J. Phycol.

*, ***–*** (2018)
© 2018 Phycological Society of America
DOI: 10.1111/jpy.12760

MORPHOLOGY AND MOLECULAR PHYLOGENY OF EPIZOIC ARAPHID DIATOMS ON

MARINE ZOOPLANKTON, INCLUDING PSEUDOFALCULA HYALINA GEN. & COMB. NOV.
(FRAGILARIOPHYCEAE, BACILLARIOPHYTA)1

o mez 2
Fernando G
Carmen Campos 3, E-11500, Puerto de Santa Mar
ıa, Spain

Lu Wang
Marine Biodiversity and Global Change Research Center and State Key Laboratory of Marine Environmental Science, Xiamen
University, Xiamen 361005, China
Department of Marine Sciences, University of Connecticut, Groton, Connecticut 06340, USA
Institute of Oceanography, Minjiang University, Fuzhou 350108, China

and Senjie Lin

Marine Biodiversity and Global Change Research Center and State Key Laboratory of Marine Environmental Science, Xiamen
University, Xiamen 361005, China
Department of Marine Sciences, University of Connecticut, Groton, Connecticut 06340, USA

Some diatoms are able to colonize as epibionts on
their potential zooplankton predators. Here, we
report Pseudohimantidium pacificum living on the
copepod Corycaeus giesbrechti and as a new finding on
Oithona nana, Protoraphis atlantica living on the
copepod Pontellopsis brevis, Protoraphis hustedtiana on
the cypris larvae of barnacles, and Falcula hyalina on
the copepod Acartia lilljeborgii. The epizoic diatoms
were able to grow as free-living forms under culture
conditions. Pseudohimantidium pacificum and
P. atlantica appeared as the most derived species
from their benthic diatom ancestors. The mucilage
pad or stalk of the strains of these species showed
important morphological distinction when compared
with their epizoic forms. Barnacle larvae explore
benthic habitats before settlement, and epibiosis on
them is an example where P. hustedtiana profits from
the host behavior for dispersal of its benthic
populations. Molecular phylogenies based on the SSU
rRNA and RuBisCO large subunit (rbcL) gene
sequences revealed F. hyalina as an independent
lineage within the Fragilariales (Tabularia,
Catacombas, and others), consistent with its
morphological distinction in the low number of rows
(≤6) in the ocellulimbus, among other features. We
propose the transfer of F. hyalina to the genus
Pseudofalcula gen. nov. Molecular phylogeny suggests
a single order for the members of the Cyclophorales
and the Protoraphidales, and that the epibioses of
araphid diatoms on marine zooplankton have been
1
Received 17 December 2017. Accepted 24 May 2018. First Pub-
lished Online 16 June 2018.
2
Author for correspondence: e-mail fernando.gomez@
fitoplancton.com.
F. G
omez and L. Wang contributed equally to this paper.
Editorial Responsibility: T. Mock (Associate Editor)
independently acquired several times. These clades
are constituted of both epizoic and epiphytic/
epilithic forms that evidence a recent acquisition of
the epizoic modus vivendi.
Key index words: araphid pennate diatom; DNA
barcoding; epibiont; epibiosis; epiphytic diatom;
symbioses; zoochory
Abbreviations: BP, Bootstrap probability; PP,
Posterior probability
Diatoms are responsible for 40% of marine pri-
mary production, and they are a primary food
source for copepods (Nelson et al. 1995). Copepods
dominate the zooplankton biomass and are consid-
ered the most abundant animal group in the ocean
and an important link between primary productivity
and higher trophic levels (Mauchline 1998). Most of
the studies treat diatoms as a food source for filter
feeders. However, some species of diatoms show a
symbiotic relationship with copepods and other zoo-
plankton groups using the host external surface as a
suitable living environment for the settlement and
growth (Hiromi et al. 1985).
The epizoic diatoms may have negative effects on
the host such as a decrease in growth, mating capa-
bilities, and swimming and an increase in suscepti-
bility to predation (Totti et al. 2011). Proliferations
of epizoic diatoms have been held responsible for
harmful events such as mortality of juvenile lobsters
(Hargraves and Maranda 2002). Some epizoic dia-
toms show a high host specificity and infection rates
(Gonz
alez and Vergara 1984). They have evolved
specialized structures such as mucilage pads or
stalks to adhere to the exoskeleton of the copepods

1
2
EZ ET AL.
F E R N A N D O G OM
(Russell and Norris 1971, Gibson 1978, 1979,
Takano 1983, Hiromi et al. 1985). These diatoms
were first classified under the genera Amphora, Cym-
bella, or Licmophora, which are common epiphytic
genera on seaweeds, rocky, and sediment substrates
(Schr€ oder 1911, Fruchtl 1924, Klevenhusen 1933).
Subsequently, the epizoic diatoms were placed in
new genera (Krasske 1941, Takano 1983). Pseudohi-
mantidium pacificum is known as epizoic on poe-
cilostomatoid copepods (Corycaeus, Farranula;
Hustedt and Krasske 1941); Protoraphis atlantica
mainly colonizes pontellid and candacid copepods
(Labidocera, Pontellopsis; Gibson 1978); Protoraphis
hustedtiana was described as epibiont on copepods,
P. hustedtiana var. nana as epizoic on a snail, and
P. hustedtiana f. latior as epiphyte on macroalgae
(Simonsen 1970, Foged 1984, Takano 1985); Falcula
hyalina mainly colonizes calanoid copepods (Acar-
tia). Russell and Norris (1971) attempted to culture
P. pacificum without success. Furthermore, authors
have suggested that the epizoic diatoms cannot
grow as free-living species or in other substrates,
suggesting an especial metabolic interaction
between the host and epibiont (Hitomi et al. 1985).
In the current taxonomic schemes, the genus
Pseudohimantidium and the two species of Protoraphis
are the only members of the family Protoraphi-
daceae and order Protoraphidales (Simonsen 1970,
Round et al. 1990, Sunesen et al. 2015). The correct
generic classification of F. hyalina has been dis-
cussed (Prasad et al. 1989, Round et al. 1990, Li
et al. 2014). The morphology of the epizoic araphid
diatoms has been illustrated by scanning electron
microscopy (Rivera et al. 1986, Prasad et al. 1989,
Hallegraeff and McWilliam 1990, Lee et al. 1993,
Fernandes and Calixto-Feres 2012, Li et al. 2014,
Sunesen et al. 2015, Donadel and Torgan 2016).
The records on these epizoic diatoms are numerous
in distinct ocean regions (G
arate-Liz
arraga and
Mu~ net
on-G
omez 2009, Padmakumar et al. 2015,
Sahu et al. 2015, G
arate-Liz
arraga and Esqueda-
Esc
arcega 2016). The molecular data are restricted
to sequences of P. pacificum, and one partial SSU
rRNA gene sequence (1,177 bp, base pairs) of P. at-
lantica (EF423413), which are not furnished with
essential information such as illustrations, geo-
graphic origin, or the host types.
The present paper describes the epibiosis of
Pseudohimantidium pacificum, Protoraphis atlantica, and
Falcula hyalina on copepods, and the previously
unreported epibiosis of Protoraphis hustedtiana on
barnacle larvae. This is the first attempt to resolve
the evolutionary origin of these symbiotic diatoms
by molecular methods.
MATERIALS AND METHODS
Sampling, cultures, and microscopy observations. Zooplankton
collected daily from surface waters with a plankton net
(80 lm mesh size) off Ubatuba, S~ao Paulo State, Brazil
(23°320 20.15″ S, 45°50 58.94″ W) in 2014 and 2015. Aliquots of
the zooplankton samples were poured into a petri dish and
examined with a stereoscope microscope. Infected zooplank-
ton was isolated with a micropipette into an Uterm€ ohl cham-
ber with filtered seawater collected that day from the same
locality. Individuals were examined with an inverted micro-
scope (Olympus IX73; Olympus Inc. Tokyo, Japan) and pho-
tographed with a digital camera (Cyber-shot DSC-W300; Sony,
Tokyo, Japan) mounted on the microscope’s eyepiece. The
epizoic diatoms were detached with the tip of a micropipette
and placed individually with a fine capillary into a clean
chamber, and washed several times into a series of drops of
0.2 lm-filtered seawater to remove other organisms. The dia-
tom cells were placed in 12-well tissue culture plate with
0.2 lm-filtered and sterilized seawater supplemented with f/2
medium with silicates and incubated at 23°C, with 100 lmol
photons  m2  s1 from cool-white tubes provided under a
12:12 h L:D cycle. Individual cells were re-isolated and placed
into a 6-well tissue culture plate and cultured in the same
conditions. The diatom cells remained attached to the bot-
tom of the culture chamber. Periodically, the water was
removed from the surface and new culture media was added.
When the diatoms fully covered bottom of the culture cham-
ber, the cells were detached by scratching the surface with a
glass Pasteur pipette and transfer to new culture chambers.
The cultures of Pseudohimantidium pacificum, Protoraphis
spp. and Falcula hyalina were used for infection experiments
on free-living copepods. The copepod genera Corycaeus and
Pontellopsis are carnivorous, and difficult to maintain under
laboratory conditions. Only omnivore-herbivore copepods
such as Acartia or Euterpina were maintained for periods of
10–15 d with microalgae as food resources.
For scanning electron microscopy, the cultures were placed
in filtered seawater with glutaraldehyde (final concentration
of 5%). The sample was filtered onto a 3 lm pore size Nucle-
pore membrane filter, fixed with osmium, dehydrated with a
graded series of ethanol (30%, 50%, 70%, 90%, 95%, 99%,
100%). Filters were mounted on stubs, sputter-coated with
gold and viewed under a field-emission scanning electron
microscope Zeiss Supra 55 (Carl Zeiss Inc., Oberkochen, Ger-
many). Images were presented on a black background using
Adobe Photoshop CS3 (Adobe Systems Inc., San Jose, CA,
USA).
DNA extraction, PCR, and sequencing. For molecular studies,
exponentially growing cultures were harvested with a micro-
pipette and concentrated by centrifugation. Finally, the pellet
was placed in a 1.5 mL Eppendorf tube filled with absolute
ethanol. The sample was kept at room temperature and in
darkness until the molecular analysis could be performed.
Prior to PCR, the sample tube was centrifuged and ethanol
was evaporated by placing the tube overnight in a desiccator
at room temperature. Four hundred microliters of DNA lysis
buffer (10 mM Tris pH 8.0; 0.1 M EDTA pH 8.0; 0.5% SDS;
200 lg  mL1 proteinase K) was used to re-suspend the cell
pellets, the resuspension was transferred into a 1.5 mL tube
and incubated at 56°C for 2 d. Ceramic beads (0.5 mm diam-
eter) were added to the lysate after incubation and the
remaining intact cells were disrupted through a FastPrep-24
bead mill (MP Biomedicals, Irvine, CA, USA) at 6 m  s1 for
1 min. Subsequent steps of DNA extraction followed a CTAB
method coupled with DNA Clean-up & Concentration col-
umn (Zymo Research, Orange, CA, USA; Zhang et al. 2005).
At the end of the extraction process, DNA was eluted in
30 lL 10 mM Tris-HCl buffer (pH 8.0). Fragments of the
SSU rDNA were amplified using the universal primers 18S-
comF1 and 18ScomR1 (Wang et al. 2016). RuBisCO large
subunit (rbcL) sequences were amplified using RubI-forward/
RubI-reverse primers (Wang et al. 2016). The PCR conditions
were 94°C for 1 min, 34 cycles of 15 s at 95°C, 30 s at 56°C,
 EPIZOIC ARAPHID DIATOMS
and 1 min at 72°C, followed by an additional step of exten-
sion at 72°C for 7 min. The resultant amplicons were purified
using Clean & Concentrator column and directly sequenced
with corresponding primers (DNA Analysis Facility, Yale
University, New Haven, CT, USA). Due to a contamination
with a thraustochytrid protist in the culture of Protoraphis at-
lantica, PCR products of SSU rRNA gene sequences were also
cloned into a T-vector (T-vector pMD19; Takara Bio Inc,
Mountain View, CA, USA) and eight clones were picked up
and sequenced. The newly generated sequences were depos-
ited in DDBJ/EMBL/GenBank under accession numbers
MH016662–MH016667 and MH305568–MH305569.
Phylogenetic analyses. Matrices comprising SSU rRNA and
RuBisCO large subunit sequences were, respectively, assembled
from most similar sequences identified using BLAST (http://bla
st.ncbi.nlm.nih.gov/Blast.cgi). Furthermore, available sequences
of araphid diatoms were included, and a sequence from the stra-
menopiles Bolidomonas spp. and Triparma strigata were used as
outgroups. Alignments of the SSU rRNA sequences were per-
formed using SSU-Align package (Nawrocki et al. 2009), which
performs secondary structure alignments based on the covari-
ance model (CM; Cannone et al. 2002, for a diatom example
see Theriot et al. 2009). The sequences were aligned to the con-
sensus CM model of Eukarya integrated in the SSU-Align pack-
age. Alignment columns with low posterior probability (PP
≤0.95), which generally corresponded to large loops for which
positional homology and covarying nucleotides were difficult to
assign, were removed using the SSU Mask routine in the SSU-
Align package. Maximum likelihood (ML) phylogenetic analysis
was performed with a PhyML web server (http://www.atgc-mon
tpellier.fr/phyml/) using the best model TN93+G+I, which was
automatically selected by smart model selection. The branch
supports of the ML tree were inferred using Shimodaira-Hase-
gawa-like (SH) aLRT and were estimated by performing 100
non-parametric bootstrap replicates. The Bayesian inference
(BI) analysis was performed with MrBayes 3.1.2 (Ronquist and
Huelsenbeck 2003) using the evolutionary model GTR+I (0) +G
(0.4564) selected under the AIC criterion by MrModeltest v.2
(Nylander 2004); Markov chain Monte Carlo simulations were
run for 1,000,000 generations, along with at least 200 samples
from the posterior probability distribution, and that diagnostics
are calculated every 1,000 generations. The average standard
deviation of split frequencies was 0.006 and the first 25% of the
states discarded as burn-in.
Alignments of the RuBisCO large subunit sequences were
performed using CLUSTALW (http://www.genome.jp/tools-
bin/clustalw#clustalw.aln) with default alignment parameters
to construct phylogenetic trees using ML and Maximum Par-
simony methods. Maximum Likelihood phylogenetic analysis
was performed with PhyML 3.0 as implemented in SeaView
4.3 (Gouy et al. 2010) using the GTR model. In addition, a
BI of phylogeny was performed with Mr. Bayes 3.1.2 (Ron-
quist and Huelsenbeck 2003), along with 1,200,000 genera-
tions of Markov chain Monte Carlo with trees sampled every
1,000 cycles, the average standard deviation of split frequen-
cies was 0.009 and the first 25% of the states discarded as
burn-in.
RESULTS
Pseudohimantidium pacificum. The most com-
mon epibiosis we found was Pseudohimantidium paci-
ficum on the exoskeleton of the copepod Corycaeus
giesbrechti (Corycaeidae, Poecilostomatoida; Fig. 1, a–
i). To date, P. pacificum is only known as an epibiont
on copepods of the family Corycaeidae (Corycaeus,
Farranula) and the harpacticoid copepod Euterpina
3
acutifrons (see Table S1 in the Supporting Informa-
tion). For the first time, we report P. pacificum on
Oithona nana (Oithonidae, Cyclopoida; Fig. 1, j–l).
Several individuals of O. nana were infected by
P. pacificum in the same plankton tow (Fig. 1j). That
epibiosis was occasional, while that of P. pacificum
on Corycaeus was common, and in some periods was
found in up to 80% of the individuals. The diatom
was more frequently attached to the urosome, the
posterior thoracic appendages and antennae, while
the dorsal cephalosome tended to be free of epi-
bionts (Fig. 1, a–l). The epibiotic burden was mod-
erate, and diatoms did not massively cover the host.
The frustules were attached at foot poles to a
dichotomously branching stalk (Fig. 1, e–f). Frus-
tules showed hyaline spine-like filaments on the
valve surface that resembled the hitchhiker seeds
(Fig. 1, g–i). The chloroplasts in the epibiotic dia-
toms were small, elliptical, and numerous (10–20;
Fig. 1, e–i).
A strain was obtained from epizoic cells on
Corycaeus giesbrechti. Under culture conditions as
free-living forms, the bifurcate stalks were not
formed. The strain grew on the bottom of the cul-
ture chamber forming clusters of dorsoventrally
joined cells, while in the epizoic stage were attached
by one valve pole. The cultured cells showed more
intense pigmentation than in the epizoic stage
(Fig. 1m). The apical axis varied from 35 to 42 lm,
while the transapical axis was 8–10 lm. Valve out-
line was slight sickle-shaped, and some cells showed
a slight gibbous mid-ventral portion (Fig. 1, n–s).
Sternum was narrow, linear, and strongly curved in
one of the valve poles (Fig. 1, t–u). Striae were unis-
eriate, 31–34 in 10 lm, parallel at the valve center
becoming radiate toward the poles (Fig. 1, t–ac).
Poroids were elongate, 2–3 in 1 lm. Each apex
showed several striae composed of smaller poroids
aligned with the rimoportulae at one apex and
bend in the other. There were six or seven rimopor-
tulae juxtaposed to each other and placed parallel
to the ventral side. The rimoportula openings were
fused in a common fissure called apical labiate
groove (Fig. 1, v and w, y and z). The apices showed
a series of 12–16 vertical slits at the mantle edge
(Fig. 1, ad–ae).
Protoraphis atlantica. The diatom was found on
the copepod Pontellopsis brevis (Pontellidae, Cala-
noida). The frustules were mainly placed in the
metasome and in lower number in the caudal setae.
Frustules were attached to the host as single cells or
pairs joined at the valve poles (Fig. 2, a–c). Long
stalks were not observed. Under culture conditions
in the free-living form, the cells developed a thick
and long stalk with distal end attached to the sub-
strate (here the bottom of the culture chamber).
The stalk was hyaline, unbranched, and uniform in
diameter throughout its length. These stalks main-
tained the frustules separated from the bottom of
the culture chamber. Consequently, the frustules
4
EZ ET AL.
F E R N A N D O G OM

FIG. 1. Pseudohimantidium pacificum. (a–i) Light micrographs of cells on the exoskeleton of the copepod Corycaeus giesbrechti. (e–f) Note
the dichotomous stalk ramification. (g–i). The arrows point the hyaline spine-like filaments on the diatom surface. (j–l). Cells on the cope-
pod Oithona nana. (m) Strain isolated from C. giesbrechti. (n–o) Several views of glutaraldehyde-fixed cells. (t–ae). Scanning electron micro-
graphs. (t–u). Valvar view. The arrow points the curved sternum near the apex in one of the poles. The insets show the striae. (v–z) Detail
of the apical labiate groove. (aa–ac). Other cells. (ad–ae) The arrows point the series of vertical slits. Scale bars: a–f, j–l, 100 lm; g–i, m–u,
x, aa–ac, 10 lm; v–w, y–z, ad–ae, 1 lm.
 EPIZOIC ARAPHID DIATOMS
were out of the focal plane of the inverted micro-
scope. After scraping the bottom, the detached frus-
tules retained the long stalks with coiled distal ends
(Fig. 2k). The stalks formed an entangled skein
(Fig. 2d). The frustules formed chains up to three
cells attached at the pole by the mucilage pad
(Fig. 2, e–p). Cells showed four elongate chloro-
plasts (Fig. 2, e–k).
Based on scanning electron microscopy observa-
tions, the sternum was straight and slightly curved
toward the apices. The apical axis was 28–30 lm
and the transapical axis was 6–8 lm. Transapical
striae were finely punctate, parallel throughout the
valve, 34–36 in 10 lm (Fig. 2, q–t). Each apex
showed a labiate groove of rimoportula openings
arranged parallel to the transapical axis. The apical
labiate groove was longer at the foot pole than at
the head pole (Fig. 2, q–t). The generic name Pro-
toraphis refers to the hypothesis that this “raphe-like
slit” could be a precursor of the raphe. The apices
contained a series of 16 vertical slits that increased
in length toward the apex. There were 3–5 rows of
small pores between the apical labiate groove and
the vertical slits (Fig. 2, u–aa). The mucilage pad
extruded from a position between the series of ver-
tical slits and the rows of pores around the apical
labiate groove (Fig. 2, z–aa). The rimoportula in
the internal valve is illustrated in the Figure 2y.
Exceptionally, some frustules showed a sternum
bifurcated at the end of the pole and two apical
labiate grooves (Fig. 2ab). The areolae were ellipti-
cal with one or two short opposed spines or bars
(Fig. 2ac).
Protoraphis hustedtiana. During the warmer sea-
son from November to April, the cypris larvae of bar-
nacles, tentatively Chthamalus (Cirripedia, Crustacea),
showed clusters of an epizoic diatom (Fig. 3, a–f).
The cypris larva is the last planktonic stage of the bar-
nacle before the settlement. The larvae explore the
potential substrates to settle and develop the sessile
adult. Then, the carapace is released during the set-
tlement of the larvae to the rocky surface. The frus-
tules of Protoraphis hustedtiana were placed in the
head shield (carapace) and also in the thoracopods.
The cells were distributed without an apparent spatial
preference. Pairs of cells attached by the valve pole
were observed (Fig. 3, d and e). Bifurcate stalks were
not observed (Fig. 3, a–f).
Under culture conditions as free-living forms, the
strain formed clusters of the cells joined by short
stalks, sometimes forming bifurcate stalks (Fig. 3j).
The cells contained four chloroplasts that were lar-
ger and more pigmented in the free-living than in
the epizoic forms (Fig. 3, g–n). The cell dimensions
were 38–40 lm in the apical axis and 6–8 lm in the
transapical axis. The shape of Protoraphis hustedtiana
was more elongated and with more pointed apices
than Protoraphis atlantica. Protoraphis hustedtiana
showed more heavily silificied valves than the other
epizoic diatoms. The sternum was narrow, straight
5
in the center of the valve and curved (in opposite
directions) at the valve end (Fig. 3, o and p). The
number of striae was 34–40 in 10 lm. The length of
the apical labiate groove was similar in both apices
(Fig. 3, o–q). Some valves showed two apical labiate
grooves in one of the poles (Fig. 3, u and v). The
apical pore field showed a series of eight vertical
slits (Fig. 3, r–v). There were one or two rows of
small circular pores between the apical labiate
groove and the series of vertical slits (Fig. 3, r–v).
Pseudofalcula hyalina gen. & comb. nov. (=Falcula
hyalina). As reported in previous studies, the spe-
cies F. hyalina showed important morphological dif-
ferences with its generic type and other congeneric
species (see Discussion). Falcula hyalina differed
from other Fragilariales in the lower number (≤6)
of rows in the ocellulimbus, among other morpho-
logical features (see Discussion). In the molecular
phylogenies based on the SSU rRNA and RuBisCO
large subunit gene sequences, F. hyalina clustered as
an independent lineage within the Fragilariales. For
these reasons, we propose the transfer of F. hyalina
into the genus Pseudofalcula gen. nov.
Pseudofalcula F. G
omez, Lu Wang & Senjie Lin,
gen. nov.
Diagnosis: Valves arcuate, linear, with round
apices. The apical pore field is located at the end of
both apices. It is composed of a distinctive ocel-
lulimbus with meshwork of the plate set into the
valve mantle and surrounded externally by a rim.
The pore field contains six or less longitudinal rows
of closely packed porelli. There is a single rimopor-
tula, situated at one of the valve poles. There are
several circular pores between the apical pore field
and rimoportula. The cells of Pseudofalcula gen. nov.
have two plate-like chloroplasts, while a single chro-
matophore in Tabularia. The valve striation of the
frustules of Pseudofalcula gen. nov. is simple, while
double-layered valve striation in Catacombas and
Hyalosynedra. The valve of Pseudofalcula gen. nov.
showed one rimoportula, while two rimoportulae in
Catacombas and Hyalosynedra. The apical pore field
showed six or less rows, while the genera Tabularia,
Catacombas, Hyalosynedra, Ctenophora, and Neosynedra
have more than six rows.
Etymology: Pseudo-, Ancient Greek weυdής
(pseudes): false, not genuine, fake. The type species
has been misplaced in the genus Falcula.
Type species: Pseudofalcula hyalina (Takano) F.
G
omez, Lu Wang & Senjie Lin, comb. nov.
Basionym: F. hyalina Takano (1983) (Bull. Tokai
Reg. Fish. Res. Lab. 111: 24, figs. 1, 4–14).
Observations. Dense clusters of P. hyalina cells were
often found associated with individuals of the cala-
noid copepod Acartia lilljeborgii (Acartiidae, Cala-
noida). The clusters of diatom cells were denser in
the urosome, the antenna and the setae of the furca,
and more disperse cells in other parts of the host
body (Fig. 4, a–j). Pseudofalcula hyalina did not
develop a stalk, and the foot pole of each cell
6
EZ ET AL.
F E R N A N D O G OM

FIG. 2. Protoraphis atlantica. (a–p) Light micrographs. (a, c) Infected copepod Pontellopsis brevis. (b) Other individual of P. brevis that
was used to establish the strain. The inset shows a structure exclusive of male copepods. (d–l) Living cultured cells with long stalk after
scraping the culture chamber. (m–p) Glutaraldehyde-fixed cells. (q–ac) Scanning electron micrographs. (q–s) Valve view. (q) The insets
show the apical labiate grooves. (t) Girdle view. Note the apical labiate groove longer at the foot pole than at the head pole. (u–w) Foot
poles with the mucilage pad extrusions. (x) Girdle view of the pole. (y) View of the rimoportulae in the internal valve. (z–aa) Detail of
the head (left) and foot pole (right). Note the apical labiate groove and the vertical slit series. (ab) Atypical apex with bifurcated sternum
and two apical labiate grooves. (ac) Detail of the areolae. Scale bars: c–d, 50 lm; e–m, 10 lm; q–aa, 5 lm; ab–ac, 1 lm.

attached to the host by a mucous material, sometimes 4–5 lm in the transapical axis. The frustules of
forming banana-like bunches (Fig. 4, d and i). Pairs Pseudofalcula hyalina were arcuate with obtuse apices
of cells attached by the poles were not observed. Cell in valve view and the ventral side slightly undulated
dimensions were 25–28 lm in the apical axis and in some valves (Fig. 4j). The frustules were
 EPIZOIC ARAPHID DIATOMS 7

FIG. 3. Protoraphis hustedtiana. (a–f) Light micrographs of cells epizoic on cypris larvae of barnacle. (d–e) The arrows point at the for-
mation of cell pairs. (g–n) Light micrographs of the strain. (j) The arrow points at the bifurcate stalk. (o–v) Scanning electron micro-
graphs. (o–p, r–t) Apex with a single apical labiate groove. (p, u–v) Apex with two apical labiate grooves. Scale bars: a–f, 100 lm; g–q,
10 lm; r–v, 1 lm.
8
EZ ET AL.
F E R N A N D O G OM
rectangular in girdle view (Fig. 4j). Anomalously,
some cultured individuals showed a straight valve
with sigmoid ends (Fig. 4n). Under culture condi-
tions as free-living forms, there were not significant
differences between the free-living and the epizoic
cells. The free-living cells also formed banana-like
bunches of 4–6 cells on the bottom of the culture
chamber (Fig. 4k). The cells showed the two plate-
like chloroplasts, and the pigmentation was more
intense in the free-living cultured cells than in the
epizoic forms (Fig. 4, k and l).
Based on scanning electron microscopy observa-
tions, the valves were weakly silicified and showed a
parallel pattern with two sectors of striae separated
by a wide sternum (Fig. 4, m–o). The transapical
striae were uniseriate parallel in the center and
sometimes radiate at the apices. The areolae were
ellipsoidal, apically oriented (Fig. 4, q and r). A sin-
gle elongate rimoportula opening was located
slightly dorsal in one of the valve poles (Fig. 4,
m–q, s, v). A sunken apical pore field, named ocel-
lulimbus, was located at the ends of both apices ori-
ented toward the ventral side (Fig. 4, r–v). The
ocellulimbus contained three or four rows of closely
packed porelli (Fig. 4, w–y). The ocellulimbus pre-
sumably secrets mucilage allowing the attachment to
the host. There were small circular pores irregularly
arranged near the ocellulimbus that were distinct
from the adjacent areolae (Fig. 4, x and y).
Molecular phylogeny of epizoic araphid diatoms. We
provide the first molecular data set of the species
Protoraphis hustedtiana, Pseudofalcula hyalina gen. &
comb. nov. (=Falcula hyalina), and additional
sequences of Pseudohimantidium pacificum. We also
provide the complete SSU rRNA gene sequence of
Protoraphis atlantica. The available sequence of P. at-
lantica (EF423413) in GenBank of only 1,177 bp was
excluded from the alignment. However, the partial
sequence of P. atlantica (EF423413), from an
unknown host and geographic origin, showed a high
similarity (99%) when compared with our sequence.
In the phylogenetic tree, the SSU rRNA gene
sequences of Pseudohimantidium and Protoraphis spp.
clustered together as a sister group of Lucanicum con-
catenatum with high support (BP, 93; PP, 1). These
genera clustered with Cyclophora spp. and Astrosyne
radiata with high support (BP, 90; PP, 0.99; Fig. 5).
The topology was similar in the RuBisCO large sub-
unit (rbcL) phylogeny where Protoraphis, Pseudohiman-
tidium, and Lucanicum concatenatum clustered
together with strong support (BP, 96; PP, 1). That
clade clustered as a sister group of Cyclophora spp. with
strong support (BP, 97; PP, 1; Fig. 6). In the SSU
rDNA phylogeny, the sequence of P. hyalina clustered
as an independent linage among two clades: one
clade for Synedra, Hyalosynedra, Thalassionema, and
another clade for Tabularia and Catacombas (Fig. 5).
In the rbcL phylogeny, P. hyalina clustered as an inde-
pendent lineage within the clades of the Fragilariales
(Fig. 6).
DISCUSSION
Taxonomical remarks. Our observations support
the generic transfer of Falcula hyalina (Figs. 4–6).
Voigt (1960) described Falcula rogallii, F. media, and
F. semiundulata from the digestive content of an her-
bivorous fish, and he also described F. paracelsiana
from the liquid used to wash algae (Voigt 1961).
These four species are supposed to be epiphytic,
while F. hyalina was described as epibiont on cope-
pods (Takano 1983). In the original description,
Takano compared the dimensions, number of
striae, and areolae of the five congeneric species.
Falcula hyalina was smaller, with a lower number of
striae and higher number of areolae than the con-
generic species. Despite these differences, Takano
placed F. hyalina into Falcula. Geissler and Gerloff
(1963) described apical pore field of F. media as
composed of longitudinal slits and lamellae. Round
et al. (1990) reported that F. hyalina is smaller, with
a wider sternum, and the apical pore field is poroi-
dal (ocellolimbus) rather than series of slits as
reported in the other congeneric species. Prasad
et al. (1989) and Li et al. (2014) compared
F. hyalina with the closer genera and have suggested
a re-classification into another genus. Li et al.
(2014) reported that F. hyalina and Catacombas pos-
sess two plate-like chromophores, while Tabularia
and Hyalosynedra have a single plate-like chro-
mophore. The valve striation of F. hyalina is a sim-
ple valve structure, while double-layered in
Catacombas and Hyalosynedra. Falcula hyalina and Tab-
ularia possess one rimoportula per valve. Snoeijs
(1992) and Kaczmarska et al. (2009) reported frus-
tules of Tabularia fasciculata with normally one rimo-
portula per valve and exceptionally without or with
two rimoportulae per valve. The genera Catacombas,
Hyalosynedra, Ctenophora, and Neosynedra have two
rimoportulae per valve. The valves of these genera
are straight, while F. hyalina possesses arcuate valves.
However, some individuals of our strain showed
straight valves with curved apices in a sigmoid fash-
ion (Fig. 4n). The ocellulimbus is known in Cata-
combas, Tabularia, and Hyalosynedra (Williams and
Round 1986). The main difference is that the ocel-
lulimbus of F. hyalina possesses a lower number of
rows of porelli (≤6) than the species of Tabularia,
Catacombas, or Hyalosynedra. Prasad et al. (1989)
noted 3–5 rows of porelli, similar to our Brazilian
strain (Fig. 4y). Donadel and Torgan (2016)
reported exceptionally six rows in some valves col-
lected from southern Brazil. Morphology and molec-
ular phylogeny based on the SSU rRNA (Fig. 5) and
rbcL (Fig. 6) gene sequences support the proposal
of a new genus, Pseudofalcula gen. nov., for the
F. hyalina.
Sunesen et al. (2015) validated the order Pro-
toraphidales as “members epizoic on marine cope-
pods.” Copepods are not the only hosts of
Pseudohimantidium and Protoraphis (Table S1), and
 EPIZOIC ARAPHID DIATOMS 9

FIG. 4. Pseudofalcula hyalina gen. & comb. nov. (=Falcula hyalina). (a–j). Light micrographs of the cells on several parts of the copepod
Acartia lilljeborgii. (k–l). Light micrographs of the strain. (m–y). Scanning electron micrographs. The arrowhead and arrow point at the
rimoportula opening and ocellulimbus, respectively. (m) Note the ventral margin slightly undulate. The inset shows the apices. (n) Atypi-
cal straight valve with curved apices. (o–p) View of the internal valve. (q–v) Apices of the external valva with or without rimoportulae.
(w–y) Detail of the apex showing the ocellulimbus with 3 or 4 rows of porelli. (x–y) Note the small circular pores around the ocellulim-
bus. Scale bars: a–d, f–l, 20 lm; m, n–o, 5 lm; p–y, 1 lm.
10
EZ ET AL.
F E R N A N D O G OM

FIG. 5. SSU rDNA phylogenetic tree of araphid diatoms. Newly sequenced species are shown in bold type. Numbers after each taxon
name are GenBank accession numbers. Support of branches is shown in the internodes as SH-aLRT values ≥0.5 (left), and the posterior
probabilities ≥0.8 (right) from Bayesian analysis. Scale bar denotes 0.02 substitutions per site.
 E P I Z O I C A R A PH I D D I A T O MS
FIG. 6. RuBisCO large subunit
(rbcL) tree of araphid diatoms.
Newly sequenced species are
shown in bold type. Numbers
after each taxon name are
GenBank accession numbers.
Support of branches is shown in
the internodes as SH-aLRT values
≥0.5 (left), and the posterior
probabilities ≥0.8 (right) from
Bayesian analysis. Scale bar
denotes 0.02 substitutions per
site.
benthic free-living forms have been described
(Foged 1984). Lobban and Ashworth (2014)
emended the order Cyclophorales to encompass the
new genus Lucanicum. In our molecular phyloge-
nies, Lucanicum clustered with a high support as
basal in clade of the Protoraphidales (Figs. 5 and
6). The sequences of species of Cyclophora, Lucani-
cum, and the Protoraphiales clustered together with
a strong support. The orders Cyclophorales and Pro-
toraphidales were proposed in Round et al. (1990).
The molecular data suggest a single order for the
members of the Cyclophorales and the Protoraphi-
dales (Figs. 5 and 6).
Morphological adaptations for the epizoic life. Molecu-
lar data revealed that the epibiosis of araphid dia-
toms on marine zooplankton have been
independently acquired several times (Figs. 5 and
6). Synedra cyclopum may constitute another lineage.
That species is known from freshwater environments
as epibiont of cladoreran and copepod crustaceans,
as well as in the free-living stage (Chiavelli et al.
11
1993, Gaiser and Bachmann 1994). Pseudohimantid-
ium pacificum, P. hyalina, and Synedra cyclopum have
arcuate valves, while straight in Protoraphis. Our
strain of P. hyalina showed straight valves with curved
apices in a sigmoidal fashion (Fig. 4n). Pseudohiman-
tidium and Protoraphis, which clustered together in
the molecular phylogenies, have different shapes.
Consequently, there is no special selection of the
valve shape in epizoic diatoms. The attachment to
the host, using the mucilage pads exuded through
the apical pore fields (Hasle 1974), is an essential
need for the survival of the epizoic diatoms. Pseudohi-
mantidium and Protoraphis showed a series of vertical
slits, and Pseudofalcula has an ocellulimbus. These
structures are similar in the epilithic and epiphytic
relatives. According to Ohtsuka et al. (2011), during
the copepod mating the frustules of P. pacificum
transfer from one to another host, using viscous
secretion released from the series of vertical slits and
this transfer takes ~10 min. After settlement, the
stalk substance is released from the rimoportula
12
F E R N A N D O G OM
openings that form the apical labiate groove (Oht-
suka et al. 2011). Under culture conditions, P. pacifi-
cum did not produce the stalk (Fig. 1, m–s), while
P. atlantica produces long stalks (Fig. 2, d–k). Infec-
tions were not observed in our experiments, except
for P. atlantica. The frustules and the long stalks of
P. atlantica entangled in the caudal appendices of
the copepod. The bottom of the culture chambers
showed fecal pellets with the digested frustules of
the epizoic species.
Protoraphis atlantica and P. hustedtiana clustered
together in the molecular phylogeny (Fig. 5). Pro-
toraphis atlantica is exclusively known as epizoic on
copepods, while P. hustedtiana has been also
described from benthic habitats (Foged 1984). In
our sampling area, P. hustedtiana was commonly
found on the barnacle larvae (Fig. 3, a–f). The
cyprid larva is the last stage before undergoes meta-
morphosis into a sessile barnacle. It is not a feeding
stage; its role is to find a suitable place to settle.
During days to weeks, the cyprid larva explores
potential surfaces with the modified antennules
(Anderson 1994). This is an example of zoochory
where the behavior of the host is used for the dis-
persal of the benthic populations of the diatom. Pro-
toraphis hustedtiana is more elongated, more heavily
silicified and with coarser pores than P. atlantica
(Figs. 2, y–ac, 3, r–v). Pseudofalcula hyalina showed
more weakly silicified valves than its relatives. Sev-
eral clades of araphid diatoms that are able of a
free-living planktonic lifestyle have reduced the
thickness of the valves with lightly silicified frustules
and more numerous small poroids (i.e., Asterionellop-
sis; Kooistra et al. 2009). The reduction of the thick-
ness of the frustules does not affect the sinking of
the epizoic diatoms because they live attached to a
motile host. However, lighter frustules will reduce
the copepod sinking rate and its effort for buoy-
ancy. The morphological adaptations to the epizoic
life could be a higher production of mucilage pads
or stalks, and lighter valves with reduction of the
thickness and increase of the poroids. Pseudohiman-
tidium pacificum also showed a spine-like structures,
or tentative tiny hooks that resemble the hitchhiker
seeds (Fig. 1, g–i). This feature, together with the
long bifurcated stalks, may favor the higher success
of Pseudohimantidium.
Host specificity of the epizoic diatoms. Copepods are
the most abundant animals in the world oceans,
and consequently the most available substrates.
However, the epibioses occur preferentially in some
copepod families (Table S1). The life of a copepod
is finite, and the epibionts must colonize free-living
or younger copepods. In addition, copepods like
other crustaceans shed periodically the exoskeleton.
The ecdysis helps to remove epibionts and ectopara-
sites. Diatoms are photosynthetic organisms that
need to live in the euphotic zone during the day
light. Copepods are known to carry out vertical
migrations. The most common form is the

EZ ET AL.
nocturnal vertical migration, where copepods
ascend to the surface around dusk remaining at the
surface for the night then migrating to depth again
around dawn. This migration does not favor the epi-
zoic diatoms. There is a reverse migration, but it is
rare and restricted to bathypelagic copepods (Bene-
detti et al. 2016). Pseudohimantidium pacificum is
responsible of the most abundant and widespread
epibiosis with Coryceaus as preferential host
(Table S1). In this study, we exceptionally found
epibioses on the copepod Oithona that belongs to
the same order as Coryceaus (Fig. 1, j–l). Gonz
alez
and Vergara (1984) found that P. pacificum infected
67% of the individuals of Corycaeus, while the other
45 species of copepods were not affected. The
epizoic diatoms are found in epipelagic copepods
with a near-surface distribution, and without or
weak vertical migrations such as Corycaeus giesbrechti,
Pontellopsis brevis, and Acartia lilljeborgii. Epibioses are
more common on carnivorous copepods such as
Corycaeus spp., and the calanoid families Candaci-
idae and Pontellidae (Table S1). Studies have
demonstrated that Pseudomantidium colonized new
hosts through the copepod mating (Russell and
Norris 1971, Gonz
alez and Vergara 1984, Ohtsuka
et al. 2011). Gibson (1979) reported that heavily
infected copepods with P. atlantica performed copu-
lation successfully. It is unclear whether the epi-
bioses affect negatively the reproduction of the
copepods. The experiments are difficult because
most of the epibioses involve carnivorous copepods
that are difficult to culture (Benedetti et al. 2016).
The mating behavior of the copepods is an essential
feature for the dispersal of the epizoic diatoms.
Consequently, a combination of factors such as the
upper distribution in the water column and the
mechanism of copula of the copepod will determine
the success of the epibiosis and the host specificity.
Interactions between epizoic diatoms and cope-
pods. Epibiosis is usually considered a commensal
relationship. There are experiments on the ciliate-
copepod epibioses based on cultures of herbivorous
or omnivorous copepods. The presence of ciliate
epibionts significantly affected the swimming activity
of males and, therefore, decreased their ability to
search for and grasp females, leading to reduced
mating success (Souissi et al. 2013, Burris and Dam
2014). In contrast, Ikeda (1977) suggested that
there might be a mutual compensational relation-
ship between epizoic diatoms and marine copepods.
The benefit of the epibiosis for the diatom is clear:
a substrate in the euphotic zone, accessing wider
nutrient space due to the host movement, including
the utilization of nutrients released from host bodies
and/or from captured prey, and the avoidance of
micrograzers (Totti et al. 2011). The epibioses would
not have succeeded if the epibiont burden were
excessive affecting the feeding, predator avoidance,
or mating success of the host. There are reports of
massive infections with harmful effects for the host,
 E P I Z O I C A R A PH I D D I A T O MS
but they are restricted to hypereutrophic environ-
ments such as aquaculture tanks (Hargraves and
Maranda 2002). The negative impacts for the cope-
pod host could be the loss of energy due to increased
drag in swimming and feeding, interference with
host mating and greater susceptibility to visual preda-
tors due to effects of increasing apparent volume
(Willey et al. 1990, McAllen and Scott 2000). One of
the greatest hazards in an epibiont’s life is the dan-
ger of falling victim to predators of the host. The epi-
zoic diatoms may distort the typical shape of the
copepod as a kind of visual camouflage, and the host
may be confused with an aggregation of diatoms.
Other predators may use chemical clues such as the
copepod exudations. The epizoic diatoms may act as
chemical camouflage. The ecological experimenta-
tion has an open field in the study of a prey (diatom)
and predator (copepod) that may evolve into mutu-
alistic symbioses.
F.G. was supported by Brazilian Conselho Nacional de Desen-
volvimento Cient
ıfico e Tecnol
ogico (grant number BJT
370646/2013-14) and a MEL Senior Fellowship at Xiamen
University, China (grant number MELRS1711). L.W. and S.L.
were supported by the Chinese Visiting Scholarship Council
and the Natural Science Foundation of China (grant no.
41330959). This is contribution #2018276 of MEL, Xiamen
University.
REFERENCES
Anderson, D. T. 1994. Barnacles: Structure, Function, Development
and Evolution. Chapman and Hall, London, 357 pp.
Benedetti, F., Gasparini, S. & Ayata, S. D. 2016. Identifying cope-
pod functional groups from species functional traits. J.
Plankt. Res. 38:159–66.
Burris, Z. P. & Dam, H. G. 2014. Deleterious effects of the ciliate
epibiont Zoothamnium sp. on fitness of the copepod Acartia
tonsa. J. Plankt. Res. 36:788–99.
Cannone, J. J., Subramanian, S., Schnare, M. N., Collett, J. R.,
D’Souza, L. M., Du, Y. S., Feng, B. et al. 2002. The compara-
tive RNA web (CRW) site: an online database of comparative
sequence and structure information for ribosomal, intron,
and other RNAs. BMC Bioinformatics 3:2.
Chiavelli, D. A., Mills, E. L. & Threlkeld, S. T. 1993. Host prefer-
ence, seasonality and community interactions of zooplankton
epibionts. Limnol. Oceanogr. 38:574–83.
Donadel, L. & Torgan, L. C. 2016. Falcula hyalina (Fragilariaceae,
Bacillariophyta) from a coastal lagoon, Southern Brazil: an
additional approach on its morphology. Phytotaxa 243:185–9.
Fernandes, L. F. & Calixto-Feres, M. 2012. Morphology and distri-
bution of two epizoic diatoms (Bacillariophyta) in Brazil.
Acta Bot. Bras. 26:836–41.
Foged, N. 1984. Freshwater and littoral diatoms from Cuba. Bibl.
Diatomologica 5:1–243.
Fruchtl, F. 1924. Beitrag zur Kenntnis der qualitativen und quan-
titativen Verbreitung Nordadriatischer Plankton-copepoden
undihrer Epibioten. Sitzungsberichte der Kais, Akademie der Wis-
senschaften in Wien, Abs. 1, 132:195–223.
Gaiser, E. E. & Bachmann, R. W. 1994. Seasonality, substrate pref-
erence and attachment sites of epizoic diatoms on clado-
ceran zooplankton. J. Plankt. Res. 16:53–68.
G
arate-Liz
arraga, I. & Esqueda-Esc
arcega, G. M. 2016. Prolifera-
tion of Falcula hyalina and Cylindrotheca closterium (Bacillario-
phyceae) on copepods in Bah
ıa de La Paz, Gulf of
California, Mexico. Rev. Biol. Mar. Oceanogr. 52:197–201.
13
G
arate-Liz
arraga, I. & Mu~ net
on-G
omez, M. S. 2009. Primer reg-
istro de la diatomea epibionte Pseudohimantidium pacificum y
de otras asociaciones simbi
oticas en el Golfo de California.
Acta Bot. Mex. 88:31–45.
Geissler, U. & Gerloff, J. 1963. Elektronenmikroskopische
Beitr€age zur Phylogenie der Diatomeenraphe. Nova Hedwigia
6:339–52.
Gibson, R. A. 1978. Pseudohimantidium pacificum, an epizoic dia-
tom new to Florida Current (Western North Atlantic Ocean).
J. Phycol. 14:371–3.
Gibson, R. A. 1979. Protoraphis atlantica sp. nov., a new marine
epizoic diatom. Bacillaria 2:109–26.
Gonz
alez, H. & Vergara, L. 1984. Interacci
on ecol
ogica entre la
diatomea epizoica Pseudohimantidium pacificum Hust. &
Krasske 1941 y cop
epodos del g
enero Corycaeus. Rev. Biol.
Mar. 20:77–90.
Gouy, M., Guindon, S. & Gascuel, O. 2010. SeaView version 4: a
multiplatform graphical user interface for sequence align-
ment and phylogenetic tree building. Mol. Biol. Evol. 27:221–
4.
Hallegraeff, G. M. & McWilliam, P. S. 1990. The complex labiate
process of the epizoic diatom Protoraphis hustedtiana Simon-
sen. Nova Hedwigia, Beih. 100:39–45.
Hargraves, P. E. & Maranda, L. 2002. Potentially toxic or harmful
microalgae from the northeast coast. Northeastern Nat. 9:81–
120.
Hasle, G. R. 1974. The “mucilage pore” of pennate diatoms. Nova
Hedwigia, Beih. 45:167–86.
Hiromi, J., Kadota, S. & Takano, H. 1985. Diatom infestation of
marine copepods (Review). Bull. Tokai Reg. Fish. Res. Lab.
117:37–46.
Ikeda, T. 1977. A pelagic marine copepod associated with dia-
toms. Bull. Plankt. Soc. Jap. 24:39–42.
Kaczmarska, I., Ehrman, J. M., Barros, M., Moniz, J. & Davidovich,
N. 2009. Phenotypic and genetic structure of interbreeding
populations of the diatom Tabularia fasciculata (Bacillario-
phyta). Phycologia 48:391–403.
Klevenhusen, W. 1933. Beitrag zur Systematik und Biologie der
Corycaeen nach dem Material der Meteorexpedition. Zool.
Jahrb. Abt. Syst. 64:583–602.
Kooistra, W. H. C. F., Forlani, G. & De Stefano, M. 2009. Adapta-
tions of araphid pennate diatoms to a planktonic existence.
Mar. Ecol. (Berl.) 30:1–15.
Krasske, G. 1941. Die Kieselalgen des Chilenischen Kustenplank-
tons. Arch. Hydrobiol. 38:260–87.
Lee, J. H., Lee, J. Y. & Kim, M. O. 1993. The fine structure of the
marine epizoic pennate diatom Pseudohimantidium pacificum
in Korean coastal waters. J. Oceanol. Soc. Korea 28:202–11.
Li, X. S., Chen, C. P., Liang, J. R., Wu, W. Z. & Gao, Y. H. 2014.
Morphology and occurrence of a marine epizoic diatom Fal-
cula hyalina Takano (Bacillariophyta) in China. Algol. Stud.
145/146:169–79.
Lobban, C. S. & Ashworth, M. P. 2014. Lucanicum concatenatum, n.
gen., n. sp., a benthic marine diatom from Guam, and a less
restrictive diagnosis for Cyclophorales (Bacillariophyta). Mar.
Biodivers. Rec. 7:1–8.
Mauchline, J. 1998. The biology of calanoid copepods. Adv. Mar.
Biol. 33:1–710.
McAllen, R. & Scott, G. W. 2000. Behavioural effects of biofouling
in a marine copepod. J. Mar. Biol. Assoc. U.K. 80:369–70.
Nawrocki, E. P., Kolbe, D. L. & Eddy, S. R. 2009. Infernal 1.0:
inference of RNA alignments. Bioinformatics 25:1335–7.
Nelson, D. M., Treguer, P., Brzezinski, M. A., Leynaert, A. & Que-
guiner, B. 1995. Production and dissolution of biogenic silica
in the ocean: revised global estimates, comparison with
regional data and relationship to biogenic sedimentation.
Global Biogeochem. Cycles 9:359–72.
Nylander, J. A. A. 2004. MrModeltest v2. Program distributed by
the author. Evolutionary Biology Centre, Uppsala University,
Sweden.
Ohtsuka, S., Horiguchi, T., Ganamura, Y., Yamaguchi, A., Shimo-
mura, M., Suzaki, T., Ishiguro, K., Hanaoka, H., Yamada, K.
& Ohtani, S. 2011. Symbiosis of planktonic copepods and
14
F E R N A N D O G OM
mysids with epibionts and parasites in the North Pacific:
diversity and interactions. In Asakura, A. [Ed.] New Frontiers
in Crustacean Biology. Koninklijke Brill NV, Leiden, pp. 1–14.
Padmakumar, K. B., Cicily Thomas, L., Vijayan, A. & Sanjeevan,
V. N. 2015. First report of epizoic diatom Pseudohimanti-
dium pacificum on copepod Corycaeus affinis from North
Eastern Arabian Sea. Indian J. Geomarine Sci. 44:856–9.
Prasad, A. K. S. K., Livingston, R. J. & Ray, G. L. 1989. The mar-
ine epizoic diatom Falcula hyalina from Choctawhatchee Bay,
the Northeastern Gulf of Mexico: frustule morphology and
ecology. Diatom Res. 4:119–29.
Rivera, P. S., Gonz
alez, H. E. & Barrales, H. L. 1986. Cingulum
and valve morphology of Pseudohimantidium Hustedt &
Krasske (Bacillariophyceae). Phycologia 25:19–27.
Ronquist, F. & Huelsenbeck, J. P. 2003. MRBAYES 3: Bayesian
phylogenetic inference under mixed models. Bioinformatics
19:1572–4.
Round, F. E., Crawford, R. M. & Mann, D. G. 1990. Diatoms: Biol-
ogy and Morphology of the Genera. Cambridge University Press,
Cambridge, 747 pp.
Russell, D. J. & Norris, R. E. 1971. Ecology and taxonomy of an
epizoic diatom. Pac. Sci. 25:357–67.
Sahu, G., Panigrahi, S., Mohanty, A. K., Achary, M. S., Samantara,
M. K., Bramha, S. N., Satpathy, K. K. & Dovgal, I. V. 2015.
New record of an epizoic diatom, Pseudohimantidium pacifi-
cum on two species of copepods from the Indian Ocean.
Indian J. Geomarine Sci. 44:1331–4.
Schr€
oder, B. 1911. Adriatisches Phytoplankton. Sitzungsberichte
der Kais. Akademie der Wissenschaften in Wien, Abt I, 120:601–
57.
Simonsen, R. 1970. Protoraphidaceae, eine neue Familie der Dia-
tomeen. Nova Hedwigia, Beih. 31:377–94.
Snoeijs, P. J. M. 1992. Studies in the Tabularia fasciculata complex.
Diatom Res. 7:313–44.
Souissi, A., Souissi, S. & Hwang, J. S. 2013. The effect of epibiont
ciliates on the behavior and mating success of the copepod
Eurytemora affinis. J. Exp. Mar. Biol. Ecol. 445:38–43.
Sunesen, I., Romero, S., Toubes, E. & Sar, E. A. 2015. Morphol-
ogy and distribution of three araphid diatoms (Fragilario-
phyceae, Bacillariophyta) from marine coastal waters of
Argentina. Iheringia S e r. Bot. Porto Alegre 70:265–78.
Takano, H. 1983. New and rare diatoms from Japanes marine
waters -XI. Three new species epizoic on copepods. Bull.
Tokai Reg. Fish. Res. Lab. 111:23–35.

EZ ET AL.
Takano, H. 1985. A small form of the marine diatom Protoraphis
hustedtiana, epizoic on a snail. Bull. Tokai Reg. Fish. Res. Lab.
117:31–3.
Theriot, E. C., Cannone, J. J., Gutell, R. R. & Alverson, A. J. 2009.
The limits of nuclear-encoded SSU rDNA for resolving the
diatom phylogeny. Eur. J. Phycol. 44:277–90.
Totti, C., Romagnoli, T., Stefano, D. M., Camillo, D. C. G. &
Bavestrello, G. 2011. The diversity of epizoic diatoms: rela-
tionships between diatoms and marine invertebrates. In Seck-
bach, J. & Dubinsky, Z. [Eds.] All Flesh Is Grass, Cellular
Origin, Life in Extreme Habitats and Astrobiology, vol. 16.
Springer, Dordrecht, pp. 323–43.
Voigt, M. 1960. Falcula, un nouveau genre de diatom
ees de la
M
editerran
ee. Rev. Algol. nov. ser. 5:85–8.
Voigt, M. 1961. Sur le genre Falcula. Rev. Algolog. nov. ser. 6:53–6.
Wang, L., Lin, X., Goes, J. I. & Lin, S. 2016. Phylogenetic analyses
of three genes of Pedinomonas noctilucae, the green endosym-
biont of the marine dinoflagellate Noctiluca scintillans, reveal
its affiliation to the order Marsupiomonadales (Chlorophyta,
Pedinophyceae) under the reinstated name Protoeuglena noc-
tilucae. Protist 167:205–16.
Willey, R. L., Cantrell, P. A. & Threlkeld, S. T. 1990. Epibiotic
euglenoid flagellates increase the susceptibility of some zoo-
plankton to fish predation. Limnol. Oceanogr. 35:952–9.
Williams, D. M. & Round, F. E. 1986. Revision of the genus Syne-
dra Ehrenb. Diatom Res. 1:313–39.
Zhang, H., Bhattacharya, D. & Lin, S. 2005. Phylogeny of
dinoflagellates based on mitochondrial cytochrome B and
nuclear small subunit rDNA sequence comparisons. J. Phycol.
41:411–20.
Supporting Information
Additional Supporting Information may be
found in the online version of this article at the
publisher’s web site:
Table S1. Records of marine araphid epizoic
diatoms on marine zooplankton.