NOVELTY AND PHYLOGENETIC RELATIONSHIPS WITHIN THE SERENDIPEIDAE (CESTODA: “TETRAPHYLLIDEA”)

The cestode family Serendipeidae Brooks and Evenhuis, 1995 remains 1 of the more intriguing clades of “Tetraphyllidea.” All 9 described species are known only from cownose rays in the genus Rhinoptera Cuvier. The group is morphologically cohesive. Similarities in proglottid anatomy among members of its 3 genera (i.e., DuplicibothriumWilliams and Campbell, 1978; Glyphobothrium Williams and Campbell, 1977, and Serendip Brooks and Barriga, 1995) are particularly striking. Recent work by Stephan and Caira (2022) revealed the existence of a considerable amount of undescribed novelty in species of cownose rays not previously examined for this family of cestodes. These authors also observed a curious scolex morphology in 1 of their new species of Duplicibothrium that raises questions about the reciprocal monophyly of the latter genus and Serendip.

The current study builds on the work of Stephan and Caira (2022) to further our understanding of the diversity and phylogenetic relationships of the Serendipeidae. It has 3 main goals. The first is to formally describe 3 of the 6 species, previously known solely by provisional names, for which sufficient material is now available for further study (i.e., Duplicibothrium n. sp. 2 of Jensen and Bullard [2010], and Duplicibothrium n. sp. 4 and Duplicibothrium n. sp. 5 of Stephan and Caira [2022]). The second is to revise the generic classification of the family to align with the morphological features and phylogenetic relationships of its members. The result was the erection of a new genus to house a morphologically cohesive and molecularly divergent subset of the species formally assigned to Duplicibothrium. The third is to explore the question about the reciprocal monophyly of Serendip and Duplicibothrium raised by Stephan and Caira (2022) based on a phylogenetic analysis that includes sequence data generated for the D1–D3 region of the 28S rDNA gene for a member of Serendip for the first time.

Each host specimen was assigned a unique collection code and number, and photographs and measurements were taken. More detailed information for these specimens can be accessed in the Global Cestode Database (Caira et al., 2022) by unique collection code and number (e.g., CH-40). Cestodes examined in this study were obtained from 2 specimens of Rhinoptera bonasus (CH-40, CH-42) collected off Charleston, South Carolina, in June 2015, and 3 specimens of Rhinoptera jayakari (MZ-1, MZ-3, MZ-4) collected off Tofo, Inhambane, Mozambique in June 2016. In all instances, the body cavity was opened with a midventral incision, and a small sample of liver tissue was taken and preserved in 95% ethanol for molecular verification of host identity. The spiral intestine was then removed, opened with a longitudinal incision, rinsed in seawater, and examined for cestodes. A subset of the cestodes found was fixed in 95% ethanol for molecular sequencing; the remaining specimens were fixed in 10% seawater-buffered formalin (9:1) for examination with light and scanning electron microscopy (SEM). The spiral intestine of each ray was then fixed in either 95% ethanol or seawater-buffered formalin. After approximately 1 wk, cestodes and spiral intestines fixed in seawater-buffered formalin were transferred to 70% ethanol for storage. Cestodes and spiral intestines fixed in 95% ethanol were transferred to new 95% ethanol and stored in a −20 C freezer. Spiral intestines were examined under an Olympus SZ-30 dissecting microscope (Olympus, Center Valley, Pennsylvania), and any additional cestodes were removed and transferred to either 70% or 95% ethanol.

Cestodes prepared for light microscopy were hydrated in a graded ethanol series, stained for 20–60 min in a working solution of Delafield's hematoxylin (1:9 mixture of hematoxylin: distilled water), differentiated in tap water, destained in acidic 70% ethanol, neutralized in basic 70% ethanol, dehydrated in a graded ethanol series, cleared in methyl salicylate, mounted in Canada balsam on glass slides under glass coverslips, and left to dry in an oven set to 55 C for 1 wk. Measurements were taken with a Zeiss Axioskop 2 Plus compound microscope (Zeiss, Thornwood, New York) using a SPOT Diagnostic Instrument Digital Camera System and SPOT software (version 4.6; SPOT Imaging Solutions, Sterling Heights, Michigan). Measurements are given in the text as the range (in micrometers unless stated otherwise). In instances in which measurements were taken from 5 or more specimens, the range is followed in parentheses by the mean, standard deviation, number of specimens measured, and total number of measurements in instances in which more than 1 measurement was made per worm. In all other cases, the range is followed in parentheses by the number of specimens measured. Shape terminology follows Clopton (2004).

Scoleces prepared for SEM were hydrated in a graded ethanol series, transferred to a 1% solution of osmium tetroxide overnight, dehydrated in a graded ethanol series, placed in hexamethyldisilazane in a fume hood for 30 min, and then allowed to air dry. The specimens were then mounted on double-sided PELCO carbon tabs (Ted Pella Inc., Redding, California) on aluminum stubs, sputter-coated with 45 nm of gold/palladium, and examined with an FEI Nova NanoSEM 450 field emission SEM (FEI, Hillsboro, Oregon) at the Bioscience Electron Microscopy Laboratory, University of Connecticut (Storrs, Connecticut). Microthrix terminology follows Chervy (2009).

Museum abbreviations used are as follows: LRP, Lawrence R. Penner Parasitology Collection, University of Connecticut, Storrs, Connecticut; NMB-P, National Museum, Bloemfontein, South Africa, Parasite Collection; USNM, National Museum of Natural History, Smithsonian Institution, Washington, D.C.

Sequence data were generated for the D1–D3 region of the 28S rDNA gene for 2 adult specimens of an undescribed species of Serendip (henceforth referred to as Serendip n. sp. 1) from Aetomylaeus asperrimus (Gilbert) collected off Mariato, Veraguas, Panama. The middle portion of each cestode specimen was removed and allowed to air dry for approximately 5 min at room temperature before extraction. The scoleces were prepared as whole mounts (as described above) to serve as vouchers for the specimens sequenced. These hologenophores (sensu Pleijel et al., 2008) were deposited in the Lawrence R. Penner Parasitology Collection (LRP 9827 and LRP 9828). Extraction, amplification, and Sanger sequencing of DNA followed Bueno and Caira (2017). The primer pairs used for amplification were LSU-5 (5′-TAGGTCGACCCGCTGAAYTTA-3′; Littlewood et al., 2000) and LSU-1500R (5′-GCTATCCTGGAGGGAAACTTCG-3′; Tkach et al., 2003). The primer pairs used for sequencing were LSU-55F (5′-AACCAGGATTCCCCTAGTAACGGC-3′) (Bueno and Caira, 2017) and LSU-1200R (5′-GCATAGTTCACCATCTTTCGG-3′; Littlewood et al., 2000).

The 2 newly generated sequences of the undescribed species of Serendip were combined with sequence data for the D1–D3 region of the 28S rDNA gene from GenBank for all 43 specimens of the Serendipeidae comprising the ingroup and both species of Caulobothrium used as members of the outgroup in the phylogenetic analysis of Stephan and Caira (2022). GenBank and LRP voucher numbers for all specimens are provided on the phylogenetic tree following the host, life-cycle stage, and unique host specimen number. When possible, a unique cestode specimen code is also provided.

Sequences were aligned and trimmed using the MAFFT (Katoh and Standley, 2013) multiple alignment plug-in in Geneious Prime 2022.0.1® (www.geneious.com). The aligned data matrix was 1,235 base pairs (bp) in length. A maximum-likelihood phylogenetic analysis was performed on the cluster in the Bioinformatics facility of the Institute of Systems Genomics at the University of Connecticut using IQ-TREE 1.6.10 (Nguyen et al., 2015). GTR+I+G with empirical base frequencies was selected as the best-ranked model of molecular evolution according to the corrected Akaike information criterion (AICc) as implemented in ModelFinder (Kalyaanamoorthy et al., 2017) in IQ-TREE and followed by tree reconstruction and 200 nonparametric bootstrap replicates with the command iqtree -s datamatrix.nex -m MFP -merit AICc -b 200.

Nanoduplicibothrium n. gen.

Worms protandrous, euapolytic, tiny; proglottids weakly craspedote. Scolex with 4 bothridia, each with apical sucker; dorsal and ventral bothridia fused lengthwise in 2 pairs. Bothridial surfaces with continuous band of loculi extending throughout bothridial margins and single central column of loculi stopping short of locular band; distinct row of posterior loculi and small, circular loculi flanking apical sucker lacking. Cephalic peduncle absent. Surfaces of bothridia with papilliform filitriches; spinitriches lacking. Testes extending into postovarian field, in 2 irregular dorso-ventral fields. Vas deferens not observed. Genital pores in anterior third of proglottid, submarginal, irregularly alternating. Vagina opening into genital atrium anterior to cirrus sac. Ovary digitiform, radiating from central isthmus. Uterus ventral, sacciform, extending to cirrus sac. Vitellarium follicular; dorsal vitelline follicles in single extensive field, interrupted or not by cirrus sac and ovary; ventral vitelline follicles lateral. Parasites of cownose stingrays (Rhinopteridae); cosmopolitan in distribution.

Nanoduplicibothrium leanneae n. gen. n. sp.

Nanoduplicibothrium jillae (Stephan and Caira, 2022) n. comb. (note that USNM number 1009862 for 1 of the paratypes is incorrect, the correct number is USNM 1660862), Nanoduplicibothrium megaphallum n. sp., Nanoduplicibothrium paulum (Ruhnke, Curran, and Holbert, 2000) n. comb.

Nanoduplicibothrium n. sp. 3 (formerly Duplicibothrium n. sp. 3 of Stephan and Caira, 2022).

urn:lsid:zoobank.org:act:38D8932A-0711-4026-B64E-EC8695CC7CD3

The name of this genus reflects the small (nano; L.) size of its members relative to species of Duplicibothrium.

Nanoduplicibothrium differs from Duplicibothrium in that, as the name suggests, members of this genus are relatively tiny worms; none of the 4 species exceed 3 mm in total length, whereas all 5 species of Duplicibothrium reach a total length of 4.6–29.2 mm. Furthermore, its species lack, rather than possess, a distinct row of posterior loculi on each of their bothridia and a cephalic peduncle. We also believe the interpretation of the bothridia of this species as possessing 5 posterior loculi is incorrect; these loculi are more appropriately assigned to the continuous band of loculi extending throughout the bothridial margins. Nanoduplicibothrium n. gen. conspicuously differs from Serendip in that the bothridia of its members each bear transverse and longitudinal septa, rather than radially diverging septa, marginal loculi, and a marginal velum. The new genus conspicuously differs from the monotypic Glyphobothrium in that its scolex consists of dorsal and ventral bothridia fused lengthwise in 2 pairs with essentially no scolex proper, rather than being globular with 4 bothridia fused to the outer surface of a sizeable scolex proper as illustrated by the longitudinal section of a scolex presented by Caira et al. (1999: fig. 59). Our attention was first drawn to the potential novelty of this genus when we mapped morphology onto the phylogenetic tree in Stephan and Caira (2022: fig. 1) and noticed that the distinguishing features outlined above were consistent with the 2 major subclades that emerged in the tree resulting from their phylogenetic analysis.

Nanoduplicibothrium leanneae n. gen. n. sp.
(Figs. 1, 2A–E)

Worms weakly craspedote, euapolytic, 630–884 (741 ± 97; 5) long; maximum width at level of scolex. Proglottids 3–4 (3 ± 1; 5) in total number. Scolex consisting of 4 bothridia arranged in 2 dorso-ventral fused pairs (Figs. 1A, 2A). Cephalic peduncle absent. Bothridia each with oval apical sucker, 198–217 (n = 3) long, 76–100 (n = 3) wide, free anteriorly, sessile posteriorly, with at least 84 loculi arranged as continuous band extending throughout bothridial margins and single central column stopping short of locular band. Distal (Fig. 2B) and proximal (Fig. 2C) bothridial surfaces covered with papilliform filitriches; strobila immediately behind scolex (Fig. 2E) and more posteriorly (Fig. 2D) covered with densely arranged capilliform filitriches.

Immature proglottids wider than long, becoming longer than wide with maturity (Fig. 1C). Mature proglottids 1 (n = 5) in number; terminal mature proglottid (Fig. 1B) 211–377 (291 ± 59; 5) long, 121–186 (161 ± 27; 5) wide, length:width ratio 1.58– 2.05 (1.8 ± 0.18; 5):1. Testes 16–21 (n = 3) in number, arranged in 2 irregular columns extending throughout proglottid length, 2 irregular rows deep, oblong, 21–40 (n = 4) long, 29–45 (38 ± 4; 4, 15) wide. Vas deferens not observed. Cirrus sac ovoid (Fig. 1D), 50–106 (n = 2) long, 33–45 (n = 4) wide; containing coiled cirrus. Cirrus weakly developed in terminal proglottids, armed with spinitriches. Genital pores slightly submarginal, 67–76% (n = 4) of proglottid length from posterior end of proglottid, irregularly alternating. Vagina extending from ovarian bridge along midline of proglottid to anterior margin of cirrus sac then along anterior margin to open into genital atrium anterior to cirrus. Ovary terminal in position, highly digitiform, 64–99 (n = 4) long, 88–112 (n = 4) wide. Vitellarium follicular; dorsal vitelline follicles arranged in single extensive field, interrupted by terminal genitalia and ovary; ventral vitelline follicles arranged in 2 lateral bands interrupted by terminal genitalia. Uterus median, ventral, sacciform, extending from ovarian bridge to level of cirrus sac. Excretory ducts 4 in number, arranged in 1 dorsal and 1 ventral pair. Gravid proglottids not observed.

American cownose ray, R. bonasus Mitchill (Myliobatiformes; Rhinopteridae).

Off Charleston (32°42′18.08″N, 79°53′18.77″W) South Carolina, Atlantic Ocean.

Gulf of Mexico (30°14′16.90″N, 89°13′35.13″W), Mississippi (based on larval stage data).

Spiral intestine.

Holotype (USNM 1678883) and 2 paratypes (USNM 1678884, 1678885); 3 paratypes (LRP 11015, 11016, 11019) and SEM vouchers (LRP 11017, 11018); scoleces prepared for SEM retained with JNC at the University of Connecticut.

urn:lsid:zoobank.org:act:B842F328-A15E-42C3-8929-44D497C070E9.

The name of this tiny worm honors Leanne Kennedy Harty in recognition of her enthusiastic assistance with multiple aspects of laboratory support over the past several years.

Nanoduplicibothrium leanneae n. gen. n. sp. differs from both N. paulum and N. jillae in terms of the number of total loculi per bothridia (84 vs. 57–63 and 59, respectively). This new species further differs from N. paulum in that it is a smaller worm (630–884 vs. 700–2,900) and it possesses a shorter terminal proglottid (211–295 vs. 350–950). Nanoduplicibothrium leanneae n. gen. n. sp. further differs from N. jillae in its possession of a narrower terminal proglottid (81–87 vs. 112–126). The description of N. leanneae n. gen. n. sp. provides a formal name for the species provisionally referred to as Duplicibothrium n. sp. 2 by Jensen and Bullard (2010) and Stephan and Caira (2022). A large number of loculi and the tendency for the bothridia to fold makes the exact number of loculi in N. leanneae n. gen. n. sp. difficult to determine.

Nanoduplicibothrium megaphallum n. sp.
(Figs. 2F–I, 3)

Worms weakly craspedote, euapolytic, 1,880–3,000 (2,350 ± 390; 9) long; maximum width at level of scolex. Proglottids 5–7 (6 ± 1; 9) in total number. Scolex consisting of 4 bothridia arranged in 2 dorso-ventral fused pairs (Figs. 2F, 3A). Cephalic peduncle absent. Bothridia each with oval apical sucker, 411 (n = 1) long, 182 (n = 1) wide, free anteriorly, sessile posteriorly, with at least 66 loculi; loculi arranged as continuous band extending throughout bothridial margins and single central column stopping short of locular band (Fig. 3B). Distal (Fig. 2G) and proximal (Fig. 2H) bothridial surfaces covered with papilliform filitriches; strobila covered with densely arranged capilliform filitriches (Fig. 2I).

Immature proglottids wider than long, becoming longer than wide with maturity (Fig. 3E). Mature proglottids (Fig. 3C) 1 (n = 9) in number; terminal mature proglottid (Fig. 3D) 763–1,088 (960 ± 115; 9) long, 275–375 (319 ± 43; 9) wide, length:width ratio 2.67– 3.61 (3.12 ± 0.31; 9):1. Testes 24–28 (n = 4) in number, arranged in 2 irregular columns extending throughout proglottid length, 2 irregular rows deep, oblong, 22–40 (n = 4) long, 25–37 (n = 4) wide. Vas deferens not observed. Cirrus sac enormous, ovoid, conspicuously tilted posteriorly, 277–419 (325 ± 44; 8) long, 119–174 (143 ± 18; 8) wide; containing coiled cirrus. Cirrus armed with spinitriches. Genital pores slightly submarginal, 67–76% (72 ± 3; 9) of proglottid length from posterior end of proglottid, irregularly alternating. Vagina extending from ovarian bridge along midline of proglottid to anterior margin of cirrus sac then along anterior margin to open into genital atrium anterior to cirrus. Ovary terminal in position, highly digitiform, 201–308 (253 ± 38; 8) long, 170–229 (197 ± 22; 8) wide. Vitellarium follicular; dorsal vitelline follicles arranged in single extensive field, interrupted by cirrus sac, distal portion of vagina and ovary; ventral vitelline follicles arranged in 2 lateral bands interrupted by terminal genitalia. Uterus median, ventral, sacciform, extending from ovarian bridge to level of cirrus sac. Excretory ducts 4 in number, arranged in 1 dorsal and 1 ventral pair. Gravid proglottids not observed.

Oman cownose ray, R. jayakari Boulenger (Myliobatiformes; Rhinopteridae).

Off Tofo, (23°47′33.02″S, 35°31'16.38"E), Mozambique, Indian Ocean.

None.

Spiral intestine.

Holotype (NMB-P 925) and 2 paratypes (NMB-P 926, NMB-P 927); 3 paratypes (LRP 11022–11024) and 2 SEM vouchers (LRP 11020, 11021); 3 paratypes (USNM 1678886–1678888); scoleces prepared for SEM retained with JNC at the University of Connecticut.

urn:lsid:zoobank.org:act:C83F6419-1749-4AB8-A89E-FEFC7DB79708.

This species is named for the conspicuously large size (mega; L.) of its cirrus (phallum; L.) compared to its congeners.

Nanoduplicibothrium megaphallum n. sp. differs from N. paulum, N. jillae, and N. leanneae in its possession of a cirrus sac that is conspicuously longer (277–419 vs. 64–144, 33–36, and 50–68, respectively) and wider (119–174 vs. 36–110, 28–37, and 39–40, respectively) as well as in its total number of loculi per bothridium (66 vs. 57–63, 59, and 84, respectively). This new species further differs from N. jillae and N. leanneae in that it is a larger worm (1,880–3,000 vs. 700–1,000 and 630–884, respectively) and has a greater number of proglottids (5–7 vs. 2–4 and 3–4, respectively). The description of this species provides a formal name for the species provisionally referred to as Duplicibothrium n. sp. 4 by Stephan and Caira (2022).

Worms protandrous, euapolytic, moderate in size; proglottids weakly craspedote. Scolex with 4 bothridia, each with apical sucker; dorsal and ventral bothridia fused lengthwise in 2 pairs. Bothridial surfaces divided into loculi by transverse septa, longitudinal septa, or a combination of both types of septa; distinct posterior row of 5 or 7 loculi present, occasionally with 4 small loculi flanking apical sucker on each side. Cephalic peduncle present, scutellate. Surfaces of bothridia with papilliform filitriches; spinitriches lacking. Testes extending into postovarian field, in 2 irregular dorso-ventral fields. Genital pores in anterior third of proglottid, submarginal, irregularly alternating. Vagina opening into genital atrium anterior to cirrus sac. Ovary digitiform, radiating from central isthmus. Uterus ventral, sacciform, extending to cirrus sac. Vitellarium follicular; dorsal vitelline follicles in single extensive field, interrupted or not by cirrus sac and ovary; ventral vitelline follicles arranged in 2 lateral bands, encroaching on midline of proglottid or not. Parasites of cownose stingrays (Rhinopteridae); cosmopolitan in distribution.

Duplicibothrium minutumWilliams and Campbell, 1978 

Duplicibothrium bilai n. sp., Duplicibothrium cairae Ruhnke, Curran, and Holbert, 2000; Duplicibothrium colossumStephan and Caira, 2022 and Duplicibothrium jeannettae,Stephan and Caira, 2022.

Duplicibothrium n. sp. 1 (of Jensen and Bullard, 2010) and Duplicibothrium n. sp. 6 (of Stephan and Caira, 2022).

The most recent diagnosis of Duplicibothrium (see Stephan and Caira, 2022) is emended to reflect the transfer of the species originally assigned to this genus that lack a distinct row of posterior loculi and a cephalic peduncle to Nanoduplicibothrium n. gen. This emendation also reflects 3 modifications in the interpretation of the morphology of Duplicibothrium minutum made following examination of paratype specimens deposited in LRP (LRP 3551, 3610–3612) and images of the holotype (USNM 1370283). First, the bothridia of D. minutum, like those of its congeners, bear, rather than lack, an apical sucker on their anterior margin (Fig. 4A). In addition, the vitelline follicles in the proglottid of the holotype are much larger than illustrated by Williams and Campbell (1978: fig. 3) and, rather than being circum-medullary, do not extend into the ventral portions of the proglottid (Fig. 4B). Major modifications of the diagnosis are indicated in bold. The new species of Duplicibothrium described below is distinguished only from the 4 described species now considered to belong to that genus.

Duplicibothrium bilain. sp.
(Figs. 5, 6)

Worms weakly craspedote, euapolytic, 4,033–5,940 (4,730 ± 680; 8) long; maximum width at level of scolex. Proglottids 11–18 (14 ± 3; 7) in total number. Scolex consisting of 4 bothridia arranged in 2 dorso-ventral fused pairs (Figs. 5A, 6A) and elongate cephalic peduncle (Fig. 5C). Bothridia each with apical sucker, pyriform, 287–384 (327 ± 32; 7, 13) long, 174–260 (206 ± 26; 8, 14) wide, free anteriorly, sessile posteriorly, with 37 loculi; loculi arranged as single apical loculus followed by 2 lateral columns of 10 loculi, 1 medial column of 9 loculi, and 1 row of 7 posteriormost elongate loculi. Cephalic peduncle 1,407–2,127 (1,732 ± 228; 8) long, 118–148 (130 ± 10; 8) wide. Distal (Fig. 6B) and proximal (Fig. 6C) bothridial surfaces covered with papilliform filitriches; cephalic peduncle scutellate; scutes consisting of densely arranged capilliform filitriches (Fig. 6D).

Immature proglottids wider than long, becoming longer than wide with maturity (Fig. 5C). Mature proglottids 1–2 (1 ± 0.4; 7) in number; terminal mature proglottid (Fig. 5B) 730–1,138 (848 ± 151; 6) long, 373–490 (429 ± 42; 6) wide, length:width ratio 1.7–2.8 (2 ± 0.4; 6):1. Testes 38–51 (45 ± 5; 6) in number, arranged in 2 irregular columns extending throughout proglottid length, including dorsal to ovary, 2 irregular rows deep, oblong, 33–65 (44 ± 8; 5, 20) long, 50–93 (72 ± 11; 5, 20) wide. Vas deferens minimal, coiled medial and posterior to cirrus sac. Cirrus sac inconspicuous, pyriform (Fig. 5D), 102–164 (n = 4) long, 42–93 (n = 4) wide; containing coiled cirrus. Cirrus armed with spinitriches. Genital pores slightly submarginal, 78–85% (82% ± 2; 5) of proglottid length from posterior end of proglottid, irregularly alternating. Vagina extending from ovarian bridge along midline of proglottid to anterior margin of cirrus sac then along anterior margin to open into common genital atrium anterior to cirrus. Ovary terminal in position, highly digitiform, 215–384 (281 ± 67; 5) long, 279–335 (306 ± 24; 5) wide. Vitellarium follicular; dorsal vitelline follicles arranged in single extensive field, partially interrupted by cirrus sac and ovary; ventral vitelline follicles arranged in 2 lateral bands. Uterus median, ventral, sacciform, extending from ovarian bridge to level of cirrus sac. Excretory ducts 4 in number, arranged in 1 dorsal and 1 ventral pair. Gravid proglottids not observed.

Oman cownose ray, R. jayakari Boulenger (Myliobatiformes; Rhinopteridae).

Off Tofo, (23°47′33.02″S, 35°31'16.38"E), Mozambique, Indian Ocean.

None.

Spiral intestine.

Holotype (NMB-P 922) and 2 paratypes (NMB-P 923, NMB-P 924); 3 paratypes (LRP 11011–11013) and 1 SEM voucher (LRP 11014); 3 paratypes (USNM 1678889–1678891); scolex prepared for SEM retained with JNC at the University of Connecticut.

urn:lsid:zoobank.org:act:834414EA-FED6-44B7-B05B-BEDD22382852.

This species honors the intrepid Dr. Sam Bila of the University of Maputo in Mozambique, without whose valuable assistance the collection of the type host of this species would not have been possible.

Duplicibothrium bilai n. sp. differs from D. cairae, D. minutum, and D. jeannettae in that it possesses a wider terminal proglottid (413–460 vs. 190–336, 128–240, and 186–316, respectively). Duplicibothrium bilai n. sp. further differs from D. minutum in that its bothridia bear both longitudinal and transverse septa rather than only transverse septa as well as in its possession of a greater number of testes (38–48 vs. 26–32). This new species further differs from D. jeannettae in that it possesses a longer cephalic peduncle (1,407–2,127 vs. 774–1,920), a posteriormost row of 7 rather than 5 loculi, and a longer terminal proglottid (735–1,138 vs. 514–656). This new species differs from D. colossum in that it is a smaller worm (4.3–5.9 vs. 10–29 mm) and possesses fewer proglottids (11–22 vs. 85–139). Duplicibothrium bilai n. sp. most closely resembles D. cairae; however, it differs in possessing fewer proglottids (11–22 vs. 20–35) and a greater number of total loculi per bothridia (37 vs. 27–33). With the description of D. bilai n. sp. we are assigning a formal name to the specimens provisionally identified as Duplicibothrium n. sp. 5 in the molecular phylogeny of Stephan and Caira (2022).

The discovery of the undescribed species of Serendip n. sp. 1 (Fig. 4C) in the rough eagleray, Aetomylaeus asperrimus, off the Pacific coast of Panama has advanced our understanding of Serendip. Beyond expanding the known host associations of this genus to include a species in the family Myliobatidae, this material provided an opportunity for Serendip to be included in a molecular phylogenetic analysis for the first time. In the tree resulting from the present maximum-likelihood phylogenetic analysis of the Serendipeidae (Fig. 7), a subclade consisting of the species of Nanoduplicibothrium grouped as the sister taxon to the subclade consisting of the 2 specimens of Serendip n. sp. 1. A subclade consisting of the specimens of Duplicibothrium grouped as the sister to that clade. All 3 groups were supported with bootstrap support values ≥80%.

The results of our molecular phylogenetic analysis have helped address the question of the reciprocal monophyly of Duplicibothrium and Serendip raised by Stephan and Caira (2022) given the close similarity in bothridial morphology between D. colossum and species of Serendip. In the tree resulting from our analysis, as was found by Stephan and Caira (2022), D. colossum was deeply embedded among species of Duplicibothrium. Consistent with recognition of Serendip as an independent genus, Serendip n. sp. 1 grouped as the sister taxon to the subclade containing species of Nanoduplicibothrium.

The tree also supports the morphological differences seen between species of Duplicibothrium and those of Nanoduplicibothrium. Given the close relationship between the latter species and those of Serendip, we initially entertained the idea of assigning these species to Serendip. However, the substantial differences in scolex morphology between the 2 groups in combination with unique similarities within the 2 groups, made the establishment of a new genus the more diagnosable option.

It is interesting that, based on the larval work of Jensen and Bullard (2010), it appears that the final larval stage of Duplicibothrium minutum parasitizes a variety of bivalves, whereas the final larval stage of Nanoduplicibothrium leanneae parasitizes a variety of gastropods. Although these data come from only a single species in each genus, it will be interesting to see if this trend toward differential use of major groups of molluscs as intermediate hosts holds for other members of each genus. Based on these associations and the diet of rhinopterid and myliobatid species, we would predict that species of Serendip and Glyphobothrium will be similarly found to employ molluscs as their final intermediate hosts.

The erection of a new genus to house a subset of species originally assigned to Duplicibothrium led us to rethink our interpretation of the facial loculi of D. colossum. Under the revised concept of Duplicibothrium, all members of the genus possess a distinct posterior row of 5 or 7 bothridial loculi. In the case of D. colossum, the 5 loculi in this row occupy much of the length of the bothridium leaving little space for the longitudinal septa and/or transverse septa forming loculi in the anterior regions of the bothridia of its congeners. The resemblance to species of Serendip is thus superficial.

With respect to global diversity, the erection of Nanoduplicibothrium brings the total number of genera in the Serendipeidae to 4 (i.e., Duplicibothrium, Glyphobothrium, Nanoduplicibothrium, and Glyphobothrium). Description of 3 new species brings the total number of species in the family to 12; with 3 provisionally identified species (i.e., Duplicibothrium n. sp. 1 of Jensen and Bullard [2010], and Duplicibothrium n. sp. 6 and Nanoduplicibothrium n. sp. 3 of Stephan and Caira [2022]) remaining to be formally described. All 15 of these species are known only from cownose rays of the genus Rhinoptera. The apparent fidelity of the Serendipeidae for species of Rhinoptera, in combination with the fact that most species in this genus of cownose rays have now been examined for cestodes, led Stephan and Caira (2022) to suggest little diversity of serendipeids likely remains to be discovered across the globe. However, the collection of specimens of Serendip n. sp. 1 in the rough eagleray, A. asperrimus, from Panama expands the potential repertoire of hosts beyond the Rhinopteridae to include the Myliobatidae, suggesting that the global diversity of this family of cestodes may be more extensive than previously thought. Unfortunately, description of this new species awaits the collection of additional material to allow characterization of its morphological features.

The erection of Nanoduplicibothrium substantially reduces the number of instances of congeners parasitizing the same host species discussed by Stephan and Caira (2022). Rhinoptera jayakari hosts N. megaphallum and D. bilai, Rhinoptera steindachneri hosts N. paulum and D. cairae, and R. bonasus hosts N. leanneae and D. minutum. Although Rhinoptera brasiliensis hosts D. minutum and Duplicibothrium n. sp. 1 (of Jensen and Bullard, 2010), it also hosts Nanoduplicibothrium n. sp. 3 (of Stephan and Caira, 2022). Similarly, Rhinoptera marginata hosts D. colossum and D. jeannettae, but also hosts N. jillae. Thus, in most cases, species parasitizing the same host species are not each other's closest relatives; rather, they belong to separate genera.

Our results support the position of Caira et al. (2017) that Glyphobothriidae, which was established by Monks et al. (2015) to house Duplicibothrium and Glyphobothrium to the exclusion of Serendip, is a junior synonym of Serendipeidae. We have provided evidence of the reciprocal monophyly of Serendip, Duplicibothrium, and Nanoduplicibothrium relative to one another, but the phylogenetic affinities of the monotypic Glyphobothrium remain unresolved. The somewhat unusual scolex morphology of Glyphobothrium zwerneri Williams and Campbell, 1977, which consists of a globular scolex with 4 sessile bothridia fused to the outer surface of the scolex proper, leads us to believe it will be found to represent a phylogenetically divergent group relative to the 3 other genera once it is included in a molecular phylogenetic analysis. However, scolex morphology can be misleading (e.g., Jensen et al., 2016). Further complicating the situation are the depictions of the scolex of specimens identified as G. zwerneri by Monks et al. (2015:fig. 1) collected from R. bonasus off Campeche, México. These specimens exhibit dorsal and ventral bothridia fused length-wise in 2 pairs and thus resemble Nanoduplicibothrium more closely than Glyphobothrium. As the oldest of the 4 generic names in the family, if Glyphobothrium is found to group among members of any of the other 3 genera of Serendipeidae, the generic assignments of the other members of the genus within which it falls will need to be revised.

We thank Kirsten Jensen of the University of Kansas for providing the image of the proglottid of the holotype of D. minutum as well as for collecting tapeworms from stingrays in South Carolina and assisting with the collection of tapeworms from stingrays in Mozambique. We also thank Kaylee Herzog and Rachel Guyer for assisting with the fieldwork in South Carolina. We are especially grateful to Fernando Marques of the University of São Paulo, Brazil for logistics and assisting with the fieldwork in Mozambique as well as for providing specimens of Serendip n. sp. 1 from Panama. We thank the associate editor and 1 reviewer for their thoughtful comments on an earlier version of this manuscript. This work was supported with National Science Foundation grants DEB 1921404 and 1921411. Any opinions, findings, conclusions, or recommendations expressed in this material are those of the authors and do not reflect the views of the National Science Foundation.