Content uploaded by G David Johnson
Author content
All content in this area was uploaded by G David Johnson on Jun 08, 2023
Content may be subject to copyright.
Vol.:(0123456789)
1 3
Ichthyological Research
https://doi.org/10.1007/s10228-023-00906-4
FULL PAPER
Cusk‑eel confusion: revisions oflarval Luciobrotula
andPycnocraspedum andre‑descriptions oftwo bythitid larvae
(Ophidiiformes)
MatthewG.Girard1,2 · BruceC.Mundy3· AiNonaka1· G.DavidJohnson1
Received: 24 November 2022 / Revised: 1 February 2023 / Accepted: 7 February 2023
© The Author(s) under exclusive licence to The Ichthyological Society of Japan 2023
Abstract
Since 2006, an ophidiiform larva with an ovoid body, elongate anterior dorsal-fin ray, and long trailing fleshy filament has
been identified as Pycnocraspedum squamipinne. Similarly, the larvae of the ophidiid genus Luciobrotula have been tenta-
tively identified since 1988, with posteriorly displaced dorsal fins and bulging or exterilium guts. However, neither of these
larval forms morphologically agree with their adult counterparts. Recently, blackwater divers captured and photographed
specimens of larval Luciobrotula and Pycnocraspedum off the coast of Hawaiʻi and Florida, making them available for both
morphological and molecular sampling. After examining these larvae and analyzing DNA barcode sequences, as well as a
newly captured and sequenced adult of Pycnocraspedum phyllosoma, we revise the previously identified “Pycnocraspedum”
larvae to species of Luciobrotula. We describe the larvae of Luciobrotula bartschi and Luciobrotula corethromycter for the
first time, highlighting an extraordinary loss of multiple anterior dorsal-fin elements in their ontogeny. We also generate the
first DNA sequences for L. corethromycter and P. phyllosoma, adding to the depauperate number of sequences available for
ophidiiforms. For the previously identified “Luciobrotula” larvae, neither morphological nor molecular characters provide
definitive identification other than recovering them among the Bythitidae. We provide new morphological observations,
revised descriptions, and generate a phylogeny of ophidiiform fishes based on COI to place these larvae in a phylogenetic
context, prompting further investigation into the relationships of the Ophidiiformes using additional genetic markers. Our
study emphasizes the importance of blackwater diving to improving our understanding of marine larval fishes and the need
for additional molecular sampling of the diverse order of brotulas, cusk-eels, pearlfishes, and their allies.
Keywords Benthocometes· Blackwater· COI· Integrative taxonomy· Ophidiidae
“Ophidiiform fishes, in general, are too poorly known ana-
tomically to resolve questions of phylogenetic relationships
(Cohen & Nielsen 1978), and details of ontogeny, includ-
ing developmental osteology, have been described for too
few genera to contribute to a resolution of these questions”
(Fahay and Hare in Richards 2005).
Introduction
In 2006, Evseenko and Okiyama described a remarkable
ophidiid larva from the New Guinean “Dana” larval flatfish
collection. The 22.5 mm standard length (SL) specimen pos-
sesses many bothid-like features, including an ovoid body,
elongate anterior dorsal-fin ray, long and thin cartilaginous
ventral process of the coracoid, and a protruding, but not
exterilium, gut (see Fraser and Smith 1974). The larva also
* Matthew G. Girard
GirardMG@si.edu
Bruce C. Mundy
mundyichthyo@gmail.com
Ai Nonaka
NonakaA@si.edu
G. David Johnson
JOHNSOND@si.edu
1 Department ofVertebrate Zoology, National Museum
ofNatural History, Smithsonian Institution, Washington,
DC20560, USA
2 Biodiversity Institute, University ofKansas, Lawrence,
KS66045, USA
3 Ocean Research Explorations, P.O. Box235926, Honolulu,
HI96823, USA
M. G. Girard etal.
1 3
has a long fleshy appendage trailing away from the body,
attaching near the anus (Evseenko and Okiyama 2006: fig.1;
Fig.1). Based on a suite of characters, the authors placed
the larva among the Ophidiiformes and within Group 1 of
Howes (1992) classification, which includes the genera
Brosmophyciops, Brotula, Cherublemma, Dicrolene, Gen-
ypterus, Glyptophidium, Hoplobrotula, Hypopleuron, Lam-
programmus, Lepophidium, Monomitopus, Neobythites,
Ogilbia, Ophidion, Parophidion, Petrotyx, Pycnocraspedum,
and Sirembo. Highlighting the anterior position of the dorsal
fin (compare Figs.1 with 2a–b), the number of elongate gill
rakers on the first arch (4), and the first rib attaching to the
vertebral centrum, Evseenko and Okiyama (2006) identi-
fied the larva as “Pycnocraspedum squamipinne” (hereafter,
taxonomic names in quotes refer to larval identifications that
are revised in this study). However, the authors noted a “lack
of agreement” in the number of precaudal vertebrae and dor-
sal-fin rays between their larva and adult Pycnocraspedum,
with the larva having a greater number for both (14 or 15 lar-
val vs. 12–13 adult precaudal vertebrae; 102 larval vs. 88–92
adult dorsal-fin rays; see Evseenko and Okiyama 2006: 195).
Counts of fin rays, myomeres, and vertebrae are critical to
the recognition and diagnosis of larval fishes (e.g.,Moser
etal. 1984; Okiyama 1988, 2014; Richards 2005; Fahay
2007), and such discrepancies in counts could indicate the
larva belongs to another genus or group of ophidiids.
In discussing Group 1 of his classification, Howes (1992:
117) noted the genus Luciobrotula (Figs.2c–d) “possibly
also belongs with this group” based on expanded second
and third ribs and connections between these ribs and the
gas bladder. The larvae of Luciobrotula were tentatively
identified by Okiyama (1988, 2014; translated by an author)
based on specimens 10.1–23.0 mm SL from the waters of
Japan and the Philippines. These described “Luciobrotula”
larvae have compressed bodies, various levels of bulging
or exterilium guts, posteriorly displaced dorsal fins, and 13
caudal-fin rays (see Okiyama 2014: 434–435; Fig.3). In
2014, Okiyama separated these larvae into four types (types
1–4) largely based on differences in dorsal- and anal-fin-ray
counts, pigmentation, and exterilium-gut length relative to
SL. Okiyama (2014) also noted that the amount of variation
seen in these four types of “Luciobrotula” larvae exceeds the
known number of species that occur in Japan (only Lucio-
brotula bartschi) and listed these larvae as “Luciobrotula?”
to indicate their tentative identification.
Blackwater diving, nighttime open water drift dives (see
Nonaka etal. 2021 for more information), and photography
have opened new opportunities to learn about marine larval
Fig. 1 Larvae of Luciobrotula
from the Hawaiian Islands, mor-
phologically similar to larval
“Pycnocraspedum” described
by Evseenko and Okiyama
(2006). a Blackwater photo
L. bartschi, USNM 454562,
captured by A. and N. Deloach,
offshore of Kona, Hawaiʻi, 11
November 2021; b preserved
USNM 454562; c preserved L.
cf. bartschi, USNM 454451.
Scale bars = 1 cm
Larvae of Luciobrotula
1 3
fishes. Photographs of these larvae have allowed new mor-
phological characters to be observed and behaviors to be
described (e.g., Nonaka etal. 2021; Pastana etal. 2022). In
2017, a larva similar to Okiyama’s (2014) “Luciobrotula”
types 2 and 3 was photographed and captured by blackwa-
ter divers off Hawaiʻi (Figs.3d–e). The Hawaiian specimen
was barcoded for cytochrome c oxidase subunit I (COI)
and included in the study by Nonaka etal. (2021) but the
sequence was only able to be verified to “Actinopterygii”
when compared to the Barcode of Life Database (BOLD; see
their table1). Recently, blackwater divers and photographers
encountered a larva matching Okiyama’s (2014) “Luciobrot-
ula” type 4 off Florida (Figs.3a–c) and larvae similar to Evs-
eenko and Okiyama’s (2006) “Pycnocraspedum” off Hawaiʻi
and Florida. After generating and comparing COI sequences
from these newly captured larvae and an adult specimen
of Pycnocraspedum phyllosoma from the Gulf of Mexico
to publicly available sequences on BOLD and GenBank,
we find these larvae have been misidentified by previous
authors. In this study, we revise the identity of larval “Pyc-
nocraspedum” to the genus Luciobrotula and describe the
larvae of L. bartschi and Luciobrotula corethromycter. With
the larvae of these two species of Luciobrotula identified, we
then examine larvae that morphologically match Okiyama’s
(2014) “Luciobrotula” types. While neither morphological
nor molecular characters could provide definitive identifica-
tions of these larvae, our phylogenetic analysis places these
larvae among the Bythitidae. We provide revised descrip-
tions of these larvae, reducing the number of larval types
originally described by Okiyama (2014) from four to two
based on new morphological observations, and highlight a
connection between the coracoid and intestine.
Materials andmethods
Classification. We follow the classifications of the Ophi-
diiformes by Møller etal. (2016) and Fricke etal. (2022).
Morphological identification, examination, and labora-
tory imaging of larvae and adults. Seven larval specimens,
four of “Pycnocraspedum” and three of “Luciobrotula”,
were examined, along with type and non-type adult speci-
mens of Luciobrotula and Pycnocraspedum. All specimens
used, their lengths, preparations, and museum catalog num-
bers are listed in the Specimens Examined section. Museum
codes follow Sabaj (2020) except for NMNH referring to
non-Fishes Division personnel and resources at the National
Museum of Natural History, Smithsonian Institution.
Fig. 2 Adult a Pycnocraspe-
dum armatum (USNM 162717
holotype) µCT scan, b P.
phyllosoma (USNM 227388), c
Luciobrotula bartschi (USNM
74151 holotype) µCT scan, d L.
corethromycter (USNM 188549
paratype). Arrows indicate
origin of dorsal and anal fins.
Scale bars = 1 cm
M. G. Girard etal.
1 3
Measurements of larval specimens were taken with a digi-
tal caliper to the nearest 0.1 mm. Measurements of adult
specimens were taken with a measuring tape to the nearest
1 mm. Larval specimens were cleared and stained following
Potthoff (1984) with the modifications listed in Girard etal.
(2020). Morphological features were documented using
equipment listed in Girard etal. (2020). Type and large spec-
imens were either x-rayed or scanned using microcomputed
tomography (µCT) to view internal osteology. Specimens
were x-rayed using a Thermo Scientific PXS5-927 MicroFo-
cus 90kV X-Ray Source and a duraSCAN 1417-NDI Digital
Flat Panel X-Ray Detector at NMNH. Specimens were µCT
scanned using either a GE Phoenix v|tome| x M 240/180kV
Dual Tube μCT at NMNH or a Nikon Metrology XT H 225
ST at the Chemical and Biophysical Instrumentation Center,
Yale University. Specimens were scanned using 110–120 kV,
90–100 µA, a 250–333 ms exposure time, and a 62.9–98.9
µm voxel size. Resulting scans were then visualized and seg-
mented using the protocol described in Girard etal. (2022a).
MorphoSource identifiers for the µCT scans can be found in
the Specimens Examined section.
DNA extraction and amplification. Protocols for tis-
sue sampling, DNA extraction, PCR, and sequencing COI
follow the methods described in Nonaka etal. (2021) and
Weigt etal. (2012a) using primers from Baldwin etal.
(2009). Sequence contigs were built, edited, and assem-
bled into FASTA files using Geneious, vers. 11.1.5 (Kearse
et al. 2012). Sequences were deposited on both Gen-
Bank (OQ359786–OQ359790) and BOLD [LUPY001-
23–LUPY005-23; see Electronic Supplementary Material
(ESM) S1].
Taxon identification and analyses of molecular data. To
identity larvae using molecular characters, we downloaded
all publicly available COI sequences of Luciobrotula (11)
and Pycnocraspedum (6) from BOLD and GenBank, as well
as an additional 131 sequences representing of all four rec-
ognized families, 61 genera (of 121, ~50%), and 131 species
(of 562, ~23%) of ophidiiforms (see ESM S1). Sequences
came from a series of published and unpublished works,
including Miya etal. (2003), Ward and Holmes (2007),
Steinke etal. (2009), Lara etal. (2010), Cawthorn et al.
(2011), Mabragana etal. (2011), Hubert etal. (2012), Weigt
etal. (2012b), McCusker etal. (2013), Chen etal. (2014),
Landi etal. (2014), Parmentier etal. (2016), Campbell etal.
(2017), Chang etal. (2017), Robertson etal. (2017 [but see
Lea etal. 2023]), Delrieu-Trottin etal. (2019), Bañón etal.
Fig. 3 Larval Bythitidae types 1 and 2. a, b Blackwater photos of
type 1 larva, USNM 465411, captured offshore of West Palm Beach,
Florida, 18 February 2022, photos © D. Devers; c preserved USNM
465411; d blackwater photo of type 2 larva, USNM 447052, captured
by J. Milisen, offshore of Kona, Hawaiʻi, 13 May 2017, figure from
Nonaka etal. (2021: fig.9C) published in Ichthyology & Herpetology
109 by the American Society of Ichthyologists and Herpetologists
(https:// www. asih. org) and is licensed under CC BY 4.0; e preserved
USNM 447052. Scale bars = 1 cm
Larvae of Luciobrotula
1 3
(2020), Nonaka etal. (2021), Jayakumar etal. (2021), Wong
etal. (2021), Marín etal. (2022), and Pham etal. (2022).
Downloaded sequences were collated into a single file with
newly sequenced larvae and adults, aligned with MAFFT
vers. 7 (Katoh and Standley 2013) within Geneious and
exported as a PHYLIP-format file for phylogenetic analy-
sis. The aligned matrix, which contained 154 terminals, 655
base pairs (bps) in length (~98% complete at the level of
individual bps), and 297 parsimony-informative sites, was
broken into three partitions, one for each of the three codon
positions in the protein-coding locus. These three parti-
tions were input for ModelFinder function within IQ-Tree
vers. 2.2.0 (i.e., MFP + MERGE, Chernomor etal. 2016;
Kalyaanamoorthy etal. 2017; Minh etal. 2020) that selected
a partitioning scheme and models based on Bayesian infor-
mation criterion. Phylogenetic analysis was performed by
10 independent analyses within IQ-Tree with the number
of unsuccessful iterations to stop (-nstop) set to 1,000 and
perturbation strength (-pers) set to 0.1. The 10 resulting trees
were then used as starting trees for a second set of 10 inde-
pendent analyses within IQ-Tree with the -nstop set to 2,000
and more-thorough nearest-neighbor interchange search
(-allnni). Support for the best-fitting topology of the data-
set was generated using 5,000 Ultrafast bootstrap replicates
(-bb, -wbtl) and reconciled with the most likely phylogeny
using IQ-Tree (-con). Analyses were rooted on species of
Polymixia (see ESM S1–2).
Results
Placement of newly sequenced larvae. The hypothesis of
relationships recovered from our analysis of COI is shown in
Fig.4 and ESM S2 and has a ln L = -27836.101. Out of 147
nodes, 96 (~65%) were supported by a bootstrap value ≥95%
and 133 (~90%) nodes had a bootstrap value ≥70% (see
ESM S2). The resulting topology recovers the specimens
sequenced in this study in the following positions: all three
larval “Pycnocraspedum” are recovered in a clade contain-
ing all specimens of adult Luciobrotula sampled. Of these
“Pycnocraspedum” larvae, the larva from Hawaiʻi (USNM
454562) is recovered in a clade with all adult samples of
Luciobrotula bartschi; the larvae from Florida (USNM
465301, USNM 465380) are recovered in a clade sister to
all samples of adult Luciobrotula coheni. The adult speci-
men P. phyllosoma is recovered sister to samples of P. squa-
mipinne. This clade of Pycnocraspedum is recovered sister
to Neobythites, not monophyletic with any “Pycnocraspe-
dum” larvae sequenced. Larval “Luciobrotula” from Hawaiʻi
(USNM 447052) is recovered in a clade of bythitid genera
Barathronus, Cataetyx, Diplacanthopoma, Paraphyonus,
and Sciadonus. Larval “Luciobrotula” from Florida (USNM
465411) is also recovered in a clade of bythitids, among the
genera Brosmophyciops, Grammonus, and Saccogaster.
Relationships among ophidiiforms. While testing the
monophyly and interrelationships among the Ophidiiformes
are beyond the focus of this study, our phylogeny reflects
previously published multi-locus phylogenies (e.g.,Møller
etal. 2016; Ghezelayagh etal. 2022) with respect to mem-
bers of the Carapidae nested within the Ophidiidae and a
monophyletic Dinematichthyidae. However, there are several
differences between our topology and those from multi-locus
datasets. These include the non-monophyly of Bythitidae,
where Dinematichthyidae is nested within the family, and
members of the former Aphyonidae in a clade separate from
Bythitidae (Fig.4; ESM S2). We also recover the follow-
ing four ophidiiform genera as non-monophyletic: Cataetyx,
Encheliophis, Lepophidium, and Ophidion. These results
may be because our analysis is limited to a single locus. We
do not modify the classification of the Ophidiiformes based
on our results because our recovered relationships should be
tested using additional data.
Revised identification of “Luciobrotula” and “Pyc-
nocraspedum” larvae. Based on our analysis of COI, dis-
tributions of ophidiiform species in the localities where the
larvae were captured, and counts of precaudal vertebrae,
total vertebrae, elongate gill rakers, anal-fin rays, pectoral-
fin rays, and caudal-fin rays, we revise the identification of
the following “Pycnocraspedum” larvae: larva from Hawaiʻi
(USNM 454562; Fig.1) is revised to Luciobrotula bartschi;
larva described and illustrated from New Guinea by Evs-
eenko and Okiyama (2006) and larva from Hawaiʻi (USNM
454451; Fig.1) are revised to L. cf. bartschi; larvae from
Florida (USNM 465301, USNM 465380; Fig.5) are revised
to L. corethromycter. Dorsal-fin-ray counts disagree with
these revised identifications and discrepancies are discussed
below.
For previously identified “Luciobrotula” larvae, we com-
bine two of the four originally described types by Okiyama
(2014) based on overlapping morphological counts and char-
acters (see Discussion section) and revise the identifications
of the following: larvae a and b described and illustrated
for “Luciobrotula” type 1 in Okiyama (2014: 434), larva
described and illustrated for “Luciobrotula” type 4 in Okiy-
ama (2014: 435), and larva from Florida (USNM 465411)
are revised to Bythitidae type 1; larvae described and illus-
trated for “Luciobrotula” types 2 and 3 in Okiyama (2014:
434–435), larva from Hawaiʻi (USNM 447052), and larva
from Japan (USNM 465770) are revised to Bythitidae type
2. Hereafter, we refer to these larvae by their revised names.
General morphology of larval Luciobrotula bartschi.
USNM 454451 and USNM 454562 (Figs.1, 6a–b), post-
flexion. Note: USNM 454451 is identified as L. cf. bartschi
but is included in this section (see Discussion). Counts
(Table1): dorsal-fin rays 102–109; anal-fin rays 68–75;
M. G. Girard etal.
1 3
Table 1 Counts for larval and adult specimens
Species Develop-
mental
stage
Museum cata-
log number
Capture local-
ity
Precaudal
vertebrae
Total verte-
brae
Dorsal-fin
rays
Anal-fin rays Pectoral-fin
rays
Pelvic-fin rays Caudal-fin
rays
Elongate gill
rakers
Luciobrotula
bartschi Adult See 2, 5, 9 Pacific 14–15 (15) 50–55 (53) 85–97 (89) 66–75 (68) 25–28 (26) 2 (2) 11–12 (11) 3–4 (3)
Luciobrotula
bartschi Larva USNM
454562
Pacific,
Hawaiʻi
15 53 102 72 28 2 11 3
Luciobrotula
cf. bartschi Larva See 4 Pacific, New
Guinea
14 or 15 54 102 68 26 2 10? 3
Luciobrotula
cf. bartschi Larva USNM
454451
Pacific,
Hawaiʻi
15 57 109 75 28 2 11 3
Luciobrotula
lineata Adult See 1, 3, 9 Pacific 15 (15) 51–56 (56) 90–92 (92) 67–76 (76) 26 (26) 2 (2) 12 (12) 3 (3)
Luciobrotula
polylepis Adult See 8 Pacific, New
Guinea
(13) (50) (86) (70) (32) (2) (11) (3)
Pycnocraspe-
dum arma-
tum
Adult See 1, 9 Pacific,
Hawaiʻi
12 (12) 47–55 (55) 90–99 (97) 61–72 (71) 26 (26) 2 (2) 10 (10) 4–6 (4)
Pycnocraspe-
dum squami-
pinne
Adult See 7 Indian 12 47–49 63–92 54–79 24 2 10 4
Luciobrotula
corethromy-
cter
Adult See 2, 5, 9 Atlantic 15–16 (16) 53–57 (56) 91–103 (93) 68–77 (70) 26–30 (28) 2 (2) 10–12 (11) 3 (3)
Luciobrotula
corethromy-
cter
Larva USNM
465301
Atlantic,
Florida
16 56 108 77 29 2 11 3
Luciobrotula
corethromy-
cter
Larva USNM
465380
Atlantic,
Florida
16 56 105 69 31 2 10 3
Pycnocraspe-
dum phyl-
losoma
Adult See 9 Atlantic 12 (12) 51–54 96–101 (97) 59–72 (71) 24–26 (26) 2 (2) 10 4
Bythitidae
type 1
Larva USNM
465411
Atlantic,
Florida
12 53 ~76 ~58 U U U U
“Luciobrotula”
type 1 and 4
Larva See 6 Pacific, Japan ? 51–54 77–80 61–62 25 1 or 2 13 ?
Bythitidae
type 2
Larva USNM
447052
Pacific,
Hawaiʻi
~11 ~55 87 70 U 1 13 3
Bythitidae
type 2
Larva USNM
465770
Pacific, Japan 12 55 88 68 U 1 13 3
Larvae of Luciobrotula
1 3
pectoral-fin rays 26–28; pelvic-fin rays 2; caudal-fin rays
10–11; precaudal vertebrae 14–15, total vertebrae 54–57.
Following description based on in-situ images as well as
ethanol and cleared-and-stained specimens. Head large,
nearly as deep as long, body broadly ovoid, tapering pos-
teriorly to a narrow hypural plate. Maxilla and premaxilla
short, at oblique upturned angle. Premaxilla and dentary
with small, distantly spaced teeth. Distal end of maxilla
dorsoventrally expanded, posterior margin convex. Posterior
tip of premaxilla nearly reaching posterior margin of max-
illa. Supramaxilla indistinguishable. Large rostral cartilage
attached to ascending process of premaxilla. Spine associ-
ated with symphysis of dentary. Eight branchiostegals (full
complement). Three elongate gill rakers present on first arch.
Other tooth plates of branchial arches not developed. Body
and head scaleless. Dorsal fin origin over supraoccipital.
Distinct notch between dorsal margin of head and first dor-
sal-fin pterygiophore. Anterior three proximal-middle radi-
als undifferentiated, cartilaginous, and anteriorly directed,
with ventral margin following contour of neurocranium.
First dorsal-fin ray elongate and robust, approaching or
exceeding SL of specimen. In-situ images of fixed speci-
men show the elongate rays were damaged and truncated
during capture and subsequent fixation (compare Figs.1a
with 1b–c, 6a). Accordingly, total length of elongate ray not
reported. Remaining dorsal- and anal-fin rays approximately
subequal in length. Pectoral fin large, fan-like, with broad
base. Four pectoral radials present, cartilaginous. Coracoid
with elongate cartilaginous ventral process, extending to
and associated with ascending loop of gut. Gut massive,
protruding ventrally below the body musculature, lacking
exterilium morphology of some larval ophidiids (e.g., Brotu-
lotaenia, Lamprogrammus, Leptobrotula; Fraser and Smith
1974; Fahay and Nielsen 2003; Okiyama and Yamaguchi
2004). Gut with single, broad intestinal loop in posterior half
of tract, near the 12th-to-13th vertebra. Tissue surrounding
gut vascularized, possibly liver. End of gut loop and anus
extends posteroventrally from gut loop, surrounded by bolus
of tissue, widely separated from body contour. Anus slightly
anterior to anal fin. Long fleshy filament present anterior to
anus, extending from body wall. No ossification present in
filament. In-situ images of fixed specimens show the elon-
gate filament was damaged during capture and subsequent
fixation (compare Figs.1a with 1b–c, 6a). Accordingly, total
length of filament not reported but approaches or exceeds
SL of specimen. Pelvic fin minute, girdle reduced and asso-
ciated with cleithral symphysis. Caudal fin small, without
procurrent rays, middle rays longest. Lower two hypurals
differentiated, upper hypurals as single element.
Melanophores few on upper part of premaxilla and elon-
gate dorsal-fin ray. Melanophores on body minute, almost
entirely restricted to areas near looped gut. Several mel-
anophores near anus. Fleshy filament near anus with dense
Table 1 (continued)
Species Develop-
mental
stage
Museum cata-
log number
Capture local-
ity
Precaudal
vertebrae
Total verte-
brae
Dorsal-fin
rays
Anal-fin rays Pectoral-fin
rays
Pelvic-fin rays Caudal-fin
rays
Elongate gill
rakers
“Luciobrotula”
type 2 and 3
Larva See 6 Pacific, Japan,
and Philip-
pines
? 57 90–92 68–72 25 1 or 2 13 ?
Values listed in paratheses indicate count from holotype
U fin rays undifferentiated, ? value unknown
In-table citations: 1, Gosline 1954; 2, Cohen 1964; 3, Prokofiev 2005; 4, Evseenko and Okiyama 2006; 5, Nielsen 2009; 6 Okiyama 2014; 7, Jayakumar etal. 2021; 8, Wong etal. 2021; 9, see
specimens examined section
M. G. Girard etal.
1 3
melanistic pigmentation, increasing in density distally to
almost completely black tip. First dorsal-fin ray directed
anteriorly when fully flexed. Fleshy filament near anus
trails in straight line behind, appearing rigid. Body almost
completely transparent. Broad blotches of reflectance or
iridescence covering ventral margin of gut, dorsal fin, and
anal fin, with these areas extending onto the body in discrete
bands (white blotches in Fig.1a). At least four broad bands
of reflectance or iridescence present on upper flank, three
on lower flank. These areas not discernable in preserved
specimens.
General morphology of larval Luciobrotula corethro-
mycter. USNM 465301, flexion (Figs.5a–b, 6c–d), and
USNM 465380, postflexion (Figs.5c–d). Counts (Table1):
dorsal-fin rays 105–108; anal-fin rays 69–78; pectoral-fin
rays 29–31; pelvic-fin rays 2; caudal-fin rays 10–11; precau-
dal vertebrae 16, total vertebrae 56–57.
General morphology same as larval L. bartschi (see
above). Differences include: melanophores present on
elongate first dorsal-fin ray, increasing in density distally
to almost completely black tip (may also be present in L.
bartschi but specimens examined missing distal part of fin
ray); fewer areas of reflectance or iridescence along the body
(compare Figs.1a with 5a–c). It is possible that this differ-
ence in reflectance or iridescence is an artifact of different
angles and/or intensities of camera strobes used. However,
given the multiple angles at which the specimens were pho-
tographed, we doubt that camera strobes are the sole reason
for these differences.
General morphology of Bythitidae type 1 larva.
USNM 465411, flexion, (Figs.3a–c; ESM S3), and Okiy-
ama (2014: 434, type 1 figs. a–b; 435 type 4, figs. a–c).
Counts (Table1): dorsal-fin rays 76–80; anal-fin rays 58–62;
pectoral-fin rays 25; pelvic-fin rays 1 or 2; caudal-fin rays
13; precaudal vertebrae 12, total vertebrae 51–54. Follow-
ing description based on in-situ images of USNM 465411
(Figs.3a–b), a cleared-and-stained specimen (ESM S3),
and specimens described and illustrated by Okiyama (2014,
see above). Head large, nearly as deep as long, body com-
pressed, tapering posteriorly to a narrow hypural plate.
Premaxilla and dentary with small, distantly spaced teeth.
Distal end of maxilla dorsoventrally expanded with concave
Fig. 4 Hypothesis of ophidii-
form relationships based on an
analysis of COI. Terminals col-
lapsed to genus level. Voucher
information can be found in
ESM S1. Phylogeny with all
terminals available in ESM S2.
Black, gray, patterned, or white
bars between phylogeny and ter-
minal names note family-level
classification (see Møller etal.
2016; Fricke etal. 2022). We
do not modify the classification
of the Ophidiiformes based on
our results as they are based on
a single locus. Terminals with
blue text highlight larval and
adult taxa targeted in this study.
“*” indicates newly sequenced
adult specimen (see Specimens
Examined). Illustrations associ-
ated with terminals USNM
447052 and USNM 465411 ©
M. Okiyama (2014) reproduced
with permission of Tokai Uni-
versity Press. Illustration associ-
ated with clade of Luciobrotula
based on USNM 454451
Larvae of Luciobrotula
1 3
posterior margin. Posterior tip of premaxilla reaching pos-
terior margin of maxilla. Supramaxilla indistinguishable.
Large rostral cartilage attached to ascending process of pre-
maxilla. Articular with acute spine-like process ventrally.
Seven branchiostegals. Gill rakers not developed on first
arch. Other tooth plates of branchial arches not developed.
Body and head scaleless. At least three supraneurals ante-
rior to first dorsal-fin pterygiophore, one in each interneural
space between first, second, and third vertebrae. Dorsal fin
origin over fifth vertebra. Dorsal- and anal-fin rays approxi-
mately subequal in length. Pectoral fin fan-like, with broad
base, rays undifferentiated. Coracoid with elongate carti-
laginous process, distally wavy, extending to and forming
symphysis at ventral margin of body. Gut bulging ventrally
to slightly below the body musculature, without exterilium
morphology. Single intestinal loop in the posterior part of
the tract, near anus. Gut tissue vascularized, possibly liver.
End of intestinal tract and anus large, slightly anterior to
anal fin, extending posteroventrally from body wall, widely
separated from body contour. Pelvic fin minute, single ray
present, girdle reduced and associated with cleithral sym-
physis. Caudal fin small, fan like.
Melanophores on head concentrated in longitudinal band
from tip of premaxilla to posterior margin of neurocranium.
Few large melanophores above brain. Melanophores on body
concentrated above anus and in vertical band near the caudal
peduncle. Melanophores in vertical band extend onto both
dorsal and anal fin. Few small melanophores speckle body.
Head with strong, horizontal strip of yellow, overlapping
areas of melanophores in fixed specimen (see above). Body
almost completely transparent, with three discrete areas of
color, one white diagonal blotch on flank above anus, one
yellow blotch near caudal peduncle, anus with strong hori-
zontal strip of yellow. Yellow coloration overlapping areas
of dense melanophores in fixed specimen (see above). Dorsal
fin with eight discrete blotches of white, anterior five verti-
cally oriented, posterior three horizontally oriented. Anal
fin with three discrete blotches of white coloration, anterior
most vertically oriented, posterior two horizontally oriented.
Slight dots of reflectance or iridescence covering exterilium
gut, dorsal fin, and anal fin, most dense on exterilium. Anus
highly reflective anterior to anal fin.
General morphology of Bythitidae type 2 larvae.
USNM 447052 and USNM 465770, postflexion (Figs.3d–e,
6e–f), and Okiyama (2014: 434, type 2 figs. a–b; 435 type
3, fig. a). Counts (Table1): dorsal-fin rays 87–92; anal-fin
rays 68–72; pectoral-fin rays 25; pelvic-fin rays 1 or 2; cau-
dal-fin rays 13; precaudal vertebrae 11–12, total vertebrae
55–57. Following description based on in-situ images of
USNM 447052 (Fig.3d), ethanol and cleared-and-stained
specimens, and specimens described and illustrated by Okiy-
ama (2014, see above). Head large, nearly as deep as long,
body compressed, tapering posteriorly to a narrow hypural
plate. Two nostrils, anterior nostril low on snout, slightly
above upper lip, characteristic of the Bythitidae (Nielsen
etal. 1999). Premaxilla with small, distantly spaced teeth.
Fig. 5 Larvae of Luciobrotula
corethromycter from Florida.
a Blackwater photo USNM
465301 captured by R. Collins,
24 June 2021; b preserved
USNM 465301; c blackwater
photo USNM 465380 captured
by N. Deloach, 4 August 2021;
d preserved USNM 465380.
Scale bars = 1 cm
M. G. Girard etal.
1 3
Distal end of maxilla dorsoventrally expanded with concave
posterior margin. Posterior tip of premaxilla nearly reaching
posterior margin of maxilla. Supramaxilla present. Large
rostral cartilage attached to ascending process of premax-
illa. Articular with acute spine-like process ventrally. Eight
branchiostegals. Three elongate gill rakers present on first
arch. Other tooth plates of branchial arches not developed.
Body and head scaleless. Four supraneurals anterior to first
dorsal-fin pterygiophore. Dorsal fin origin over fourth or
fifth vertebra. Dorsal- and anal-fin rays approximately sub-
equal in length. Pectoral fin large, fan-like, with broad base,
rays undifferentiated. Four pectoral radials present. Coracoid
with highly elongate cartilaginous posterior process, extend-
ing along the entire anterior margin of exterilium gut. Distal
tips of processes loop near ventral margin of gut, attaching
separately to ascending portion of intestine, do not form
discrete symphysis. Gut with single, broad intestinal loop
in middle of tract, near base of exterilium. Intestine mus-
cular, extending throughout exterilium in USNM 447052,
contracted in USNM 465770, not extending to exterilium
ventral margin (compare Figs.3d–e with 6e–f). Tissue sur-
rounding gut loop vascularized, possibly liver. Anus above
exterilium, slightly anterior to anal fin, extending poster-
oventral from body wall, widely separated from body con-
tour. Pelvic fin minute, girdle reduced and associated with
cleithral symphysis. Caudal fin small, without procurrent
rays, middle rays longest. Upper two hypurals differentiated
completely, lower two hypurals separated at proximal end.
Melanophores on body and fins few, minute, almost
entirely restricted to medial fins and upper flank (Fig.3e).
Body almost completely transparent. Dorsal fin with three
discrete blotches of pigment. Slight dots of reflectance or
Fig. 6 Cleared-and-stained
larval specimens. a Luciobro-
tula bartschi USNM 454562. b
Focused view of anterior dorsal-
fin rays in larval L. bartschi;
“*” highlights sixth precaudal
vertebra; arrow points to 13th
dorsal-fin ray. c Luciobrotula
corethromycter USNM 465301.
d Focused view of anterior dor-
sal-fin rays in larval L. bartschi;
“*” highlights sixth precaudal
vertebra; arrow points to 14th
dorsal-fin ray. e Bythitidae type
2 USNM 465770, note elongate
coracoid cartilage within
exterilium. f Focused view
of looping coracoid cartilage
within Bythitidae type 2 USNM
465770; arrows highlight distal
end of right cartilage attaching
to intestinal loop. Scale bars =
1 cm
Larvae of Luciobrotula
1 3
iridescence covering exterilium gut, dorsal fin, and anal fin,
most dense on exterilium (Fig.3d). Anus highly reflective
anterior to anal fin.
Discussion
Larvae of Luciobrotula. Despite the bothid-like appearance
of their specimen, Evseenko and Okiyama (2006) correctly
identified their larva from New Guinea as an ophidiiform.
However, counts of precaudal vertebrae and dorsal-fin rays
disagree with their identification of “Pycnocraspedum squa-
mipinne” (see above, Table1). Our morphological exami-
nation and molecular sequencing of newly captured “Pyc-
nocraspedum” larvae from Hawaiʻi and Florida identifies
them as species of Luciobrotula. Phenotypically, counts of
precaudal vertebrae, total vertebrae, elongate gill rakers,
anal-fin rays, pectoral-fin rays, and caudal-fin rays over-
lap between these larvae and adults of Luciobrotula (see
Table1; Cohen 1964; Nielsen 2009). Genotypically, pre-
viously generated COI sequences of adult specimens of L.
bartschi are identical to those generated from our Hawaiian
larva (USNM 454562), and all sequences of this species
analyzed were recovered as monophyletic in our phylog-
eny (Fig.4, ESM S2). One Hawaiian larva was unable to
be sequenced (USNM 454451) and we identify it as L. cf.
bartschi, recognizing that a second species in the genus,
Luciobrotula lineata, also occurs in Hawaiʻi (Gosline 1954).
As these two Hawaiian species are primary differentiated
by the length of the lateral line (see Nielsen 2009), which is
not developed in the larva, it is difficult to confidently assign
USNM 454451 to one of these species. Ultimately, we iden-
tified this larva as L. cf. bartschi until additional specimens
of L. lineata can be captured and examined. We also tenta-
tively revise the identity of the larva described by Evseenko
and Okiyama (2006) to L. cf. bartschi despite two species
in the genus occurring near or in New Guinea. Aside from
its type locality of Hawaiʻi, L. lineata is also known from
Kyushu-Palau Ridge (Prokofiev 2005) and may have a wider
range that includes New Guinea. This species has more cau-
dal-fin rays (12; Prokofiev 2005) than the larva described
by Evseenko and Okiyama (2006). Additionally, Luciobro-
tula polylepis from the Solomon Sea near New Guinea has
fewer precaudal vertebrae (13) and more pectoral fin rays
(32, Wong etal. 2021) than the larva described by Evseenko
and Okiyama (2006, see Table1). Although our sequences
from the Floridian larvae were identical to each other, they
did not match any previously sequenced ophidiiform taxon
and were recovered within the larger clade of Luciobrot-
ula in our phylogeny. Only one species of Luciobrotula, L.
corethromycter, is known from the waters of Florida (Cohen
1964; Nielsen 2009), however, a COI sequence has yet to be
publicly released for this taxon. Counts from the two larval
specimens are within the ranges of precaudal vertebrae,
total vertebrae, anal-fin rays, pectoral-fin rays, and caudal-
fin rays of adult L. corethromycter (see Table1; Cohen
1964; Nielsen 2009). Therefore, we identify these larvae as
L. corethromycter, with the genetic resources generated in
this study being the first available for this taxon.
While we have identified the larvae of L. bartschi and L.
corethromycter using both genotypic and phenotypic char-
acters, characteristics of the larval dorsal fin disagree with
these identifications in three key areas: the number of dorsal-
fin rays are greater in the larvae than the adult (L. bartschi:
102–109 vs. 85–97; L. corethromycter: 105–108 vs. 91–103;
see Table1; Cohen 1964; Nielsen 2009); the dorsal fin origi-
nates above the neurocranium in the larvae compared to the
middle of the abdomen in adults; the elongate first dorsal-
fin ray is present in the larvae but absent in adults. For the
counts and dorsal-fin morphology to agree between develop-
mental stages, the larvae of Luciobrotula would lose ~2–24
anterior dorsal-fin elements through ontogeny. While reduc-
tions of dorsal-fin-ray lengths between larvae and adults are
common among marine fishes broadly (e.g., Bathysauridae,
Bothidae, Carangidae, Macrouridae; Fahay 2007; Okiyama
2014), including members of ophidiiforms (e.g., Brotulo-
taenia, Lamprogrammus, see Fahay and Nielsen 2003),
losses of entire elements and associated posterior shifts of
fins are limited to a few groups. One example of such a
loss occurs the ophidiiform family Carapidae, that lose the
vexillum anterior to the adult first dorsal-fin ray ontogeneti-
cally (Olney and Markle 1979; Markle and Olney 1990).
Nielsen and Evseenko (1989) highlighted an unusual loss
of dorsal-fin elements and consequent posterior shift of fin
origin in the ontogeny of the ophidid Benthocometes robus-
tus. Approximately 10 pterygiophores and elongate anterior
dorsal-fin rays that insert above the neurocranium in larval
B. robustus become lost or reduced, represented by rudi-
mentary ossifications between the posterior margin of the
neurocranium and anterior margin of the dorsal fin in the
adult (compare Nielsen and Evseenko 1989: figs.4 and 6).
The larval dorsal fin of B. robustus is similar to the anterior
dorsal-fin morphology of larval Luciobrotula described here,
but with more elongate anterior rays and fin elements well
ossified (compare Nielsen and Evseenko 1989: figs.3, 4 with
Figs.6a–d). Further, our analysis of COI recovers Benthoc-
ometes as the sister genus to Luciobrotula (Fig.4). Based on
the counts and insertion of the dorsal-fin elements between
larvae and adults and a putative close ally, Benthocometes,
having ontogenetic reductions in dorsal-fin elements, we
hypothesize that an extraordinary number of dorsal-fin ele-
ments are lost in the ontogeny of Luciobrotula. While we
did not find cartilaginous or ossified elements that resembled
pterygiophores between the posterior margin of the neuro-
cranium and anterior-most dorsal-fin pterygiophore in adult
specimens of Luciobrotula, we found the dorsal fin inserts
M. G. Girard etal.
1 3
between the fifth and eighth vertebrae in the adult specimens
(Fig.2). Our cleared-and-stained larvae have 13–14 dor-
sal-fin elements anterior to the sixth vertebra (Figs.6a–d),
17–18 elements anterior to the eighth vertebra, and these
counts are within the range needed to be lost through ontog-
eny to conform with adult dorsal-fin-ray counts (i.e., 2–24;
see Table1). While specimens of transitioning Luciobrotula
will need to be captured and examined to fully understand
the ontogenetic changes occurring in the dorsal fins of these
fishes, this is only the second example of 10+ fin elements
being lost through ontogeny, along with a putative close
ophidiid relative Benthocometes (Nielsen and Evseenko
1989). Our findings call into question the ubiquitous utility
of fin-ray counts when identifying larval ophidiids and we
encourage researchers to use multiple data types to identify
larvae going forward.
Unknown larvae of Pycnocraspedum. Considering the
revisions made above, the larvae of Pycnocraspedum are
now unknown. We recover our newly sequenced P. phyl-
losoma and previously generated sequences of P. squami-
pinne in a distantly related clade from Luciobrotula, sister
to species of Neobythites (Fig.4). The larvae of Neoby-
thites are similar in overall physiognomy to the adult form
(Fahay 2007; Okiyama 2014) and the unknown larvae of
Pycnocraspedum may also be more similar in overall appear-
ance to their adult counterparts. We encourage subsequent
sampling to use both genotypic and phenotypic characters
to correctly identify the larvae of these fishes.
Larval bythitid types 1 and 2. Okiyama (2014) tenta-
tively identified the larvae of “Luciobrotula”, separating
them into four types (types 1–4) largely based on differences
in exterilium-gut length and dorsal- and anal-fin-ray counts.
However, the substantial overlap in counts (see Table1) and
descriptions for these larvae by Okiyama (2014) suggest
they are more similar than different. For example, three of
the four larval descriptions direct the reader to the descrip-
tion of type 1, with the type 4 identification section noting
“See type 1… type 4 shares main characters/counts includ-
ing caudal fin counts (13)” (Okiyama 2014: 435, translated
by an author). The length of the exterilium appears to be the
most important character in differentiating the four types
described by Okiyama (2014). “Luciobrotula” types 1 and
4 (Okiyama 2014) have the shortest guts among the four
types, with only slight bulging or exterilium guts illustrated,
and are listed as having 77–80 dorsal-fin rays, 61–62 anal-
fin rays, 25 pectoral-fin rays, 13 caudal-fin rays, and 51–54
total vertebrae (see Table1). These ranges are all within
expected variation for an ophidiiform species (see Table1
and references therein). Furthermore, the specimens are
listed as having two pelvic-fin rays, but the illustrations show
highly reduced pelvic girdles and a single ray (see Okiyama
2014). Although the newly captured larva (USNM 465411)
is from the Atlantic Ocean—Okiyama’s (2014) larvae were
from the Pacific—the larva has fin-ray and vertebral counts
within the ranges of the types 1 and 4 larvae (see Table1), a
reduced pelvic girdle, and is highly similar in overall physi-
ognomy and pigment. Despite being from different oceans,
we interpret the newly captured larva from Florida, as well
as those in types 1 and 4 described by Okiyama (2014 see
above) to be closely related, possibly of the same genus (i.e.,
Bythitidae type 1). For the remaining specimens we exam-
ined representing larval “Luciobrotula”, we interpret the
larvae captured from Japan (USNM 465770) and Hawaiʻi
(USNM 447052), as well as those in types 2 and 3 described
by Okiyama (2014), to be closely related (i.e., Bythitidae
type 2) given their overlapping counts and overall physiog-
nomy (see Table1; Figs.3d–e, 4, 6e–f).
Although recovered among a larger clade of the Bythiti-
dae, species-level identifications of larval types 1 and 2 are
not possible based on molecular characters and the current
number of ophidiiform barcodes available on public reposi-
tories. Only 66 genera (~54%), and 135 species (~24%) of
ophidiiforms currently have barcodes available on BOLD
or GenBank, with the most-recent targeted molecular study
on ophidiiforms (Møller etal. 2016) not including COI
sequences despite sampling other mitochondrial markers
(i.e., 16s, ND4). Without greatly increasing the number of
taxa sequenced, we cannot confidently identify these and
other ophidiiform larvae using standard barcoding meth-
ods. As for morphological characters, the dorsal- and anal-
fin ray counts of both larval types are not exclusive to any
bythitid in these localities. However, caudal-fin-ray counts in
these larvae are an unusual 13. A few species in the bythitid
genus Tuamotuichthys have 13 caudal-fin rays (Nielsen
and Møller 2008). However, species of Tuamotuichthys are
currently known to only occur in the western and southern
Pacific Ocean. While the genus Parasaccogaster occurs in
both Atlantic and Pacific Oceans, only one species, Par-
asaccogaster normae of the southeast Pacific, is known to
have 13 caudal-fin rays (Nielsen etal. 2012). It is worth
highlighting that the species in both Parasaccogaster and
Tuamotuichthys are represented by one to a few specimens
and the fin-ray counts may be broader in range than cur-
rently known. The 3–4 supraneurals present in these larvae
may also be helpful for identification. Patterson and Rosen
(1989) noted that a single ossified supraneural anterior to
the second neural spine is the primitive condition for the
Ophidiiformes. Bythitids have a variety of conditions, from
no supraneurals (e.g., Dermatopsis and Ogilbia) to as many
as six cartilaginous elements (e.g., Brotulina [currently Din-
ematichthys], Calamopteryx, Grammonus, Lucifuga, and
Ogilbia; Patterson and Rosen 1989; Carnevale and Johnson
2015). Although three of these genera are now classified
in the Dinematichthyidae (i.e., Dinematichthys, Dermatop-
sis, and Ogilbia [see Møller etal. 2016]), our phylogeny
includes representatives from these genera, none of which
Larvae of Luciobrotula
1 3
are recovered closely related to larval Bythitidae types 1
and 2. Without a detailed survey of supraneural and ptery-
giophore patterns across the Bythitidae, the utility of these
larval supraneurals are limited. There is also the possibility
that these larvae represent an undescribed lineage of ophidii-
form, as sampling these fishes from their often deep, rugged,
and rocky habitats is difficult. Although our understanding
of ophidiiform fauna continues to increase, additional work
is needed to fully understand the biodiversity of brotulas,
cusk-eels, pearlfishes, and their larvae.
Fahay and Nielsen (2003) suggested that an early-form-
ing, elongate coracoid process descending along the exter-
ilium gut is a character that supports a sister-group rela-
tionship between Brotulotaenia and Lamprogrammus and
diagnostic to an expanded Brotulotaeniinae. This character
has also been described in the larva of Leptobrotula (Okiy-
ama and Yamaguchi 2004), which was also included in the
expanded Brotulotaeniinae. Both types of bythitid larvae
examined in this study are similar to these brotulotaeniins,
with the cartilaginous coracoid having a ventral process with
a distal tip that extends near the ventral margin of the gut.
In Bythitidae type 2, the cartilaginous coracoid processes
form a loop within the exterilium gut (Figs.3d–e, 6e–f) and
the intestine is differentially expanded or contracted (com-
pare Figs.3d–e with 6e–f). These coracoids and differences
in intestine length have not been previously described. The
tissue surrounding the expanded intestine in USNM 447052
is taut when compared to the loose tissue surrounding the
contracted intestine in USNM 465770 (compare Figs.3d–e
with 6e–f). While we did not find any muscular attach-
ments to the coracoid processes, they are attached to the
otherwise muscular intestine and surrounded by vascular
tissue (Fig.6f) that may represent the liver. Given the dif-
ferences in appearance of the exterilium tissue relative to the
length of the muscular intestine, we suspect that the larvae
can manipulate the exterilium and intestine length through
expansion or contraction. The intestines are stippled in the
illustrations by Okiyama (2014: 434–435), highlighting that
those with the most-elongate gut have an intestinal loop that
reaches the ventral margin of the exterilium, similar to the
larva in Figs.3d–e. Such expansions or contractions would
explain the overall variation in exterilium length observed
in these larvae. Given this, we question the utility of exter-
ilium gut length as a diagnostic feature among these bythitid
type 2 larvae. Further, our phylogeny (Fig.4) shows the
larval elongate coracoid character occurring in multiple
separate lineages of ophidiiforms. As this larval character
was diagnostic for the Brotulotaeniinae, but phylogenetic
hypotheses (e.g., Møller etal. 2016; Ghezelayagh etal.
2022) have found non-monophyly of ophidiiform subfami-
lies recognized by Nielsen etal. (1999), Fahay and Nielsen
(2003), and others, we emphasize that the classification of
the order should be re-evaluated using both morphological
and molecular characters.
Conclusion
In-situ photos and specimens captured by blackwater divers
and photographers allowed for the examination and sequenc-
ing of newly captured ophidiiform larvae, revision of larval
Luciobrotula morphology, and redescription of two larval
bythitids. With newly revealed morphological features in
these larvae, such as the elongate coracoid processes and
the exceptional losses of anterior dorsal-fin elements, and
larvae of many species unknown, now including the larvae
of Pycnocraspedum, we hope that this study will encourage
blackwater divers to continue to capture additional speci-
mens, images, and video footage of ophidiiform larvae to
further understand their morphology and diversity. The pho-
tographs and specimens captured by the divers continue to
increase our understanding of the biology and natural his-
tory of marine fish larvae at an accelerated rate (see Nonaka
etal. 2021; Pastana etal. 2022). However, our ability to
identify these exceptional larvae using molecular characters
is directly related to the accuracy and completeness of bar-
code reference libraries (Pentinsaari etal. 2020; Girard etal.
2022b; Mulcahy etal. 2022; Philips etal. 2022; Lea etal.
2023). Given that less than one quarter of ophidiiform spe-
cies diversity has been barcoded and even fewer additional
mitochondrial and nuclear loci have been made publicly
available to date, this study highlights the continual need for
generating sequences from vouchered museum specimens of
brotulas, cusk-eels, pearlfishes, and allies. The sequences of
L. corethromycter and P. phyllosoma are the first to be gen-
erated for these taxa and data from many other species are
needed. Such efforts will greatly enhance our understanding
of ophidiiform species diversity, their evolutionary history,
and the under-explored morphological diversity of their
larvae.
Specimens Examined Specimens are adults unless oth-
erwise denoted as “Larva-” preceding specimen prepara-
tion type. Specimens examined as cleared and stained are
denoted “CS”; specimens examined as whole ethanol speci-
mens are denoted “ET” with a “*” indicating the specimen
was also x-rayed or scanned using a µCT. Image stacks of
µCT scans have been uploaded to MorphoSource, with asso-
ciated accession numbers listed in brackets following the
preparation type. All measurements listed are SL.
Benthocometes robustus: NHMD P77784, 1 Larva-CS,
20 mm, 27 September 1921, Mediterranean Sea; NHMD
M. G. Girard etal.
1 3
P77785, 1 Larva-CS, 39 mm, 3 October 1921, Alboran Sea;
NHMD P77786, 1 CS, 96 mm, 2 March 1968, Brazil.
“Bythitidae type 1”: USNM 465411, 1 Larva-CS, 9.1
mm, captured and photographed by D. Devers, 18 February
2022, West Palm Beach, Florida.
“Bythitidae type 2”: USNM 465770, 16.6 mm, 1 Larva-
CS, 9 May 1987, Japan; USNM 447052, 24.0 mm, 1 Larva-
ET, captured and photographed by J. Milisen, 13 May 2017,
Kona, Hawaiʻi.
Luciobrotula bartschi: USNM 74151 holotype, 1 ET*
[491478], 260 mm, 27 December 1908, Palawan, Philip-
pines; USNM 179900, 1 ET, 116 mm, 3 June 1909, Samar,
Philippines; USNM 454562, Larva-1 CS, 22.0 mm, photo-
graphed and captured by A. Deloach, N. Deloach, and S.
Kovacs, 11 November 2021, Kona, Hawaiʻi.
Luciobrotula cf. bartschi: USNM 454451, Larva-1 ET,
22.9 mm, captured by S. Yano, 28–29 September 1988,
Kona, Hawaiʻi.
Luciobrotula coheni: USNM 421217, 1 ET, 128 mm, 25
November 2010, Costa Rica; USNM 421356, 1 ET, 208 mm,
19 November 2010, Panama; USNM 421491, 1 ET, 206 mm,
12 November 2010, Panama; USNM 421528, 1 ET, 167 mm,
12 November 2010, Panama; USNM 422550, 1 ET, 99 mm,
24 November 2010, Costa Rica.
Luciobrotula corethromycter: USNM 188547 holotype, 1
ET*, 534 mm, 25 May 1962, Panama; USNM 188548 para-
type, 1 ET, 295 mm, 26 July 1962, Alabama; USNM 188549
paratype, 1 ET* [491484], 310 mm, 14 December 1962,
Florida; USNM 188550 paratype, 1 ET, 390 mm, 31 May
1962, Panama; USNM 188551 paratype, 1 ET*, 500 mm, 22
March 1963, French Guiana; USNM 334068, 1 CS, 58 mm,
10 June 1985; USNM 395816, 1 CS, 158 mm, 2 June 1964,
Columbia; USNM 465301, Larva-1 CS, 15.0 mm, captured
and photographed by R. Collins, 24 June 2021, West Palm
Beach, Florida; USNM 465380, Larva-1 ET, 15.4 mm, cap-
tured and photographed by A. Deloach, N. Deloach, and S.
Kovacs, 5 August 2021, West Palm Beach, Florida.
Luciobrotula lineata: USNM 162716 holotype, 1 ET*,
267 mm, 3 June 1950, Hawaiʻi.
Pycnocraspedum armatum: USNM 162717 holotype,
1 ET* [491490], 302 mm, 2 June 1950, Hawaiʻi; USNM
227411, 1 ET, 124 mm, 20 November 1968, Hawaiʻi; USNM
227412, 2 ET*, 248–250 mm, 24 September 1972, Hawaiʻi;
USNM 395796, 1 CS, 123 mm, 1 November 1967, Hawaiʻi;
USNM 395797, 1 CS, 117 mm, 19 November 1968, Hawaiʻi.
Pycnocraspedum phyllosoma: UF 233512, 1 ET, 94
mm, 23 July 1969, Anegada, British Virgin Islands;
USNM 227388, 1 ET, 124 mm, 19 November 1968, Nic-
aragua; USNM 227413, 1 ET*, 233 mm, 27 May 1965,
Turks and Caicos Islands; USNM 421586, 1 ET, 352 mm,
2013, Curaçao; USNM 421587, 1 ET, 301 mm, 2013,
Curaçao; DEEPEND PC12-B0923-2790-MTSW6-SN-
325A1-PS3661, tissue G176, 90 mm, 23 September 2011,
Gulf of Mexico; YPM ICH 2902 holotype, 1 ET * [491493],
100 mm, 4 April 1927, Turks and Caicos Islands.
Supplementary Information The online version contains supplemen-
tary material available at https:// doi. org/ 10. 1007/ s10228- 023- 00906-4.
Acknowledgements We thank blackwater divers R. Collins, A.
Deloach, N. Deloach, D. Devers, S. Kovacs, and J. Milisen for captur-
ing larval specimens and allowing us to use some of their exceptional
photographs in this publication; L. Ianniello for her continued enthusi-
asm, useful discussions, and contributions to the USNM collections; R.
Smetana and K. Bemis for reading and providing feedback on this man-
uscript; J. Hill (NMNH) and G. Watkins-Colwell, T. Wu, and A. Zhang
(YPM) for assistance with µCT scanning; K. Murphy and D. Pitassy
(USNM) and A. Reft (NOAA National Systematics Laboratory) for
facilitating the loan of specimens; J. Moore (FAU), J. Kojima (Marine
Ecology Research Institute), N. Schnell (MNHN), M. A. Krag and P.
R. Møller (NHMD), A. Bernard, A. Cook, H. Johnson, M. Shivji, and
T. Sutton (Nova Southeastern University), K. Matsuura, M. Nakae, G.
Shinohara (NSMT), D. Kobayashi (Pacific Islands Fisheries Science
Center); R. Robins (UF), C. Baldwin, J. Clayton, K. Murphy, D. Pitassy
(USNM), G. Watkins-Colwell (YPM), and members of the DEEPEND
project for providing data, support, and/or access to specimens and tis-
sues in their care. Finally, we thank W. Schwarzhans (NHMD) and uni-
dentified reviewers for their comments and suggestions on this study.
Extractions and sequencing of DNA were conducted at the NMNH
Laboratories of Analytical Biology. Analyses were conducted using the
Community Cluster at the University of Kansas. MGG, AN, and GDJ
were supported in part by the Herbert R. and Evelyn Axelrod Endow-
ment for Systematic Ichthyology at NMNH. MGG was supported in
part by the NMNH Office of the Associate Director for Science. BCM
was supported by the National Marine Fisheries Service Pacific Islands
Fisheries Science Center. This is Ocean Research Explorations Con-
tribution ORE-13.
Data availability The data generated during and/or analysed in this
article are available in the article, the Supplementary Information,
GenBank, or MorphoSource.
Declarations
Conflicts of interest The authors declare no conflicts of interest.
Ethics approval Species of fishes in this study are not listed as threat-
ened or endangered by the IUCN Red List or CITES. All methods
of capture and preservation conform to the Guidelines for the Use
of Fishes in Research established by the American Fisheries Society,
American Institute of Fishery Research Biologists, and American Soci-
ety of Ichthyologists and Herpetologists. Larvae collected off West
Palm Beach, Florida, were acquired under Florida permit SAL-21-
2155A-SR. Permit not required for larval collections off Kona, Hawaiʻi.
References
Baldwin CC, Mounts JH, Smith DG, Weigt LA (2009) Genetic iden-
tification and color descriptions of early life-history stages of
Belizean Phaeoptyx and Astrapogon (Teleostei: Apogonidae)
with comments on identification of adult Phaeoptyx. Zootaxa
2008:1–22
Bañón R, de Carlos A, Ruiz‐Pico S, Baldó F (2020) Unexpected deep‐
sea fish species on the Porcupine Bank (NE Atlantic): biogeo-
graphical implications. J Fish Biol 97:908–913
Larvae of Luciobrotula
1 3
Campbell MA, Nielsen JG, Sado T, Shinzato C, Kanda M, Satoh
TP, Miya M (2017) Evolutionary affinities of the unfathomable
Parabrotulidae: molecular data indicate placement of Parabro-
tula within the family Bythitidae, Ophidiiformes. Mol Phylo-
genet Evol 109:337–342
Carnevale G, Johnson GJ (2015) A Cretaceous cusk-eel (Teleostei,
Ophidiiformes) from Italy and the Mesozoic diversification of
Percomorph fishes. Copeia 103:771–791
Cawthorn DM, Steinman HA, Corli Witthuhn R (2011) Establish-
ment of a mitochondrial DNA sequence database for the identi-
fication of fish species commercially available in South Africa.
Mol Ecol Resour 11:979–991
Chang CH, Shao KT, Lin HY, Chiu YC, Lee MY, Liu SH, Lin PL
(2017) DNA barcodes of the native ray‐finned fishes in Taiwan.
Mol Ecol Resour 17:796–805
Chen W-J, Santini F, Carnevale G, Chen JN, Liu SH, Lavoué S,
Mayden RL (2014) New insights on early evolution of spiny-
rayed fishes (Teleostei: Acanthomorpha). Front Mar Sci 1:53
Chernomor O, von Haeseler A, Minh BQ (2016) Terrace aware data
structure for phylogenomic inference from supermatrices. Syst
Biol 65:997–1008
Cohen DM (1964) A review of the ophidioid fish genus Luciobrotula
with the description of a new species from the Western North
Atlantic. Bull Mar Sci 14:387–398
Cohen DM, Nielsen JG (1978) Guide to the identification of genera
of the fish order Ophidiiformes with a tentative classification of
the order. NOAA Tech Rep NMFS Circ 417:1–72
Delrieu-Trottin E, Williams JT, Pitassy D, Driskell A, Hubert N,
Viviani J, Cribb TH, Espiau B, Galzin R, Kulbicki M, de Loma
TL, Meyer C, Mourier J, Mou-Tham G, Parravicini V, Plan-
tard P, Sasal P, Siu G, Tolou N, Veuille M, Weight L, Planes S
(2019) A DNA barcode reference library of French Polynesian
shore fishes. Sci Data 6:114
Evseenko SA, Okiyama M (2006) Remarkable ophidiid larva (Neo-
bythitinae) from New Guinean waters. Ichthyol Res 53:192–196
Fahay MP (2007) Early stages of fishes in the Western North Atlan-
tic Ocean (Davis Strait, Southern Greenland and Flemish Cap
to Cape Hatteras). Northwest Atlantic Fisheries Organization,
Dartmouth
Fahay MP, Nielsen JG (2003) Ontogenetic evidence supporting a
relationship between Brotulotaenia and Lamprogrammus (Ophi-
diiformes: Ophidiidae) based on the morphology of exterilium
and rubaniform larvae. Ichthyol Res 50:209–220
Fraser TH, Smith MM (1974) An exterilium larval fish from
South Africa with comments on its classification. Copeia
1974:886–892
Fricke R, Eschmeyer WN, Van der Laan R (eds) (2022) Eschmeyer's
catalog of fishes: genera, species, references. Electronic version,
updated 4 October 2022. http:// resea rchar chive. calac ademy. org/
resea rch/ ichth yology/ catal og/ fishc atmain. asp. Accessed 14 Octo-
ber 2022
Ghezelayagh A, Harrington RC, Burress ED, Campbell MA, Buckner
JC, Chakrabarty P, Glass JR, McCraney WT, Unmack PJ, Thacker
CE, Alfaro ME, Friedman ST, Ludt WB, Cowman PF, Friedman
M, Price SA, Dornburg A, Faircloth BC, Wainwright PC, Near TJ
(2022) Prolonged morphological expansion of spiny-rayed fishes
following the end-Cretaceous. Nat Ecol Evol 6:1211–1220
Girard MG, Davis MP, Baldwin CC, Dettaï A, Martin RP, Smith WL
(2022b) Molecular phylogeny of threadfin fishes (Polynemidae)
using ultraconserved elements. J Fish Biol 100:793–810
Girard MG, Davis MP, Smith WL (2020) The phylogeny of carangi-
form fishes: morphological and genomic investigations of a new
fish clade. Copeia 108:265–298
Girard MG, Davis MP, Tan HH, Wedd DJ, Chakrabarty P, Ludt WB,
Summers AP, Smith WL (2022a) Phylogenetics of archerfishes
(Toxotidae) and evolution of the toxotid shooting apparatus. Integr
Org Biol 4:obac013
Gosline WA (1954) Fishes killed by the 1950 eruption of Mauna Loa
II. Brotulidae. Pac Sci 8:68–83
Howes GJ (1992) Notes on the anatomy and classification of ophidii-
form fishes with particular reference to the abyssal genus Acantho-
nus Günther, 1878. Bull Br Mus (Nat Hist) 58:95–131
Hubert N, Meyer CP, Bruggemann HJ, Guerin F, Komeno RJ, Espiau
B, Causse R, Williams JT, Planes S (2012) Cryptic diversity in
Indo-Pacific coral-reef fishes revealed by DNA-barcoding pro-
vides new support to the centre-of-overlap hypothesis. PLoS ONE
7:e28987
Jayakumar TKT, Murugan A, Kumar ATT, Lal KK (2021) Redescrip-
tion of a rare cusk eel, Pycnocraspedum squamipinne (Actinop-
terygii, Ophidiiformes, Ophidiidae), from Bay of Bengal. Acta
Ichthyol Piscat 51:77–83
Kalyaanamoorthy S, Minh BQ, Wong TKF, von Haeseler A, Jermiin
LS (2017) ModelFinder: fast model selection for accurate phylo-
genetic estimates. Nat Methods 14:587–589
Katoh K, Standley DM (2013) MAFFT multiple sequence alignment
software version 7: improvements in performance and usability.
Mol Biol Evol 30:772–780
Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S,
Buxton S, Cooper A, Markowitz S, Duran C, Thierer T, Ashton
B, Meintjes P, Drummond A (2012) Geneious basic: an integrated
and extend-able desktop software platform for the organization
and analysis of sequence data. Bioinformatics 28:1647–1649
Landi M, Dimech M, Arculeo M, Biondo G, Martins R, Carneiro M,
Carvalho GR, Brutto SL, Costa FO (2014) DNA barcoding for
species assignment: the case of Mediterranean marine fishes.
PLoS ONE 9:e106135
Lara A, Ponce de León JL, Rodriguez R, Casane D, Cote G, Bernatchez
L, García‐Machado E (2010) DNA barcoding of Cuban freshwater
fishes: evidence for cryptic species and taxonomic conflicts. Mol
Ecol Resour 10:421–430
Lea RN, Frable BW, Robertson DR (2023) Misidentification of Ophid-
ion imitator Lea, 1997 as Otophidium indefatigabile Jordan &
Bollman, 1890 (Ophidiiformes: Ophidiidae: Ophidiinae). Zootaxa
5230:95–96
Mabragana E, Díaz de Astarloa JM, Hanner R, Zhang J, Gonzalez
Castro M (2011) DNA barcoding identifies Argentine fishes from
marine and brackish waters. PLoS ONE 6:e28655
Marín A, Gozzer-Wuest R, Grillo-Nuñez J, Alvarez-Jaque IB, Riveros
JC (2022) DNA barcoding reveals overlooked shark and bony fish
species in landing reports of small-scale fisheries from northern
Peru. Mar Fish Sci 35:307–314
Markle DF, Olney JE (1990) Systematics of the pearlfishes (Pisces:
Carapidae). Bull Mar Sci 47:269–410
McCusker MR, Denti D, Van Guelpen L, Kenchington E, Bentzen
P (2013) Barcoding Atlantic Canada's commonly encountered
marine fishes. Mol Ecol Resour 13:177–188
Minh BQ, Schmidt HA, Chernomor O, Schrempf D, Woodhams MD,
von Haeseler A, Lanfear R (2020) IQ-TREE 2: new models and
efficient methods for phylogenetic inference in the genomic era.
Mol Biol Evol 37:1530–1534
Miya M, Takeshima H, Endo H, Ishiguro NB, Inoue JG, Mukai T,
Satoh TP, Yamaguchi M, Kawaguchi A, Mabuchi K, Shirai SM,
Nishida M (2003) Major patterns of higher teleostean phylog-
enies: a new perspective based on 100 complete mitochondrial
DNA sequences. Mol Phylogenet Evol 26:121–138
Moser HG, Richards WJ, Cohen DM, Fahay MP, Kendall AW Jr,
Richardson SL (eds) (1984) Ontogeny and systematics of fishes.
Special Publication No. 1. American Society of Ichthyologists and
Herpetologists, Lawrence
Møller PR, Knudsen SW, Schwarzhans W, Nielsen JG (2016) A new
classification of viviparous brotulas (Bythitidae) – with family
M. G. Girard etal.
1 3
status for Dinematichthyidae – based on molecular, morphological
and fossil data. Mol Phylogenet Evol 100:391–408
Mulcahy DG, Ibáñez R, Jaramillo CA, Crawford AJ, Ray JM, Gotte
SW, Jacobs JF, Wynn AH, Gonzalez-Porter GP, McDiarmid RW,
Crombie RI, Zug GR, de Queiroz K (2022) DNA barcoding of
the National Museum of Natural History reptile tissue holdings
raises concerns about the use of natural history collections and
the responsibilities of scientists in the molecular age. PLoS ONE
17:e0264930
Nielsen JG (2009) A revision of the bathyal genus Luciobrotula
(Teleostei, Ophidiidae) with two new species. Galathea Rep
22:141–156
Nielsen JG, Cohen DM, Markle DF, Robins CR (1999) FAO species
catalogue. Volume 18. Ophidiiform fishes of the world (Order
Ophidiiformes). An annotated and illustrated catalogue of pearl-
fishes, cusk-eels, brotulas and other ophidiiform fishes known to
date. FAO, Rome
Nielsen JG, Evseenko SA (1989) Larval stages of Benthocometes
robustus (Ophidiidae) from the Mediterranean. Cybium 13:7–12
Nielsen JG, Møller PR (2008) New and rare deep-sea ophidiiform
fishes from the Solomon Sea caught by the Danish Galathea 3
Expedition. Steenstrupia 30:21–46
Nielsen JG, Schwarzhans W, Cohen D (2012) Revision of Hastato-
bythites and Saccogaster (Teleostei, Bythitidae) with three new
species and a new genus. Zootaxa 3579:1–36
Nonaka A, Milisen JW, Mundy BC, Johnson GD (2021) Blackwa-
ter diving: an exciting window into the planktonic arena and its
potential to enhance the quality of larval fish collections. Ichthyol
Herpetol 109:138–156
Okiyama M (ed) (1988) An atlas of early stage fishes in Japan. Tokai
University Press, Tokyo
Okiyama M (ed) (2014) An atlas of early stage fishes in Japan, second
edition. Tokai University Press, Hadano
Okiyama M, Yamaguchi M (2004) A new type of exterilium larva
referable to Leptobrotula (Ophidiiformes: Ophidiidae: Neoby-
thitinae) from tropical Indo-West Pacific. Ichthyol Res 51:77–80
Olney JE, Markle DF (1979) Description and occurrence of vexillifer
larvae of Echiodon (Pisces, Carapidae) in the western north-Atlan-
tic and notes on other carapid vexillifers. Bull Mar Sci 29:365–379
Parmentier E, Lanterbecq D, Eeckhaut I (2016) From commensalism
to parasitism in Carapidae (Ophidiiformes): heterochronic modes
of development? PeerJ 4:e1786
Pastana MNL, Girard MG, Bartick MI, Johnson GD (2022) A novel
association between larval and juvenile Erythrocles schlegelii
(Teleostei Emmelichthyidae) and pelagic tunicates. Ichthyol Her-
petol 110:675–679
Patterson C, Rosen DE (1989) The Paracanthopterygii revisited: order
and disorder. In: Cohen DM (ed) Papers on the systematics of
gadiform fishes. Science Series 32. Natural History Museum of
Los Angeles County, Los Angeles, pp 5–36
Pentinsaari M, Ratnasingham S, Miller SE, Hebert PDN (2020) BOLD
and GenBank revisited – Do identification errors arise in the lab
or in the sequence libraries? PLoS ONE 15:e0231814
Pham MH, Hoang DH, Panfili J, Ponton D, Durand JD (2022) Diversity
of fishes collected with light traps in the oldest marine protected
area in Vietnam revealed by DNA barcoding. Mar Biodivers 52:30
Philips MJ, Westerman M, Cascini M (2022) The value of updating
GenBank accessions for supermatrix phylogeny: the case of the
New Guinean marsupial carnivore genus Myoictis. Mol Phylo-
genet Evol 166:107328
Potthoff T (1984) Clearing and staining techniques. In: Moser HG,
Richards WJ, Cohen DM, Fahay MP, Kendall AW Jr, Richardson
SL (eds) Ontogeny and systematics of fishes. Special Publication
No. 1. American Society of Ichthyologists and Herpetologists,
Lawrence, pp 35–37
Prokofiev A (2005) On some rare ophidiiform fishes from the South
Atlantic and Indo - West Pacific, with erection of a new genus,
Megacataetyx gen. novum (Teleostei: Ophidiiformes). Estestven-
nye i Tekhnicheskie Nauki 2:111–128
Richards WJ (ed) (2005) Early stages of Atlantic fishes: an identifica-
tion guide for the Western Central Atlantic. Vol. I & II. CRC
Press, Boca Raton
Robertson DR, Angulo A, Baldwin CC, Pitassy D, Driskell A, Weigt
LA, Navarro IJF (2017) Deep-water bony fishes collected by the
B/O Miguel Oliver on the shelf edge of Pacific Central America:
an annotated, illustrated and DNA-barcoded checklist. Zootaxa
4348:1–125
Sabaj MH (2020) Codes for natural history collections in ichthyology
and herpetology. Copeia 108:593–669
Steinke D, Zemlak TS, Herbert PD (2009) Barcoding nemo: DNA-
based identifications for the ornamental fish trade. PLoS ONE
4:e6300
Ward RD, Holmes BH (2007) An analysis of nucleotide and amino acid
variability in the barcode region of cytochrome c oxidase I (cox1)
in fishes. Mol Ecol Notes 7:899–907
Weigt LA, Baldwin CC, Driskell A, Smith DG, Ormos A, Reyier EA
(2012b) Using DNA barcoding to assess Caribbean reef fish bio-
diversity: expanding taxonomic and geographic coverage. PLoS
ONE 7:e41059
Weigt LA, Driskell AC, Baldwin CC, Ormos A (2012a) DNA bar-
coding fishes. In: Kress WJ, Erickson DL (eds) DNA barcodes:
methods and protocols. Humana Press, Totowa, pp 109–126
Wong M-K, Lee M-Y, Chen W-J (2021) Integrative taxonomy reveals a
rare new cusk-eel species of Luciobrotula (Teleostei, Ophidiidae)
from the Solomon Sea, West Pacific. Eur J Taxon 750:52–69
Publisher's Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Springer Nature or its licensor (e.g. a society or other partner) holds
exclusive rights to this article under a publishing agreement with the
author(s) or other rightsholder(s); author self-archiving of the accepted
manuscript version of this article is solely governed by the terms of
such publishing agreement and applicable law.