Molecular assessment of three species of Anilocra (Isopoda, Cymothoidae) ectoparasites from Caribbean coral reef fishes, with the description of Anilocra brillae sp. n.

Abstract A morphological review and molecular characterization of Anilocra haemuli Bunkley Williams & Williams, 1981, were completed using specimens collected from Haemulon flavolineatum Desmarest, 1823 (French grunt) and Epinephelus guttatus Linnaeus, 1758 (red hind). Molecular and morphological data suggest that the isopods parasitizing H. flavolineatum and E. guttatus are different species. The specimens collected from E. guttatus are recognized as a new species, Anilocra brillae sp. n. Differences between Anilocra brillae sp. n. and A. haemuli include but are not limited to the pleonites 1–3 of A. brillae sp. n. being wider than 4–5 and 4–5 subequal, whereas the pleonites 1–2 of A. haemuli are wider than 3–5, and 3–5 are subequal. The seventh pereopod of A. brillae sp. n. is proportionally larger, has more robust setae, and the setae are distributed more extensively over the articles when compared to A. haemuli. Additionally, this study provides the first genetic characterization of three Anilocra spp. from the Caribbean, and is based on mitochondrial cytochrome c oxidase subunit gene (COI) for A. haemuli from H. flavolineatum, A. brillae sp. n. from E. guttatus, and A. chromis Bunkley Williams & Williams, 1981 from Chromis multilineata Guichenot, 1853.


Introduction
In the past half-century, taxonomic studies on the fish parasitic isopod genus Anilocra Leach, 1818, have reported nine species from the Caribbean (Bunkley Williams and Williams 1981) and 12 species from Australia (Bruce 1987). This genus of parasite parasitizes the external surfaces of marine fish hosts that inhabit subtropical, tropical, and temperate waters . Host specificity of species of Anilocra is highly variable, such that different Caribbean Anilocra have been identified as family, genus, and species specific (i.e. Bunkley Williams 1981, Bruce 1987). For example, Anilocra holocentri Bunkley Williams & Williams, 1981 has been reported only to infest Holocentrus adscensionis Osbeck, 1765, whereas Anilocra chaetodontis Bunkley Williams & Williams, 1981 has been reported to infest four members of the genus Chaetodon Linnaeus, 1758. Anilocra haemuli Bunkley Williams & Williams, 1981 is the only Caribbean species reported to infest fishes from two families: Haemulidae and Serranidae. Anecdotal accounts from both parasitologists and ecologists suggest that records of A. haemuli from Haemulids and Serranids may in fact be two species given the differences in the biology and ecology of these host fishes.
To evaluate this claim a review of Anilocra haemuli morphology using specimens from both the Haemulidae and Serranidae families is warranted. The original description of A. haemuli was published before molecular approaches were used to aid in confirming the morphological classification of organisms. In the original description, careful attention was taken to describe A. haemuli as type specimens were collected from the same host and locality (Bunkley Williams and Willams 1981). Nevertheless, multiple morphologically similar species of Anilocra may have been identified as A. haemuli because there was no other method to verify if these specimens represented multiple species.
An increasing number of ecological studies are using Anilocra to understand trophic level dynamics , Binning et al. 2014, and A. haemuli infestation has been associated with altering H. flavolineatum behavior and condition Sikkel 2014, 2015, Welicky et al. in press). To facilitate future ecological and evolutionary studies on Anilocra-host interactions, the identity of Anilocra haemuli is here validated using both a morphological redescription and a molecular analysis.

Molecular analysis
Of the specimens collected, genomic DNA was extracted from eight Anilocra from E. guttatus, seven A. haemuli from H. flavolineatum, and eight A. chromis from C. multilineata using a rapid DNA extraction method as described in the KAPA Express Extract Kit (Kapa Biosystems, Cape Town, South Africa). Polymerase chain reactions (PCR) were used to amplify a 710 basepair fragment of the mitochondrial cytochrome c oxidase subunit gene (COI) using the primer sets LCO 1490 and HCO 2198 (Folmer et al. 1994). PCR was performed using 12.5 μl Thermo Scientific DreamTaq PCR master mix (2×) (2× DreamTaq buffer, 0.4 mM of each dNTP, and 4 mM MgCl2), 1.25 μl of each primer, 1 μl DNA, and 9 μl of PCR-grade nuclease free water (Thermo Scientific, Vilnius, Lithuania). Total volume per reaction was 25 μl, and PCR reactions were conducted using a ProFlex™ PCR thermal cycler (Applied Biosystems by Life Technologies). Reactions were amplified under the following PCR conditions: Stage 1, 94°C for 5min, Stage 2, 36 cycles of 94°C for 30s, 47°C for 50s, 72°C for 2min, and Stage 3, 72°C for 10min. PCR products were sent to a commercial sequencing company (Inqaba Biotechnical Industries (Pty) Ltd, Pretoria, South Africa) for purification and sequencing in both directions. Obtained sequences were assembled, and chromatogram-based contigs were generated using Geneious Ver. 9.1. Sequences were aligned and trimmed to the length of the shortest sequence using MEGA 7 bioinformatics software program (http://www.megasoftware.net) Using BLASTn (Basic Local Alignment Search Tool; http://www.ncbi.nlm. nih.gov/blast), the obtained sequences were verified as belonging to the Isopoda. Pair-wise distance (p-distance) using the Kimura 2-parameter model and nucleotide differences were determined in MEGA7. Supplemental comparisons among the sequences of this study and those available for Anilocra sp. from GenBank were also determined. Newly-generated sequences for Anilocra spp. were deposited in Gen-Bank under the accession numbers: A

Morphological data
Anilocra haemuli from Haemulon flavolineatum and Anilocra from Epinephelus guttatus were examined using material previously collected by Ernest Williams and Lucy Bunkley-Williams during 1976-1977and 1983 and reported in Bunkley- Williams and Williams (1981). Additionally, specimens from each host were collected using the aforementioned methods as part of other studies conducted in the US Virgin Islands (USVI) and British Virgin Islands (BVI) during 2011-2016. Isopods were processed according to the techniques de scribed in Hadfield et al. (2010Hadfield et al. ( , 2013. Descriptions were prepared using DELTA (Descriptive Language for Taxonomy, Coleman et al. 2010) using a general character set for the Cymothoidae (Hadfield et al. , 2016. Ratios and measurements were rounded off to one decimal place and were made using maximum values of the specific measured article. Ratios and measurements were taken from the female (♀) and transitional stage (TS) specimens used for the drawings and presented herein as figures. Pleotelson length (TL) and width (W) for all specimens examined are reported. All measurements are reported in milliimeters (mm). Classification follows Brandt and Poore (2003).
Remarks. The body of female Anilocra is dorsally symmetrical and strongly vaulted. The posterior margins of their cephalon are smooth and straight, and the rostrum is more blunt than pointed. The rostrum folds into the area between the antennula bases. The antennula is shorter than the antenna. The posterolateral margins of the pereonites are not produced. Coxae 1-3 are short, posteriorly rounded and do not form a rounded point posteriorly, whereas coxae 4-6 are longer, less rounded and more elongate than coxae 1-3, and form a rounded point posteriorly. The pereopods gradually increase in size towards the posterior.
In the Cymothoidae, the external-attaching genera include but are not limited to Anilocra, Nerocila Leach, 1818, Renocila Miers, 1880, Creniola Bruce, 1987, andPleopodias Richardson, 1910. Anilocra can be distinguished from Nerocila by the posterior margin of the cephalon, which is conspicuously trilobed in Nerocila, whereas the posterior margin of the cephalon of Anilocra is not tri-lobed to weakly tri-lobed. The posterolateral pereonite margins of Nerocila are more produced, elongate and pointed than that of Anilocra. In the Caribbean, some species of Anilocra and Renocila share numerous similarities, but in Anilocra pereopod 6 is shorter in length than pereopod 7, whereas in Renocila pereopods 6 and 7 are of similar length. To date the genera Creniola and Pleopodias have not been reported from the Caribbean.  PC Sikkel and/or ER Brill: ♀ (30,11;31,12 Ovigerous female. Size intact (29, 13). Body weakly ovoid, 2-2.6 times as long as greatest width, dorsal surfaces smooth and polished in appearance, widest at pereonite 5, most narrow at pereonite 1, lateral margins mostly ovate posteriorly. Cephalon 0.5-0.7 times longer than wide, visible from dorsal view, weakly trapezoid shaped. Frontal margin rounded to form blunt rostrum or simple, not folded. Eyes oval with distinct margins, one eye width 0.1-0.2 times width of cephalon; one eye length 0.4-0.5 times length of cephalon. Pereonite 1 smooth, anterior border straight, anterolateral angle narrowly rounded, not produced. Posterior margins of pereonites smooth and slightly curved laterally. Coxae 2-3 wide; with posteroventral angles rounded; 4-7 rounded and curved; not extending past pereonite posterior margin. Pereonites 1-5 increasing in length and width; 6-7 decreasing in length and width; 1-4 narrower. Pleon with pleonite 1 wider than pleonites 2-5, visible in dorsal view; pleonites posterior margin 1-3 posteriorly weakly concave, 4-5 mostly straight. Pleonite 2 not overlapped by pereonite 7; posterolateral angles of pleonite 2 narrowly rounded. Pleonite 1 similar in form to pleonite 2. Pleonite 5 free, not overlapped by lateral margins of pleonite 4, posterior margin straight. Pleotelson 0.9 times as long as anterior width, dorsal surface smooth. Pleotelson lateral margins convex, posterior margin narrowly rounded.

Anilocra haemuli
Antennula consisting of 7-8 articles; peduncle articles 1 and 2 distinct and articulated; article 2 0.8 times as long as article 1; article 3 0.9 times as long as wide, 0.4 times as long as combined lengths of articles 1 and 2; flagellum with 5 articles, extending to posterior margin of eye. Terminal article with 2 short simple terminal setae. Antenna consisting of 10 articles; article 3 1.6 times as long as article 2; article 4 1.2 times as long as wide, 1.5 times as long as article 3; article 5 1.3 times as long as wide, 1.1 times as long as article 4; flagellum with 5 articles, terminal article terminating in 5 short simple setae, extending to middle of pereonite 1. Mandibular molar process ending in an acute incisor; mandibular palp article 3 with 7 simple setae. Maxillula simple with 4 terminal robust setae. Maxilla mesial lobe partly fused to lateral lobe; lateral lobe with 2 recurved robust setae; mesial lobe with 2 recurved robust setae. Maxilliped weakly segmented, with lamellar oostegite lobe, article 3 with 3 small robust setae.  Pereopod 1 basis 1.7 times as long as greatest width; ischium 0.7 times as long as basis; merus proximal margin without bulbous protrusion; carpus with straight proximal margin; propodus 1.3 times as long as wide; dactylus stout, 2.7 times as long as propodus, 3.8 times as long as wide. Pereopod 2 propodus 2.1 times as long as wide; dactylus 2.2 as long as propodus. Pereopod 6 basis 2.6 times as long as greatest width; ischium 0.5 times as long as basis; propodus 1.3 times as long as wide; dactylus 2.5 times as long as propodus. Pereopod 7 basis 3.2 times as long as greatest width; ischium 0.7 times as long as basis, without protrusions; merus proximal margin without bulbous protrusion; merus 1.1 times as long as wide, 1.6 times as long as ischium; carpus 1.5 times as long as wide, 0.5 times as long as ischium, without bulbous protrusion; propodus 2.6 times as long as wide, 0.8 times as long as ischium; dactylus slender, 1.8 times as long as propodus, 5.0 times as long as wide. Pereopod 7 with few setae on propodus, carpus, and merus. Pleopods without setae, exopod larger than endopod. Pleopod 1 exopod 1.5 times as long as wide, lateral margin weakly convex, distally narrowly rounded, medial margin weakly oblique, mesial margin weakly convex; endopod 1.6 times as long as wide, lateral margin weakly convex, distally narrowly rounded, mesial margin slightly convex; peduncle twice as wide as long, without retinaculae, pointed projection on lateral margin. Pleopods 2-5 similar to pleopod 1. Pleopods 3-5 endopods proximal borders do not extend below exopod to peduncle, fleshy lobes and medial lobes present. Peduncle lobes absent.
Transitional stage. Size (12, 6). Similar to female but smaller. Body 2.5 times as long as wide. Antennula bases separated, consisting of 8 articles, extending to posterior margin of eye. Antenna consisting of 10 articles, extending to middle of pereonite 1. Mandibular molar process ending in an acute incisor; mandibular palp article 3 with 11 simple setae. Maxillula simple with 4 terminal robust setae. Maxilla mesial lobe partly fused to lateral lobe; lateral lobe with 2 recurved robust setae; mesial lobe with 2 recurved robust setae. Maxilliped weakly segmented, with lamellar oostegite lobe, article 3 with 3 small recurved robust setae. Pereopod 7 with few small robust setae on carpus, merus and propodus. Pleopod 2 appendix masculina absent.
Distribution Remarks. The description of A. haemuli from H. flavolineatum given above is in agreement with the original description in Bunkley Williams and Williams (1981). We supplement the original species diagnosis by now providing drawings and measurements of the antenna and antennula articles, additional pereopods, and pleopods.
Antennula bases separated, shorter than antenna, consisting of 7-9 articles; peduncle articles 1 and 2 distinct and articulated; article 2 1.5 times as long as article 1; article 3 0.9 times as long as wide, 0.5 times as long as combined lengths of articles 1 and 2; flagellum with 4 articles, extending to posterior margin of eye. Terminal arti-  cle terminating in 1 short simple seta. Antenna comprised of 9-10 articles, peduncle article 3 1.5 times as long as article 2; article 4 1.3 times as long as wide, 1.1 times as long as article 3; article 5 1.6 times as long as wide, 1.1 times as long as article 4; flagellum with 4 articles, terminal article with 5 short simple setae, extending to posterior of pereonite 1. Mandibular molar process ending in an acute incisor; mandibular palp article 3 with 8 simple setae. Maxillula simple with 4 terminal robust setae. Maxilla mesial lobe partly fused to lateral lobe; lateral lobe with 2 recurved robust setae; mesial lobe with 1 recurved robust seta. Maxilliped weakly segmented, with lamellar oostegite lobe, article 3 with 3 recurved robust setae.
Pereopod 1 basis 1.8 times as long as greatest width; ischium 0.23 times as long as basis; merus proximal margin without bulbous protrusion; carpus with straight proximal margin; propodus 1.9 times as long as wide; dactylus moderately slender, 1.8 times as long as propodus, 3.7 times as long as wide. Pereopod 2 propodus 1.7 as long as wide; dactylus 2.7 times as long as propodus, 4.9 times as long as wide. Pereopods gradually increasing in size towards posterior. Pereopod 6 basis 1.7 times as long as greatest width; ischium 0.7 times as long as basis; propodus 1.5 times as long as wide, dactylus 2.3 times as long as propodus, 3.8 times as long as wide. Pereopod 7 basis 3.0 times as long as greatest width; ischium 0.7 times as long as basis, without protrusions; merus proximal margin without bulbous protrusion, 2.0 times as long as wide, 0.7 times as long as ischium; carpus 1.5 times as long as wide, 0.6 times as long as ischium, without bulbous protrusion; propodus 3.2 times as long as wide, 0.8 times as long as ischium; dactylus moderately slender, 0.9 times as long as propodus, 3.5 times as long as wide. Pereopod 7 with many setae on propodus, carpus, and merus.
Uropod more than half the length of pleotelson, peduncle 0.7 times longer than rami, peduncle lateral margin without setae; rami not extending beyond pleotelson, marginal setae absent, apices broadly rounded. Endopod apically rounded, 2.2 times as long as greatest width, lateral margin weakly convex, mesial margin weakly convex, terminating without setae. Exopod not extending to end of endopod, 2.6 times as long as greatest width, apically rounded, lateral margin convex, mesial margin weakly convex or weakly concave, terminating without setae.
Transitional stage. Size (11,6). Similar to female but smaller. Body 2.6 times as long as wide. Antennula bases separated, consisting of 8 articles, extending to middle of eye. Antenna consisting of 10 articles, extending to middle of pereonite 1. Mandibular molar process ending in an acute incisor; mandibular palp article 2 with 2 simple setae, article 3 with 7 simple setae. Maxillula simple with 4 terminal robust setae. Maxilla mesial lobe partly fused to lateral lobe; lateral lobe with 2 recurved robust setae; mesial lobe with 2 recurved robust setae. Maxilliped weakly segmented, with lamellar oostegite lobe, article 3 with 3 recurved robust setae. Pereopod 7 with several small robust setae on carpus, merus and propodus. Pleopod 2 appendix masculina absent.
Etymology. This species is named in honor of Elizabeth R. Brill for her dedication to studying the ecology of A. haemuli, and for collecting many of the A. haemuli and A. brillae sp. n. specimens used in this study.
Distribution. Known from St. John and St. Thomas, USVI, Guana Island, BVI, and islands of Puerto Rico, Spanish Virgin Islands. Expected distribution throughout the Caribbean Sea, where fish of the Serranidae family inhabit.
Hosts. Known from Epinephelus guttatus (Linnaeus, 1758). Remarks. Previously, A. brillae sp. n. was identified as A. haemuli. Compared to A. haemuli, the outer margins of the cephalon and pereonites 1-4 of A. brillae sp. n. form a more pronounced trapezoid shape and the remaining portion of the body is ovoid. A. brillae sp. n has more strongly narrowed pleonites than A. haemuli. Pleonites 1-3 of A. brillae sp. n. are wider than 4-5 and 4-5 are subequal, whereas the pleonites 1-2 of A. haemuli are wider than 3-5, and 3-5 are subequal. Pleonite 5 is more posteriorly rounded in A. brillae sp. n, but this is somewhat variable among individuals. Another more variable feature is A. brillae sp. n. has a more caudomedially pointed pleotelson than A. haemuli. Typically, the seventh pereopod of A. brillae sp. n. is proportionally larger, has more robust setae, and the setae are distributed more extensively over the articles when compared to A. haemuli. The antennula peduncle of A. brillae sp. n. is regularly observed as shorter and more robust than that of A. haemuli. With respect to attachment, both species infest the subocular region, and if infested by two parasites, one parasite typically attaches under each eye. Infestation by a third A. brillae sp. n. on a single host seems to occur with more frequency than tertiary infestation by A. hae-muli on a single host. The third parasite is typically attached between the eyes on the head of the host, or adjacent to one of the other parasites (RLW, pers obs).
Anilocra brillae sp. n. can be distinguished from all other Caribbean species except Anilocra haemuli using the same morphological comparisons described between A. haemuli and other Anilocra spp. given in Bunkley Williams and Williams (1981). Additionally, the body of A. brillae sp. n. is not expanded and is more elongate compared to the bodies of A. holocanthi and A. chaetodontis.

Discussion
The results of this study provide the first reliable COI sequences for species of Anilocra, and confirm that A. haemuli from H. flavolineatum is morphologically and genetically different than the Anilocra specimens collected from E. guttatus, and are here described as A. brillae sp. n. Our morphological data suggest there are two different species given the number of differences consistently observed, and our molecular analyses demonstrate a 4% difference between A. haemuli and A. brillae sp. n. This difference is less than half of that observed between A. brillae sp. n. and A. chromis, which are more conspicuously morphologically different. Our supplemental analyses were conducted utilizing the available Anilocra sp. COI sequences on GenBank, and there was a high level of interspecific divergence of these sequences compared with our dataset. The large differences in interspecific divergence between the specimens of this study and those provided on GenBank may be explained by the fact that the GenBank specimens may have been misidentified or not identified at all, as no morphological identification was described in Ketmaier et al. (2007). Thus, further interspecific comparisons cannot be assessed at this time.
Anilocra spp. have been reported to influence the fitness Lester 2004, Fogelman et al. 2009) and behavior (Meadows andMeadows 2003, Welicky andSikkel 2015, Welicky et al. in press) of their fish hosts, and Anilocra brillae sp. n. infests E. guttatus, a grouper species that is currently recovering from previously intense fishing pressure (Nemeth et al. 2005). There is limited knowledge on the biotic stressors that influence E. guttatus population dynamics, and thus the effects of A. brillae sp. n. on E. guttatus should be examined as a potential stressor. Moreover, by studying this hostparasite interaction, further insight into variations in life histories of Anilocra spp. may be gained, if the life cycle of the parasite coincides with that of their host. The only complete description of an Anilocra spp. life cycle is of a species that infests an egg laying/guarding fish species (Adlard and Lester 1995), whereas many Anilocra spp. infest broadcast spawners. Interestingly, A. brillae sp. n. infests a fish species that undergoes an annual long distance migration to spawn in an aggregation (Nemeth 2011). Given that Anilocra spp. infection has been reported to influence host swimming performance in some fish (e.g., Binning et al. 2013), A. brillae sp. n. infection may indirectly influence the reproductive success of their hosts. This study exemplifies that there is an incomplete but growing knowledge of cymothoid life histories, genetics, and morphology, and how these disciplines relate to host-parasite ecology. Continued efforts to conduct studies in these disciplines are necessary to better understand one of the least understood parasite families.