Type specimens of the new species Neoniphon pencei from Rarotonga, Cook Islands have been deposited in the Bernice P. Bishop Museum, Honolulu (BPBM); the California Academy of Sciences, San Francisco (CAS); and the U.S. National Museum of Natural History, Washington, D.C. (USNM).

Measurements and counts given here follow the methods outlined in Randall (1998). Lengths of specimens are given as ratios of: standard length (SL) measured from the tip of the snout to the base of the caudal fin at the end of the hypural plate; body depth, taken at the point of maximum depth; or head length, measured from the median anterior point of the upper lip to the end of the longest opercular spine. Meristics and measurements were compared with data obtained from the literature for all five currently recognized species (Shimizu and Yamakawa 1979).

Tissue samples were obtained from each of the thirteen individuals of N. pencei collected at Rarotonga, Cook Islands by spear at 90-115 m. Tissue samples were also obtained from twenty-two specimens of the five other species of Neoniphon: N. sammara (n=6) collected from Diego Garcia, British Indian Ocean Territory; N. opercularis (n=2) collected from Moorea, French Polynesia; N. aurolineatus (n=7) collected from Oahu, Hawaii; N. marianus (n=1) collected from the Commonwealth of the Bahamas; and N. argenteus (n=6) collected from the Republic of Kiritimati. Total genomic DNA was extracted from each sample using the 'HotSHOT' protocol (Meeker et al. 2007). A 577-bp fragment of the mtDNA cytochrome b (Cyt b>) region was amplified using modified primers from Song et al. (1998) (5’-TGAAGTTGTCGGGATCTCCT-3’) and Taberlet et al. (1992) (5’-TGCCGTGACGTAAACTATGG-3’). Polymerase chain reaction (PCR) was performed in a 15 µl reaction containing 7.5 µl BioMix Red (Biolone Inc., Springfield, NJ, USA), 0.2 µM of each primer, 5-50 ng template DNA, and nanopure water (Thermo Scientific* Barnstead, Dubuque, IA, USA) to volume. PCR cycling parameters were as follows: initial 95°C denaturation for 10 min. followed by 35 cycles of 94°C for 30 sec, 60°C for 30 sec, and 72°C for 30 sec, followed by a final extension of 72°C for 10 min. PCR products were visualized using a 1.5% agarose gel with GelStarTM (Cambrex Bio Science Rockland, Inc., Rockland MA, USA) and then cleaned by incubating with 0.75 units of Exonuclease and 0.5 units of Shrimp Akaline Phosphate (ExoSAP; USB, Cleveland, OH, USA) per 7.5 µl of PCR product for 30 min. at 37°C followed by 85°C for 15 min. Sequencing was conducted in the forward direction and reverse direction when needed using a genetic analyzer (ABI 3130XL, Applied Biosystems, Foster City, California) at the Hawai'i Institute of Marine Biology EPSCoR Sequencing Facility. The sequences were aligned, edited and trimmed to a common length using Geneious Pro (v.5.6.6) DNA analysis software (Drummond et al. 2012). Twelve representative Cyt b> haplotypes were deposited in GenBank (accession numbers KJ188431-188436 and KJ201921-201926). jModelTest v.2.1.4 (Darriba et al. 2012, Guindon and Gascuel 2003) was used with an Akaike information criterion (AIC) test to determine the best nucleotide substitution model for the data. The GTR+G model with gamma parameter 0.1840 was identified to be the best suited model for phylogenetic inference. Maximum Likelihood, Neighbor-Joining, and Maximum Parsimony tree-building methods were implemented using Mega v.5.2.2 (Tamura et al. 2011). Sargocentron rubrum (Genbank accession number AP004432.1) was used to root a maximum likelihood phylogenetic reconstruction. Clade support was evaluated by bootstrapping 1,000 replicates in all cases (Felsenstein 1985).

A DNA barcode (cytochrome c oxidase I; COI) was completed for the holotype and one paratype (BPBM 41196) using the primers from Baldwin et al. (2009), Fish-BCH (5'-ACTTCYGGGTGRCCRAARAATCA-3') and Fish-BCL (5'-TCAACYAATCAYAAAGATATYGGCAC-3') using the following PCR protocol: initial 95°C denaturation for 10 min. followed by 35 cycles of 94°C for 30 sec, 55°C for 30 sec, and 72°C for 30 sec, followed by a final extension of 72°C for 10 min. All other procedures were as described above. Both individuals possessed the same COI haplotype, so only one record was deposited in GenBank (http://www.ncbi.nlm.nih.gov/; accession number KJ188437) and BOLD (www.boldsystems.org; dx.doi.org/10.5883/DS-NPE511).

Genetic results

After alignment and editing, a 377-bp partial sequence of Cyt b was obtained for all thirty-five Neoniphon samples, resulting in twelve unique haplotypes. All three phylogenetic methods used resulted in congruent tree topologies and are presented as a Maximum Likelihood reconstruction (Fig. 4). Phylogenetic reconstruction recovered strong support for clades corresponding to known Neoniphon species. The species N. pencei showed strong clade support (100% bootstrap support for all three methods) for belonging to a single clade distinct from currently described Neoniphon species. There was not enough signal to resolve the sister relationship between some members within the genus Neoniphon; however, this description is not necessary for the goals of this study. Neoniphon pencei shows 9-12.5% uncorrected sequence divergence and 34-47 mutations between all other known Neoniphon species and posesses 8 diagnostic sites unique from all other species of Neoniphon within this this region of Cyt b. This is consistent with species level sequence divergence found in other fish taxa (Bellwood et al. 2004, Fessler and Westneat 2007, Randall and Rocha 2009, Rocha 2004, Rocha et al. 2008).

Figure 4.  

Maximum likelihood phylogenetic reconstruction for the genus Neoniphon based on Cyt b sequences from 35 individuals, yielding 12 unique haplotypes, rooted with Sargocentron rubrum. Branch support values are Maximum Likelihood, Neighbor-Joining, and Maximum Parsimony bootstrap percent values respectively. Triangles at branch termini represent multiple haplotypes; vertical bars at branch termini represent multiple individuals with identical haplotypes.

Most recent authors who have reported on Neoniphon (e.g., Randall and Heemstra 1985, Randall and Heemstra 1986, Kotlyar 1996, Kotlyar 1998, Randall and Greenfield 1999, Greenfield 2003, Satapoomin 2009) consider it to be a valid genus (a senior synonym of Flammeo), distinct from other genera in the subfamily Holocentrinae (particularly Sargocentron; Fowler 1904), primarily on the basis of the position of the last dorsal-fin spine (relative to the penultimate dorsal-fin spine and first dorsal-fin ray), and the protruding lower jaw in species of Neoniphon (Randall and Heemstra 1985). A more recent phylogenetic analysis of holocentrids by Dornburg et al. (2012), however, reported evidence that Sargocentron and Neoniphon are paraphyletic. Specifically, they found that four of the five species of Neoniphon (they did not include N. aurolineatus in their analyses) cluster among several subclades that include nine of the seventeen species of Sargocentron they analyzed (S. coruscum, S. diadema, S. inaequalis, S. ittodai, S. microstoma, S. punctatissimum, S. suborbitalis [=suborbitale], S. vexillarium and S. xantherythrum). The other eight species of Sargocentron they analyzed (S. caudimaculatum, S. cornutum, S. melanospilos, S. praslin, S. rubrum, S. seychellense, S. spiniferum and S. tiere) form a separate clade (their "Sargocentron group 1"). They argue that the characters used to differentiate these species are ecologically plastic and therefore current relationships represent ecotypes rather than their evolutionary relationships. We acknowledge the results of this study and welcome a new comprehensive analysis of the entire Holocentrinae in light of new genetic evidence. However, in the absence of observed morphological characters that are consistent with the genetic results, we choose to retain these six species within the genus Neoniphon, to the exclusion of Sargocentron, thereby maintaining nomenclatural stability. Neoniphon pencei clearly differs from all species placed in the genus Sargocentron on the basis of a closer association of the last dorsal-fin spine with the first soft-ray rather than the penultimate spine and the strongly protruding lower jaw (Randall and Heemstra 1985) as well as life color.

Meristic data of the type specimens of Neoniphon pencei are included in Table 1, and proportional measurements are included in Table 2. Neoniphon pencei is distinctive from all other species of holocentrids, both morphologically and genetically. Table 3 summarizes morphological differences between N. pencei and other species in the genus. It differs most substantially from all other Neoniphon in number of lateral line scales (48-52, compared with 38-47 among all other species), number of scales above the lateral line to the origin of the dorsal fin (5, compared with 2.5-3.5) and number of scales below the lateral line to the origin of the anal fin (6-7, compared with 7-9). It also differs from N. aurolineatus, N. opercularis, and N. argenteus in proportional length of the upper-jaw (2.3-2.6 in head length, compared with 2.0-2.3), and proportional length of of the third and fourth anal spines (1.1-1.3 and 1.7-2.0, compared with 1.4-1.9 and 1.9-2.7, respectively). It is further distinguished from N. aurolineatus in total number of gill rakers (19-20, compared with 15-17); from N. opercularis in head length (2.6-2.9 in SL, compared with 2.9-3.1), orbit diameter (1.2-1.4 in head length, compared with 3.0-3.5), snout length (1.2-1.4 in orbit diameter, compared with 1.2-1.5), and interorbital width (1.7-1.9 in orbit diameter, compared with 1.2-1.5); from N. argenteus in number of pectoral rays (14, compared with 12-13), interorbital width (1.7-1.9 in orbit diameter, compared with 1.2-1.7), and first dorsal-spine length (3.3-4.1 in head length, compared with 2.4-3.1); and from N. sammara in number of soft dorsal-fin rays (13, compared with 11-12), interorbital width (1.7-1.9 in orbit diameter, compared with 1.3-1.6), and first dorsal-spine length (3.3-4.1 in head length, compared with 2.2-3.0). In addition to these morphometric characters, N. pencei differs from all other species of Neoniphon in life color, particularly in the pattern of white spots on the dorsal fin and overall body color, and the lack of yellow coloration on the body (as in N. aurolineatus and N. marianus). Genetically, it differs in its Cyt b sequence from N. argenteus by 9.8%, N. aurolineatus by 9-9.6%, N. marianus by 11.7%, N. opercularis by 9.8%, and N. sammara by 12-12.5%.

Neoniphon pencei appears most similar to N. aurolineatus and N. marianus, based on having the the fewest number of differences in morphometrics, greatest genetic similarity, and most similar aspects of life coloration with these two species. It is also similar to N. aurolineatus in the depth and habitat it occupies. However, the differences between N. pencei and these two species as noted above clearly warrant recognition of N. pencei as a distinct species. A more comprehensive phylogenetic analysis of the species of Neoniphon and related genera based on both morphology and genetics (with verified voucher specimens) is beyond the scope of this work.