Toward a revision of the bamboo corals: Part 3, deconstructing the Family Isididae

Abstract

Bamboo corals are distinguished from most other octocorals by an articulated skeleton. The nodes are proteinaceous and sclerite-free while the internodes are composed of non-scleritic calcium carbonate. This articulation of the skeleton was thought to be unique and a strong synapomorphy for the family Isididae. Our phylogeny, based on the amplification of mtMutS and 18S, shows an articulating skeleton with sclerite-free nodes has arisen independently at least five times during the evolutionary history of Octocorallia rather than being a synapomorphy characteristic of a monophyletic bamboo coral clade. The family Isididae is currently composed of four subfamilies (Circinisidinae, Isidinae, Keratoisidinae, and Mopseinae). Not only is the family polyphyletic, but our genetic analyses suggest also the subfamily Isidinae is polyphyletic based on current taxonomic classifications, and Mopseinae is not monophyletic. The type, Isis, is found outside of the well-supported Calcaxonia – Pennatulacea clade where the other members of Isididae cluster. The current classification of the family Isididae does not reflect the evolutionary history of an articulated skeleton. To better reflect the evolutionary history of these taxa we propose that three of the four the subfamilies, the genus Isidoides, and genera within the subfamily Isidinae, be elevated to family level to produce a classification with five families with a bamboo-like skeleton: Chelidonisididae, Isididae, Isidoidae, Keratoisididae, and Mopseidae.

Full text

ZOOTAXA
ISSN 1175-5326 (print edition)
ISSN 1175-5334 (online edition)
Accepted by M. Daly: 5 Sept. 2021; published: 30 Sept. 2021 247
Zootaxa 5047 (3): 247–272
https://www.mapress.com/j/zt/
Copyright © 2021 Magnolia Press Article
https://doi.org/10.11646/zootaxa.5047.3.2
http://zoobank.org/urn:lsid:zoobank.org:pub:072B07D8-324A-412E-A76E-C39067AC77AE
Toward a revision of the bamboo corals: Part 3, deconstructing
the Family Isididae
ESPRIT HEESTAND SAUCIER1, SCOTT C. FRANCE2 & LES WATLING3*.
1Faculty of Science, Brigham Young University - Hawaii, 55-220 Kulanui St, Laie, Hawaii, 96762, USA
�
esprit.saucier@byuh.edu; https://orcid.org/0000-0002-2156-0926
2Department of Biology, University of Louisiana at Lafayette, Lafayette, Louisiana, USA
�
scott.france@louisiana.edu; https://orcid.org/0000-0002-2114-5968
3School of Life Sciences, University of Hawaii at Manoa, Honolulu, Hawaii 96822
�
watling@hawaii.edu; https://orcid.org/0000-0002-6901-1168
*Corresponding author:
�
watling@hawaii.edu
Abstract
Bamboo corals are distinguished from most other octocorals by an articulated skeleton. The nodes are proteinaceous and
sclerite-free while the internodes are composed of non-scleritic calcium carbonate. This articulation of the skeleton was
thought to be unique and a strong synapomorphy for the family Isididae. Our phylogeny, based on the amplification of
mtMutS and 18S, shows an articulating skeleton with sclerite-free nodes has arisen independently at least five times during
the evolutionary history of Octocorallia rather than being a synapomorphy characteristic of a monophyletic bamboo
coral clade. The family Isididae is currently composed of four subfamilies (Circinisidinae, Isidinae, Keratoisidinae,
and Mopseinae). Not only is the family polyphyletic, but our genetic analyses suggest also the subfamily Isidinae is
polyphyletic based on current taxonomic classifications, and Mopseinae is not monophyletic. The type, Isis, is found
outside of the well-supported Calcaxonia – Pennatulacea clade where the other members of Isididae cluster. The current
classification of the family Isididae does not reflect the evolutionary history of an articulated skeleton. To better reflect
the evolutionary history of these taxa we propose that three of the four the subfamilies, the genus Isidoides, and genera
within the subfamily Isidinae, be elevated to family level to produce a classification with five families with a bamboo-like
skeleton: Chelidonisididae, Isididae, Isidoidae, Keratoisididae, and Mopseidae.
Key words: Octocorallia, Isididae, Chelidonisididae, Isidoidae, Keratoisididae, Mopseidae, phylogenetic analysis
Introduction
The term “bamboo coral” commonly refers to the family Isididae in the cnidarian anthozoan subclass Octocoral-
lia. Bamboo corals are colonial animals with an internal skeleton made up of alternating calcareous internodes and
sclerite-free proteinaceous nodes. The nodes can be slightly flexible while the internodes are rigid. The alternating
nodes and internodes give the skeleton the impression of a bamboo stalk, hence the common name for the family
(Figure 1). Linnaeus was the first to formally name and describe the jointed skeleton when describing Isis hippuris
Linnaeus, 1758, the type species of the family. Bamboo corals are found in all the major ocean basins, from as shal-
low as 3 m depth (Rowley et al. 2015) to over 4700 m (Lapointe and Watling, 2015), although the majority of the
species live in the deep sea (>200 m). Most bamboo corals attach to hard rocky substrates via a holdfast; this limits
their distribution to habitats such as seamounts, rocky outcrops, canyon ledges, and other areas of ocean floor where
hard substrate is exposed. However, there are some species that have adapted to living anchored in soft sediments,
e.g. Acanella arbuscula (Johnson, 1862), Lepidisis caryophyllia Verrill, 1883.
Isidids have long been recognized because of their unique skeleton. Following Linnaeus, the Isididae quickly
became a repository for a wide range of jointed-skeleton coral species and included bamboo corals currently clas-
sified in the genera Mopsea Lamouroux, 1816, Isidella Gray, 1857, and Primnoisis Studer, 1887, as well as taxa

HEESTAND SAUCIER ET AL.
248 · Zootaxa 5047 (3) © 2021 Magnolia Press
now in genera of the scleraxonian family Melithaeidae (Melithaea Milne Edwards and Haime 1857). Although me-
lithaeids also have a jointed skeleton of alternating flexible nodes and rigid internodes, they are distinct from isidids
in that both nodes and internodes are composed of sclerites. However, classification on the basis of skeleton articu-
lation has remained an important character for diagnosing the isidids in the intervening two and a half centuries. In
1812, Lamouroux placed the genus Isis into the family Isidae (= Isididae) along with Melitaea (=Melithaea) and
then added Mopsea in 1816. Lamarck (1836), in “Histoire Naturelle des Animaux sans Vertèbres,” recognized the
family but included only Isis and Mopsea. Milne Edwards and Haime (1857) included within the family Gorgonidae
(=Gorgoniidae) the subfamily Isidinae, comprising the genera Isis, Mopsea, and Melithaea. Gray (1857) erected the
suborder Lithophyta, composed of two families: Isidae (=Isididae) and Melitaeadae (=Melithaeidae). In the Isidae,
Gray included Isis, Mopsea, and a newly created genus, Isidella Gray, 1857, and in the Melitaeadae he included
Melitaea and Solanderia (since shown to be a hydroid) and the new genus Mopsella. Kölliker (1865) retained the
family Gorgoniidae and the subfamily Isidinae. A new genus, Keratoisis, was added to the subfamily Isidinae by
Wright (1869), who also moved Mopsea out of the subfamily. Gray (1870) recognized four families that included
bamboo corals, grouped in an un-named division of his suborder Lithophyta: Mopseadae (Mopsea), Acanelladae
(Acanella Gray, 1870; Isidella), Keratoisidae (Keratoisis), and Isidae (Isis); however, Studer (1878) recognized
only Isidae. Verrill (1883) placed Ceratoisis (=Keratoisis), Acanella, Isidella, Callisis (a genus name not adopted
by subsequent workers; = Keratoisis) and Lepidisis in the family Ceratoisidae and left Isis as the sole genus in the
family Isidae. Studer (1887) continued to recognize Isidae, which he divided into three subfamilies: Ceratoisidi-
nae (=Keratoisidinae), Primnoisidinae and Isidinae; later, Wright and Studer (1889) replaced Primnoisidinae with
Mopseinae, Kükenthal (1915) added Muricellisidinae and Grant (1976) added Circinisidinae and Peltastisidinae.
Kükenthal (1915) added the Muricellisidinae for a small, spiky colony fragment collected off Japan. The genus
Muricellisis was discussed by Bayer & Stefani (1987) and Bayer (1990), who retained the subfamily Muricel-
lisidinae, but this subfamily was not recognized in Williams and Cairns (2019) “Systematic List of Valid Octo-
coral Genera”, and instead the genus was included in the subfamily Isidinae (http://researcharchive.calacademy.
org/research/izg/OCTOCLASS.htm). Recently, Watling (2020) showed that the two species assigned to the genus
Muricellisis were misidentified members of the Anthothelidae and Melithaeidae. Grant (1976) also proposed a fifth
subfamily, Peltastisidinae, for the genera Peltastisis, Chathamisis, and Minuisis, but Alderslade (1998) determined
that Peltastisidinae could not be sustained because the diagnostic feature of the subfamily (specifically, an opercu-
lum of 8 scales) was not correct. Kükenthal (1915) corrected the family name from Isidae to Isididae, and Bayer
(1956) corrected the spelling of Keratoisidinae. The monographs on the Gorgonacea by Kükenthal (1919, 1924) are
the most comprehensive studies of Isididae, although Bayer commented on different aspects of isidid classification
in a series of papers (Bayer and Stefani, 1987a,1987b; Bayer, 1990).
Thus, there are four subfamilies currently accepted within the Isididae: Isidinae, Circinisidinae, Keratoisidinae,
and Mopseinae (Bayer, 1956; Alderslade, 1998; Williams and Cairns, 2001,-2019, unpublished; Alderslade and
McFadden, 2012), and two genera (viz., Australisis and Caribisis) that are unassigned (Bayer and Stefani, 1987;
although in Williams and Cairns online “Systematic List of Valid Octocoral Genera” they are classified as Kera-
toisidinae). However, to date there have been no phylogenetic studies that examine the monophyly of the family
Isididae or attempt to understand the relationships of the subfamilies to each other.
A few bamboo corals have been included in broader phylogenetic analyses of Octocorallia (e.g. McFadden
et al. 2006) and there have been genetic studies focused within the subfamily Keratoisidinae (e.g. France, 2007;
Brugler and France, 2008; Van der Ham, 2009; Dueñas and Sanchez, 2009; Alderslade and McFadden, 2012). Mc-
Fadden et al. (2006) used mitochondrial (mt) genes ND2 and mtMutS (formerly referred to as msh1) to construct a
phylogeny of 103 genera representing twenty-eight families of octocorals, including two keratoisidin specimens,
which were the only representatives from Isididae used in the analysis. Using the gene ITS2 to examine intraspecific
variation, Rowley et al. (2015) found Isis to be phylogenetically more related to species in the Alcyoniidae than to
other bamboo corals that were also included as part of the outgroup. In an examination of the evolution of corals
and their calcareous skeletons through deep time, Quattrini et al. (2020), using ultra-conserved elements (UCEs),
found the genus Isis allied with Rumphella in a clade far removed from species in the Keratoisidinae. Within the
Keratoisidinae, Brugler and France (2008) discovered a novel mitochondrial genome arrangement compared to
other known octocoral mt genomes (more recently, Hogan et al (2019) reported the same arrangement is seen in
some species of the sea pen, Anthoptilum). Van der Ham et al. (2009) explored the utility of a mt intergenic region
(igr4) as a genetic barcode for Isididae instead of the commonly used cox1 gene (http://www.boldsystems.org/) be-

DECONSTRUCTING THE ISIDIDAE Zootaxa 5047 (3) © 2021 Magnolia Press · 249
cause of the >67-fold increase in genetic diversity seen in that region compared to cox1. They found the use of igr4
yielded fewer haplotypes than the mitochondrial mismatch repair gene mtMutS, although when both gene regions
were used in tandem an increase in haplotype diversity was revealed. Alderslade and McFadden (2012), describing
a new keratoisidin genus, Jasonisis, used mtMutS and cox1 to build a phylogeny of thirty-one specimens from the
subfamilies Keratoisidinae (n=27) and Mopseinae (n=4), and presented revised descriptions of these subfamilies.
There have been numerous papers using morphology (e.g. Bayer and Stefani, 1987a; Etnoyer, 2008) and/or genetics
(e.g. Watling and France, 2011; Dueñas et al. 2014; Watling, 2015) to describe new species in the subfamily Kera-
toisidinae or Mopseinae (Moore et al. 2016), but none that examined the Isididae as a whole.
Our objective was to use DNA sequence data to build a phylogeny to evaluate the monophyly of the Isididae
and the relationship of the component subfamilies to each other and to other members of the Octocorallia. Addition-
ally, we explore the use of alternative characters, such as the mt genome arrangement, as a diagnostic tool for the
subfamilies. The analysis incorporated sixty-one taxa from across the subclass Octocorallia, including taxa from all
four currently accepted isidid subfamilies (Circinisidinae, Isidinae, Keratoisidinae, Mopseinae). This paper is the
third in our series revising the bamboo corals. Part 1 redescribed and reassigned to other families the two species
in the subfamily Muricellisidinae (Watling, 2020) and part 2 redescribed the genus Lepidisis (Watling and France,
in press). Part 3 (this paper) will elevate the existing subfamilies to family level and create new families for two
genera. Part 4 will detail the clades comprising the family Keratoisididae and future papers will describe new genera
within the Keratoisididae.
Material and methods
Taxon sampling. Over 400 samples were collected world-wide by the authors and colleagues using deep submer-
gence vehicles (DSV), remotely operated vehicles (ROV), dredges, or by SCUBA. The condition of the specimens
varied based on the method of collection. Collected specimens were photographed and then preserved for genetic
analysis in 95% or higher ethanol; voucher specimens for morphological analysis collected after 2004 were fixed in
4% formalin when possible and then stored in 80% ethanol. Additional samples were contributed by colleagues or
procured from museum collections; details can be found in Table 1.
DNA extraction, PCR amplification, and sequencing. DNA was extracted from preserved specimens using a
CTAB protocol (Berntson and France, 2001) or using the MasterPureTM DNA Purification kit (Epicentre® Biotech-
nologies). DNA stocks were diluted for polymerase chain reaction (PCR) to a working concentration of ≈40ng/µL.
The 5’ and 3’ ends of the mitochondrial gene mtMutS (formerly referred to as msh1; see Bilewitch et al. 2011)
and the middle of the nuclear 18S gene were targeted for amplification and sequencing. The 5’ and 3’ ends of mt-
MutS refer to the first ~700 bp (hereafter mtMutS -5’) and last ~800 bp (mtMutS -3’), respectively, of the ≈3 kbp-
long mtMutS gene. For both of these targets, amplicons extended across the mtMutS gene boundary to the adjacent
gene. The middle of 18S (hereafter simply 18S) refers to a ~600bp region of the nuclear ribosomal 18S gene cor-
responding to positions 568-1194 of GenBank Acc. #EF622534, Keratoisidinae sp. BAL208-1. PCRs were used to
amplify the gene regions with primers specifically designed to target those regions (Table 2). Half-volume (25µL)
reactions were used for all PCR amplifications following the manufacturer’s protocols. Because species in the sub-
family Keratoisidinae have a different mitochondrial gene arrangement in the upstream region of mtMutS compared
to other octocorals (Brugler and France, 2008), a forward primer anchored in the cox3 gene was used to amplify
the mtMutS-5’ rather than the nad4L primer used to amplify the same region for other taxa; phylogenetic analyses
used only the mtMutS sequence and not the upstream gene region sequence. TaKaRa Ex TaqTM polymerase and as-
sociated reagents (TaKaRa Bio Inc.) were used for most PCRs. For PCR using poor-quality DNA (e.g. very highly
sheared), Restorase® DNA Polymerase (Sigma-Aldrich Chemicals) was used. PCR products were cycle-sequenced
using BigDye Terminator v.1.1 cycle sequencing kit (Applied Biosystems) following manufacturer’s protocols with
the same primers as used in the PCR reaction, then visualized on the ABI PRISM® 3130xl Genetic Analyzer at the
University of Louisiana at Lafayette, or sent to a commercial facility (e.g. Beckman Coulter Genomics). All raw
DNA sequence traces were examined and edited using Sequencher v. 4.7 (Gene Codes). All novel haplotype se-
quences derived from this work, and haplotypes from taxa not yet represented in online databases, were submitted
to GenBank (Table 3).

HEESTAND SAUCIER ET AL.
250 · Zootaxa 5047 (3) © 2021 Magnolia Press
FIGURE 1. Morphological characteristics of Isididae sensu lato. Scale bars represent 1 mm unless otherwise indicated. A)
Isidinae: Isis hippuris skeleton with tissue removed. B) Isis hippuris branchlet with tissue and polyps. C) SEM images of Isis
hippuris sclerites. D) Circinisidinae: Zignisis sp. skeleton with tissue removed. E) Zignisis sp. branchlet covered with tissue
and polyps. F) SEM images of Zignisis sp. sclerites. G) Mopseinae: Minuisis sp. skeleton with tissue removed. H) Minuisis sp.
branchlet covered with tissue and polyps. I) SEM images of Minuisis sp. sclerites. J) Keratoisidinae: Keratoisis grayi skeleton
with tissue removed. K) Keratoisis grayi branchlet covered with tissue and polyps. L) SEM images of Keratoisis grayi scler-
ites.

DECONSTRUCTING THE ISIDIDAE Zootaxa 5047 (3) © 2021 Magnolia Press · 251
TABLE 1. Collection location data for specimens used in this study. Taxa for which only DNA sequences (mtMutS-5’) were available (and no physical specimen) are shown by
GenBank accession number and an asterisk (*). Sequences downloaded from GenBank and not generated by the authors may not have collection information available.
CMNI = Canadian Museum of Nature; DFO = Fisheries and Oceans Canada – Offshore Benthic Ecology Lab, Bedford Institute of Oceanography; MAGNT = Museum and Art
Gallery of the Northern Territory; MNHN = Museum National d’Histoire Naturelle, Paris; NIWA = National Institute of Water and Atmospheric Research (New Zealand); OCDN
= Coral Reef Research Foundation (Palau); RMNH = Netherlands Centre for Biodiversity; SAM = South African Museum; SBMNH = Santa Barbara Museum of Natural History;
UF = Florida Natural History Museum; USNM = National Museum of Natural History (Smithsonian Inst.); YPM = Yale Peabody Museum.
Taxon LAB ID or GenBank number Voucher Number Collection date Depth (m) Latitude Longitude
Acanthogorgiidae:
Acanthogorgia sp. BI1043 USNM 94442 25-Aug-1993 1295 18.8135 -159.0587
Calcigorgia sp. 62242C1 USNM 1133587 6-Jul-2004 96 51.9717 -173.9465
Alcyoniidae:
Alcyonium digitatum GQ342466* SBMNH 360700 17-Jul-1991 5 54.064 -4.738
Heteropolypus ritteri DQ302816* RMNH Coel. 40802
Parasphaerasclera rotifera GQ342472* UF 3890
Anthothelidae:
Anthothela grandiflora OCE1025 YPM IZ 044560 10-Sep-2001 814 40.2836 -68.1187
Arulidae:
Arula petunia JX203774* USNM 1178392 19-Mar-2008 21 -30.2811 30.8092
Chrysogorgiidae:
Chrysogorgia artospira KEL4072 YPM IZ 38596 19-May-2004 2253 38.7830 -63.9628
Helicogorgia flagellata SAMH 3929 SAM-H3929 8-Jul-1985 125 -31.1100 30.3000
Iridogorgia magnispiralis KEL4032 YPM IZ 038580 19-May-2004 2311 38.7791 -63.9628
Metallogorgia melanotrichos MAN7062 YPM IZ 044562 15-May-2004 1847 38.1488 -61.1022
Clavulariidae:
Phenganax parrini GQ342490* MAGNT C015597
Telestula sp. DQ302802* MAGNT C014984 758 -26.4323 167.1812
Coralliidae:
Hemicorallium ducale CR2022 USNM 94456 29-Aug-1993 1370 18.6450 -158.2800
Hemicorallium niobe KEL2092 YPM IZ 028622 16-Jul-2003 1859 38.8498 -63.9262
Ellisellidae:
Ellisella schmitti JAS238 21-Jul-2007 13.8
Ellisella sp. DFH118B YPM IZ 44566 13-Sep-2005 99.97 27.8707 -93.6176
......continued on the next page

HEESTAND SAUCIER ET AL.
252 · Zootaxa 5047 (3) © 2021 Magnolia Press
TABLE 1. (Continued)
Taxon LAB ID or GenBank number Voucher Number Collection date Depth (m) Latitude Longitude
Junceella fragilis NC024181*
Funiculinidae:
Funiculina sp. Fel808611 24-Aug-2008 875-884 27.5996 -92.7690
Gorgoniidae:
Leptogorgia virgulata LV10212 YPM IZ 44567 12-Apr-2001 4.5 32.7780 -79.8850
Antillogorgia bipinnata DQ640646*
Incertae sedis
Isidoides armata haplotype A CP38321 MNHN-IK-2009–1632 8-Sep-2011 645 -21.9930 167.1200
Isidoides armata haplotype B TER20516 MNHN-IK-2008-1510 17-Oct-2008 840-800 -23.9583 169.7217
Isididae: Circinisidinae:
Gorgonisis sp. C14412 MAGNT C014412 22-May-2007 292-330 -29.2187 159.0075
Plexipomisis sp. C12715 MAGNT C012715 16-May-2004 72 -32.9228 130.7260
Zignisis sp. C012403 MAGNT C012403 23-Oct-1996 20 -31.9833 115.5000
Zignisis sp. C013153 MAGNT C013153 18-Mar-2001 15 -33.95 134.2667
Isididae: Isidinae:
Chelidonisis sp. 1 KC788274* USNM 1478212 27-Aug-2009 246 27.4229 -87.4029
Chelidonisis sp. 2 LII10402 USNM 1478373 19-Oct-2010 539 27.5979 -91.8250
Isis hippuris OCDN9247L OCDN 9247L 25-Oct-2005 3
Isis sp. OCDN9770F OCDN 9770F 16-Sep-2008 4
Isididae: Keratoisidinae:
Acanella sp. MIL1011 YPM IZ 044537 18-Aug-2009 1688 34.8178 -50.5062
Eknomisis dalioi KEL4061 YPM IZ 044542 17-Jul-2007 1859 38.8498 -63.9262
Isidella sp. 1 BAL2081 YPM IZ 044539 2-Sep-2009 1815 39.4147 -65.4108
Cladarisis sp. 1 JAC2016 YPM IZ 44568 17-Nov-2010 482 30.0809 -79.9494
Isidella sp. 3 MAN1052 YPM IZ 028549 14-Jul-2003 1510.2 38.2648 -60.5474
Cladarisis sp. 2 P42251 YPM IZ 044562 29-Nov-2013 1385 23.0491 -163.1542
Isidella sp. 5 R133902 DFO R1339-2 17-Jul-2010 2346 48.2285 -43.9254
Jasonisis sp. KEL6042 YPM IZ 044543 31-Aug-2005 2554 38.7583 -64.0915
Keratoisis sp. 1 BEA5051 YPM IZ 034868 12-May-2004 1478 39.8842 -67.4718
Keratoisis grayi 5663 CMNI 2021-0118 5-Dec-2009 612 44.81 -55.642
......continued on the next page

DECONSTRUCTING THE ISIDIDAE Zootaxa 5047 (3) © 2021 Magnolia Press · 253
TABLE 1. (Continued)
Taxon LAB ID or GenBank number Voucher Number Collection date Depth (m) Latitude Longitude
Keratoisis sp. 2 PIC1051 YPM IZ 044545 29-Oct-2009 1970 39.1538 -65.9483
Lepidisis sp. 1 KEL1041 YPM IZ 028700 16-Jul-2007 2015 38.7887 -64.1315
Lepidisis sp. 2 MAN7071 YPM IZ 044563 16-May-2008 1826 38.1487 -61.1018
Lepidisis sp. 3 PIC1072 YPM IZ 044546 29-Oct-2009 1946 39.6533 -65.9483
Orstomisis sp. P56961 YPM IZ 044564 9-Nov-2011 1498 24.3996 -166.0727
Isididae: Mopseinae:
Minuisis sp. TER10038 MNHN-IK-2008-1753 26-Oct-2012 410-390 -22.2498 167.2170
Notisis elongata TER11082 MNHN-IK-2008-1783 28-Oct-2012 370-440 -23.0300 168.3550
Primnoisis sp. TER13018 31-Oct-2012 320-360 -22.7302 167.2758
Lissopholidisis nuttingi TER5015 MNHN-IK-2008-1590 21-Oct-2012 240-260 -23.6705 168.0000
Sclerisis sp. NIWA38301 NIWA 38301 5-Mar-2012 1145-1200 -67.7233 -179.7122
Nephtheidae:
Dendronephthya mollis HQ694725*
Paragorgiidae:
Paragorgia coralloides BB1 USNM 98785 29-Oct-1995 1950 12.7333 -102.6000
Parisididae:
Parisis sp. 1 TER7013 MNHN-IK-2008-1595 22-Oct-2012 300-320 -23.3848 168.0232
Parisis sp. 2 TER10044 MNHN-IK-2008-1786 25-Oct-2012 260 -22.2983 167.1533
Pennatulidae:
Ptilosarcus gurneyi 599113B3 1-Jul-2003 60 51.9088 -177.2126
Plexauridae:
Paramuricea cf hawaiiensis BI1032 USNM 94441 25-Aug-1993 1370 18.8135 -159.0587
Swiftia sp. J2095293 YPM IZ 044566 25-Jul-2004 843 51.8114 -173.8345
Primnoidae:
Narella dichotoma LAD12 USNM 98831 20-Sep-1996 1451 20.7826 -157.1489
Paracalyptrophora hawaiinensis PBS09 USNM 98823 19-Sep-1996 427 20.9896 -157.3195
Thouarella grasshoffi MAN8081 USNM 1078188 16-May-2004 1461 38.1457 -61.0914
Protoptilidae:
Protoptilum sp. BI1013 USNM 94465 25-Aug-1993 1440 18.8135 -159.0587
Victorgorgiidae:
Victorgorgia nuttingi BI1071 USNM 94435 25-Aug-1993 1010 18.8135 -159.0587

HEESTAND SAUCIER ET AL.
254 · Zootaxa 5047 (3) © 2021 Magnolia Press
TABLE 2. Primers used in successful PCR and cycle sequencing. Most primer combinations were selected to cross a gene boundary, shown in the first two columns.
Primers that were previously published are denoted by footnotes.
Primer 1 Primer 2
Gene 1 Gene 2 Name Sequence Name Sequence Annealing Temperature
cox1 12S COILA9312F gcggcttgacacccckatgttgtgggctattgg 12S12935R acgacggccatgcaatacc 53°C
12S nad1 12S12935F ggtattgcatggccgcgt nd1Bam11227R atcagctagtacgaacctggct 52°C
12S nad1 12S-306F gcaaacaggattagataccc nd1Bam11227R atcagctagtacgaacctggct 50-51°C
nad1 nad6 nd13209F1ggtrcgagcttctttcccyag nd64699R1ggattcggcgtagatataacc 60-62°C
nad1 nad4 nd13209F1ggtrcgagcttctttcccyag nad4gen4974F2taggyttatttactcatacwat 49-50°C
nad1 cob nd13209F1ggtrcgagcttctttcccyag cybBAM1279R ggttcctctactggattggctcct 48-50°C
cob nad6 cybGen402f gattaccggaatattcttagctatgc nd64699R1ggattcggcgtagatataacc 61°C
nad6 nad4L nd6BAM1878F gctaagcagccaaatwcacct nd4Lp6109R ccataatggctaagccgatag 50°C
nad6 nad3 nd6BAM1878F gctaagcagccaaatwcacct nd3genR aggmtcaaayccacaytcataarc 56°C
nad6 nad3 nd6BAM1878F gctaagcagccaaatwcacct ND32280R ataacaacccaatatccaaa 48-50°C
nad3 nad4L nd3genF gyttatgartgtggrtttgakcc nd4Lp6109R ccataatggctaagccgatag 53-55°C
nad3 nad4L nd32280F tttggatattgggttgttat nd4Lp6109R ccataatggctaagccgatag 45-49°C
nad4L mtMutS nd4L2475F2tagttttactggcctctac mutChry3458R3tgaagyaaaagccactcc 48-55°C
cox3 mtMutS CO3Bam5657F2gctgctagttggtattggcat mutChry3458R3tgaagyaaaagccactcc 50-55°C
mtMutS 16S MutS4759F4tgtagctcatgatattag 16S5PR4tcacgtccttaccgatag 50-55°C
16S nad2 16S874F cttctggctgctgcaaag nd2genR gcccacatatgaaanggagc 51°C
nad2 nad5 nd2genF gctccntttcatatgtgggc nd5Bam2494R gggtcwtctctctcatataacttgta 58-59°C
nad2 nad5 nd2BAM1248F gtgttaggggcactwtcttcgggg nd5BAM2638R ctgccttgttagcttgtatacgtg 50°C
nad5 nad4 nd5Bam3091F agttattgcttattctacttgtagtca nd4Bam4479R cagctttatyttctctgccc 58-59°C
......continued on the next page

DECONSTRUCTING THE ISIDIDAE Zootaxa 5047 (3) © 2021 Magnolia Press · 255
TABLE 2. (Continued)
Primer 1 Primer 2
Gene 1 Gene 2 Name Sequence Name Sequence Annealing Temperature
nad5 nad4 nd53gen13736F agtaayttacaatctggtytagtt nd4gen4953R atwgtatgagtaaataarccta 47-50°C
nad5 nad4 nd5-136F tatgttatgaaatccttattaatggg nd4-1092R agaaaataggggcatagtcataac 48-50°C
nad4 cox3 nd4Bam5446F gagttarttcgtgtatgtttgcct CO3Bam5657F2gctgctagttggtattggcat 61°C
nad4 nad4L nd42599F5gccattatggttaactattac nd4Bam5446F gagttarttcgtgtatgtttgcct 50°C
cox3 cob CO3Bam5657F2gctgctagttggtattggcat cytb3741R1gtccgttttcatgtagcttcc 55°C
atp6 cox3 ATP66458F cctaaccgatggcaagc CO3Bam5876R gcttgtagggctgtaaaytgaacc 50-53°C
atp6 cox3 ATP6-285F gtatttaccccaactgccc CO3Bam5876R gcttgtagggctgtaaaytgaacc 50°C
cox2 atp6 CO2Bam7585F atggaygctgtrcctggacgtc atp66458R gcttgccatcggttagg 50°C
cox2 atp6 CO2Bam7585F atggaygctgtrcctggacgtc atp6-392R ccactaggcatcatcatt 49-51°C
cox2 cox3 CO2Bam7585F atggaygctgtrcctggacgtc CO3Bam5876R gcttgtagggctgtaaaytgaacc 58°C
cox2 cox1 COII8068xF6ccataacaggrctwgcagcatc COIoctR6atcatagcatagaccatacc 47-50°C
cox2 cox1 COII8068xF6ccataacaggrctwgcagcatc CO1p269R gaaaggccatatcgggtgcac 51°C
18S 18S 18S-Cf7cggtaattccagctccaatag 18S-Yr7cagacaaatcgctccaccaac 52-55°C
1Pante et al. 2013; 2 Brugler & France, 2008; 3 Pante et al. 2012; 4 France, 2007; 5 France & Hoover, 2002; 6 McFadden et al. 2011; 7 Apakupakul et al. 1999.

HEESTAND SAUCIER ET AL.
256 · Zootaxa 5047 (3) © 2021 Magnolia Press
TABLE 3. GenBank accession numbers for specimens used in the tree; numerical values show the unaligned sequence length used in the analyses. DNA sequences generated for
this study have GenBank numbers with prefix KX. ‘NS’ indicates not sequenced for that gene region.
mtMutS 18S
Taxon LAB ID 5’ end 3’ end GenBank ID 18S GenBank ID
Acanthogorgiidae:
Acanthogorgia sp. BI1043 939 956 AY268461 674 AF052907
Calcigorgia sp. 62242C1 939 NS GU563308 674 FJ389255
Alcyoniidae:
Alcyonium digitatum GQ342466 939 956 GQ342466 NS
Anthomastus ritteri DQ302816 939 NS DQ302816 NS
Eleutherobia rotifera GQ342472 939 NS DQ302803 NS
Anthothelidae:
Anthothela grandiflora OCE1025 927 NS DQ297415 NS
Arulidae:
Arula petunia JX203774 939 NS JX203774 NS
Chrysogorgiidae:
Chrysogorgia artospira KEL4072 724 NS GQ180132 NS
Helicogorgia flagellata SAMH 3929 901 NS JN227929 674 JN227971
Iridogorgia magnispiralis KEL4032 939 956 JN227997 674 FJ526216
Metallogorgia melanotrichos MAN7062 940 956 GQ868340 NS
Clavulariidae:
Phenganax parrini CSM-NB2 939 NS GQ342490 NS
Telestula sp. NTMC 14984 939 NS DQ302803 NS
Coralliidae:
Hemicorallium ducale CR2022 939 956 EU293805 674 AF052919
Hemicorallium niobe KEL2092 847 642 EF060051 674 FJ643592
Ellisellidae:
Ellisella schmitti JAS238 939 956 JN227995 NS
Ellisella sp. DFH118B 939 956 JN227994 NS
Junceella fragilis NC024181 939 956 NC024181 NS
Funiculinidae:
......continued on the next page

DECONSTRUCTING THE ISIDIDAE Zootaxa 5047 (3) © 2021 Magnolia Press · 257
TABLE 3. (Continued)
mtMutS 18S
Taxon LAB ID 5’ end 3’ end GenBank ID 18S GenBank ID
Funiculina sp. Fel808611 939 NS JN227941 674 JN227965
Gorgoniidae:
Leptogorgia virgulata AY268458 939 956 AY268458 674 AY268464
Antillogorgia bipinnata DQ640646 939 956 DQ640646 NS
Incertae sedis
Isidoides armata haplotype A CP38321 939 915 KX362340 616 KX362353
Isidoides armata haplotype B TER20516 939 956 JN227946 674 JN227972
Isididae: Circinisidinae:
Gorgonisis sp. C14412 939 NS JN227917 NS
Plexipomisis sp. C12715 939 NS EU268052 NS
Zignisis sp. C012403 939 NS EU268053 NS
Isididae: Isidinae:
Chelidonisis sp. 1 LII09170 939 NS KC788274 NS
Chelidonisis sp. 2 LII10402 939 NS KX362350 NS
Isis hippuris OCDN9247L 800 NS JN383338 NS
Isis sp. OCDN9770F 699 NS KX362351 NS
Isididae: Keratoisidinae:
Acanella sp. MIL1011 939 956 NC011016 674 GU206549
Eknomisis dalioi KEL4061 939 904 EU268066 629 KX362288
Isidella sp. 1 BAL2081 882 956 EF622534 457 FJ358837
Cladarisis sp. 1 JAC2016 939 951 KX362341 672 KX362354
Isidella sp. 3 MAN1052 939 956 EU293798 674 GU188958
Cladarisis sp. 2 P42251 939 934 KX362343 371 KX362358
Isidella sp. 5 R133902 939 920 KX362365 642 KX362360
Jasonisis sp. KEL6042 937 956 EF060022 NS
Keratoisis sp. 1 BEA5051 939 696 GU933629 NS
Keratoisis sp. 2 PIC1051 939 956 JN228007 674 GU188957
Lepidisis sp. 1 KEL1041 939 947 EU268024 665 KX362355
Lepidisis sp. 2 MAN7071 931 956 EF060036 518 KX362356
......continued on the next page

HEESTAND SAUCIER ET AL.
258 · Zootaxa 5047 (3) © 2021 Magnolia Press
TABLE 3. (Continued)
mtMutS 18S
Taxon LAB ID 5’ end 3’ end GenBank ID 18S GenBank ID
Lepidisis sp. 3 PIC1072 918 889 EF060039 674 FJ358838
Orstomisis sp. P56961 939 921 KX362366 672 KX362359
Isididae: Mopseinae:
Minuisis sp. TER10038 939 947 KX362344 NS
Notisis elongata TER11082 939 920 KX362346 616 KX362361
Primnoisis sp. TER13018 839 943 KX362347 425 KX362362
Lissopholidisis nuttingi TER5015 939 936 KX362348 672 KX362363
Sclerisis sp. NIWA38301 939 900 KX362342 194 KX362357
Nephtheidae:
Dendronephthya mollis HQ694725 939 956 HQ694725 NS
Paragorgiidae:
Paragorgia coralloides BB1 939 NS JX128350 NS
Parisididae:
Parisis sp. 1 TER7013 930 NS KX362352 638 KX362364
Parisis sp. 2 TER10044 939 933 KX362345 NS
Pennatulidae:
Ptilosarcus gurneyi 599113B3 939 NS JN227904 NS
Plexauridae:
Paramuricea cf. hawaiiensis BI1032 939 956 EU293799 NS
Swiftia sp. J2095293 936 NS GU563302 674 FJ389258
Primnoidae:
Narella dichotoma LAD12 939 705 EF060048 661 HM590862
Paracalyptrophora hawaiinensis PBS09 931 NS DQ297429 NS
Thouarella grasshoffi MAN8081 939 NS GQ868334 674 FJ526218
Protoptilidae:
Protoptilum sp. BI1013 939 956 EU293804 674 AF052911
Victorgorgiidae:
Victorgorgia sp. BI1071 939 NS GU563313 674 FJ389266

DECONSTRUCTING THE ISIDIDAE Zootaxa 5047 (3) © 2021 Magnolia Press · 259
Mitochondrial gene map. Nine specimens (one Circinisidinae, three Isidinae, three Keratoisidinae, and two
Mopseinae) were selected to PCR amplify across all gene boundaries in the mitochondrial genome. We hypoth-
esized that the mt genomes would consist of thirteen protein-coding genes, two ribosomal RNA genes, a transfer
RNA, and the mismatch repair gene, similar to mt genomes of Isididae reported in Brugler and France (2008),
and thus designed PCR reactions to target the junctions between these genes. We recognized the arrangement of
genes may differ among taxa and so where initial gene junction-spanning PCR reactions failed, we tested alternate
primer combinations reflecting different octocoral gene orders (e.g. Figueroa and Baco, 2015). PCR products were
sequenced and compared using the BLAST algorithm on the NCBI GenBank database against existing complete
mt genomes to ensure that gene boundaries were crossed. Primer combinations used to cross all gene boundaries
are shown in Table 2. PCR amplicons ranged from ~300bp to ~800bp length. Sequences were aligned to BAL208-1
(GenBank Acc. # EF622534) to help delineate gene boundaries.
Phylogenetic analyses. Four hundred thirty-eight Isididae samples were amplified for mtMutS. A subset of
those was selected to represent the diversity of Isididae such that one representative of each unique haplotype was
included in the phylogenetic analyses. Preferential treatment was given to specimens that had sequences generated
by our lab and used in previous analyses (e.g. France, 2007; Pante et al. 2012), thus reducing the number of ampli-
fication and cycle sequencing reactions required.
Sequences from other octocorallian families were generated de novo or obtained from GenBank. Some taxa
were specifically selected for comparative purposes because they possess isidid-like qualities (e.g. the genus Pari-
sis, which also has a node – internode skeletal structure) or because of prior knowledge of phylogenetic relation-
ships to the isidid subfamilies (e.g. Chrysogorgiidae and Primnoidae; Pante et al. 2012). In the case of Paramuricea
cf. hawaiiensis, it was not possible to obtain sequences for all gene regions (mtMutS-5’, mtMutS-3’, 18S) from a
single specimen, so a chimeric concatenated sequence was made from different specimens identified as the same
species (Table 3).
The edited mtMutS and 18S sequences were aligned using MAFFT ( Katoh et al. 2002; Katoh and Standley,
2013) with the L-INS-i algorithm. The alignment of mtMutS was checked by translating DNA sequences to amino
acids and adjusted where necessary to ensure that the reading frame was maintained. An Incongruence Length Dif-
ference (ILD) test (Farris et al. 1994) was performed and it was determined that the different gene markers could
be concatenated. RAxML 7.2.8 BlackBox was run in the CIPRES Portal (Miller et al. 2010. http://www.phylo.org),
using the default settings, to generate a maximum likelihood tree with 500 bootstrap replicates. Bootstrap values
above 50% are presented on the maximum likelihood trees. MrBayes 3.2.7a on XSEDE was run in the CIPRES Por-
tal to generate Bayesian inference phylogenies (Ronquist and Huelsenbeck, 2003) using 6 nucleotide substitution
types, 4x4 substitution model with I+G among-site rate variation; 5 x 106 generations, 2 runs, 4 chains, sampling
every 1000 generations; and a burn-in of 1250. The Bayesian likelihood scores were transposed onto the maximum
likelihood tree.
Results
Taxon sampling. Through the collection and identification of specimens for this analysis, we found expanded rang-
es for members of Circinisidinae and Mopseinae. Our specimens of Notisis elongata (Roule, 1908) and Primnoisis
sp. represent geographic range extensions for these genera. They were collected on the MNHN/IRD-sponsored
expedition Terrasses, which sampled the deep sea south of New Caledonia (Table 1) (Location and other details for
these cruises can be found at:
http://expeditions.mnhn.fr/campaign/terrasses;jsessionid=3823489BC062907B77182FFF98CCCDA9). Both
species were previously known only from off Antarctica and the Australian continental shelf as far north as 38°S,
respectively (Alderslade 1998, Moore et al. 2016). Similarly, Lissopholidisis nuttingi (Grant, 1976) (240-260 m)
and Minuisis sp. (390-410 m) were collected from shallower depths than previously recorded (978 m and 500 m,
respectively), and Notisis elongata (370-440 m) and Sclerisis sp. (1145-1200 m) were collected from deeper depths
than previously recorded (150 m and 684 m, respectively). Gorgonisis, known previously from one location off the
coast of western Australia, was collected off eastern Australia, and Plexipomisis, known only from southeastern
Australia, was collected in the Great Australian Bight.

HEESTAND SAUCIER ET AL.
260 · Zootaxa 5047 (3) © 2021 Magnolia Press
PCR amplification and sequencing. The mtMutS-5’ region was successfully PCR amplified and sequenced
from all specimens selected for this analysis (n=61). However, the success rate was lower for the other two targeted
gene regions (Table 3) perhaps due to poor quality DNA, particularly in specimens collected > ten years ago and
borrowed from museums, or those collected from shallower depths (primarily circinisidins and Isis). Thirty-one
specimens were also successfully amplified for the mtMutS-3’ region. This region generally shows a lower rate of
substitution than the mtMutS-5’ region and the addition of mtMutS-3’ sequence typically does not increase the num-
ber of observed haplotypes; however, for some specimens the mtMutS-3’ region reveals novel mtMutS haplotypes
not seen in comparisons of just mtMutS-5’ sequences. Twenty-nine specimens were successfully amplified for the
nuclear ribosomal gene 18S. Among the bamboo corals, the middle fragment of 18S was more variable than the
beginning or ending fragment, and although it did not reveal new haplotypes relative to the mtMutS data, the varia-
tion seen in 18S coincides with the variation seen in mtMutS and its addition to the analysis helped bolster bootstrap
support for some of the deeper nodes in the resulting trees (Figures 2 and 3). PCR amplification and sequencing of
mtMutS-5’, mtMutS-3’ and 18S, yielded a concatenated alignment of 2638 bp length including indels (1013 bp from
the mtMutS-5’ [n=61, unaligned length from 602 – 822 bp] + 949 bp from the mtMutS-3’ [n=31, unaligned length
from 552 – 857 bp] + 674 bp from the 18S [n=29, unaligned length from 368 – 670 bp]); 61 individual sequences
downloaded from GenBank were also included in the alignment (Table 3). Molecular phylogenies were constructed
using 26 unique haplotypes obtained from specimens classified in the family Isididae and representing all four
subfamilies, selected to provide coverage of the observed genetic diversity seen in each subfamily. To assess the
monophyly of the Isididae, 35 representative taxa from 18 additional families of octocorals were also included.
FIGURE 2. Maximum likelihood tree showing the divergence among the subfamilies of the family Isididae sensu lato. The tree
is based on a concatenated alignment of mtMutS-5’, mtMutS-3’, and 18S and includes 61 total taxa (26 isidids). The clades that
include taxa currently classified as Isididae are indicated in bold font.
Phylogenetic analyses. A RAxML tree for all taxa, based on mtMutS-5’ sequences only, was generated in order
to determine if the missing data from the other two gene regions altered the relationships among the taxa in the tree.
The mtMutS-5’ tree topology (data not shown) was identical to the tree generated based on the concatenated align-
ment of all three gene regions (3GR). However, stronger bootstrap support for the deep nodes was recovered using
3GR versus mtMutS-5’-only. Only trees using data from the 3GR alignment are presented.

DECONSTRUCTING THE ISIDIDAE Zootaxa 5047 (3) © 2021 Magnolia Press · 261
Because two-thirds of the dataset comprises mtMutS, a gene that is only found within the mitochondria of
taxa of the subclass Octocorallia, we could not use a non-octocoral outgroup in our analyses; we present unrooted
trees.
The family Isididae was not recovered as a monophyletic clade in the maximum likelihood tree. Taxa from three
of the isidid subfamilies (Keratoisidinae, Mopseinae, and Circinisidinae) were included in the Calcaxonia – Pen-
natulacea clade (Figure 2), one of the two major clades of Octocorallia recognized in previous studies of the subclass
(McFadden et al. 2006). However, Isis (Isidinae), the type genus of the family, is nested within the Holaxonia – Al-
cyoniina clade. Chelidonisis, which is also currently included in the subfamily Isidinae, forms a weakly supported
clade with pennatulaceans (n=3) and ellisellids (n=3). Circinisidinae, Keratoisidinae and Mopseinae do not group
with each other as a monophyletic clade (Figure 3). Instead, there are two well-supported clades that are not sister to
each other: one includes all taxa currently assigned to the subfamily Keratoisidinae and the other includes Mopsei-
nae and Circinisidinae, which did not resolve as separate subclades.
FIGURE 3. Midpoint rooted RAxML cladogram. Bootstrap values > 50 are shown at the nodes. Baysian posterior probability
values >0.5 are shown below the node. The tree is based on a concatenated alignment of mtMutS-5’, mtMutS-3’, and 18S, and
includes 61 total taxa (26 isidids). Clades with an articulated skeleton are marked with a large asterisk (*) on the branch. The
type genus of the Isididae is indicated by the black rectangle and Isididae sensu lato are indicated by the purple rectangles. Light
yellow circles indicate clades that have the “ancestral” mt genome arrangement (as defined by Brockman & McFadden, 2012).
The dark blue circle indicates the clade that has the Keratoisidinae mt genome arrangement mapped by Brugler and France
(2008). Only isidid mt genomes are mapped on the tree. Small asterisks (*) following taxon names indicate specimens used to
test mt gene boundaries. The “S1 clade” of Keratoisidinae is the one with taxa Cladarisis sp. 1 and sp. 2.
Mitochondrial gene arrangement. All expected mt gene junctions based on the typical complement of genes
found in other octocorallian mt genomes were successfully PCR amplified and sequenced for targeted members of
the Keratoisidinae, Mopseinae, and Chelidonisis (Isidinae) (Figure 3, Table 4). Eight out of the sixteen expected
gene boundaries for Zignisis sp. (Circinisidinae) were crossed using PCR but only seven of the eight were success-

HEESTAND SAUCIER ET AL.
262 · Zootaxa 5047 (3) © 2021 Magnolia Press
fully sequenced (Table 4). This is most likely due to the poor quality of the amplicon resulting from weak amplifica-
tion for this specimen, despite multiple extractions and amplification attempts with different primer combinations
(Table 2). To enhance amplification we tried increasing the amount of template per reaction, increasing primer con-
centration per reaction, and adding bovine serum albumin (BSA). Six out of sixteen gene boundaries for Isis (Isidi-
nae) were successfully crossed using PCR and sequenced (cox2-atp8-atp6-cox3, nd2-16S, mtMutS-nd4L, nd3-nd6)
(Table 4); despite numerous attempts and different primer combinations we were not able to amplify the remaining
regions. DNA extractions from five different Isis specimens all yielded low molecular weight DNA, which likely
contributed to our inability to successfully amplify all target gene boundaries.
TABLE 4. GenBank accession numbers for each gene boundary sequenced for the mt genome gene boundary mapping
study. ‘NA’ represents non-existent gene boundaries in the specified taxon due to gene order rearrangement. ‘xx’ indicates
not sequenced for that gene region.
Genome arr. #1 Genome arr. #2
Circinisidinae:
Zignisis sp.
(C013153)
Chelidonisis sp.
(LII10402)
Isidinae:
Isis hippuris
(OCDN9247L)
Mopseinae:
Notisis elongata
(TER11082)
Keratoisidinae:
Keratoisis grayi
(5663)
Mt Gene Junction
cox2-cox1 KX530104 KX530105 xx KX530103 KX530101
cox1-rns xx xx xx KX530083 KX530085
rns-nad1 KX530096 KX530097 xx KX530098 KX530095
nad1-cob xx xx xx xx KX530087
cob-nad6 xx xx xx KX530091 KX530090
nad6-nad3 KX530133 KX530134 KX530135 KX530136 KX530132
nad3-nad4L NS KX530118 xx KX530119 KX530117
nad4L-mtMutS (#1 only) KX530081 KX362350 JN383338 KX362346 NA
mtMutS-rnl xx xx xx KX362346 KX362308
rnl-nad2 KX530110 KX530111 KX530112 KX530113 KX530109
nad2-nad5 xx KX530141 xx KX530139 KX530140
nad5-nad4 xx xx xx xx KX530148
nad4-tRNA-cox3
(#1 only)
KX530126 KX530127 xx KX530128 NA
cox2-atp8-atp6-cox3 xx xx KX530121 KX530123 KX530124
nad4L-nad4 (#2 only) NA NA NA NA KX530137
cox3-tRNA-mtMutS
(#2 only)
NA NA NA NA KX362308
Most genes are separated by an intergenic region (IGR) ranging from 2-556 bp in length and averaging 44 bp
(Table 5), although two pairs of genes overlap (nad2-nad5 and rns-cox1) (rns=small subunit RNA). These IGR
lengths are comparable to those described by Brugler and France (2008) and van der Ham et al. (2009) for bamboo
corals. The longest IGR (556 bp) is found between cob-nad6 in Keratoisidinae (igr4 of van der Ham et al., 2009),
but the same region is only 27 bp in Mopseinae. The genus Chelidonisis has a longer IGR between rns–nad1 com-
pared to that of Circinisidinae, Keratoisidinae, and Mopseinae (144 bp vs. 46 bp length). It remains unclear whether
there are IGRs flanking the rnl (large subunit RNA), as a 16S RNA transcript has not been sequenced, and so the
exact start and stop positions of the gene remain unknown. Until that work is done, the default assumption is to
record no IGRs in these regions (rnl-nad2 and mtMutS-rnl).
Our data show two different mt genome arrangements among Isididae taxa tested (Table 4). With the exception
of the “S1 clade” (sensu Pante et al. 2013), keratoisidin taxa have a mt genome arrangement as reported by Brugler
and France (2008: EF622534) and van der Ham et al. (2009) for Keratoisidinae. This is in contrast to Mopseinae,
Chelidonisis, and the keratoisidin “S1 clade”, which have gene arrangements that match with all other taxa from the
Holaxonia-Alcyoniina clade so far sequenced for their mt genome – that proposed to be the ‘ancestral’ octocoral mt
gene order by Brockman and McFadden (2012). Our data are limited for Circinisidinae and Isis, but what we were

DECONSTRUCTING THE ISIDIDAE Zootaxa 5047 (3) © 2021 Magnolia Press · 263
able to obtain is consistent with the hypothesis that they also have this ‘ancestral’ mt gene arrangement (Figure 3,
Table 4).
TABLE 5. Length of IGR regions (in base pairs). Negative numbers indicate overlapping gene regions. ‘NA’ signifies
gene junctions that are not present in the specified mt genome arrangement. ‘xx’ signifies gene junctions that we were
unable to amplify or sequence. Genes were compared to those in the published mt genome GenBank Accession no.
EF622534.
Genome arr. #1 Genome arr. #2
Circinisidinae:
Zignisis sp.
(C013153)
Chelidonisis sp.
(LII10402)
Isidinae:
Isis hippuris
(OCDN9247L)
Mopseinae:
Notisis elongata
(TER11082)
Keratoisidinae:
Keratoisis grayi
(5663)
Mt Gene Junction
cox2-cox1 103 163 xx 137 109
cox1-rns xx xx xx -7 -7
rns-nad1 46 144 xx 46 46
nad1-cob xx xx xx xx 2
cob-nad6 xx xx xx 27 556
nad6-nad3 17 43 55 43 16
nad3-nad4L xx 25 xx 35 20
nad4-nad4L NA NA NA NA 196
nad4-nad5 xx xx xx xx 57
nad2-nad5 xx -13 xx -13 -13
rnl-nad2 0 0 0 0 0
mtMutS-rnl xx xx xx 0 0
nad4L-mtMutS 11 14 14 14 NA
mtMutS-trnM NA NA NA NA 11
nad4-trnM 27 82 xx 57 NA
trnM-cox3 34 44 xx 55 29
cox3-atp6 xx xx 60 44 32
atp6-atp8 xx xx 15 4 12
atp8-cox2 xx xx 21 21 4
Discussion
Phylogeny. The primary characteristic that has been used to define the Isididae has been the presence of an articu-
lated skeleton composed of sclerite-free gorgonin nodes separating calcareous internodes. However, our genetic
analyses did not recover a monophyletic clade including all bamboo corals, suggesting that the gross morphological
similarities of the articulated skeleton in the subfamilies of the Isididae are the result of convergent evolution, hav-
ing arisen independently several times on the octocoral tree, e.g. an articulated skeleton with a sclerite-free node is
homoplastic. Across Octocorallia there are several other instances of a skeleton with some form of nodal-internodal
structure, though the nodes differ with regard to their composition and how they are laid down. In Parisis Verrill,
1864 (Parisididae: Aurivillius, 1931) the nodes are made of gorgonin and have sclerites embedded in them (Bayer,
1955; Grasshoff and Bargibant, 2001). In Mirostenella Bayer, 1988 (Primnoidae: Milne-Edwards, 1857) strands of
calcified skeleton extend from the internodes into the gorgonin of the nodes (Bayer, 1988; Zapata-Guardiola et al.
2013). Members of the family Melithaeidae Gray, 1870 possess nodes that contain sclerites held together with gorgo-
nin forming a proteinaceous network filled with mesoglea, which causes the nodes to be swollen (Alderslade, 2006).
Likewise, the construction of the calcareous part of the octocorallian axis can vary considerably. Kölliker
(1865) and Kükenthal (1919) noted calcareous inclusions in the axes of some holaxonian corals as well as the scler-
axonian families. In the scleraxonian genus Corallium, the axis is made of concentric layers of organic matter and
Mg/Ca crystalline units that bear protuberances, thus giving the concentric layers a wavy appearance (Vielzeuf et

HEESTAND SAUCIER ET AL.
264 · Zootaxa 5047 (3) © 2021 Magnolia Press
al., 2008). The outlines of the organic material are suggestive that the calcareous units are modified sclerites.
Among calcaxonian families, axis construction is variable. Bayer (1955) noted that the calcareous units of
the axis (which he referred to as “sclerodermites,” a term borrowed from scleractinian skeletal morphology) of
Ellisellidae “are built up in layers concentric with the axis core, they are arranged in a distinctly radial fashion…”
(p. 215). A radially arranged pattern was also seen in the Keratoisidinae (Acanella and Keratoisis) and Mopseinae
(Primnoisis), although the individual units were obscured. Noe and Dullo (2006) showed that these concentric rings
of calcareous units contained decreasing amounts of organic matter as the axis grew outward, and in Keratoisis, at
least, the calcareous units were arranged tangentially. However, in the Isidinae, the calcareous units of the axis of the
genus Isis “are united into bundles, much like the sclerodermites of stony corals, radiating outward from the central
core”, and Bayer (1955) suggested that “the sclerodermites of Isis are actually highly modified spicules.” (p. 216).
Calcaxonian families Chrysogorgiidae, Pleurogorgiidae Cairns et al. 2021, and Huziogorgiidae Lopez-Gonzalez
2020, have axis compositions that consist of alternating non-undulating concentric rings of dense organic matter
and calcareous material, while in Ifalukellidae and Primnoidae the concentric organic matter rings are slightly to
strongly undulatory (Alderslade, 1986; Kükenthal, 1919).
In our phylogeny, Isis, the type species for both the subfamily Isidinae and the family Isididae, is nested outside
the Calcaxonia - Pennatulacea clade, whereas all other members of the family included in the analysis were nested
well within that clade (Figure 2). Rowley et al. (2015) generated a phylogram based on analyses of the nuclear ITS2
region and likewise found Isis grouped with holaxonian, rather than calcaxonian, taxa, a result more recently sup-
ported by Quattrini et al. (2020) and McFadden et al. (2021) using ultra-conserved elements (UCEs). The subfam-
ily Isidinae currently comprises two genera (Isis and Chelidonisis), which do not form a monophyletic clade in our
analyses (Figures 2, 3).
The subfamily Isidinae has been diagnosed as Isididae “having sclerites in the form of six or eight radiates,
clubs [or] tuberculate spindles” (Alderslade, 1998), with their axial skeletons ridged to varying degrees. However,
there are significant morphological, geographical, and depth differences among the two genera. Isis is found in
shallow (2-10 m), warm seas, whereas Chelidonisis is found in deep (> 200 m) cold waters. The axis of Isis, as
noted above, consists of calcareous units resembling sclerites and externally the internodes are heavily fluted, the
coenenchyme is thick, into which the polyps are able to completely retract, the polyps are devoid of sclerites except
for tiny rods in the tentacles, and the coenenchyme sclerites are clubs, capstans, and tuberculate spindles (Bayer
and Stefani, 1987b) (Figure 1). In contrast, in the species of Chelidonisis, the axis is solid and bears multiple small
spines, the coenenchyme is thin, the polyps contract into small hemispherical or conical verrucae, and the sclerites
are predominantly 6-radiates (Bayer and Stefani, 1987a). In our cladogram, Chelidonisis resolves as sister to a clade
of pennatulaceans + ellisellids, albeit without any meaningful support for this relationship (Figure 3). However, in
a more rigorous analysis using UCE and exon loci McFadden et al. (2021) showed Chelidonisis to be related to
several holaxonians. Given the incomplete taxon sampling across Octocorallia, the relationship of Chelidonisis to
other octocorals remains unknown; however, it is clearly well diverged from Isis. Given the morphological and mi-
tochondrial genetic differences between these two genera, we suggest Chelidonisis should be removed from Isidinae
(family Isididae) and placed in a new family to reflect its evolutionary path independent from Isis.
Sequences representing the subfamilies Mopseinae and Circinisidinae cluster together in the tree, forming
a well-supported Mopseinae-Circinisidinae monophyletic clade near to Primnoidae and Chrysogorgiidae. The
Mopseinae in this phylogeny are paraphyletic, with the circinisidin taxa forming an embedded monophyletic clade
within the mopseins, albeit with only weak support. The lack of separation between these two subfamilies could
be due to relatively poor taxon sampling in our tree, with only three of seven Circinisidinae genera and five of
nineteen Mopseinae genera included (Table 3). Additionally, we were only able to generate sequences for the 5’
end of mtMutS for the circinisidins. Although our data did not resolve reciprocally monophyletic clades for the two
subfamilies, Alderslade (1998) distinguished Circinisidinae and Mopseinae on the basis of morphological features
of the sclerites. More recently, Alderslade and McFadden (2012) reaffirmed Mopseinae as a valid subfamily based
on morphology, depth, and distribution; however, they did not discuss Circinisidinae in that paper. Both circinisidins
and mopseins have a thin coenenchyme covering smooth to slightly ridged internodes and light golden to brown
nodes (Figure 1). The polyps are non-retractile but highly contractile, and when compared to keratoisidins they are
smaller in size (<1 mm compared to 1.5-6 mm). Circinisidin polyps are covered with “transversely arranged scler-
ites in the form of mostly smooth oval scales whose distal margin is entire but often undulated; [and the] surface
sclerites of the coenenchyme are in the form of rooted heads or smooth ovals” (Alderslade, 1998: 20). Mopsein

DECONSTRUCTING THE ISIDIDAE Zootaxa 5047 (3) © 2021 Magnolia Press · 265
polyps have “smooth, tuberculate, or thorny scales or plates… When arranged transversally they generally have a
distal margin that is dentate, tuberculate, scalloped, thorny or thorn-like… the proximal margin generally has lobes
or tuberculate root-like processes” (Alderslade and McFadden, 2012: 37).
Mopseins and circinisidins have overlapping geographic distributions. Mopseins are more widespread (as might
be expected having more generic and species diversity than circinisidins), ranging across the southern hemisphere,
from New Caledonia to the Antarctic Peninsula, to the tips of South America and South Africa. The majority of
mopseins are known from depths shallower than 270 m, but three genera are found from 500 - 1000 m, one of
which, Sclerisis, is included in this study. Circinisidins have a narrower distribution, found only around southern,
south-western, and south-eastern regions of Australia at depths between 20-330 m. Although we were only able to
sequence fewer than half the described genera in the mopsein and circinisidin subfamilies, the 8 genera analyzed
captured much of the range of the geographic, bathymetric, and sclerite diversity seen in each of the subfamilies.
The five mopsein genera used for this analysis (Lissopholidisis, Minuisis, Notisis, Primnoisis, and Sclerisis) are
known from the New Caledonia region, off the coast of Australia and New Zealand, the Antarctic Peninsula, Ross
Sea, and the Southern Ocean. The narrower distribution of circinisidins was covered by the three genera used in
this analysis (Gorgonisis, Plexipomisis, Zignisis) in terms of both geography and bathymetry. Sclerites in the five
sampled mopsein genera are scale-like and range in shape from crescent-like to triangular to large irregular shapes.
The textures of the sclerites range from entirely smooth to thorny, and some of the sclerites are smooth on one side
and tuberculate on the other. For example, Sclerisis has highly tuberculate thorny sclerites, while Lissopholidisis has
smooth triangular-shaped sclerites. Indeed, because of this difference in texture of sclerites, it wasn’t until recently
(Alderslade and McFadden, 2012) that Sclerisis was classified within Mopseinae. Our genetic analysis supports
Alderslade and McFadden’s (2012) hypothesis about this classification of Sclerisis.
Keratoisidinae are found in all major oceans but strictly in the deep sea (>200 m), with the deepest published re-
cord from 4700 m (Lapointe and Watling, 2015); the axial skeleton is smooth, with brown to black nodes; the polyps
are non-retractile and are protected by sclerites in the form of needles, spindles, rods, or scales (Figure 1). Two mt
genome arrangements are found in the subfamily, each of which are associated with one of the two major subclades
(Figure 3). Keratoisidinae comprises nine currently recognized genera, but in our phylogeny these described genera
do not form monophyletic clades (these genera will be dealt with in a forthcoming companion paper). For example,
species that exhibit internodal branching vs non-branching do not form monophyletic clades (see also France,
2007). We included specimens from 8 of the 9 described genera in this analysis (Acanella Gray, 1870; Cladarisis
Watling, 2015; Eknomisis Watling and France 2011; Isidella Gray, 1857; Jasonisis Alderslade and McFadden, 2012
Keratoisis Wright, 1869; Lepidisis Verrill, 1883; and Orstomisis Bayer, 1990; only Bathygorgia Wright, 1885 is
missing). Our samples were carefully chosen from specimens across the keratoisidin clade to represent the diversity
seen in the group. Additionally, the samples encompass the large geographic range (samples from both the Atlantic
and Pacific Ocean), depth distribution (482 – 2346 m), and sclerite morphology found in Keratoisidinae. Acanella,
Isidella, Keratoisis, and Lepidisis and a wide depth distribution (300 – 2525 m). On the other hand, Eknomisis,
Jasonisis, and Orstomisis are all known currently as monotypic genera with narrow geographic and depth distribu-
tions, although our genetic data suggest additional undescribed species. Keratoisidins have needle-like sclerites,
which can range in size and ‘sharpness,’ from acutely pointed to bluntly rounded (like a rod). Our samples spanned
this continuum from the “elongated ovals” of Jasonisis (Alderslade and McFadden, 2012) to the long needle-like
sclerites of Lepidisis caryophyllia, which extend the length of the polyp and protrude between the tentacles.
Our phylogenetic analysis suggests the genus Isidoides to be closely related to the genera of the Keratoisidinae,
but it differs from the latter in having a unique mitochondrial gene order (Pante et al., 2013) and a combination of
distinct morphological features. The axis of Isidoides appears to be composed of concentric layers of gorgonin and
calcite, as is seen in primnoids and chrysogorgiids, and lacks the node-internode structure of keratoisids (there are
unarticulated keratoisids but from their position in the phylogeny of that group it seems clear the lack of articula-
tions is a secondarily derived feature). The sclerites are small biscuit-like structures, thicker than the flat rods, but
not cylindrical as in rods characteristic of keratoisids, and also thicker than the scales seen in chrysogorgiids and
primnoids.
This study is the first phylogenetic analysis to include species from all four subfamilies of the family Isididae,
and the results suggest that bamboo corals are polyphyletic. The Keratoisidinae form a well-supported monophy-
letic clade while species of Circinisidinae and Mopseinae are intermingled on a clade that is not sister to the Kera-
toisidinae. The two species of Isidinae in our dataset were well separated from each other, and from all other isidids,

HEESTAND SAUCIER ET AL.
266 · Zootaxa 5047 (3) © 2021 Magnolia Press
on the phylogeny. Several characters, including sclerite morphology, polyp arrangement, coenenchyme thickness,
contractile polyps, and crystal structure of the skeleton (Grant, 1976), as well as geographic distribution and depth
range inhabited, support the results from our genetic analysis that these are independently evolving lineages. These
results also provide additional examples of evolutionary radiations and diversification in the deep sea, independent
of shallow water, reinforcing the conclusions of Pante et al. (2012). Our results suggest that three of the four sub-
families, and Chelidonisis and Isidoides, should be raised to family level, and the Circinisidinae should be merged
into the Mopseidae. We provide diagnoses for these proposed families below.
Mitochondrial gene arrangement. Among the Isididae taxa examined, two different mt genome arrangements
were observed (Table 4). Both arrangements are seen among Keratoisidinae genera. Our phylogeny indicates an
early divergence among Keratoisidinae taxa, yielding two clades, each of which is characterized by one of the two
genome arrangements (Figure 3). Taxa in the more diverse lineage (in terms of described genera) have the mt ge-
nome arrangement first described by Brugler and France (2008), and subsequently by van der Ham et al. (2009), as
unique to Keratoisidinae. However, Hogan et al. (2019) have found the same arrangement independently evolved
in the sea pen genus Anthoptilum, and so the hypothesis that it is a synapomorphy of this lineage is no longer sup-
ported. The “S1 clade” of Keratoisidinae (that includes Cladarisis nouvianae Watling, 2015), as well as selected
Isidinae (Chelidonisis) and Mopseinae (Notisis and Sclerisis), have the mt gene arrangement that Brockman and
McFadden (2012) describe as the ‘ancestral’ octocoral mt gene order. We were able to obtain only limited data on
gene boundaries for Zignisis (Circinisidinae), but enough to show it differs from Keratoisidinae and is consistent
with the mt genome arrangement determined for Mopseinae. Likewise, multiple attempts to extract DNA from eight
different Isis specimens all yielded only low quality, highly sheared DNA and few successful PCR amplifications.
However, what data we were able to collect is consistent with Isis having the same arrangement as the predicted
‘ancestral’ octocoral mt gene order, which is shared by all species so far tested in the Alcyoniina-Holaxonia clade
(e.g. Figueroa and Baco, 2015).
Taxonomy
Subclass Octocorallia
Order Alcyonacea
Suborder Holaxonia
Family Isididae Lamouroux, 1812
Type genus: Isis Linnaeus, 1758
Diagnosis. Colony with articulated skeleton of gorgonin nodes alternating with calcareous internodes that are
solid and composed of radiating bundles of sclerite-like units. Branches originate from the internodes, always begin-
ning with a node at the base of the branch. Coenenchyme thick; polyps completely retractile into small pockets in
the coenenchyme, leaving only small pores visible on the surface. Polyps mostly devoid of sclerites, with small rods
only in the tentacles. Coenenchyme sclerites mostly small clubs, but may include 6-, 7- or 8-radiates, warty rods,
spindles and crosses, heavily sculpted capstans, dumbbells, and double cones.
Included genera: Isis Linnaeus, 1758
Remarks. The genus Chelidonisis has routinely been included in the Isididae subfamily Isidinae with Isis,
chiefly on the basis of the presence of 6-radiate sclerites in the coenenchyme. However, there are several mor-
phological differences that distinguish the two genera. The axis of Isis is thick with smooth longitudinal ridges
on the internodes, whereas in Chelidonisis the axis is slender and the ridges on the internodes are festooned with
small spines. The coenenchyme of Isis is very thick and encompasses retractile polyps as compared to the thin
coenenchyme of Chelidonisis, wherein the polyps contract into armored verrucae on the branch. In addition, the
coenenchyme sclerites of Isis are very diverse capstans, tuberculate spindles, and clubs, whereas in Chelidonisis the
sclerites are predominantly 6-radiates and spindles, and capstans are never present. Finally, the two genera occur in
widely separated clades on the molecular phylogeny (Figures 2, 3), as also shown by Quattrini et al. (2020).

DECONSTRUCTING THE ISIDIDAE Zootaxa 5047 (3) © 2021 Magnolia Press · 267
Family Chelidonisididae n. family
Type genus: Chelidonisis Studer, 1890
Diagnosis. Modified from Bayer and Stefani (1987a). Colony branched in one plane, occasionally anastomos-
ing, branching from distal end of internodes; internodes longitudinally ridged, the ridges armed with spines; polyps
forming hemispherical or bluntly conical verrucae distributed mostly on two sides of the branches in the plane of
ramification; coenenchyme thin; sclerites predominantly in the form of 6-radiates and with small tuberculate plates
in the tentacles.
Included genera: Chelidonisis Studer, 1890.
Remarks. See the Remarks under the family Isididae.
Suborder Calcaxonia
Family Mopseidae Gray, 1870 n. rank
Circinisidinae Grant, 1976
Peltastisidinae Grant, 1976
Type genus: Mopsea Lamouroux, 1816
Diagnosis. After Alderslade and McFadden (2012). Colony branched or unbranched, branches arising pre-
dominantly from internodes but also from nodes. Axial internodes with longitudinal ridges, sometimes ornamented
with large or small spines, denticles, or granules. Coenenchyme thin. Sclerites of coenenchyme spindles, platelets,
crosses, or more rarely nodules or goblets. Polyps contractile, sometimes angled to the branch surface, adaxially re-
duced, with or without adaxial sclerites, or more-or-less erect and completely covered with sclerites. Body sclerites
in the form of smooth, tuberculate, or thorny scales or plates arranged either transversely or longitudinally, rarely in
a disorganized pattern. Anthopomal (distal body) sclerites intermesenterially situated and forming a protective cover
over the contracted tentacles. Tentacle rachis with crescentic scales.
Included genera: Acanthoisis Studer [& Wright], 1887; Chathamisis Grant, 1976; Echinisis Thomson & Ren-
net, 1931; Iotisis Alderslade, 1998; Jasminisis Alderslade, 1998; Ktenosquamisis Alderslade, 1998; Lissopholidisis
Alderslade, 1998; Minuisis Grant, 1976; Mopsea Lamouroux, 1816; Myriozotisis Alderslade, 1998; Notisis Gravier,
1913; Oparinisis Alderslade, 1998; Paracanthoisis Alderslade, 1998; Peltastisis Nutting, 1910; Primnoisis Studer
[& Wright], 1887; Pteronisis Alderslade, 1998; Sclerisis sensu Bayer & Stefani, 1987(a); Sphaerokodisis Alder-
slade, 1998; Tenuisis Bayer & Stefani, 1987(a); Tethrisis Alderslade, 1998.
Remarks. Grant (1976) erected the subfamilies Circinisidinae and Peltastisidinae in the family Isididae to dis-
tinguish a small group of genera from the members of the Mopseinae. Alderslade (1998) accepted the Circinisidinae
on the basis of the distinctly smooth body sclerites (that is, lacking spines or complex tubercles on the exposed face
and never having a dentate or thorny free margin), but rejected the Peltastisidinae, suggesting that Grant’s interpreta-
tion of the sclerite arrangement forming the opercular covering was in error. We have combined the Circinisidinae
into the Mopseidae since both groups form one distinct clade in the molecular tree. Further molecular work may
show the two groups to form subfamilies within the Mopseidae, but at present that cannot be supported because
genera currently assigned to the Mopseinae would make the subfamily paraphyletic.
Family Keratoisididae Gray, 1870, n. rank
Ceratoisidae Verrill 1883
Type genus: Keratoisis Wright (1869)
Diagnosis. Colonies typically with an articulated skeleton of hollow or solid calcium carbonate internodes
interrupted by brown to dark brown proteinaceous and sclerite-free nodes. unbranched (“whip-like”) or branched
with branches originating at the nodes, or from the internodes, either immediately distal to the nodes, or from mid-
way along the internode. Coenenchyme usually thin but sometimes slightly thickened. Colonies may be covered
in a fleshy tegument containing nematocysts. Polyps are contractile to varying degrees, but never within the coe-

HEESTAND SAUCIER ET AL.
268 · Zootaxa 5047 (3) © 2021 Magnolia Press
nenchyme, or only the tentacles contract over the polyp body oral area. Sclerites are only needles, spindles, rods,
or scales that are arranged longitudinally, transversely, or obliquely along the polyp body and in the coenenchyme.
One or a cluster of mesenterially arranged needle-like sclerites protrude between the bases of the tentacles in many
species (in contrast to the intermesenterially arranged anthopomal sclerites of the Mopseidae); other species may not
possess needles but a mesenterial arrangement of rod-like sclerites may still be present. Pharyngeal sclerites usually
present; include tuberculated or spiny rodlets, and double stars.
Included genera: Acanella Gray (1870), Bathygorgia Wright (1885), Cladarisis Watling 2015, Eknomisis
Watling & France (2011), Isidella Gray (1857), Jasonisis Alderslade & McFadden (2012), Keratoisis Wright (1869),
Lepidisis Verrill (1883), Orstomisis Bayer (1990).
Remarks. The Keratoisididae is the most morphologically and genetically diverse of the families resulting
from the revision of the Isididae. That diversity is manifested in colonies that range in size from a few centimeters
to several meters; from tall spindly whips to large bushes; from polyps that contract to small mounds on the branches
to tall polyps where only the tentacles contract over the mouth; from polyps with large numbers of needles, rods, or
scales to polyps with barely any sclerites but a thick outer integument; or from polyps covered with a nematocyst
impregnated integument to polyps where the outer layer of cells is barely discernable. Some of the variation seems
to be derived secondarily. For example, small, bramble-like colonies that we have observed on both Atlantic and
Pacific seamounts seem to possess an unarticulated axis that, based on the placement of those species in a more
complete phylogenetic analysis of Keratoisididae, may represent a loss of that feature (Heestand Saucier, 2016). All
of these variations characterize a large number of clades within the family that will be dealt with in a companion
paper.
Brugler and France (2008) identified a novel (at the time) mitochondrial gene arrangement for keratoisidids,
and this, along with our preliminary molecular phylogeny results, drew our attention to the idea that the family Isi-
didae was polyphyletic, as suggested more than a century ago by Kükenthal (1919). Subsequent analyses showed
the “S1 clade” of Keratoisididae (that includes Cladarisis Watling 2015) to possess the presumed ancestral octocoral
mitochondrial gene order (Pante et al. 2013), and, more recently, Hogan et al. (2019) have reported the interesting
finding that species of Pennatulacea in the genus Anthoptilum have the same “keratoisidin” mitochondrial genome
arrangement found by Brugler and France (2008). Thus, a specific mitochondrial gene arrangement cannot be used
as a synapomorphy of this newly erected family.
Family Isidoidae, n. family
Type genus: Isidoides Nutting (1910)
Diagnosis. Colonies are planar with pseudo-dichotomous branching. Axis solid, calcareous, with concentric
layers that can be white to dark golden-brown in color but not articulated. Polyps are mostly bi-serially or tri-serially
arranged on the axis and are non-retractile, but in a preserved state the tentacles are tightly contracted and overlay
the oral disk. The sclerites are smooth, flattened, finger biscuit-shaped rods of relatively uniform size that are abun-
dant and tightly packed throughout the surface tissue of the colony, but rarely in the pharynx. Cross-shaped sclerites
are present but few in number.
Included genera: Isidoides
Remarks. The Isidoides clade comprises a single species, Isidoides armata, that was described by Nutting
(1910) based on specimens collected from the southwestern Pacific Ocean. There were no new reports of Isidoides
until 2003, when independent collections on the Norfolk Ridge and off New Caledonia in 2008 and 2011 led to a
re-description of the genus by Pante et al. (2013). That work also noted I. armata has a mitochondrial gene arrange-
ment that is (so far as is known) unique among Octocorallia. In the original description of Isidoides, Nutting (1910)
noted that the smooth sclerites resembled those of the bamboo coral genus Bathygorgia, although closer examina-
tion reveals the crystal arrangement in the sclerites of the two genera is quite different (unpublished observation).
However, because Isidoides lacked the characteristic articulated skeleton of Isidae (=Isididae), Nutting placed the
genus in the Gorgonellidae (=Ellisellidae). Bayer (1979) included Isidoides in a key to the Chrysogorgiidae, but
with neither justification nor explanation of what characters suggested that relationship. Later, Bayer & Grasshoff
(1994) stated that Isidoides “appears to be” a chrysogorgiid and not an ellisellid, but without further elaboration on
the placement of the genus. Our phylogenetic results and those of Pante et al. (2013) do not support a close relation-

DECONSTRUCTING THE ISIDIDAE Zootaxa 5047 (3) © 2021 Magnolia Press · 269
ship of Isidoides with chrysogorgiids or ellisellids but instead find them to be sister to Keratoisididae. Bathygorgia,
resurrected by Lapointe and Watling (2015), has “obliquely or longitudinally placed spicules [=sclerites], club- or
biscuit-like in shape” (Wright 1885), which would seem to be similar to Isidoides sclerites as originally noted by
Nutting (1910), but an examination of the type specimen of B. profunda shows the sclerites to be “biscuit-like” only
in their elongate form but in fact are all small rods (Lapointe and Watling, 2015). Thus, except for the lack of an
articulated skeleton, Isidoides was thought to fit within the definition of the old Keratoisidinae. On the other hand,
the axis of Isidoides is often a dark brown (see Pante et al. 2013) suggesting that the axis may be constructed more
like that of primnoids, with concentric layers of gorgonin alternating with calcareous material. The structure of the
axis has not yet been examined in further detail. Some consideration was given to including Isidoides in the family
Keratoisididae, but to do so would have meant the latter family could not be characterized by any synapomorphy.
Acknowledgments
The authors wish to thank Aude Andouche and Sarah Samadi (Muséum National d’Histoire Naturelle) for help
with the collection and generous donations of material; Cathy McFadden for sharing of samples collected from the
Museum and Art Gallery of the Northern Territory, the Santa Barbara Museum of Natural History, and the Coral
Reef Research Foundation; Andrea Quattrini for samples of Chelidonisis; and the Smithsonian Institution’s National
Museum of Natural History and the Rijksmuseum van Natuurlijke Historie for the loan and use of their material.
Specimens from New Caledonia and Papua New Guinea were provided by the Muséum national d’Histoire naturelle
(MNHN; Paris, France), and were collected during the “Terrasses” and “BioPapua” cruises respectively, undertaken
by the MNHN and the Institut de Recherche pour le Développement (IRD). The “Terrasses” cruise, organized by
Sarah Samadi (IRD/MNHN), is part of the “Tropical Deep-Sea Benthos” research program (formerly MUSOR-
STOM) and the “BioPapua” cruise, organized by Laure Corbari (MNHN) and Sarah Samadi, were spearheaded by
the MNHN and the IRD (Bouchet et al. 2008). We are grateful to the crew of the R/V Alis for contributing to the
success of this cruise. A special thank you goes to the crews of the NOAA Ship Ronald H. Brown and ROV Jason
II, ROV Hercules, R/V Atlantis and HOV Alvin, R/V Kaimikai-o-Kanaloa and DSV Pisces V, R/V F.G. Walton
Smith and ROV Global Explorer. Thank you to Mercer Brugler, Eric Pante, and Jana Thoma for their moral sup-
port and laboratory help. This work was partially funded by the Louisiana Board of Regents Fellowship to EHS,
University of Louisiana Graduate Student Organization and Student Government Association, and NOAA Office of
Exploration and Research. This is publication number 144 from the School of Life Sciences, University of Hawaii
at Manoa.
References
Alderslade, P. (1986) New species of the gorgonian genus Plumigorgia (Coelenterata: Octocorallia) with a review of the family
Ifalukellidae. The Beagle, Occasional Papers of the Northern Territory Museum of Arts and Sciences, 3, 95–122.
https://doi.org/10.5962/p.260877
Alderslade, P. (1998) Revisionary systematics in the gorgonian family Isididae, with descriptions of numerous taxa (Coelen-
terata: Octocorallia). Records of the Western Australian Museum, Supplement, 55, 1–359.
Alderslade, P. (2006) New subfamilies and a new genus and species of Melithaeidae (Coelenterata: Octocorallia: Alcyonacea)
with comparative data on the structure of both melithaeid and suberogorgiid axes. Zootaxa, 1199 (1), 1–47.
https://doi.org/10.11646/zootaxa.1199.1.2
Alderslade, P. & McFadden, C.S. (2012) A new genus and species of the family Isididae (Coelenterata: Octocorallia) from a
CMAR Biodiversity study, and a discussion on the subfamilial placement of some nominal isidid genera. Zootaxa, 3154
(1), 21–39.
https://doi.org/10.11646/zootaxa.3154.1.2
Apakupakul, K., Siddall, M.E. & Burreson, E.M. (1999) Higher level relationships of leeches (Annelida: Clitellata: Euhiru-
dinea) based on morphology and gene sequences. Molecular phylogenetics and evolution, 12, 350–359.
https://doi.org/10.1006/mpev.1999.0639
Bayer, F.M. (1955) Contributions to the nomenclature, systematics, and morphology of the Octocorallia. Proceedings of the
United States National Museum, 105, 207–220.
https://doi.org/10.5479/si.00963801.105-3357.207
Bayer, F.M. (1956) Octocorallia. In: Moore, R.C. (E d.), Treatise on Invertebrate Paleontology. Part F. Coelenterata. Geo-

HEESTAND SAUCIER ET AL.
270 · Zootaxa 5047 (3) © 2021 Magnolia Press
logical Society of America and University of Kansas Press, Lawrence, Kansas, pp. 166–231.
Bayer, F.M. (1988) Mirostenella articulata, a remarkable new genus and species of primnoid octocoral (Gorgonacea) with un-
calcified axial nodes. Proceedings of the Biological Society of Washington, 101, 251–256.
Bayer, F.M. (1990) A new isidid octocoral (Anthozoa: Gorgonacea) from New Caledonia, with descriptions of other new species
from elsewhere in the Pacific Ocean. Proceedings of the Biological Society of Washington, 103, 205–228.
Bayer, F.M. & Stefani, J. (1987a) New and previously known taxa of isidid octocorals (Coelenterata: Gorgonacea), partly from
Antarctic waters. Proceedings of the Biological Society of Washington, 100, 937–991.
Bayer, F.M. & Stefani, J. (1987b) Isididae (Gorgonacea) de Nouvelle-Calédonie: nouvelles clé des genres de la famille. Bulletin
du Muséum national d’histoire naturelle, 4e Series, Section A, 9, 47–106.
Bayer, F.M. & Stefani, J. (1988) Primnoidae (Gorgonacea) de Nouvelle-Calédonie. Bulletin du Muséum national d'histoire
Naturelle, 4e Series, Section A, 10, 449–518.
Bilewitch, J.P. & Degnan, S.M. (2011) A unique horizontal gene transfer event has provided the octocoral mitochondrial genome
with an active mismatch repair gene that has potential for an unusual self-contained function. BMC Evolutionary Biology,
11, 228.
https://doi.org/10.1186/1471-2148-11-228
Bouchet, P., Héros, V., Lozouet, P. & Maestrati, P. (2008) A quarter-century of deep-sea malacological exploration in the South
and West Pacific: where do we stand? how far to go? Tropical Deep-Sea Benthos, 25, 9–40.
Brockman, S.A. & McFadden, C.S. (2012) The mitochondrial genome of Paraminabea aldersladei (Cnidaria: Anthozoa: Octo-
corallia) supports intramolecular recombination as the primary mechanism of gene rearrangement in octocoral mitochon-
drial genomes. Genome biology and evolution, 4, 994–1006.
https://doi.org/10.1093/gbe/evs074
Brugler, M.R. & France, S.C. (2008) The mitochondrial genome of a deep-sea bamboo coral (Cnidaria, Anthozoa, Octocorallia,
Isididae): genome structure and putative origins of replication are not conserved among octocorals. Journal of Molecular
Evolution, 67, 125–136.
https://doi.org/10.1007/s00239-008-9116-2
Dueñas, L. & Sánchez, J.A. (2009) Character liability in deep-sea bamboo corals (Octocorallia, Isididae, Keratoisididae) Marine
Ecology Progress Series, 307, 11–23.
https://doi.org/10.3354/meps08307
Dueñas, L., Alderslade, P. & Sánchez, J.A. (2014) Molecular systematics of the deep-sea bamboo corals (Octocorallia: Isidi-
dae: Keratoisidinae) from New Zealand with descriptions of two new species of Keratoisis. Molecular Phylogenetics and
Evolution, 74, 15–28.
https://doi.org/10.1016/j.ympev.2014.01.031
Etnoyer, P.J. (2008) A new species of Isidella bamboo coral (Octocorallia: Alcyonacea: Isididae) from northeast Pacific sea-
mounts. Proceedings of the Biological Society of Washington, 121, 541–553.
https://doi.org/10.2988/08-16.1
Fabricius, K. & Alderslade, P. (2001) Soft corals and sea fans. Australian Institute of Marine Sciences, and the Museum and Art
Gallery of the Northern Territory, Queensland, 264 pp. [ISBN 0642322104]
Farris, J.S., Kallersjö, M., Kluge, A.G. & Bult, C. (1994) Testing significance of incongruence. Cladistics, 10, 315–319.
https://doi.org/10.1111/j.1096-0031.1994.tb00181.x
Figueroa, D.F. & Baco, A.R. (2015) Octocoral mitochondrial genomes provide insights into the phylogenetic history of gene
order rearrangements, order reversals, and cnidarian phylogenetics. Genome Biology and Evolution, 7, 391–409.
https://doi.org/10.1093/gbe/evu286
France, S.C. (2007) Genetic analysis of bamboo corals (Cnidaria: Octocorallia: Isididae): does lack of colony branching distin-
guish Lepidisis from Keratoisis? Bulletin of Marine Science, 81, 323–333.
France, S.C. & Hoover, L.L. (2002) DNA sequences of the mitochondrial COI gene have low levels of divergence among deep-
sea octocorals (Cnidaria: Anthozoa). Hydrobiologia, 471, 149–155.
https://doi.org/10.1023/A:1016517724749
Grant, R.B. (1976) The marine fauna of New Zealand: Isididae (Octocorallia: Gorgonacea) from New Zealand and the Antarctic.
Memoirs of the New Zealand Oceanographic Institute, Wellington, 66, 1–56.
Grasshoff, M. & Bargibant, G. (2001) Coral Reef Gorgonians of New Caledonia. 3rd Edition. Institut de Recherche Por le Dével-
oppement, Paris, 335 pp. [ISSN: 1286-4994]
Gray, J.E. (1858) Synopsis of the families and genera of axiferous zoophytes or barked corals. Proceedings of the Zoological
Society of London for 1857, 1858, 278–294.
https://doi.org/10.1111/j.1096-3642.1857.tb01242.x
Gray, J.E. (1870) Catalogue of the lithophytes or stony corals in the collection of the British Museum. British Museum, London,
vi + 51 pp.
https://doi.org/10.5962/bhl.title.11307
Heestand Saucier, E. (2016) Phylogenetic studies of the deep-sea bamboo corals (Octocorallia: Isididae: Keratoisidinae). PhD
Dissertation, University of Louisiana, Lafayette, Louisiana, 410 pp.
Hickson, S.J. (1907) Coelenterata. I. –Alcyonaria. Report of the Scientific Investigations of the National Antarctic Expedition,
1901-1904. British Museum (Natural History), London, England, 3, 1–15.

DECONSTRUCTING THE ISIDIDAE Zootaxa 5047 (3) © 2021 Magnolia Press · 271
Hogan, R.I., Hopkins, K., Wheeler, A.J., Allcock, A.L. & Yesson, C. (2019) Novel diversity in mitochondrial genomes of deep-
sea Pennatulacea (Cnidaria: Anthozoa: Octocorallia). Mitochondrial DNA Part A, 30, 764–777.
https://doi.org/10.1080/24701394.2019.1634699
Katoh, K., Misawa, K., Kuma, K. & Miyata, T. (2002) MAFFT: a novel method for rapid multiple sequence alignment based on
fast Fourier transform. Nucleic Acids Research, 30, 3059–3066.
https://doi.org/10.1093/nar/gkf436
Katoh, K. & Standley, D.M. (2013) MAFFT multiple sequence alignment software version 7: improvements in performance and
usability. Molecular Biology and Evolution, 30, 772–780.
https://doi.org/10.1093/molbev/mst010
Kölliker, R.A. von (1865) s.n. In: Icones histologicae. Oder Atlas der vergleichenden Gewebelehre. Zweite Abtheilung. Der fei-
nere Bau der Höheren Thiere. Erste Heft. Die Bindesubstanz der Coelenteraten. Verlag von Wilhelm Bngelmann, Leipzig,
pp. i–iv + 87–181.
https://doi.org/10.5962/bhl.title.11946
Kükenthal, W. (1915) System und Stammesgeschichte der Isididae. Zoologischer Anzeiger, 46, 116–126.
Kükenthal, W. (1919) Gorgonaria. Wissenchaftliche Ergebnisse der Deutschen Tiefsee-Expedition auf dem Dampfer “Valdivia”
1898–1899, 13, 1–946.
Kükenthal, W. (1924) Gorgonaria. In: Das Tierreich. Vol. 47. Walter de Gruyter and Co, Berlin and Leipzig, pp. 1–478.
Lamarck, J.B.D. (1816) Histoire naturelle des Animaux sans Vertèbres, présentant les caractères généraux et particuliers de ces
animaux, leur distribution, leurs classes, leurs familles, leurs genres, et la citation des principales espèces qui s'y rappor-
tent; précédée d'une Introduction offrant la Détermination des caractères essentiels de l'Animal, sa distinction du Végétal
et des autres corps naturels, enfin, l'exposition des principes fondamentaux de la Zoologie. Déterville and Verdière, Paris,
462 pp.
https://doi.org/10.5962/bhl.title.40014
Lamouroux, J.V.F. (1812) Extrait d’un mèmoire sur la classification des Polypiers coralligènes non entièrement pierreux. Nou-
veau Bulletin des Sciences, par la Société Philomatique, Paris, 3, 181–188.
Lamouroux, J.V.F. (1816) Histoire des Polypiers coralligènes flexibles, vulgairement nommès Zoophytes. F. Poisson, Caen,
lxxxiv + 560 pp.
https://doi.org/10.5962/bhl.title.11172
Lapointe, A. & Watling, L. (2015) Bamboo corals from the abyssal Pacific: Bathygorgia. Proceedings of the Biological Society
of Washington, 128, 125–136.
https://doi.org/10.2988/0006-324X-128.2.125
McFadden, C.S., Benayahu, Y., Pante, E., Thoma, J.N., Nevarez, P.A. & France, S.C. (2011) Limitations of mitochondrial gene
barcoding in Octocorallia. Molecular Ecology Resources, 11, 19–31.
https://doi.org/10.1111/j.1755-0998.2010.02875.x
McFadden, C.S., France, S.C., Sanchez, J.A. & Alderslade, P. (2006) A molecular phylogenetic analysis of the Octocoral-
lia (Cnidaria: Anthozoa) based on mitochondrial protein-coding sequences. Molecular Phylogenetics and Evolution, 41,
513–527.
https://doi.org/10.1016/j.ympev.2006.06.010
McFadden C.S. & van Ofwegen, L.P. (2012) Stoloniferous octocorals (Anthozoa, Octocorallia) from South Africa, with de-
scriptions of a new family of Alcyonacea, a new genus of Clavulariidae, and a new species of Cornularia (Cornulariidae).
Invertebrate Systematics, 26, 331–356.
https://doi.org/10.1071/IS12035
McFadden, C.S., Quattrini, A.M., Brugler, M.R., Cowman, P.F., Dueñas, L.F., Kitahara, M.V., Paz-García, D.A., Reimer, J.D.
& Rodríguez, E. (2021) Phylogenomics, origin, and diversification of anthozoans (Phylum Cnidaria). Systematic Biology,
70, 635–647.
https://doi.org/10.1093/sysbio/syaa103
Miller, M.A., Pfeiffer, W. & Schwartz, T. (2010) Creating the CIPRES Science Gateway for inference of large phylogenetic
trees. Proceedings of the Gateway computing environments Workshop, New Orleans, Louisiana, 2010, 1–8.
https://doi.org/10.1109/GCE.2010.5676129
Milne Edwards, H. & Haime, J. (1857) Histoire Naturelle des coralliaires ou Polypes proprement dits. Vol. 1. Librairie Ency-
clopèdique de Roret, Paris, 326 pp.
https://doi.org/10.5962/bhl.title.11574
Moore, K., Alderslade, P. & Miller, K. (2016) A taxonomic revision of the genus Primnoisis Studer [& Wright], 1887 (Coelen-
terata: Octocorallia: Isididae) using morphological and molecular data. Zootaxa, 4075 (1), 1–141.
https://doi.org/10.11646/zootaxa.4075.1.1
Pante, E., France, S.C., Couloux, A., Cruaud, C., McFadden, C.S., Samadi, S., & Watling, L. (2012) Deep-sea origins and in-situ
diversification of chrysogorgiid octocorals. PLoS ONE, 7, e38357.
https://doi.org/10.1371/journal.pone.0038357
Pante, E., Heestand Saucier, E., & France, S.C. (2013) Molecular and morphological data support reclassification of the octo-
coral genus Isidoides. Invertebrate Systematics, 27, 365–378.
https://doi.org/10.1071/IS12053

HEESTAND SAUCIER ET AL.
272 · Zootaxa 5047 (3) © 2021 Magnolia Press
Quattrini, A.M., Rodríguez, E., Faircloth, B.C., Cowman, P.F., Brugler, M.R., Farfan, G.A., Hellberg, M.E., Kitahara, M.V.,
Morrison, C.L., Paz-García, D.A. & Reimer, J.D. (2020) Palaeoclimate ocean conditions shaped the evolution of corals and
their skeletons through deep time. Nature Ecology & Evolution, 4, 1531–1538.
https://doi.org/10.1038/s41559-020-01291-1
Ronquist, F. & Huelsenbeck, J.P. (2003) MrBayes 3: Bayesian phylogenetics inference under mixed models. Bioinformatics,
19, 1572–1574.
https://doi.org/10.1093/bioinformatics/btg180
Rowley, S.J., Pochon, X. & Watling, L. (2015) Environmental influences on the Indo–Pacific octocoral Isis hippuris Linnaeus
1758 (Alcyonacea: Isididae): genetic fixation or phenotypic plasticity? PeerJ, 3, e1128.
https://doi.org/10.7717/peerj.1128
Studer, T. (1878) Übersicht der Anthozoa Alcyonaria, welche während der Reise S.M.S. Gazelle umdie Erde gesammelt wurden.
Monatsberichte. Konigliche Preussiche Akademie der Wissenschaften, Berlin, 1878, 632–688.
Studer, T. & Wright, E.P. (1887) Versuch eines Systemes der Alcyonaria. Archiv für Naturgeschichte, 53, 1–74.
van der Ham, J.L., Brugler, M.R. & France, S.C. (2009) Exploring the utility of an indel-rich, mitochondrial intergenic region
as a molecular barcode for bamboo corals (Octocorallia: Isididae). Marine Genomics, 2, 183–192.
https://doi.org/10.1016/j.margen.2009.10.002
Verrill, A.E. (1883). Report on the Anthozoa, and on some additional species dredged by the “Blake” in 1877–1879, and by the
U.S. Fish Commission steamer “Fish Hawk” in 1880-82. Bulletin of the Museum of Comparative Zoology, 11, 1–72.
Vielzeuf, D., Garrabou, J., Baronnet, A., Grauby, O. & Marschal, C. (2008). Nano to macroscale biomineral architecture of red
coral (Corallium rubrum). American Mineralogist, 93, 1799–1815.
https://doi.org/10.2138/am.2008.2923
Watling, L. (2015) A new genus of bamboo coral (Octocorallia: Isididae) from the Bahamas. Zootaxa, 3918 (2), 239–249.
https://doi.org/10.11646/zootaxa.3918.2.5
Watling, L. (2020) Toward a revision of the bamboo corals: Part 1, species in the Muricellisidinae (Octocorallia: Isididae).
Zootaxa, 4881 (2), 361–371.
https://doi.org/10.11646/zootaxa.4881.2.9
Watling, L. & France, S.C. (2011) A new genus and species of bamboo coral (Octocorallia: Isididae: Keratoisidinae) from the
New England Seamounts. Bulletin of the Peabody Museum of Natural History, 52, 209–220.
https://doi.org/10.3374/014.052.0202
Watling, L. & France, S.C. (2021) Toward a revision of the bamboo corals: Part 2, untangling the genus Lepidisis (Octocorallia:
Isididae). Bulletin of the Peabody Museum of Natural History. [in press]
Williams G.C. & Cairns, S.D. (2019) “Systematic List of Valid Octocoral Genera.” Octocoral Research Center. California
Academy of Sciences. Copyright 2001–2019. Available from: http://researcharchive.calacademy.org/research/izg/OCTO-
CLASS.htm (accessed 26 April 2021)
Wright, E.P. (1869) On a new Genus of Gorgonidae from Portugal. Annals and Magazine of Natural History, 4, 23–26.
https://doi.org/10.1080/00222936908695873
Wright, E.P. & Studer, T. (1889) Report on the Alcyonaria collected by H.M.S. Challenger during the years 1873-1876. Report
on the Scientific Results of the Voyage of H.M.S. Challenger during the Years 1873–76. Zoology, 31, i–lxxvii + 1–314.
Zapata-Guardiola, R., López-González, P. J. & Gili, J. M. (2012) A review of the genus Mirostenella Bayer, 1988 (Octocorallia:
Primnoidae) with a description of a new subgenus and species. Helgoland Marine Research, 67, 229–240.
https://doi.org/10.1007/s10152-012-0318-z