Several starfish (Echinodermata, Asteroidea) are keystone species of marine ecosystems, but some of the species are difficult to identify using morphological criteria only. The common sunstar, Crossaster papposus (Linnaeus, 1767), is a conspicuous species with a wide circumboreal distribution. In 1900, a closely similar species, C. squamatus (Döderlein, 1900) was described from the NE Atlantic Ocean, but subsequent authors have differed in their views on whether this is a valid taxon or rather an ecotype associated with temperature variations. We assessed the differentiating morphological characters of specimens from Norwegian and Greenland waters identified as C. papposus and C. squamatus and compared their distributions in the NE Atlantic as inferred from research cruises. The field data show that C. papposus is found mainly in temperate and shallow waters, whereas C. squamatus resides on the shelf-break in colder, mixed water masses. Intraspecific diversity and interspecific genetic differentiation of the two putative species, and their phylogenetic relationships to several Crossaster congeners worldwide, were explored using mitochondrial and nuclear DNA sequences. The molecular evidence suggests that C. papposus is the more diverse and geographically structured taxon, in line with its wide distribution. C. papposus and C. squamatus are closely related, yet clearly distinct taxa, while C. papposus and C. multispinus H.L. Clark, 1916, the latter from the South Pacific Ocean, are closely related, possibly sister taxa.
Citation: Ringvold H, Moum T (2020) On the genus Crossaster (Echinodermata: Asteroidea) and its distribution. PLoS ONE 15(1): e0227223. https://doi.org/10.1371/journal.pone.0227223
Editor: Geerat J. Vermeij, University of California, UNITED STATES
Received: October 4, 2019; Accepted: December 13, 2019; Published: January 7, 2020
Copyright: © 2020 Ringvold, Moum. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript.
Funding: The Authors received no specific funding for this work.
Competing interests: The Authors have declared that no competing interests exist.
Many starfish (Asteroidea) play important ecosystem roles as top predators, with some acting as keystone species, capable of structuring the communities in which they occur [1–4]. The common sun star Crossaster papposus (Linnaeus, 1767) is a typical representative, which belongs to the family Solasteridae and has a wide circumboreal distribution . Within the Crossaster genus a total of ten species and four subspecies (including one nomen nudum) are currently accepted by the World Register of Marine Species . Clusius  made one of the earliest records of Crossaster, which he described as “Stella tredecim radiorum”, later synonymized with Crossaster papposus. Historically, the generic designation of this species and some of its allies has alternated between Crossaster and Solaster (family Solasteridae). The Solasteridae family appears in the fossil record during the Lower Jurassic and with fairly clear generic characters today, according to Blake . Despite this, there have been several disagreements concerning the genera Solaster and Crossaster. Agassiz proposed two genera: Solaster Forbes, 1839 and Crossaster  (a genus already erected by Müller and Troschel in 1840), for the two species Solaster endeca (Linnaeus, 1771) and Solaster papposus (later Crossaster papposus) . A number of researchers disagreed, e.g. Viguier, Danielssen and Koren , Fisher (in , and Mortensen . E.g. Fisher considered Crossaster a junior synonym of Solaster, despite the different character of the marginals, the abactinal skeleton and spinelets. Today, both genera are accepted by Clark and Downey  and international expert groups .
Crossaster papposus, being a common and widely distributed species in the North Atlantic, was recognised by Carl von Linné at an early point in history. Much later, in 1900, Döderlein described a variety that differed slightly from C. papposus, and he tentatively termed it Solaster papposus var. squamata (later Crossaster squamatus (Döderlein, 1900) ). However, researchers have been unable to reach a consensus on whether C. squamatus should be considered a valid taxon or rather a morphotype of C. papposus [e.g. 13, 14]. So far, discrimination between the papposus and squamatus varieties has been based on morphological characteristics only, which could be strongly influenced by the organism’s environmental and ecological contexts. An integrated approach, considering both morphological and molecular evidence, can refine estimates of differentiation and potentially resolve taxonomic disagreements. Previously, allozyme analysis was successfully used on asteroids to separate species groups within the Henricia genus . Resolution is further improved by DNA sequence analysis, and the so-called DNA barcoding gene (the mitochondrial cytochrome oxidase subunit I gene; COI) provides a convenient target across the animal kingdom due to the simple inheritance pattern of mitochondrial DNA, and the comprehensive data available for comparison . DNA sequence analyses of COI were used to study starfish phylogeny , and Ward et al.  were able to distinguish 187 of 191 echinoderm species by their COI-based barcodes. Mitochondrial markers could be biased, however, due to introgression, lineage sorting, and selective sweeps. Thus, phylogenetic relationships are more reliably recovered by inclusion of nuclear encoded markers.
In the present study, we assessed potentially differentiating morphological characters between C. squamatus and C. papposus, based on specimens collected in the North Atlantic Ocean. The distributional patterns of the two types of Crossaster in the North Atlantic are discussed in relation to temperature and other environmental parameters. The genetic diversity of the two putative species, and the genetic differentiation between them, were evaluated based on specimens from across the Atlantic, using mitochondrial and nuclear ribosomal DNA (rDNA) sequences. We analysed samples from an additional four congeneric species collected worldwide, to allow for a more representative phylogenetic reconstruction of Crossaster relationships.
Materials and methods
Specimens and distributional data
Materials for the present study were mainly collected under the auspices of the ongoing Marine area database for Norwegian waters (MAREANO) program (www.mareano.no) in the NE Atlantic Ocean. The MAREANO program conducts physical, biological, and environmental mapping along the Norwegian coast, based on biannual research cruises. Introduction to, and results from, the 10 first years of the program are given in Buhl-Mortensen et al. , and detailed methodologies in Ringvold et al. . While echinoderms collected by MAREANO are normally stored in formalin, ethanol preserved samples for DNA analyses were provided for the present study. In addition, samples were collected during the Ecosystem survey (Institute of Marine Research) in 2016, using methods described in Jørgensen et al. , and the Greenland Initiating North Atlantic Benthos Monitoring (INAMon) program, from Melville Bay, NW Greenland in the NW Atlantic Ocean. Three Crossaster specimens were collected by HR, scuba diving in shallow waters at Gravdal, Tælavåg and Tellnes, near the city of Bergen, in 2016 (Fig 1). Close-up pictures of the asteroids' dorsal structure were taken with Andonstar 2MP USB Digital Microscope. The specimens are deposited in the collections of the University Museum of Bergen, Norway. The California Academy of Sciences (CAS), USA, and the National Institute of Water and Atmospheric Research Ltd (NIWA), New Zealand, provided ethanol preserved tissue samples (tube feet tissue or whole specimens) for DNA analysis of C. papposus, C. borealis Fisher, 1906, C. penicillatus Sladen, 1889, C. multispinus, and C. campbellicus McKnight, 1973 specimens from the Northern and Southern Pacific Ocean (Table 1).
Sampling stations in the Norwegian Sea where Crossaster papposus (red dots) and C. squamatus (green dots) specimens were recorded, and MAREANO video recordings of Crossaster spp. (black dots). Specimens for DNA analysis were also collected from NW Greenland in the NW Atlantic, and North and South Pacific Ocean. Fig1 is constructed with Quantum GIS (version 12.2.3).
Data on locations, depth and temperature were available for the Crossaster samplings mentioned above. In addition, we made use of the corresponding data associated with Crossaster specimens collected by the Marine benthic fauna of the Faroe Islands program (BIOFAR) and the Benthic Invertebrates of Icelandic Waters program (BIOICE). Information on sampling methods for these cruises is given in Ringvold and Andersen , Dauvin et al.  and Ringvold et al. (In prep.). Identification was based on Mortensen  and Clark and Downey .
DNA sequence analyses
Genomic DNA was extracted from tube feet of ethanol-preserved specimens using the DNeasy Blood & Tissue Kit (QIAGEN) according to the manufacturer’s instructions. An 841 bp fragment from the 5’ end of COI was PCR amplified using primers EchinoF1  and COIer . A fragment of the rDNA array was amplified using echinoderm targeting primers 18d9 and 5.8Sr, described by Petrov et al. , which we subsequently redesigned for increased specificity towards Crossaster: 18Scro1f (GTAGGTGAACCTGCGGAAGGATC) and 5.8Scro1rev (ATGTCGATGATCACTGCGTTCTGC). The resulting ~ 500 bp PCR product contains the variable internal transcribed spacer 1 (ITS1) of approximately 390 bp, and short flanking rDNA sequences (partial 18S, ~20 bp; partial 5.8S, ~100 bp); gene borders inferred from sequence comparisons to Asterias amurensis Lutken, 1871 (GenBank KX592567). PCR was performed in 20 μl volumes using the AmpliTaq Gold 360 system, containing 0.25 μM of each primer and 2.5 mM MgCl2. Cycling parameters for the amplification reactions were 95°C for 3 min, followed by 35 cycles of denaturation at 95°C for 0.5 min; annealing at 50°C (COI) or 62°C (rDNA) for 1 min; extension at 72°C for 1 min, and a final elongation step at 72°C for 10 min. Amplification products were sequenced on both strands using the BigDye v3.1 kit and Applied Biosystems 3500xL Genetic Analyzer.
Given the history of shifting taxonomic designations among Crossaster and Solaster species, we compiled available COI sequences (≥ 841 nucleotides) assigned to the two genera, as well as Heterozonias alternatus [originally Crossaster alternatus (Fisher, 1906)], for phylogenetic analysis. Phylogenetic relationships among the species were inferred using a representative COI sequence from each taxon and Lophaster furcilliger Fisher, 1905 (Solasteridae) for outgroup rooting. The resulting phylogeny confidently grouped C. papposus, C. multispinus and C. squamatus to the exclusion of other Crossaster and Solaster species. Thus, the phylogenetic relationships within and among C. papposus, C. multispinus and C. squamatus were further analysed using COI sequences from all specimens, and C. borealis for outgroup rooting.
Sequences were aligned using Muscle  with default parameters as implemented in MEGA X version 10.0.5 , and MEGA was further used to recover phylogenetic relationships based on mitochondrial and nuclear DNA sequences by the Maximum Likelihood method. The most appropriate model of sequence substitution for each phylogenetic analysis was determined based on the lowest Bayesian Information Criterion (BIC) score among 24 alternative models. For the interspecific phylogenetic representation based on COI, the GTR (General Time Reversible) model was selected, with 58% invariable sites and non-uniformity of evolutionary rates among variable sites modelled using a discrete gamma distribution with 5 rate categories and a gamma parameter of 0.9279. Phylogenetic trees incorporating inter-individual relationships for C. papposus and C. squamatus, were recovered separately based on COI and rDNA data. The Tamura 3-parameter model was selected for both data sets, using a gamma distribution with 5 rate categories and gamma parameters of 0.0933 and 0.1435, respectively. For each phylogenetic tree, bootstrap support values were calculated using 1,000 replicates. Intraspecific nucleotide diversities and the number of nucleotide substitutions per site between C. papposus and C. squamatus were estimated in DnaSP version 6.12.01  using the Jukes-Cantor model.
Among 26 ethanol conserved Crossaster specimens from Norwegian and Greenland waters that were chosen for morphological analysis, 6 and 20 were identified as C. papposus and C. squamatus, respectively (Tables 1 and 2). A total of 6 C. multispinus was received from New Zealand. Representative Crossaster species from around the world are shown in Fig 2.
2a. C. borealis (dorsal side) from Alaska, Bering Sea. 2b. C. borealis (ventral side, same specimen as 2a). 2c. Live C. squamatus from West of Shetland, September 2009, at 1050 m depth. Identified from video image by Daniel Jones. 2d. C. japonicus (Fisher, 1911) from NW Westport, New Zealand. 2e. Live C. papposus from Gravdal, near the city of Bergen, Norway. 2f. Ethanol preserved C. campbellicus from South New Zealand. 2g. Ethanol preserved C. multispinus from East New Zealand. 2h. Live C. papposus from Tellnes, near Bergen, Norway. 2i. Frozen C. squamatus from Barents Sea, IMR/ Ecocruise, st. 737. (Photo credits: 2a and 2b by Roger Clark, 2c by Daniel Jones/ SERPENT Project, National Oceanography Centre, 2d by Geoff Lemmey, CC license/ South Australian Museum and 2e-2i by Halldis Ringvold/ Sea Snack Norway.).
Class Asteroidea de Blainville, 1830
Superorder Valvatacea Blake, 1987
Order Valvatida Perrier, 1884
Family Solasteridae Viguier, 1878
Crossaster Müller & Troschel, 1840
Crossaster is a genus of Solasteridae with 8–15 tapering arms, moderate to large disc, and single series of single conspicuous marginals visible from dorsal view . Crossaster papposus (Fig 3A) and C. squamatus (Fig 3B) differ morphologically in several structures. In C. papposus, the dorsal skeleton consists of narrow bars forming an irregular reticulum of plates . All our purported C. papposus exhibited this structure (Fig 4). Large membranaceous spaces are formed within the reticulum, and in these spaces, several papulae can be found. The dorsal paxillae are unequal in size , and marginal paxillae largest. According to Mortensen  there are 3–5 furrow spines, which is in agreement with our specimens. Specimens of Crossaster are variable in colouration, however, the predominant aboral colour of C. papposus is purple-red, arms sometimes having a whitish and/or dark red band(s), and the oral side is usually white, which is the case for our live specimens (Fig 2E and 2H).
a (left). Whole, conserved specimens of small Crossaster papposus, and b. C. squamatus. They are recorded from MAREANO stations 1218–471 (R = 2 cm), and 1086–438 (R = 1,8 cm), respectively. (Photo credit: Arne Hassel/ Institute of Marine Research.).
Specimen recorded at MAREANO station 1218–471 (R = 2 cm). The dorsal skeleton sample is cut out and photographed from below. The arrows show papulae within membranaceous space. (Photo credit: Halldis Ringvold/ Sea Snack Norway.).
In C. squamatus, the dorsal skeleton is scale-like, formed by irregularly shaped plates with little or no membranaceous spaces, as seen in the MAREANO specimens (Fig 5), and with only singular papula. The aboral paxillae are equal in size, and shorter than for C. papposus . There are 5–7 furrow spines. The aboral color is usually orange-red, and arms sometimes with orange or red colored bands (Fig 2C and 2I), and the oral side yellowish-white .
Specimen recorded at MAREANO station 1086–438 (R = 1,8 cm). The dorsal skeleton sample is cut out and photographed from below. The arrows show papulae within membranaceous space. (Photo credit: Halldis Ringvold/ Sea Snack Norway.).
The dorsal skeleton of the borrowed C. multispinus consists of narrow bars forming an irregular reticulum of plates with large membranaceous spaces, as in C. papposus, and also described by Clark . However, the papulae are few and isolated. Marginal paxillae of C. multispinus are largest, and dorsal paxillae are unequal in size, as for C. papposus. C. multispinus has 8–9 adambulacral spines and 10 furrow spines, whereas C. papposus has 6–7 adambulacral spines and 3–5 furrow spines. According to Clark  C. multispinus has 11 arms with R = 4 cm and r = 2 cm, whereas C. papposus, according to Clark & Downey , is a larger species with 11–14 arms and R = 5,5 cm and r = 3 cm.
The most striking feature of C. papposus and its closest relatives (C.papposus/ C. squamatus/ C. multispinus), is the shape of the paxillae (Table 3). They all have what could be referred to as high metapaxillae, that is paxillae with high columnar plate. This is in contrast to the phylogenetic cluster containing Heterozonias alternatus and associated species, of which species have paxillae with low columnar plate. Paxillae for several of these latter species are in literature named pseudopaxillae, low metapaxillae or small paxillae [e.g. 5, 31, 32].
Results from the three marine surveillance programs MAREANO, BIOICE and BIOFAR indicate that C. papposus is mainly recorded from the shelf, in temperate water, whereas C. squamatus occurs at the shelf-break in colder water (Figs 6 and 7).
Distribution of Crossaster papposus (red dots) and C. squamatus (green dots) recorded by the BIOFAR and BIOICE programs [22, Ringvold et al. In prep.].
Phylogenetic reconstruction of Crossaster and Solaster species based on COI sequences from representative specimens of each species, suggested two clearly defined main groups with high bootstrap support (Fig 8). Firstly, C. squamatus, C. papposus and C. multispinus formed a highly supported clade, and a sister group relationship between C. papposus and C. multispinus was supported by a bootstrap value of 85. Secondly, H. alternatus, C. penicillatus, C. borealis, and C. campbellicus constituted a robust group, while the relationships within this group were less confidently resolved (bootstrap values ≤ 64). The phylogeny of Solaster species was not confidently resolved by this analysis due to low bootstrap support values.
The phylogeny was inferred from mitochondrial COI sequences using the Maximum Likelihood method and Lophaster furcilliger as the outgroup. The tree with the highest log likelihood (-3528,66) is shown. The percentage of bootstrap replicates in which the associated taxa clustered together is shown next to the branches. The tree is drawn to scale, with branch lengths measured by the number of substitutions per site.
The relationships between Crossaster species were further scrutinized using both mitochondrial COI sequences and nuclear rDNA sequences, taking the inter-individual variations of C. squamatus and C. papposus into account. The phylogenetic reconstruction based on COI and outgroup rooted by C. borealis recovered C. papposus and C. squamatus as clearly separate units with high bootstrap support, and the grouping of C. multispinus and C. papposus gained further support by a bootstrap of 84. While C. squamatus exhibited low genetic differentiation among individuals, C. papposus showed evidence of phylogeographic structuring (Fig 9).
Relationships among Crossaster papposus and C. squamatus individuals, and C. multispinus, as inferred from mitochondrial COI sequences using the Maximum Likelihood method and C. borealis as the outgroup. The tree with the highest log likelihood (-1972,69) is shown. The percentage of bootstrap replicates in which the associated taxa or individuals clustered together is shown next to the branches. The tree is drawn to scale, with branch lengths measured by the number of substitutions per site.
The phylogenetic reconstruction based on nuclear rDNA sequences recovered C. papposus and C. squamatus as clearly separate taxa, in line with the evidence based on mitochondrial gene sequences. Again, C. multispinus clustered closely with C. papposus, but in contrast to the results of the mitochondrial gene based analyses, C. multispinus grouped among the C. papposus specimens rather than branching off as a separate lineage. Interestingly, it clustered most closely with the C. papposus specimen from the Pacific Ocean, though at a low bootstrap value (Fig 10). The other main group among Crossaster species, consisting of C. borealis, C. campbellicus, and C. penicillatus, was retained in the rDNA based phylogeny, though with a different internal branching order.
Phylogenetic relationships among C. papposus and C. squamatus individuals, and representative C. multispinus, C. borealis, C. penicillatus, and C. campbellicus, was inferred from nuclear rDNA sequences using the Maximum Likelihood method. The tree with the highest log likelihood (-12678,68) is shown. The percentage of bootstrap replicates in which the associated taxa or individuals clustered together is shown next to the branches. The tree is unrooted and drawn to scale, with branch lengths measured by the number of substitutions per site.
We found 3 COI haplotypes among 13 C. squamatus individuals, compared to 13 haplotypes among 23 C. papposus individuals, and nucleotide diversities of 0.15% and 0.78%, respectively. The amphioceanic C. papposus had similar nucleotide diversities in the Atlantic (0.29%, N = 13) and the Pacific (0.30%, N = 8), and a higher nucleotide diversity (0.57%, N = 5) than C. squamatus in their common distributional area in the NE Atlantic. The net number of nucleotide substitutions per site between C. papposus from the Pacific and the Atlantic/Arctic was 0.91%. The alignment of rDNA sequences for Crossaster species contained a total of 537 nucleotide positions and arrived at 413 nucleotide positions excluding gaps. Considering the 413 positions only, there were 4 genotypes among C. papposus, while a single genotype only was observed for C. squamatus. There were 15 fixed nucleotide differences between C. papposus and C. squamatus. Overall, the number of net nucleotide substitutions per site between the two putative species was 0.063 and 0.039 for COI and rDNA, respectively.
All of the Crossaster DNA sequences recovered by the present study are referred to in Table 1. Sequences were deposited in GenBank with the accession numbers KX451838-KX451847; MK270376-MK270394; MK203712-MK203739.
Species delineation and phylogeny
Among the species currently assigned to the genus Crossaster, three species were originally described in the Atlantic: C. papposus, C. helianthus and C. penicillatus. As C. penicillatus was found in the SE Atlantic and Southern Ocean only, and the only record of C. helianthus to date is that of the holotype from Georges Bank in 1880, it would seem that C. papposus is the predominant representative of the genus in the North Atlantic, and the only one in the NE Atlantic. On the other hand, observations made by several authors suggested that variation could be contained within the C. papposus clade itself. Sladen  described the variety C. papposus var. septentrionalis, but based on a single specimen only, recorded from the Faroe Channel (-0,5⁰C). In 1900, Döderlein described individuals from Eggakanten in northern Norway, differing slightly from C. papposus in external morphology, which he tentatively termed Solaster papposus var. squamata (later Crossaster squamatus (Döderlein, 1900)). Subsequent researchers have held differing opinions as to whether the papposus and squamatus varieties should be considered valid species [11, 12, 13, 14, 39] or rather morphotypes associated with temperate (C. papposus) and colder (C. squamatus) waters [13, 14]. C. squamatus was maintained as a valid taxon in the North Atlantic Ocean (Rockall Trough) by A. M. Clark in Gage et al.  and Clark , but in Clark and Downey’s  compilation “Starfishes of the Atlantic”, it is omitted. The reason it was not included in this book could be that the authors considered it primarily as an Arctic species and, given the geographic constraint of the book, would avoid including species that did not occur in the Atlantic. C. squamatus is currently listed as an accepted species by WoRMS .
Here, we examined and unambiguously classified Crossaster specimens from the NE Atlantic as either C. squamatus or C. papposus based on external characteristics. Correspondingly, molecular markers representing both the mitochondrial and nuclear genomes clearly demonstrated that the two belong to separate lineages. The mitochondrial COI showed the typical level of divergence to be expected from a between species comparison. This is in agreement with the findings of Ward et al.  that the intraspecific divergence of echinoderm species ranged from 0 to about 3% with a mean of 0.62%, while congeneric divergence averaged 15.33%, based on COI sequences.
Nucleotide sequences of the 18S and 28S rDNA genes are traditionally utilized for phylogenetic inference and are able to resolve distant relationships due to their highly conserved primary sequence across metazoans. For the same reason, however, they are less informative for phylogenetic inference of closely related species. The variable ITS sequences contained within rDNA gene arrays have been less extensively used for phylogenetic inference, but the phylogenetic information content we were able to extract from ITS1 proved useful for investigating the closer relationships within the genus Crossaster. For C. papposus and C. squamatus, the present results based on the nuclear rDNA markers are in line with those based on mitochondrial sequence data, exhibiting some intraspecific variation and a clear phylogenetic resolution of the two putative species.
We analyzed six out of ten currently accepted Crossaster species worldwide, and recovered two major clades. The analysis suggests that C. papposus and C. squamatus belong to the same clade, as expected, while C. papposus and C. multispinus is the more closely related, possibly sister species. Heterozonias alternatus, C. penicillatus, C. borealis, and C. campbellicus constitute a robust group, which suggest that taxonomic classification within the same genus may be warranted. According to this analysis, neither Crossaster nor Solaster, as currently classified, constitutes natural clades. Instead, it suggests that H. alternatus, C. penicillatus, C. borealis, and C. campbellicus constitute a group nested among Solaster species, while the placement of Solaster species in the phylogeny remains uncertain due to low bootstrap support. A more comprehensive sampling of the genus, including the Pacific C. scotophilus (Fisher, 1913), C. japonicus, and C. diamesus (Djakonov, 1932), as well as other solasterid species, and a multigene approach, will be needed to further resolve the phylogenetic relationships of Crossaster species.
So far, molecular data on Crossaster species are scarce. A previous study  based on partial sequences of two mitochondrial rDNA genes (12S and 16S) and one nuclear protein coding gene (early stage histone H3) failed to recover the close relationship between C. papposus and C. multispinus that we identified in the present study. This previous study, however, was aimed at resolving higher order relationships among asteroids rather than the finer twigs of the phylogenetic tree. Indeed, upon reanalysis of the available 12S, 16S and histone H3 sequence data of Crossaster species only (C. papposus, C. multispinus and C. borealis), and using Lophaster furcilliger as the outgroup, we found a close relationship between C. papposus and C. multispinus, and a branching order of the species included in perfect agreement with the results of the current study.
C. multispinus has been recorded from the South Pacific Ocean only, in specific from South and Southeast Australia (Gabio Island and Disaster Bay), Tasmania, Macquarie Island and New Zealand [30, 44, www.iobis.org]. Thus, it is evident from the currently available data that the widely distributed amphioceanic C. papposus, has at least one closely related representative, C. multispinus and C. squamatus, in each of the Pacific and Atlantic Oceans. It would be interesting to further investigate the population histories and differentiation of the closely related C. papposus, C. multispinus and C. squamatus, and to identify their adaptative genetic variation, using a population genomic approach.
Biogeography of C. papposus and C. squamatus
Although the current sampling of C. papposus and C. squamatus remains limited both in terms of geographic extent and the number of individuals, it seems evident that the two species differ widely in their distribution, as well as their population genetic parameters. C. papposus is genetically more diverse than C. squamatus in terms of the number of haplotypes and nucleotide diversities. The molecular phylogenies showed evidence of geographic structure for C. papposus in that specimens from the Pacific Ocean clustered separately with strong bootstrap support and specimens from Baffin Bay/Greenland clustered tightly with those from the Barents Sea. Also, the branching patterns of specimens from the Arctic and Norwegian Sea are compatible with geographic structuring, but further sampling of individuals would be required to establish a proper phylogeography of the species.
Our estimate of 0.91% sequence divergence of COI between trans-Arctic C. papposus is in line with other recent estimates, 1.24% and 1.03%, obtained by  and , respectively. The level of divergence is relatively low compared to several other trans-Arctic sister clades and suggests a recent separation dating back some 3–400 000 years with a divergence rate of 2.8%/million years . The observation made by Loeza-Quintana & Adamowicz  that trans-Arctic interchange seems to be favoured by taxa that have shallow versus those that have deep water distributions is in line with the depth distributions of C. papposus and C. squamatus. Thus, C. papposus would seem to be a highly abundant and widely distributed species, with a higher potential for dispersal, but able to adapt locally and diversify, which could entail incipient speciations. In contrast, C. squamatus seems to lack in numbers and genetic diversity, maybe due to more restricted habitats, lack of dispersal capabilities, special adaptations, and a competitive disadvantage compared to C. papposus. Also, we note that branch lengths of the molecular phylogenies suggest that C. papposus experiences higher molecular evolutionary rates than those of its close relatives, C. squamatus and C. multispinus. Higher rates in C. papposus could be related to shorter generation times, which in turn could be due to higher temperatures and concomitant increase in developmental rates, but we are not aware of any data on generation times in Crossaster species so far, to support or contradict such a speculation.
Based on the specimens analysed here, we found a consistent morphological differentiation between C. papposus and C. squamatus and a corresponding genetic differentiation, with no evidence of introgression between the two. We did identify a few specimens, however, with a combination of dense dorsal structures (as for C. squamatus) and several papulae within each membranaceous space (as for C. papposus), generally with 5–7 and occasionally 3–5 furrow spines, during our previous examination of Crossaster specimens from the three marine surveillance programs mentioned (BIOFAR, BIOICE and MAREANO). Such a combination of morphological characteristics from both taxa is suggestive of hybridization. Further molecular analyses, targeting both mitochondrial and nuclear genes, would be required to resolve the issue, but the specimens currently available are preserved in formalin and therefore less appropriate for DNA analyses.
The geographic distribution of Crossaster papposus and C. squamatus overlap. The distribution of Crossaster papposus in the Atlantic Ocean is south along the east coast of north America, from Newfoundland and Labrador to about 40°N; Spitsbergen, north to Nordaustlandet (Barents Sea); in the NE Atlantic from Scandinavia (Finnmark, including Tromsøflaket, and south all along the Norwegian coast) to the southern North Sea, all around the British Isles, Iceland south to northern Brittany, all around the Faroe Islands on the shelf [5, 11, 22, 47–50, Ringvold et al. In prep.; this study]. According to records from CASIZ database it also occurs at the strait of Gibraltar in Spain. C. papposus is also widely distributed in the North Pacific .
C. squamatus has been recorded in Norwegian waters from Finnmark and Eggakanten (Fig 1), south to the border of Nordland and Trøndelag Counties (65°N) [12, this study], by Hansson  stated as “NW-NE Finnmark, slope of Norway S., south to 60° N”. It is also distributed all around the Faroe Islands, including the Faroe Channel; Iceland; north to Nordaustlandet in the archipelago of Svalbard; western Barents Sea; east and west of Greenland and south to the Hebridean slope (56°N) [11, 14, 22, 29, 40, 49, 53–59, Ringvold et al. in prep.]. It has also been recorded from the NW Atlantic, Newfoundland, Baffin Bay and Smith Sound [48, 60]. Worldwide, C. papposus is recorded from 0–1200 m depth [5, 11], and C. squamatus from cold water areas, from 100–1600 m depth [11, 50]. However, since previous opinions have differed with respect to taxonomic assignment of Crossaster found in the NE Atlantic, and misidentifications might have occurred, our understanding of the geographic distribution, as well as the depth distribution, might be subject to change in the future.
While there is a distributional overlap of C. papposus and C. squamatus, the abundance-weighted mean depth is shallower for C. papposus than for C. squamatus, based on morphologically identified, mainly formalin preserved materials, from the Faroe Islands (BIOFAR), Iceland (BIOICE), and Norway (MAREANO). C. papposus was found on the shelf in temperate water masses, whereas C. squamatus showed abundance-weighted mean depth below the shelf break (below 500 m depth), in the transition zone with mixed, colder water masses, including negative temperatures [22, Ringvold et al. In prep.] (Fig 7). At a few MAREANO stations, specimens were identified as C. cf. papposus. If omitting these questionable specimens, the average depth for C. papposus would have been even shallower, as observed in BIOFAR and BIOICE data. Zoogeographical analysis in Einarsson  rests mainly on several of Th. Mortensens publications (e.g. Mortensen ), supporting our findings, in placing C. papposus as part of the arctic-boreal fauna, and C. squamatus as part of the Arctic deep basin fauna.
Correspondingly, Döderlein’s  recordings of C. papposus from 11 stations (mainly Olga expedition) and the holotype of C. squamatus (North-Sea Expedition, st. 200) show that C. papposus at Svalbard is also distributed close to the shore (36–200 m depth), whereas the one recording of C. squamatus was at 1134 m depth, at -1°C (Fig 11). Jones et al.  recorded only C. squamatus (by uv-photo and video images) in the deep Faroe Channel, at stations ranging from ~ 1000 m to 1200 m depth. The bottom waters in the channel at depths below 800 m is ~ -1° C .
Recordings of Crossaster papposus (red dots) and C. squamatus (green dots), the former species sampled by the Olga expedition, and the latter from the North-Sea Expedition . C. papposus is distributed close to the Svalbard shore, whereas the one recording of C. squamatus was at the shelf-break.
Several features, such as distribution, shape of calcareous ossicles, and genetic differentiation, have been associated with temperature for several other species as well, and studies have related the distribution of macro-invertebrates to water mass as defined by both temperature and salinity [64–66]. Previous studies from the Norwegian Sea have shown that the transition zone, an area with mixture of water masses, represents maximum species diversity, and a major shift in benthic species composition (e.g. [20, 67, 68]). Temperature alone is also an important abiotic factor regarding distribution of benthic species, including Asteroidea (e.g. [69–71]). Important faunal boundaries, found globally, are believed to occur around the shelf/slope break at 200–500 m, and around 1000–1400 m depth (e.g. [20, 67, 72]). Gage  found a comparative echinoderm faunal boundary at 800–1000 m in the Rockall Trough, and Howell et al.  asteroid faunal boundaries at Porcupine Seabight at 110 m, ~700 m, and 1700 m. In both studies the boundaries were related to both depth of the thermocline and water mass structure.
A study of the deep-water amphipod Eurythenes gryllus (Lichtenstein in Mandt, 1822) suggests that depth (or pressure), together with topography, is a significant driver in allopatric (= geographic) speciation where populations become separated and isolated over a long period, and interfering with genetic interchange due to e.g. different selective pressures or mutations of the different populations. Three distinct morphological forms of the species have been detected, varying in terms of pereonites and pleonites, the shape of coxa 2, and the first and second gnathopods . Temperature has been discussed as a controlling ecological factor in the deep sea [74–76], also to E. gryllus, and may lead to genetic differentiation and speciation . This conclusion is reasonable when comparing bathyal to abyssal populations due to distinct bottom temperatures, but not regarding populations in the abyssal and hadal trenches with more similar temperatures . The distribution of Crossaster papposus and C. squamatus overlap to some extent, but they seem to prefer different depth zones with different temperatures, hence both abiotic factors (temperature and depth) may have contributed to the differentiation of the two.
Temperature is also suggested to cause changes in the calcareous skeleton/plates in e.g. Bryozoa. This is seen in the species complex Watersipora spp., generating phylogroup-specific fragments. Warm water colonies show irregular, multilobed morphology compared to cold-water colonies, which are more regular and circular in shape . The same was observed in the asteroid genus Bathybiaster where the dorsal plates of the warm water form are described as star-shaped and overlap, whereas the cold water form shows round plates which do not overlap. Grieg  therefore suggested a warm- and cold-water species within the asteroid genus Bathybiaster, namely B. vexillifer (W. Thomson, 1873) and B. robustus (Verrill, 1894). The two species were synonymized by Koehler , and followed by Mortensen  and Fisher , but maintained by Clark . Today, B. robustus is synonymized with B. vexillifer (www.marinespecies.org). Similarly, it has been suggested that the morphological differentiation between C. papposus and C. squamatus might be caused by temperature [11–14, 40, 83], but the genetic distances and phylogeny revealed by the present study makes it evident that they belong to clearly divergent lineages, consistent with species status.
While Ocean Acidification (OA) is generally considered a major threat to marine ecosystems, lowered pH was shown to increase growth of C. papposus developmental stages . In contrast, one of its main prey species, Asterias rubens Linnaeus, 1758, which has also been shown to be more prone to diseases during OA due to immune suppression, will be negatively impacted . At the Faroe Islands, C. papposus and its prey A. rubens, are found mainly on the shelf, in relatively warm water and strong currents (40–90 cm s-1) , as supported by Gale et al.  in finding a maximum density depth for both species between 0 and 100 m in a study from Atlantic Canada. In the same area, C. squamatus occurs mostly at deeper and colder waters with weaker currents (12–41 cm s-1), indicating it may have different food preferences. Bearing these differences in mind, it should be interesting to explore niche separation between the species, and whether they are differentially affected by OA and other human induced environmental changes. Changes in the population sizes of these species, which are both top predators, could have important consequences for community structure and functioning.
This article is dedicated to the German zoologist and palaeontologist Professor Ludwig Heinrich Philipp Döderlein (1855–1936). He was one of the foremost echinoderm researchers of his time, publishing 43 papers on the taxon Echinodermata .
The Institute of Marine Research (IMR), the Geological Survey of Norway and the Norwegian Mapping Authority comprise the Executive Group, which is responsible for carrying out the MAREANO field sampling and other scientific activities. We thank IMR/MAREANO and Eco cruises, Bergen University Museum, Sten-Richard Birkely, Arne Hassel, Kjell Bakkeplass, Anne Kari Sveistrup, Trude H. Thangstad, Heidi Gabrielsen, and the crew on G. O. Sars for providing specimens and station data from the Barents Sea. We also thank BIOFAR, BIOICE, CAS, NIWA, Martin Blicher and INAMon program for providing specimens. INAMon was founded at the Greenland Institute of Natural Resources (GINR) in 2014 in collaboration with partners from Canada, Iceland, Faroe Islands, Norway and Russia, with the specific aim to map and monitor benthic invertebrates, and to stimulate buildup and exchange of knowledge between Arctic countries (e.g. ). Vigdis Edvardsen performed DNA isolations, PCR, and DNA sequencing, which we highly appreciate. We thank Andrey Voronkov for translation of Russian text, and Daniel Jones, Bill Frank and Roger Clark for pictures. Sincere thanks to Paul E. Renaud for improving an earlier draft of the manuscript.
- 1. Paine RT. Food web complexity and species diversity. Am Nat. 1966;100: 65–75.
- 2. Paine RT. A short-term experimental investigation of resource partitioning in a New Zealand rocky intertidal habitat. Ecology. 1971;52: 1096–1106.
- 3. Lawrence M. Starfish, Biology and Ecology of the Asteroidea. Baltimore: The Johns Hopkins University Press; 2013.
- 4. Menge BA, Sanford E. Ecological role of sea stars from populations to meta-ecosystems. Starfish: Biology and Ecology of the Asteroidea. In: Lawrence JM, editor. Starfish, Biology and Ecology of the Asteroidea. Baltimore: The Johns Hopkins University Press; 2013. pp. 67.
- 5. Clark AM, Downey ME. Starfishes of the Atlantic. London: Chapman & Hall; 1992.
- 6. Mah C.L. (2019). World Asteroidea Database. Crossaster Müller & Troschel, 1840. Accessed through: World Register of Marine Species at: http://www.marinespecies.org/aphia.php?p=taxdetails&id=123336 on 2019-12-08
- 7. Clusius C de [l’Escluse]. Exoticorum libri decem: quibus Animalium, Plantarum–describuntur; 1605.
- 8. Blake DB. A new asteroid genus from the Jurassic of England and its functional significance. Palaeontology. 1993;36: 147–154.
- 9. Agassiz A. North American Starfishes. Memoirs of the Museum of Comparative Zoology at Harvard College. Cambridge: Welch, Bigelow, and Company; 1877.
- 10. Danielssen DC, Koren J. Remarks on the genus Solaster. Ann Mag Nat Hist. 1882;10: 436–443.
- 11. Mortensen T. Handbook of the Echinoderms of the British Isles. London: Oxford University Press; 1927.
- 12. Döderlein L. Die Echinodermen. Zoologische Ergebnisse einer Üntersuchungsfahrt des deutschen Seefischerei-Vereins nach der Bäreninsel und Westspitzbergen ausgeführt im Sommer 1898 auf S.M.S. “Olga”. Wiss Meeresunters. 1900;4: 195–248.
- 13. Östergren H. Über die arktischen Seesterne. In: Korschelt E, editor. Zool Anz 27. Leipzig: Verlag von Wilhelm Engelmann; 1904. pp 614–616.
- 14. Grieg JA. Echinodermen von dem norwegischen Fichereidampfen “Michael Sars” in den Jahren 1900–1903 gesammelt. Bergens Mus Aarb. 1906;13: 1–87.
- 15. Ringvold H, Stien J. Biochemical differentiation of two groups within the species-complex Henricia Grey, 1840 (Echinodermata, Asteroidea) using starch-gel electrophoresis. Hydrobiol. 2001;459: 57–59.
- 16. Hebert PDN, Cywinska A, Ball SL, DeWaard JR. Biological identifications through DNA barcodes. Proc R Soc B Biol Sci. 2003;270: 313–321.
- 17. Knott KE, Wray GA. Controversy and Consensus in Asteroid Systematics: New Insights to Ordinal and Familial Relationships. Amer Zool. 2000;40: 382–392.
- 18. Ward RD, Holmes BH, O’Hara TD. DNA barcoding discriminates echinoderm species. Mol Ecol Res. 2008;8: 1202–1211.
- 19. Buhl-Mortensen L, Hodnesdal H, Thorsnes T. The Norwegian Sea Floor. New Knowledge from MAREANO for Ecosystem-Based Management. Trondheim: Skipnes Kommunikasjon AS; 2015.
- 20. Ringvold H, Hassel A, Bamber RN, Buhl-Mortensen L. Distribution of sea spiders (Pycnogonida, Arthropoda) off northern Norway, collected by MAREANO. Mar Biol Res. 2015;11: 62–75.
- 21. Jørgensen LL, Ljubin P, Skjoldal HR, Ingvaldsen RB, Anisimova N, Manushin I. Distribution of benthic megafauna in the Barents Sea: baseline for an ecosystem approach to management. ICES Journ Mar Sci. 2015;72: 595–613.
- 22. Ringvold H, Andersen T. Starfish (Asteroidea, Echinodermata) from the Faroe Islands; spatial distribution and abundance. Deep Sea Res. 1. 2015;107: 22–30.
- 23. Dauvin J-C, Alizier S, Weppe A, Gudmundsson G. Diversity and zoogeography of Icelandic deep-sea Ampeliscidae (Crustacea: Amphipoda). Deep-Sea Res 1. 2012;68: 12–23.
- 24. Arndt A, Marquez C, Lambert P, Smith MJ. Molecular phylogeny of Eastern Pacific sea cucumbers (Echinodermata: Holothuroidea) based on mitochondrial DNA sequence. Mol Phyl Evol. 1996;6: 425–437.
- 25. Petrov NB, Vladychenskaya IP, Drozdov AL, Kedrova OS. Molecular genetic markers of intra- and interspecific divergence within starfish and sea urchins (Echinodermata). Biochemistry (Moscow). 2016;81: 972–980.
- 26. Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research. 2004; 32: 1792–1797. pmid:15034147
- 27. Kumar S, Stecher G, Li M, Knyaz C, Tamura K (2016) MEGA X: Molecular Evolutionary Genetic Analysis across computing platforms. Molecular Biology and Evolution. 2018;35: 1547–1549. pmid:29722887
- 28. Rozas J, Ferrer-Mata A, Sánchez-DelBarrio JC, Guirao-Rico S, Librado P, Ramson-Onsins SE, et al. DnaSP v6: DNA sequence polymorphism analysis of large datasets. Mol Biol Evol. 2017;34: 3299–3302. pmid:29029172
- 29. Djakonov AM. Starfish of the Soviet Union. Tabl Anal Faune URSS. 1950;34: 1–203.
- 30. Clark HL. Report on the sea-lilies, starfishes, brittle-stars and sea-urchins obtained by the F.I.S. "Endeavour" on the coasts of Queensland, New South Wales, Tasmania, Victoria, South Australia, and Western Australia. Biological Results of the Fishing experiments carried on by the F.I.S. Endeavour 1909–1914. Sydney: WE Smith Ltd; 1916.
- 31. Fisher WK. New starfishes from the Pacific Coast of North America. Proceedings of the Washington Academy of Sciences. 1906:8: 111–139.
- 32. Verrill AE. Monograph of the shallow-water starfishes of the North Pacific coast of America from the Arctic Ocean to California. Harriman Alaska Series of the Smithsonian Institute 1914:14: 1–408.
- 33. McKnight DG. The Marine Fauna of New Zealand Echinodermata: Asteroidea (Sea-stars). 3 Orders Velatida, Spinulosida, Forcipulatida, Brisingida with addenda to Paxillosida, Valvatida. Wellington: NIWA (National Institute of Water and Atmospheric Research; 2006.
- 34. McKnight DG. Additions to the asteroid fauna of New Zealand: families Radiasteridae, Solasteridae, Pterasteridae, Asterinidae, Ganeriidae, and Echinasteridae. New Zealand Oceanographic Institute Records. 1973:2: 1–15.
- 35. Sladen WP. Report on the Asteroidea. Report on the Scientific Results of the Voyage of H.M.S. Challenger during the years 1873–1876. Zoology. 30, 51. 1889.
- 36. Fisher WK. New starfishes from the Philippine Islands, Celebes, and the Moluccas. Proceedings of the United States National Museum. 1913:46: 201–224.
- 37. Fisher WK. New starfishes from deep water off California and Alaska. Bulletin of the Bureau of Fisheries. 1905:24: 291–320.
- 38. Sladen WP. Asteroidea dredged during the cruise of the «Knight Errant» in July and August 1880. Proceedings of the Royal Society of Edinburgh. 1882:11: 698–707.
- 39. Ringvold H. Artsbestemming og utbredelse av sjøstjerner (Echinodermata: Asteroidea) rundt Færøyene. M.Sc. thesis (In Norwegian), Norwegian University of Science and Technology 1996.
- 40. Gage JD, Pearson M, Clark AM, Paterson LJ, Tyler PA. Echinoderms of the Rockall Trough and adjacent areas. 1. Crinoidea, Asteroidea and Ophiuroidea. Bull Br Mus Nat Hist Zool. 1983;45: 263–308.
- 41. Clark AM. 1996. An index of names of recent Asteroidea: Part 3. Velatida and Spinulosida. In: Jangoux M, Lawrence JM, editors. Echinoderm Studies Vol. 5. Rotterdam: M. A. A. Balkema; pp.183–250.
- 42. Mah, CL. World Asteroidea Database. Crossaster squamatus (Döderlein, 1900). Accessed through: World Register of Marine Species at: http://www.marinespecies.org/aphia.php?p=taxdetails&id=124155 on 2019-12-08
- 43. Mah CL, Foltz D. Molecular phylogeny of the Valvatacea (Asteroidea: Echinodermata). Zool Journ Linn Soc. 2011;161: 769–788.
- 44. Mah CL, McKnight DG, Eagle MK, Pawson DL, Améziane N, Vance DJ, et al. Phylum Echinodermata: sea stars, brittle stars, sea urchins, sea cucumbers, sea lilies. In: Gordon DP, editor. New Zealand inventory of biodiversity: 1. Kingdom Animalia: Radiata, Lophotrochozoa, Deuterostomia. Christchurch: Canterbury University Press; 2009. pp. 371–400.
- 45. Layton KKS, Corstorphine EA, Hebert PDN. Exploring Canadian echinoderm diversity through DNA barcodes. PLOS ONE. 2016; 1–16.
- 46. Loeza-Quintana T, Adamowicz SJ. Iterative calibration: A novel approach for calibrating the molecular clock using complex geological events. J Mol Evol. 2018;86: 118. pmid:29429061
- 47. Dons C. Norges strandfauna. Asteroider. Kgl Norske Vidensk Selsk Forh. 1935;8: 29–32.
- 48. Grainger EH. Sea stars (Echinodermata: Asteroidea) of arctic North America. In: Stevenson JC, editor. Fisheries Research Board of Canada Bull nr. 152; 1966. pp. 1–70.
- 49. Gulliksen B, Palerud R, Brattegard T, Sneli JA. Distribution of marine benthic macro-organisms at Svalbard (including Bear Islands) and Jan Mayen. Research report for DN 1999–4. Trondheim: Directorate for Nature Management; 1999.
- 50. Anisimova NA, Cochrane SJ. An annotated checklist of the echinoderms of the Svalbard and Franz Josef Land archipelagos and adjacent waters. Sarsia. 2003;88: 113–135.
- 51. Fisher WK. Asteroidea of the North Pacific and adjacent waters. 1. Phanerozonia and Spinulosa. Bull US Natn Mus. 1911; 76.
- 52. Hansson HG. European Echinodermata Check-List. A draft for the European Register of Marine Species (part of "Species 2000") compiled at TMBL (Tjärnö Marine Biological Laboratory). 1999.
- 53. Bell FJ. Catalogue of the British Echinoderms in the British Museum (Natural History). London; 1892.
- 54. Grieg JA. Echinodermata. Report of the Second Norwegian Arctic Expedition in the “Fram”, 1898–1902. 1907;2: 1–28.
- 55. Murray J, Hjort J. The Depths of the Ocean. London: Mackmillan & Co; 1912.
- 56. Hofsten N. Die Echinodermen des Eisfjords. Zoologische Ergebnisse der Schwedischen Expedition nach Spitzbergen 1908. K Sven vetensk akad handl. 1915;54: 1–282.
- 57. Schorygin AA. Die Echinodermen des Barentsmeeres. Trudy Plovychego Morskogo Instituta. 1928;3: 1–131.
- 58. Lieberkind I. Echinoderma. Zool Faroes. 1929;3: 1–20.
- 59. Harvey R, Gage JD, Billett DSM, Clark AM, Paterson GLJ. Echinoderms of the Rockall Trough and adjacent areas. 3. Additional records. Bull Br Mus Nat Hist Zool. 1988;54: 153–198.
- 60. Haedrich RL, Maunder JE. The echinoderm fauna of the Newfoundland continental slope. In: Keegan BF and O’Connor BDS, editors. Echinodermata. Proceedings of the fifth international Echinodermata conference, Galway 24–29 September 1984. Rotterdam: AA Balkema; 1984. pp. 37–46.
- 61. Einarsson H. Echinoderma. In: Fridriksson A, Tuxen SL, editors. The Zoology of Iceland Vol 4. Copenhagen and Reykjavik: Ejnar Munksgaard; 1948.
- 62. Jones DOB, Bett BJ, Tyler PA. Megabenthic ecology of the deep Faroe-Shetland channel: A photographic study. Deep Sea Res 1. 2007;54: 1111–1128.
- 63. Turrell WR, Slesser G, Adams RD, Payne R, Gilibrand PA. Decadal variability in the composition of Faroe Shetland Channel bottom water. Deep-Sea Res 1. 1999;46: 1–25
- 64. Stewart PI, Pocklington P, Cunjak RA. Distribution, Abundance and Diversity of Benthic Macroinvertebrates on the Canadian Continental Shelf and Slope of Southern Davis Strait and Ungava Bay. Arctic. 1985;38: 281–291.
- 65. Anisimova NA. Distributional pattern of Echinoderms in the Eurasian Sector of the Arctic Ocean. In: Herman Y, editor. The Arctic Seas, Climatology, Oceanography, Geology, and Biology. New York: Van Nostrand Reinhold; 1989. pp. 281–301.
- 66. Copley JTP, Tyler PA, Sheader M, Murton BJ, German CR. Megafauna from sublittoral to abyssal depths along the Mid-Atlantic Ridge south of Iceland. Oceanol Acta (0399–1784) (Gauthier-Villars). 1996;19: 549–559.
- 67. Howell KI, Billett DSM, Tyler PA. Depth-related distribution and abundance of seastars (Echinodermata: Asteroidea) in the Porcupine Seabight and Porcupine Abyssal Plain, N.E. Atlantic. Deep-Sea Res 1. 2002;49: 1901–1920.
- 68. Gebruk AV, Budaeva NE, King NJ. Bathyal benthic fauna of the Mid-Atlantic Ridge between the Azores and the Reykjanes Ridge. J Mar Biol Assoc UK. 2010;90: 1–14.
- 69. Apellöf A. Int. Congr Zool Graz. Jena: G Fisher; 1912.
- 70. Franz DR, Worley EK, Merrill AS. Distribution patterns of common seastars of the middle atlantic continental shelf of the northwest atlantic (Gulf of Maine to Cape Hatteras). Biol Bull. 1981;160: 394–418.
- 71. Gage JD. The benthic fauna of the Rockall Trough: regional distribution and bathymetrical zonation. Proc Soc Edinb. 1986;88: 159–174.
- 72. Sanders HL, Hessler RR. Ecology of the deep-sea benthos. Science. 1969;163: 1419–1424. pmid:5773106
- 73. Jamieson A. The Hadal zone. Life in the Deepest Oceans. Cambridge: Cambridge University Press; 2015. pp 243–244.
- 74. France SC, Kocher TD. Geographic and bathymetric patterns of mitochondrial 16S rRNA sequence divergence among deep-sea amphipods, Eurythenes gryllus. Marin Biol. 1996;126: 633–643.
- 75. Wilson GDF, Hessler RR. Speciation in the deep sea. A Rev Ecol Syst.1987;18: 185–207.
- 76. France SC. Genetic population structure and gene flow among deep–sea amphipods, Abyssorchomene spp., from six California continental borderland basins. Mar Biol. 1994;118: 67–77.
- 77. Palumbi SR. Genetic divergence, reproductive isolation, and marine speciation. Annu Rev Ecol Syst. 1994;25: 547–572.
- 78. Korcheck K. Variation in temperature tolerance in a widely invasive Bryozoan species complex (Watersipora spp.). M.Sc. Thesis, The Faculty of Humboldt State University. 2015.
- 79. Grieg JA. Remarks on the Age of some Arctic and North-Atlantic Starfishes. Ann Mag Nat Hist. 1919;3: 400–408.
- 80. Koehler R. Echinodermes provenants des campagnes du yacht U Hirondelle u´Res. camp sci Monaco, Fasc. XXXIV. 1909. pp. 34:57.
- 81. Fisher WK. Asteroidea. Discovery Rep. 1940;20: 69–306.
- 82. Clark HL. The distribution and derivation of some New England echinoderms. Amer Nat. 1923;57: 229–237.
- 83. Brandt JF. Prodromus descriptionis animalum ab H. Mertensio observatorum. Petropoli; 1835.
- 84. Dupont A, Lundve B, Thorndyke M. Near Future Ocean Acidification Increases Growth Rate of the Lecithotrophic Larvae and Juveniles of the Sea Star Crossaster papposus. J Exp Zool (Mol Dev Evol). 2010;314B: 382–389.
- 85. Hernroth B, Baden S, Thorndyke M, Dupont S. Immune suppression of the echinoderm Asterias rubens (L.) following long-term ocean acidification. Aquat Toxicol. 2011;103: 222–224. pmid:21473849
- 86. Gale KS, Gilkinson K, Hamel JF, Mercier A. Patterns and drivers of asteroid abundances and assemblages on the continental margin of Atlantic Canada. Mar Ecol. 2014;36: 734–752.
- 87. Scholz J, Hoeksema BW, Pawson DL, Ruthensteiner B. Ludwig Döderlein (1855–1936): Some aspects of his life, research, and legacy. Spixiana. 2012;35: 177–191.
- 88. Birk MH, Blicher ME, Garm A. Deep-sea starfish from the Arctic have well-developed eyes in the dark. Proc R Soc B. 2018;285: 20172743. pmid:29436504