length of sector between terminations of macronucleus 3.8–38.5 µm (21.1±13.7) with no micronucleus were observed (Fig. 2). The adoral spiral followed a course of 370°–405°, which falls within the variation for the genus Trichodina. Based on the comparative morphological dimensions of all six populations, with only the minimum and maximum values given, the trichodinids had a convex body diameter ranging from 43.5–62.6 µm; adhesive disc diameter between 37.3–51.2 µm with a poorly to welldeveloped border membrane with a width of 2.7–5.7 µm (Table 5). The denticle ring diameter is between G. P. de Jager et al. 144 Table 2. List of 18S SSU rDNA mobiline sequences from GenBank (Clark et al. 2016) used for phylogenic inference. Species selected Accession nr. Author/Collector (Year) Locality Trichodina acuta KX904932 Wang et al. (2017) Freshwater fish T. bellotti MH730162 Marcotegui et al. (2018) Freshwater fish T. centrostrigeata KP295473 Wang et al. (2015) Freshwater fish T. compacta MF135183 Abdelkhalek et al. (2018) Freshwater fish T. domerguei KY596037 Irwin et al. (2017) Euryline fish T. heterodentata (syn.) AY88099 Gong et al. (2006) Freshwater fish T. hyperparasitis KX904933 Wang et al. (2017) Freshwater fish T. hypsilepis (syn.) EF524274 Gong et al. (unpublished) Freshwater amphibians T. meretricis FJ499387 Zhan et al. (2013) Marine molluscs T. modesta GU906245 Tang et al. (2013) Freshwater fish T. nobilis AY102172 Gong et al. (2006) Freshwater fish T. paraheterodentata GU906244 Tang et al. (2013) Freshwater fish T. paranigra MG198569 Wang et al. (2018) Freshwater fish T. pectenis JQ663868 Zhan et al. (2013) Marine molluscs T. pseudoheterodentata JQ821348 Tang et al. (2017) Freshwater fish T. reticulata MG198568 Wang et al. (2018) Freshwater fish T. ruditapicis FJ499385 Zhan et al. (2009) Marine molluscs T. sinipercae EF599255 Gong et al. (unpublished) Freshwater fish T. sinonovaculae FJ499386 Zhan et al. (2013) Marine molluscs T. tenuidens GU906245 Irwin et al. (2017) Euryline fish T. truttae LC186029 Mizuno et al. (unpublished) Freshwater fish T. uniforma HQ407383 Tang et al. (2013) Freshwater fish T. unionis KY596041 Irwin et al. (2017) Freshwater molluscs T. unionis MN08236 Wiroonpan and Purivirojkul (2019) Freshwater molluscs Urceolaria korschelti JQ663870 Zhan et al. (2013) Marine molluscs U. parakorschelti KP698205 Irwin and Lynn (2015) Marine molluscs U. urechi FJ499388 Zhan et al. (2009) Marine molluscs Fig. 1. Map of the Okavango River System in southern Africa, including the Nxamasere Floodplain where Sclerophrys gutturalis (Power, 1923) and S. poweri (Hewitt, 1953) were collected (redrawn and adapted from West et al. 2015) (scale = 200 km). New Anuran Trichodinid from Botswana 145 Figs 2–5. Micrographs of representative Trichodina koloti sp. nov. specimens from each of the six populations measured from the Nxamasere Floodplain; 2 – Haematoxylin stained nuclear material; 3–5 – collected from the skin and gills of Sclerophrys gutturalis (Power, 1927) (scale = 10 µm). 21.3–35.3 µm, consisting of 20–28 denticles and 8–13 radial pins per denticle. The denticle blades are strong, semi-circular with a prominent apophysis on the anterior side with a length from 4.3–7.1 µm, tapering off towards a pointing tip. The ray of the denticle is strong, generally straight and tapers off towards the tip, it has a length ranging from 4.5–7.4 µm. The central part width is from 1.2–2.7 µm and the total denticle span is between 5.9–16.0 µm. There is little to no biometric variation between trichodinids collected from the different pools and anuran hosts, also not for seasonality (Figs 3–9). G. P. de Jager et al. 146 Figs 6–9. Micrographs of representative Trichodina koloti sp. nov. specimens from each of the six populations measured from the Nxamasere Floodplain; 6–8 – collected from the skin and gills of Sclerophrys gutturalis (Power, 1927); 9 – collected from the skin and gills of S. poweri (Hewitt, 1935) tadpoles during the 2016 winter (July to August) expedition (scale = 10 µm). By analysing the denticle dimensions for the Botswanan tadpole ecto trichodinids, according to the van As and Basson (1989) method (Figs 10a–f): The blade region was the most constant with minor differences between the specimens examined. Almost all had large, broad blades filling a large part of the section between the y and y+1 axes with the tangent point being slightly more proximal than the distal blade margin. The distal blade margin (surface) was generally curved, gradually sloping towards the proximal direction and parallel to the border membrane. Most of the posterior blade margins (surface) were smoothly curved in a shallow L-shape, New Anuran Trichodinid from Botswana 147 with a few exceptions that were more deeply curved. In all the specimens the deepest point of the curve was more proximal than the apex of the blade, although some were almost on the same plain. The apexes of the blades were generally rounded, with some being slightly more pointed, and most of them extended past the y+1 axis. The anterior blade apophysis had slightly varying degrees of prominence, with the majority of specimens being somewhat prominent. No posterior projections on any specimens were observed. The central parts were slender and elongated. The distal surface was smaller, sloping more than the proximal. For the majority of the specimens the central part extended halfway past or more than halfway to the y axes. The form of the ray was mostly delicate. Almost all rays were of straight and equal width for the whole length, ending in a blunt tip. The majority of rays touched or ran parallel to the y–1 axis, in some specimens (Figs 10a, b) the rays were angled in a posterior direction, extending past the y–1. In most cases a delicate ray apophysis was present and in some specimens the ray apophyses were slightly more prominent (Figs 10b, c). The ray connections of the examined specimens were either marginally narrower than the width of the rest of the ray or of equal thickness of the ray. The ratio of the denticle above and below the x axis fell in a range between 0.5 and a maximum of 0.74, with most of the ranges clustering around 0.54. Remarks. All tadpole trichodinid populations presented by Arthur and Lom (1984), Kruger et al. (1993a), Dias et al. (2009), Pala et al. (2018) and the populations from the current study show similarities in the biometrical data (Tables 4 and 5) and denticle shape (Figs 10 and 11). All of the above-mentioned populations represent the same species and based on the comparative similarities the denticle plan of this typical tadpole trichodinid is as follows: Rays are always delicate and of equal thickness throughout the length of the ray. Ray connections are thin, of the same thickness or slightly narrower than the rest of the ray. Rays all terminate in rounded points. The y axes’ relationship to the rays vary from touching the complete proximal side of the ray (Figs 11c, d, f) to running parallel, but not touching the ray (Figs 10 c, e, f; 11a, b, e). Variation noted in some individuals of the Botswana population includes rays angled posteriorly where some cross the y axes (Figs 10a,b) and angled anteriorly where some touch the y axes (Fig. 10e). However in all cases, the ray connection and base of the ray never cross the y axes posteriorly. The central parts of all these trichodinid populations are narrow, elongated and of equal width throughout. The base of the central part is of the same width as the central part for all populations. Posterior projections are observed in the population described by Pala et al. (2018) (Figs 11e, f), however, the above mentioned characteristics for the central part still apply. The posterior termination of the central part is distinctly rounded. In the majority of cases, bar the population of Kruger et al. (1993a) (Fig. 11c) the central part leans in a proximal direction. In contrast, the typical denticle plan for T. hypsilepis from fish hosts is as follows: Rays are always robust and never of equal thickness throughout the length of the ray (Figs 12a–f). Ray connections vary from broad, narrowing after the connection (Figs 12a, c, f) and ray connection is well developed, but still narrower than the base of the ray (Figs 12b, e). Both of these characteristics can sometimes be observed in the same individual, but in different denticles (Fig. 12d). Rays all narrow perceptibly along their length ending in a narrow/sharp point. The y axes’ relationship to the rays vary from touching the complete proximal side of the ray (Figs 12b, c, f), running parallel, but not touching the ray (Figs 12a, d); to running though the midsection of the ray (Fig. 12e). Again, these characteristics can vary in different denticles of the same individual (Figs 12d, e, f). The central parts of all T. hypsilepis populations are squat, varying in shape. The base of the central part is always significantly wider than the central part proper. The central part in the majority of cases is triangular. In the majority of cases, bar the population of Pala et al. (2018) (Fig. 12e) the central part exhibits no slant. Molecular phylogeny Because phylogenetic inferences from ML and BI analysis were very similar, an amalgamated tree was constructed based on the ML tree (Fig. 13). The consensus sequence from the current study illustrates that T. koloti sp. nov. from Botswana clustered with strong support (95%) in the ML tree and even stronger support (100%) in the BI tree with the T. hypsilepis sequence deposited by Gong et al. (unpublished), while both these sequences form a fully supported clade (100% for both trees) with T. bellotti, as deposited by Marcotegui et al. (2018). The statistical p-values (genetic distances) G. P. de Jager et al. 148 Table 3. P-values for 18SSU sequences for representative cloned inserts from trichodinids collected from different localities and hosts in the Nxamasere Floodplain, modelled for Trichodina heterodentata Duncan, 1977 and T. hypsilepis Wellborn, 1967 sequences obtained from the NCBI database (p > 0.05 is significantly different). (*T. heterodentata as synonym toT. hypsilepis; **T. hypsilepis as synonym to T. koloti) Tadpole host Population T. heterodentata* T. hypsilepis** Sclerophrys gutturalis NX4 0.091 0.002 S. gutturalis NX5 0.093 0.004 S. gutturalis NX 6 0.091 0.002 S. powerii NX 6 0.090 0.001 Table 4. Biometrical data (in µm) of all published tadpole host populations of Trichodina hypsilepis Wellborn, 1967 (*syn. T. heterodentata Duncan, 1977), Trichodina koloti sp. nov. (**syn. T. hypsilepis Wellborn, 1967) and T. koloti (ADD – Adhesive disc diameter, BD – Body diameter, BMW – Border membrane width, CL – Collection locality, DBL – Denticle blade length, DCPW – Denticle central part width, DL – Denticle length, DRD – Denticle ring diameter, DRL – Denticle ray length, DS – Denticle span, HS – Host species, LoH – Location on host, n – population size, nD – Number of denticles, nRP/D – Number of radial pins per denticle) (*pertains to the mode, rather than the mean). Trichodina hypsilepis Wellborn, 1967* Trichodina koloti sp. nov.** Trichodina koloti sp. nov. Kruger et al. (1993a) Dias et al. (2009) Pala et al. (2018) Arthur and Lom (1984) Current study CL South Africa Brazil Brazil Cuba Botswana (Nxamasere 4) LoH Skin and Gills Skin Skin Skin and Gills HS Xenopus laevis laevis Rhinella pombali R. schneideri Unknown tadpole Sclerophrys gutturalis BD – (55.1) 38.6–59.3 (54.5±6.2) 55.1–85.7 (67.4 ± 6.9) 47.0–57.4 (53.3 ± 2.6) ADD 41.1–64.3 (49.3±5.6) 38.9–60.0 (49.8±5.3) 39.5–54.1 (48.0±5.4) 39.8–56.1 (47.6 ± 3.9) 40.4–48.2 (44.7 ± 2.2) DRD 24.4–41.2 (31.4±3.8) 24.6–33.2 (29.2±3.2) 25.0–32.3 (30.3±2.5) 25.5–34.2 (29.4 ± 2.3) 24.5–32.4 (28.0 ± 1.8) DL 7.4–9.9 (8.4±0.8) 5.4–8.7 (6.9±0.7) 5.9–8.6 (6.5±1.1) 10.2–14.8 (12.4 ± 1.1) 5.7–7.8 (6.7 ± 0.5) DBL 3.9–6.2 (5.1±0.6) 3.0–5.2 (4.2±0.4) 4.0–6.2 (4.4±0.7) 4.5–5.6 (5.2 ± 0.8) 5.0–6.6 (5.7 ± 0.4) DRL 5.9–9.2 (7.2±0.8) 4.9–8.8 (6.7±0.7) 1.3–8.6 (6.2±0.9) 6.6–11.2 (8.2 ± 0.8) 5.3–8.0 (6.7 ± 0.6) DCPW 1.8–3.3 (2.6±0.4) 2.03–3.72 (2.7±0.3) 2.0–3.1 (2.1±0.3) 2–3.1 (2.5 ± 0.4) 1.4–2.3 (1.9 ± 0.2) DS 11.6–18.7 (14.9) 10.4–14.6 (12.5) 10.2–13.3 (12.0±0.8) 14.3–19.4 (15.9 ± 1.0) 12.0–14.8 (13.3 ± 0.7) BMW 3.5–6.1 (4.7) 2.9–5.1 (4.2±0.4) 3.2–7.4 (4.3 ±0.7) 4.1–6.1 (4.9 ± 0.5) 3.1–5.7 (4.5 ± 0.7) nD 21–25 (24*) 19–24 (22*) 19–23 (22*) 20–23 (21.0 ± 0.8) 20–24 (22* ± 1) nRP/D 7–13 (11*) 7–10 (7*) 7–11 (8*) 9–12 (commonly 10–11) 10–12 (11* ±1) n 35 50 60 25 25 calculated for the number of base pair differences per site on the above mentioned sequences falls between 0.001 and 0.004 for T. koloti, while substantially larger when compared with T. hypsilepis (syn. T. heterodentata) from fish hosts (Table 4). New Anuran Trichodinid from Botswana 149 Figs 10a–f. Denticle dimensions, as by van As and Basson (1989, 1992), of Trichodina koloti sp. nov. representatives from six different populations collected at the Nxamasere Floodplain, where a–e from Sclerophrys gutturalis (Power, 1927) and f – S. poweri (Hewitt, 1935) tadpoles during the 2016 winter (July to August) expedition (scale = 10 µm). Figs 11a–f. Denticle dimensions, as proposed by van As and Basson (1989, 1992), of Trichodina koloti sp. nov. as recorded by and redrawn from: a and b – Arthur and Lom (1984), c – Kruger et al. (1993a), d – Dias et al. (2009), e and f – Pala et al. (2018). Figs 12a–f. Denticle dimensions, as proposed by van As and Basson (1989; 1992), of Trichodina hypsilepis Wellborn, 1967 (syn. T. heterodentata) as recorded by and redrawn from: a – Population A of Duncan (1977), b – Population B of Duncan (1977), c – from van As and Basson (1989), d – from Tang and Zhao (2007), e – from Pádua et al. (2012), f – from Valladão et al. (2013). DISCUSSION Arthur and Lom (1984) described T. hypsilepis from unknown tadpoles in Cuba, comparing it to the description of the same species by Wellborn (1967) from a freshwater fish, the highscale shiner (Notropis hypsilepis Suttkus and Raney, 1955), even though these authors commented on similarities with T. heterodentata in the same article. When Duncan (1977) described T. heterodentata from cichlid hosts, he either did not know about the publication of Wellborn (1967), an unlikely possibility, or decided that because of the large difference in size (both body and adhesive disc diameter) between his populations and those of Wellborn (1967), T. heterodentata is a separate species, even though denticle morphology appears strikingly similar. The similarities of three of Wellborn’s (1967) species, T. hypsilepis being one of them, and Duncan’s (1977) T. heterodentata have been discussed and it was proposed that all teleost hosted T. heterodentata-like species become a synonym to T. hypsilepis (de Jager and Basson 2020). However, T. heterodentata-like mobilines have also been described and recorded from anuran tadpoles in the Southern Hemisphere and originally described as T. heterodentata (Kruger et al. 1993a, 1993b, 1995;