Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Deep sea fish wikipedia , lookup
The Marine Mammal Center wikipedia , lookup
Marine life wikipedia , lookup
Marine pollution wikipedia , lookup
Marine microorganism wikipedia , lookup
Marine biology wikipedia , lookup
Marine habitats wikipedia , lookup
Ecosystem of the North Pacific Subtropical Gyre wikipedia , lookup
Journal of Plankton Research plankt.oxfordjournals.org Unexpected prevalence of parasite 18S rDNA sequences in winter among Antarctic marine protists ALISON C. CLEARY1,2†* AND EDWARD G. DURBIN1 1 2 GRADUATE SCHOOL OF OCEANOGRAPHY, UNIVERSITY OF RHODE ISLAND, NARRAGANSETT, RI, USA AND UNIVERSITY CENTER IN SVALBARD, LONGYEARBYEN, SVALBARD, NORWAY † PRESENT ADDRESS: MARINE BIOLOGICAL ASSOCIATION, PLYMOUTH, UK *CORRESPONDING AUTHOR: [email protected] Received August 7, 2015; accepted January 12, 2016 Corresponding editor: John Dolan Parasites are not typically considered to be important components of polar marine ecosystems. It was therefore surprising when 18S rDNA surveys of protists in the West Antarctic Peninsula in winter revealed high abundances of parasite sequences. Parasite sequences made up, on average, over half (52%) of sequence reads in samples from deep water in winter. Winter surface water and sediment samples contained relatively fewer, but still strikingly high, parasite sequence reads (13 and 9%, respectively), while surface water samples in summer contained fewer parasite sequences (1.8%). A total of 1028 distinct parasite Operational Taxonomic Units were observed in winter, with the largest abundances and diversities within Syndiniales groups I and II, including Amoebophrya. Less abundant parasite sequence groups included Apicomplexa, Blastodinium, Chytriodinium, Cryptocaryon, Paradinium, Perkinsidae, Pirsonia and Ichthyophonae. Parasite sequence distributions suggested interactions with known hosts, such as diatom parasites which were mainly in the sediments, where resting spores of Chaetoceros spp. diatoms were abundant. Syndiniales sequences were correlated with radiolarian sequences, suggesting parasite–host interactions. The abundant proportions of parasite sequences indicate a potentially important role for parasites in the Antarctic marine ecosystem, with implications for plankton population dynamics, the role of the microbial loop, carbon flows and ecosystem responses to ongoing anthropogenic climate change. KEYWORDS: 18s rDNA; parasites; protists; plankton; West Antarctic Peninsula; Syndiniales I N T RO D U C T I O N The West Antarctic Peninsula (WAP) has traditionally been thought of as a comparatively simple ecosystem, with the food web dominated by relatively few types of organisms and interactions; large phytoplankton being consumed by krill, which were in turn consumed by penguins, seals available online at www.plankt.oxfordjournals.org # The Author 2016. Published by Oxford University Press. All rights reserved. For permissions, please email: [email protected] Featured Article J. Plankton Res. (2016) 38(3): 401–417. First published online Febraury 24, 2016 doi:10.1093plankt/fbw005 JOURNAL OF PLANKTON RESEARCH j VOLUME and whales (Huntley et al., 1991). More recently, the roles of smaller phytoplankton, microzooplankton, bacteria and the ensuing microbial loop have been increasingly recognized (Garrison, 1991; Sailley et al., 2013). Alternative carbon pathways to the diatom–krill–penguin model are thought to be particularly important outside of the productive, and well-studied, spring bloom period (Ducklow et al., 2007). The microbial loop is now known to play an important role in carbon cycling and food web dynamics in the region, with bacterioplankton converting dissolved organic matter (DOM) into biomass, which is in turn consumed by small flagellates and other heterotrophic protists, creating an alternative, and more complex, pathway for carbon to travel before reaching a size available to krill and other macrofaunal predators (Ducklow et al., 2007; Sailley et al., 2013). Even as the role of many of these smaller organisms, such as bacterioplankton and heterotrophic flagellates, in the WAP marine ecosystem has been increasingly recognized, one category of small marine organisms which continues to escape attention are the parasitic protists. The abundance of protistan parasites is thought to be under-accounted for in marine planktonic and sediment systems generally (Sousa, 1991; Skovgaard, 2014). Protistan parasites can be difficult to observe, as they are small and spend much of their lives hidden within their hosts. These parasites may be invisible when analyzing preserved plankton samples (Skovgaard and Daugbjerg, 2008). Estimates of protistan parasite abundance or biomass in natural ecosystems are scarce. Analysis of metazoan parasites in estuaries showed they accounted for 2 – 3% of the total biomass (Kuris et al., 2008), while in modeling studies, when included at all, total parasite abundance has been much lower, at 0.01– 0.02% of total biomass (Arias-Gonzáles and Morand, 2006). Parasite diversity is also poorly known, particularly for protistan parasites, and particularly in the marine environment. For many parasitic protist groups, there are only rough estimates of the total species diversity and abundance. Even for described parasite organisms, the range of possible hosts and the complete life cycle are often unknown. It has been suggested that parasitic lifestyles should encourage species diversity, such that there may be an equal or greater diversity of parasites than of traditional predators in natural ecosystems (Toft, 1986). Many protist groups which have been identified as obligate parasites include only a handful of described species (Chambouvet et al., 2014). Yet, 18S ribosomal DNA (rDNA) molecular surveys suggest these groups may be highly diverse and contain hundreds or more species (Chambouvet et al., 2014; Skovgaard, 2014). Protistan parasites in the marine environment include a diverse range of organisms from several major eukaryotic j NUMBER j PAGES – j lineages. The Alveolata are particularly rich in marine parasites. Marine Alveolate groups include the Syndiniales and Amoebophrya which infect a variety of marine plankton, including radiolarians, dinoflagellates and copepods, and Euduboscquella which parasitizes radiolarians. Other alveolate parasites in the marine environment include Perkinsidae which parasitize shellfish and dinoflagellates, and parasitic ciliates which infest crustaceans, fish and gelatinous plankton, as well as other parasite groups (Skovgaard, 2014). Apicomplexan parasites, including gregarines, parasitize marine crustaceans, polycheates, mollusks and others, while Rhizaria parasites include Paradinium, which infects copepods (Skovgaard, 2014). Additional parasites from the amoebozoa, fungi, stramenopiles and euglenozoa have also been observed in the marine environment (Skovgaard, 2014). In the Antarctic, clone library studies have shown large numbers of potential parasite sequences, with the “vast majority of sequences” from deep water plankton samples belonging to Marine Alveolate (MALV) groups I and II (López-Garcı́a et al., 2001). Although these protistan parasites are poorly known, they may play important roles in ecosystem dynamics. Protistan parasites can reduce the abundances of their hosts, affecting their population dynamics, and inducing mortality rivaling in magnitude that from zooplankton grazing (Sousa, 1991; Coats, 1999; Coats and Park, 2002). Such reductions in the populations of hosts can also have potential impacts on the predators of these hosts. By killing, or reducing the fitness of, hosts in a speciesspecific and density-dependent manner, selective protistan parasites may also play a role in maintaining ecosystem diversity (Hudson et al., 2006). Protistan parasite activity can lead to increased release of DOM and particulate organic matter (POM) from their hosts as they put them under increased physiological stress, or cause them to lyse (Skovgaard, 2014). Such organic matter may be taken up by bacteria or small eukaryotes, fueling the microbial loop. Thus, understanding the roles of these parasites is important to our overall understanding of carbon flows and trophic interactions. Newly generated next-generation sequencing (NGS) data from the WAP marine ecosystem offered an opportunity to investigate the diversity, spatial and temporal distributions, and relative abundance of protistan parasites. Millions of 18S rDNA barcode sequences from the microeukaryote (0.2– 5000 mm) communities of the water column and sediments in winter and the water column in summer were used to obtain a first glimpse into the types of parasites present in the region, how these parasites are distributed in this ecosystem and in relation to potential hosts, how parasite distributions change seasonally and how much of the protist DNA in this ecosystem might be attributed to parasites. 402 A. C. CLEARY AND E. G. DURBIN j ABUNDANT PARASITE 18S IN ANTARCTIC WINTER METHOD Field collections Samples were collected in winter between 18 May and 3 June 2013 on the Research Vessel/Ice Breaker Nathaniel B. Palmer cruise NBP1304. In summer, samples were collected between 10 and 21 December 2014 on the same vessel on cruise NBP1410. Water samples were collected in 12 L Niskin bottles on a CTD rosette. Surface water was collected from the surface mixed layer at 20 m depth in winter and at 5 – 10 m depth in summer (Table I). Deep water was collected at 10 m above the seafloor within fjordic-bays and straits, and at 600 m in Palmer Deep basin (bottom depth 1345 m) (Table I). In winter, for each water sample, triplicate 2 L subsamples of whole seawater were filtered by a peristaltic pump onto 0.2 mm membrane filters, thus collecting all organisms or pieces of organisms between 0.2 mm and 5 mm diameter (the diameter of the pump tubing). In summer, for each water sample, duplicate 100 or 200 mL subsamples of whole seawater were filtered by gentle vacuum onto 0.2 mm membrane filters. Much smaller volumes were used in summer due to higher phytoplankton biomass. In winter, surface and deep water samples were collected from Flandres Bay, Gerlache Strait and Palmer Deep; surface water only was collected in Andvord Bay (Fig. 1, Table I). In summer, only surface water samples were collected, and these were analyzed from Andvord, and Flandres’ Bays, and from Palmer Deep and the Gerlache Strait. Filters were placed in individual cryovials and immediately frozen at 2808C. Temperature and salinity were recorded simultaneously with sample collection with an SBE 911plus CTD (SeaBird). Chlorophyll a was measured fluorometrically from extracted pigments in triplicate on a TD-700 fluorometer (Turner Designs) (Jespersen and Christoffersen, 1987). Sediment samples were collected in winter with a megacorer (Ocean Scientific Instruments Limited). In order to expose the sediment surface, overlaying water was gently removed by a peristaltic pump and cores were extruded to just below the level of the sediment surface. The surface-most layer of sediment (3 mm) was sampled using sterile disposable scrapers, placed in cryovials and immediately frozen at 2808C. Sediment samples collected all organisms less than 5 mm diameter. Sediment was sampled from two locations in Wilhelmina Bay, two locations in Andvord Bay, one location in Flandres Bay and one location in Palmer Deep (Fig. 1, Table I). Individual samples were collected from three separate cores from one of the corings in Andvord Bay, and triplicate samples were collected from a single core in each of the other corings. All samples were shipped from Chile to Rhode Island on dry ice and stored at 2808C until analysis. Bottom depth was recorded by shipboard Chirp 3260 echo sounder (Knudsen). Laboratory processing Total DNA was extracted from filters with the DNeasy blood and tissue kit (Qiagen). Volumes of the initial lysis buffers were all doubled to ensure the filter was submerged and all material was lysed. Total DNA was extracted Table I: Sampling locations and metadata Samples Type Date Time (local) Depth Latitude Longitude Location Map icon Temp (8C) Salinity (psu) 1, 2, 3 4, 5, 6 7, 8, 9, 79 10, 11, 12 13, 14, 15 16, 17, 18 19, 20, 21 22, 23, 24 25, 26, 27 28, 29, 30 31, 32, 33 34, 35, 36 37, 38, 39 66, 67 68, 69 70, 71 72, 73 74, 75 Water Water Water Water Water Water Water Sediment Sediment Sediment Sediment Sediment Sediment Water Water Water Water Water 24 May 29 May 29 May 1 June 1 June 3 June 3 June 18 May 22 May 26 May 27 May 30 May 1 June 10 December 17 December 18 December 19 December 21 December 23:05 11:15 11:15 10:40 10:40 00:50 00:50 18:25 18:00 20:20 16:00 16:30 13:20 09:00 09:40 09:00 00:25 14:20 20 250 20 600 20 300 20 520 628 545 356 725 1345 5 5 10 5 10 264.8469 265.0468 265.0468 264.9668 264.9668 264.7949 264.7949 264.6857 264.5350 264.8123 264.8105 265.0030 264.9667 264.8245 264.9364 264.8954 265.0567 264.8149 262.6105 263.3014 263.3014 264.3549 264.3548 263.1214 263.1214 262.2350 262.2350 262.7333 262.7175 263.3110 264.3548 262.6412 264.3579 263.719 263.2035 262.6721 Andvord Bay Flandres Bay Flandres Bay Palmer Deep Palmer Deep Gerlache Strait Gerlache Strait Wilhemina Bay Wilhemina Bay Andvord Bay Andvord Bay Flandres Bay Palmer Deep Andvord Bay Palmer Deep Gerlache Strait Flandres Bay Andvord Bay 1 2 2 3 3 4 4 5 6 7 7 8 3 9 10 11 12 13 20.89 0.72 20.77 1.43 20.86 20.24 20.58 33.66 34.47 33.64 34.66 33.65 34.51 33.84 0.26 0.02 0.18 0.01 0.26 0.02 0.24 0.19 21.33 21.03 0.20 0.51 33.86 33.33 33.39 33.55 34.07 6.8 0.42 1.42 13.14 Chlorophyll Sample numbers correspond to independent water filters or sediment samples, except in the case of 79 which is a technical replicate for 9. Dates in May and June are in 2013, dates in December are in 2014, time is in Chilean local 24 h time, depth is in meters, map symbol corresponds to Fig. 1, chlorophyll a is in mg m23. Salinity and temperature data are not available for sediment samples. 403 JOURNAL OF PLANKTON RESEARCH j VOLUME j NUMBER j PAGES – j Fig. 1. Map of sampling locations. Numbers in the figure correspond to map icon numbers in Table I. from 0.25 g of each sediment sample using the PowerSoil kit (MoBio) with the bead-vortex lysis option as per the manufacturer’s instructions. Only samples of the same type (water/sediment) were extracted on the same day. All extractions were conducted in a sterilized laminar flow hood with project-dedicated pipettes, tips and chemicals, to minimize possible contamination. Partial 18S rDNA was amplified over the v7 region using universal eukaryotic primers (960f and 1200r of Gast et al., 2004), with the following slight modifications. One variable position (Y) was substituted for a T in the Read1 primer because the primers as originally published have been observed not to efficiently amplify Mesodinium spp. and related ciliates due to a 1 bp mismatch (A.C.C., unpublished data). This modification has been applied in previous analyses amplifying a wide range of marine protists (Cleary et al., 2015). Adaptors for Illumina sequencing were added to the primers, along with a variable number (0 – 3) of ambiguous bases to offset the amplicons and increase the variability at each read position for improved base calling, with the read 1 adaptor on the reverse primer and the read 2 adaptor on the forward primer (Read1 primer: TCGTCGGCAG CGTCAGATGTGTATAAGAGACAGfNNNgGGCTY AATTTGACTCAACRCG; Read2 primer: GTCTCGT GGGCTCGGAGATGTGTATAAGAGACAGfNNNg GGGCATCACAGACCTG). In summer, the variable number of NNNs was not included in the primers as advances in illumina chemistry and base calling had rendered these obsolete. For winter samples, each reaction contained a final concentration of 1 Pfu Ultra II clear buffer (Agilent), 1 bovine serum albumin (New England Biolabs), 0.25 mM equimolar mixture of all four deoxynucleotide triphosphates (dNTPs) (Promega), 0.1 mM each primer (forward and reverse), 1 Pfu Ultra II polymerase (Agilent) and 20% by volume DNA template at extracted concentrations. Summer samples were amplified with 1 GoTaq green master mix, 0.1 mM each primer (forward and reverse) and 20% by volume DNA template at extracted concentrations. Differences in the fidelities of Taq and Pfu polymerase mean comparisons of OTU richness should be viewed with caution; however, these polymerase differences are unlikely to affect the proportions of sequence reads recovered from broad taxonomic groups. Thermocycling for all samples consisted of 958C for 30 s, followed by cycles of 948C for 30 s, 588C for 45 s and 728C for 30 s, with a final extension of 728C for 5 min. Samples were amplified for the minimum number of cycles necessary to obtain sufficient amplicon DNA for 404 A. C. CLEARY AND E. G. DURBIN j ABUNDANT PARASITE 18S IN ANTARCTIC WINTER sequencing, in order to reduce amplification biases and over-representation of abundant targets. Winter water samples were all amplified for 35 cycles, as were sediment samples 28, 29, 30, 35 and 37, while the remaining sediment samples and the summer water samples were amplified for 30 cycles. Amplicon presence and size (230–280 bp, excluding primers and adaptors) was confirmed with gel electrophoresis and UV visualization. Amplifications of no-template blanks included in each PCR showed no signs of contaminating DNA in gel images. Amplicon purification, tag addition and sequencing were done at the URI Genomics and Sequencing Center. Amplicons were AmPure (Agencourt) cleaned, re-amplified in six cycles of PCR to add unique sample identification tags and quantified on a BioAnalyzer (Agilent). All winter amplicons were pooled into one half Illumina Miseq run, and sequenced for 2 250 cycles with V2 chemistry, allowing for almost complete overlap of the amplicon. Summer amplicons were included in a separate sequencing run (along with 75 other samples) with the same chemistry and parameters as the winter samples. Each filter and each 0.25 g sediment sample were sequenced separately, and results are shown for all filters individually, with biological replicates shown side by side. Data analysis Data were demultiplexed in the Illumina default pipeline, and analyses commenced from these demultiplexed reads. Paired ends of reads were joined if the entire overlap region was 100% identical in both read directions; if the overlap region was not identical, both reads were discarded. Amplicons were then assigned sample-specific names for each of the demultiplexed samples, and all samples within a season pooled for further analysis. At this stage, additional quality control on the sequences was applied, following default parameters in split_libraries.py. Primers, and any sequence data beyond the end of the amplicons were trimmed by removing the reverse primer and any following sequence, then reverse complementing the sequences and repeating the procedure, for a final file of trimmed, 50 -30 oriented sequences. Any sequences in which both the exact primer sequences were not found were again discarded. This fairly stringent approach to quality control likely eliminated most sequencing errors, as such errors are unlikely to occur identically in the two directions of sequencing. Amplicons were then clustered into 97% sequence identity Operational Taxonomic Units (OTUs) with Uclust (Edgar, 2010). In analysis of environmental 18S sequences, 97% is a common clustering threshold (Sogin et al., 2006). Any OTU with only a single sequence in it was discarded, as such sequences may be erroneous, and even if real, add significantly to computation time without typically changing the overall results of NGS data analysis. OTUs with chimeric sequences were detected with the blast-fragments approach with a two fragment search parameter (Altschul et al., 1990), and removed from the data. All of the above analyses were conducted in Qiime v.1.8 with default parameters unless otherwise specified above (Caporaso et al., 2010). Taxonomic identity was assigned to each OTU through automated comparison in Qiime with the Silva database v. 111 (Wang et al., 2007; Quast et al., 2013) and confirmed for parasite OTUs by comparison with the PR2 database (Guillou et al., 2012). By both database comparisons, 94% of parasite OTUs sequence reads were classified into the same groups and only results from Silva are presented here. Taxonomic level of identification varied between OTUs, due to variation in the completeness of reference sequence sets, and the genetic variability within the amplicon. Selected BLAST searching of the full GenBank database was used to confirm taxonomic assignments (Altschul et al., 1990; Morgulis et al., 2008). Because multicellular organisms are unlikely to be quantitatively sampled in the small volumes of water and sediment analyzed here, and thus their appearance in samples was noisy and stochastic, these OTUs, which accounted for 9% of the initial winter data, were also removed from the final data set. Removing chimeras and metazoans reduced the overall winter data set by 3085 OTUs and 826 742 sequence reads. All OTUs were next manually classified as parasitic, non-parasitic or unknown based on literature. Literature searches were run on taxa name and “parasite”, beginning with kingdom level groups, and proceeding to lower taxonomic levels. OTUs were only considered parasitic if all available literature on any organism in the group agreed with the designation of parasitism, or if literature explicitly stated that all members of a group were parasitic. If no reports of parasitism within a group could be found, or if even at the lowest taxonomic level sequences were resolved to, known members of the group had both parasitic and other lifestyles, organisms were considered free living. This approach leads to a conservative estimate of parasitism as it excludes parasites which are closely related phylogenetically to non-parasites. To compare the communities of parasites and free living organisms, a dissimilarity matrix of all winter samples was constructed using the Bray– Curtis metric (Bray and Curtis, 1957) based on the sequence read counts for each of the parasite OTUs, normalized by the overall ( parasite and free-living) total sequence read for each sample. Likewise, a dissimilarity matrix was constructed for all free-living OTUs. Parasite and free-living communities were compared in a side-by-side cluster analysis 405 JOURNAL OF PLANKTON RESEARCH j VOLUME based on Bray– Curtis distances. Principal coordinates for the parasite assemblages in each sample (metric multidimensional scaling) were calculated and visualized in MatLab. ANOSIM was used to determine the significance of groupings of parasite OTU assemblages by source type (surface water, deep water, sediment) as implemented by the Fathom toolbox in MatLab (Jones, 2014). Interactions between parasite OTUs and potential host OTUs were explored using an OTU-wise Bray– Curtis metric for all winter samples. Read counts were normalized by the total for each OTU across all samples (as opposed to the total for each sample across all OTUs used above), in order to reduce the effects of variations in copy number or organism size This approach creates in effect a co-occurrence network of OTUs, with no a priori assumptions of host– parasite interactions, and is similar to that taken by Lima-Mendez et al. (Lima-Mendez et al., 2015). Parasite – host interaction analysis was limited to the 500 most abundant OTUs overall to avoid artifacts due to stochasticity at low abundance. Parasite – host analysis also excluded organisms for which taxonomy was not resolved below kingdom level, or with very phylogenetically close parasites and hosts (such as within the dinoflagellates) to avoid erroneous correlations due to potentially imperfect OTU picking. Selected parasite – host pairs identified in this exploratory analysis were further investigated with linear correlations of OTU read counts normalized by sample. R E S U LT S A total of 1028 parasite-associated OTUs which encompassed 309 192 sequence reads were recovered from the Antarctic coastal environment in winter, out of a total sample of 8539 OTUs and 1 990 675 cleaned sequence reads. In the much smaller summer surface water data set, only 184 parasite OTUs were identified, encompassing 11 723 sequences, out of a total sample of 3249 OTUs and 632 351 reads. In winter, 393 of these OTUs were found in surface water samples, 592 were found in deep water samples and 564 were found in sediment samples. Winter sequence reads from parasite-associated OTUs made up between 5.6% of the total sequence reads in Sample 22 (Wilhelmina Bay sediment) and 73.0% of the total sequence reads in Sample 11 (Palmer Deep deep water), with an overall winter mean of 21.4% and median of 13.8% of the sequence reads in a sample. In summer, parasites ranged in relative abundance from 0.9% (Flandres Bay surface waters) to 2.5% (Palmer Deep surface water), with an overall summer mean and median of 1.8%. Only two parasite OTUs were found in every winter sample and both were classified as j NUMBER j PAGES – j Syndiniales I (MALV I). A total of 108 OTUs were present in all sample types in winter (surface water, deep water, sediment), 83 of which were in MALV I and II. Deep water winter samples showed the highest percentages of parasite-associated OTU reads, with 52.3% of the deep water reads falling into these groups. About 13.1% of winter surface water sequences were classified as parasite-associated OTUs. Sediments showed the lowest relative abundance of parasite reads in winter, with 9.1% of total reads classified into parasite-associated OTUs (Table II). Parasite OTUs included a diverse range of organisms. Most of the parasite OTUs belonged to the alveolates. The most abundant group of sequences belonged to the Syndiniales, making up 11.3% of the total sequences, and 62.2% of the parasite sequences. Within the syndinians, organisms classified solely as belonging with Syndiniales group II made the greatest contributions to both number of reads and number of OTUs. None of these parasites could be classified to lower taxonomic levels, suggesting a large biodiversity which remains underexplored. Also within this group, Amoebophyra spp. (MALV II) were present mainly in water samples, but also in sediments. In the MALV I, Euduboscquella spp. were found mainly in deep water, and Haplozoon spp. were found to be most abundant in surface waters (Table II, Fig. 2). In MALV IV, we observed Hematodinium spp. (MALV IV), found mainly in surface water samples. Dinoflagellate parasites observed included: Chytriodinium spp., found mainly in water samples, and Blastodinium spp., found in all sample types (Table II, Fig. 2). The related Perkinsidae were observed mainly in the sediment. Apicomplexa OTUs were predominantly gregarines, with water samples containing mainly a Lankesteria spp. eugregarine, while sediment samples, where Apicomplexa were most prominent, were more mixed in their apicomplexan assemblage with notably more Lecudina spp. eugregarines, and more Cryptosporida spp. Apicomplexa also included three OTUs which could not be classified to lower taxonomic levels, again suggesting potentially currently unknown diversity within this group. Other alveolate parasites found included Cryptocaryon spp. ciliates which were observed in all sample types, and ellobiopsids which were found mainly in water samples (Table II, Figs 2 and Fig. 3). Some parasites were also observed from Rhizaria. Rhizarian parasites included Paradinium spp. which was observed across all sample types, and phytomyxea which was found mainly in sediments. Two groups of parasites belonging to the stramenopiles were observed; Pirsonia spp. was found mainly in deep water and sediments, and Solenicola spp. was found mainly in surface waters. The only parasite group found which did not fall within the 406 A. C. CLEARY AND E. G. DURBIN j ABUNDANT PARASITE 18S IN ANTARCTIC WINTER Table II: Parasite groups encountered in 18S rDNA sequences from the WAP Winter Group #OTUs % Surface % Deep % Sediment Summer % Surface Amoebophrya (MALV II) Apicomplexa Blastodinium Chytriodinium Cryptocaryon Euduboscquella (MALV I) Ellobiopsidae Haplozoon Hematodinium (MALV IV) Ichthyophonae Paradinium Perkinsidae Phytomyxea Pirsonia Solenicola Syndiniales, unclassified Syndiniales group I (MALV I) Syndiniales group II (MALV II) Totals 124 216 4 3 17 5 4 9 9 8 19 88 19 7 1 54 137 304 1028 2.766 0.058 0.047 0.001 0.002 0.003 0.002 1.185 0.006 0.004 0.029 0.003 0.001 0.001 0.001 0.683 4.835 3.499 13.125 3.687 0.130 0.008 0.001 0.015 0.377 0.340 0.390 0.210 0.020 0.036 0.062 0.002 0.032 0.000 1.401 9.562 36.063 52.337 0.013 3.004 0.003 0.000 0.032 0.024 0.000 0.000 0.002 0.027 0.055 0.589 0.132 0.037 0.000 0.463 1.645 0.277 9.144 0.376 0.003 0.000 0.000 0.000 0.000 0.000 0.017 0.000 0.003 0.000 0.000 0.000 0.001 0.000 0.051 1.055 0.349 1.828 Example known hosts Ref Dinoflagellates, ciliates, Arthropods, polychaetes, Chaetognaths Copepods Copepod eggs Fish Cilliates Crustaceans Marine worms, mollusks, larvacea Crustaceans Fish, bivalves Copepods Mollusks, dinoflagellates Diatoms Diatoms Diatoms Copepods, radiolarians, amphipods, larvacea, 1, 15, 16, 18 1, 2, 19 1, 3, 13,17 1 5 1, 16,17 1, 3 6, 14, 17 3 1, 7, 17 1, 8 9, 2 10 1, 4, 11 1 1, 12, 17 # of OTUs is the total number of observed distinct OTUs for each group. % indicates the percent of the total sequence reads in each sample type which were attributable to each group. References: (1) Skovgaard (2014); (2) Moreira and López-Garcı́a (2003); (3) Konovalova (2008); (4) Tillmann et al. (1999); (5) Wright and Colorni (2002); (6) Leander et al. (2002); (7) Glockling et al. (2013); (8) Skovgaard et al. (2008); (9) Chambouvet et al. (2014); (10) Neuhauser et al. (2011); (11) Kühn et al. (2004); (12) Guillou et al. (2008); (13) de Vargas et al. (2015); (14) Sousa (1991); (15) Lima-Mendez et al. (2015); (16) Chambouvet (2009); (17) Coats (1999); (18) Cachon (1964); (19) Takahashi et al. (2011). Fig. 2. Distribution of sequence reads for parasite and free-living organism OTUs across all winter samples. Upper rectangles indicate sample source (surface water, deep water, sediment); lower rectangles indicate sample location (Andvord Bay, Flandres Bay, Gerlache Strait, Palmer Deep), sample numbers along the x-axis correspond to Table I and are arranged within sample type by increasing bottom depth. The high percentage of parasites in sequencing reads from deep water samples is striking. Stramenopiles –Alveolata – Rhizaria (SAR) complex was the holozoan protist Ichthyopohonae spp. which was observed mainly in deep water and sediment samples (Table II, Figs 2 and 3). Parasites showed different assemblages in winter between surface waters, deep waters and sediments. These differences were statistically significant with ANOSIM P ¼ 0.001, for all possible pairwise comparisons. Surface 407 JOURNAL OF PLANKTON RESEARCH j VOLUME j NUMBER j PAGES – j Fig. 3. Parasite sequence reads across all winter samples. Formatting follows Fig. 2. The high contribution from Syndiniales and related organisms (Amoebophrya and Hematodinium) is striking, particularly in surface and deep water samples. waters contained large fractions of Syndiniales and subgroups, as well as noticeable contributions from haplozoons (Fig. 3). Deep water parasite assemblages were dominated by Syndiniales and subgroups, particularly Group II, with relatively low contributions from other groups. Sediment parasite assemblages showed much lower abundances of syndinians than was observed in the water column, with the assemblage composed largely of Apicomplexa, particularly eugregarines, with contributions from Perkinsidae as well as small contributions from Syndiniales and subgroups (Fig. 3). The notable exception to this trend is Palmer Deep, where samples showed relatively more Syndiniales, particularly Group I. Palmer Deep also showed differences within the Apicomplexa, with all three surface water samples containing eugregarine sequences of an OTU, not prominent in other samples, which showed taxonomic affinities to eugregarines isolated from flatfish. These differences in assemblages are evident in a principal coordinates analysis (Fig. 4). Samples cluster most strongly by sample type, but within sample type also show clustering by location. Parasite assemblages are closely tied to non-parasite assemblages. Parasite OTU assemblages exhibited very similar clustering patterns to non-parasite OTU assemblages (Fig. 5). Summer surface water parasite prevalences were over a factor of 6 lower than those observed for similar samples in winter. The observed assemblage of parasite types, however, was similar (Fig. 6, Table II). Both summer and winter surface water sample parasite assemblages were dominated by syndinians. In winter, there were roughly equal contributions from Group I, Group II and Amoebophrya spp., while in summer, a lower relative abundance of Amoebophrya was observed, and Syndiniales Group I made the largest fraction of the parasite sequences, particularly in the samples within fjordic-bays. Haplozoon spp. were at lower relative abundances in the parasite assemblage in summer when compared with winter. Considering relationships between specific parasites and potential hosts, the strongest correlations observed were between syndinians and radiolarians. Twelve syndinian OTUs correlated with r 2 .0.9 with individual radiolarian OTUs (Fig. 7). Considering all syndinians (Groups I and II) and all radiolarians as groups, the positive linear correlation between the percent abundances of these groups over all samples showed a slope of 9.4 and an r 2 of 0.94 (Fig. 7). DISCUSSION Parasite OTU spatial distributions indicate active ecological interactions between specific parasite groups and their hosts in the WAP. Parasite OTU assemblages clustered in a nearly identical pattern to free-living OTU assemblages, potentially indicating close biological interactions between parasite and host species or groups (Fig. 5). The Syndiniales (Groups I and II), the most abundant parasite sequence groups observed, were correlated with distributions of radiolarians, with both of these groups having their highest relative abundance in deep 408 A. C. CLEARY AND E. G. DURBIN j ABUNDANT PARASITE 18S IN ANTARCTIC WINTER Fig. 4. Principal coordinates analysis (metric multidimensional scaling) of the winter parasite communities sampled. Shape indicates the location of each sample, with shade indicating the sample type. Sample type is clearly the dominant structuring factor among these assemblages. water samples. Summer research at a station within a few miles of our sampling region has also shown tentative correlations between poorly known alveolates and radiolarians, adding strength to the argument that this is a biologically meaningful interaction (Luria et al., 2014), and suggesting it may be a year-round relationship. Sediment parasite assemblages were dominated by apicomplexans, mainly gregarines, and particularly OTUs affiliated with the eugregarines. Gregarines may be parasitizing groups which were found mainly in the sediment, such as nematodes and Platyhelminthes. It has been suggested that annelids are important hosts for gregarines (Leander, 2007), but we observed annelid sequences across all sample types, while gregarines were restricted to the sediment, suggesting annelids may not be the most important host for these groups in the WAP. Phytomyxea and Pirsonia, both groups known to parasitize diatoms (Tillmann et al., 1999), also had their highest abundances in the sediment. Although initially counter-intuitive, these diatom parasites in the sediment actually correlated well with the distribution of diatom sequences, as diatom biomass in the water column was very low in winter, but sequences for what are assumed to be resting spores of certain Chaetoceros groups were abundant in sediment samples (Cleary, 2015). Parasites have been previously observed within resting cysts of dinoflagellates, with the ability to survive periods of unfavorable conditions, and then become infective again when the host cell leaves dormancy and returns to vegetative growth (Chambouvet et al., 2011). The diatom parasite sequences observed in sediments may suggest a similar life history strategy for these organisms potentially parasitizing Chaetoceros hosts. Parasite abundances in surface waters showed striking differences between summer and winter, with ,2% of summer sequences classified as parasite, while in winter, this proportion was over 13%. All groups of parasites showed this decline, with many groups which were present in winter not detected in summer, although this may be partially due to a much smaller sample size. Summer abundances of syndinians observed here (total 1.73%) show very similar average relative abundance to early results for summer in the adjacent Scotia Sea region (2% MALV II) from DGGE (Diez et al., 2004). NGS data provide only information on relative abundance, not total abundance. Therefore, these sequence data alone cannot determine whether parasites had lower overall abundances in summer, or were just a lower proportion of the total planktonic biomass. Previous analyses on Syndiniales in an Arctic fjordic-bay with qPCR have also seen a decrease from winter to spring, which was attributed to different water masses (Thomson, 2014). However, there is no clear separation of our summer versus winter samples in temperature– salinity (T– S) space, suggesting different water masses may not be the most likely explanation for the lower summer parasite abundances observed here. 409 JOURNAL OF PLANKTON RESEARCH j VOLUME j NUMBER j PAGES – j Fig. 5. Cluster analysis of all winter samples for both parasite and non-parasite OTUs. Samples show strong clustering by sample type (surface water, deep water, sediment), as well as clustering by location. Parasite and free-living communities show very similar clustering, suggesting biological interactions. The low summer relative abundance of parasites is somewhat surprising, since it is during the summer months when planktonic biomass peaks and biological activity is at its greatest. The abundances of many potential hosts, such as diatoms, photosynthetic dinoflagellates and copepods, were much higher in summer, particularly around the seasonal spring bloom, than they were in winter, as can be partly seen in the chlorophyll a values (Table II). Protistan parasites have rapid generation times, on the order of 24–48 h (Coats, 1999), and so should be able to keep pace with the seasonal increases in available hosts. Three potential explanations for the higher relative abundance of parasite sequences in winter in the WAP are worth considering. First, parasites may be targeting some of the select species or groups which are more abundant in the fjordic-bays in winter. The one host group identified in our parasite – host exploratory analysis, the radiolarian, also showed a lower relative abundance in surface waters in summer (0.010%) than in winter (0.467%). However, Radiolaria in winter were more abundant in the deeper water column, and this environment was not analyzed in summer, so this comparison should not be over-interpreted. One prominent pelagic species which was much more abundant in winter than in summer was Euphausia superba, the Antarctic krill (Cleary et al., in review). Apicomplexan parasites are known from krill hosts (Takahashi et al., 2011), and their lower summer abundance within fjordic-bays may be related to the lower krill abundance in summer. Syndinians are known to infect a wide range of crustaceans (Skovgaard, 2014), and although they have not been described from krill, it is possible that they too infect these keystone zooplankters. However, the spatial distribution of syndinians in winter does not support the idea that their abundances 410 A. C. CLEARY AND E. G. DURBIN j ABUNDANT PARASITE 18S IN ANTARCTIC WINTER Fig. 6. Summer distributions of parasite sequences. (A) All summer sequences, in which free-living organisms dominate the assemblage. (B) Parasite sequences alone, showing the prevalence of Syndiniales, particularly Syndiniales I, similar to the observed distribution of parasites in winter. are driven by abundances of E. superba. Syndinians were of equal (surface) or greater (deep) relative abundance in the offshore Gerlache and Palmer Deep samples when compared with the within-bay samples, whereas krill were much more abundant within bays than in the more open waters sampled (Cleary et al., in review). Secondly, parasites may be more successful at infecting physiologically stressed hosts. Protistan parasite infection rates have previously been observed to be higher in photosynthetic hosts which are under nutrient stress (Coats, 1999). Phytoplankton hosts in winter are likely to be experiencing physiological stress due to very low light levels, and zooplankton may also be experiencing stress due to low availability of phytoplankton prey in this dark season. The observed ratio of diatom – parasites to diatoms was highest in winter sediments, and lowest in summer surface waters. This distribution potentially corresponds to the physiological stress of diatom cells, as cells which have sunk down to the sediment in winter are likely stressed, while diatoms in the sunlight surface waters of summer may be much less so. A third potential explanation of the lower observed relative abundances of parasites in summer is differences in the time required for parasites to locate a new host. This explanation may be particularly relevant for parasites which infect motile metazoan hosts, such as the dominant syndinians, which infect copepods and potentially other crustacean zooplankton. Water sampling protocols used here were designed to catch free-living protists with relatively low motility and escape responses, and are unlikely to have quantitatively sampled larger, motile metazoans such as copepods. Copepods have welldeveloped escape responses and the ability to detect sheer flows, and may therefore avoid the Niskin bottles, or if captured, may avoid the outflow to the Niskin spigot. That we did not quantitatively sample mesozooplankton is further suggested by the low (9%) abundance of metazoan sequences in our data set. Given that we may not be truly capturing the mesozooplankton, the parasites which are inside these mesozooplankton hosts may also not have been detected. Instead, the sequences observed may be derived largely from spores or infective stages of parasites. In summer, copepods, particularly Metridia spp., were highly abundant in the sampled fjordic-bays (E.G.D., data not shown), whereas in winter, copepods were very sparse, with only a few individuals per m3 (Cleary et al., in review). The distance between evenly distributed potential hosts is proportional to (1/ hosts-per-volume)(1/3), so as the density of hosts increases in the summer, the distance, and assuming a constant swimming/drifting speed, time, for a parasite to find a new host declines rapidly. With less time required to find 411 JOURNAL OF PLANKTON RESEARCH j VOLUME j NUMBER j PAGES – j Fig. 7. Correlations between Radiolaria and Syndiniales. All plots show radiolarian hosts on the x-axis, with syndinian parasites on the y-axis, with both axis indicating sequence read abundance as a percent of total sequence reads per sample. (A) Various Syndiniales OTUs correlated with Radiolaria OTU 13310; (B) various Syndinales OTUs correlated with Radiolaria OTU 15634; (C) all Syndiniales OTUs when compared with all Radiolaria OTUs. a new suitable host, each parasite would spend a lower fraction of its life freely floating in the water column, and by extension, at any given point in time, fewer of the parasites would be in the water column, and available to our filtering and sequencing analysis. Further research with more quantitative and targeted approaches, such as qPCR or FISH of parasites, or transcriptomic analyses of hosts will be needed to assess the merits of these possible explanations. The high proportions of parasite sequences reads observed in these winter samples from the Antarctic beg the question: Are these observed high abundances of parasite OTU sequences indicative of seasonally high parasitic activity in Antarctic marine ecosystems? There 412 A. C. CLEARY AND E. G. DURBIN j ABUNDANT PARASITE 18S IN ANTARCTIC WINTER are several potential explanations for the observed high abundances: (i) the sequence abundances are an artifact of the rDNA sequencing approach; (ii) groups classified as parasitic in fact contain organisms with other lifestyles; (iii) parasites have long-lived spores or resting stages; and (iv) parasite abundance in the Antarctic in winter is high. Each of these explanations likely plays a part in explaining the overall results, and the merits and likely impact of each are discussed below. As mentioned above, 18S NGS does not provide data on the overall abundance of organisms, but it has traditionally been used to infer relative abundances of different organisms and groups of organisms (Not et al., 2009). There is potential for biases in preservation efficiency, 18S copy numbers per cell, DNA extraction efficiency, PCR primer binding, polymerase extension, amplicon purification, sequencing and quality control. Despite these potential sources of bias, studies with bacterial mock communities have shown close resemblances of sequence read proportions with true cell abundance proportions (Jumpstart Consortium, 2012), and limited data available for eukaryotes suggest that with appropriate primers and data analysis, 18S sequence proportions are similar to community biomass proportions (Egge et al., 2013). Steps were taken to minimize the effect of all of these potential sources of bias in this analysis. PCR primers used have been optimized to efficiently amplify all major groups of eukaryotes (Cleary et al., 2015). Differences in the number of copies of the 18S gene per cell or per unit bio-volume have been observed in some groups (Zhu et al., 2005). However, some of the groups notorious for high copy numbers were observed at low abundances in these Antarctic samples. For example, the free-living dinoflagellates have been suggested to be a group prone to high 18S copy numbers, yet they only made up 1.5% of the total sequences observed; if copy number variations were the cause of the observed high abundance of parasite sequences, one would expect these closely related free-living organisms to also be abundant in the data set. This suggests that while 18S copy variations are certainly present, they are unlikely to be the full explanation for the observed high proportions of parasite sequences. The relative abundance of parasite sequences observed in summer falls within the range of parasite biomass as a proportion of total biomass which has been observed in other marine systems (Kuris et al., 2008). That the summer, productive season shows values consistent with other methodologies suggests that the high winter observations are unlikely to be fully explained as an artifact of the 18S rDNA NGS sequencing approach. Many of the groups of parasites found in this study are very poorly known. Little is known about their diversity, abundances, distributions, host preferences, morphology or ecology (Skovgaard, 2014). This is particularly true of the largest group of parasite sequences observed, the Syndiniales (MALV Groups I and II), for which there are many uncertainties (Bråte et al., 2012). Here, organisms were classified as parasites if literature described the group to which they belong as obligately parasitic, or if all described organisms within the narrowest group to which the OTU could be identified were obligate parasites. However, it may be possible that some of these groups contain as-yet undiscovered species with other lifestyles, such as free-living or mutualistic symbiosis. This has been previously suggested for some radiolarianassociated alveolates (Bråte et al., 2012). Misclassification of such hidden free-living organisms as parasites might explain some of the high winter abundance of parasite sequences we observed; however, the inverse is also true, that parasites within groups categorized as freeliving or mixed have been excluded as parasites, and might be artificially depressing the abundance of parasite sequences. Many parasitic organisms have some form of spore or infective stage, allowing them to spend time outside of a host in the process of finding a new host (Hudson et al., 2006). It is possible that some fraction of the parasite sequences observed came not from active parasites but from some type of spore. However, many of these spores are very short-lived, suggesting that they would be unlikely to form a large reservoir of DNA sequences in the environment. For example, syndinians and related organisms are thought to have very short-lived spores, and to require new hosts within a matter of days (Coats and Park, 2002). Amoebophyra spp. (MALV I) spores show exponential declines in abundance, with most spores disintegrated within 3 – 13 days, and even among surviving spores, ability to infect declined rapidly over time since production (Coats and Park, 2002). It may be possible that some of the parasite organisms observed in this study have longer-lived spores, potentially as an adaptation to the extreme seasonality in the abundance and biomass of potential hosts in Antarctic marine ecosystems. This has been suggested for other dinoflagellate parasites, as a mechanism to survive periods of low host abundance, and such cyst or cyst-like cells have been observed for three parasitic dinoflagellate groups (Paulsenella, Disodinium, Euduboscquella) (Coats, 1999). The production of long-lived resting cysts or spores may partially explain the observed seasonal pattern, with parasites, potentially in the form of overwintering spores, abundant in winter and much less abundant during the productive summer season. In austral winter in the Antarctic Peninsula, phytoplankton biomass is very low, with measured values during the time of sampling consistently ,0.5 mg chlorophyll a L21, 413 JOURNAL OF PLANKTON RESEARCH j VOLUME whereas in summer in these same regions, chlorophyll a concentrations can exceed 13 mg L21 (Table II). It might thus be advantageous for protistan parasites which rely on planktonic hosts, such as diatoms or copepods, to have the capacity to survive as a spore over the winter period with low abundances of hosts in this ecosystem. Correlations between parasite and host communities, and between individual parasite and host groups, suggest these parasites are likely to be active components of the ecosystem, rather than simply a collection of resting spores. The distribution of parasite sequences also may argue against these sequences deriving from resting spores; resting spores might be expected to sink out of the water column, and yet the sediment actually showed the lowest percentage of sequences belonging to parasite groups and a very distinct assemblage from that in the water column (Fig. 3). Previous analysis of Perkinsidae parasite sequences have been shown to represent ribosomally active cells in marine sediments, suggesting their DNA abundance is indicative of an ecologically important role (Chambouvet et al., 2014). It appears likely that protist parasites are more abundant, and more ecologically important than has been realized in the WAP coastal waters in winter, even given all of the above potential secondary explanations and caveats. Models of Antarctic marine food webs have not typically included parasites (Melbourne-Thomas et al., 2013), yet they are potentially important. Parasitism can divert carbon and energy out of the classic phytoplankton – krill –whale food chain, and into the microbial loop as POM and DOM released from ailing and dying hosts. The microbial food web is thought to be increasing in importance in this northern WAP region, as a result of ongoing anthropogenic change (Sailley et al., 2013). Previous sequencing in austral summer near our sampling site also observed a sequence assemblage dominated by alveolates related to dinoflagellates, which pigment analysis suggested were not photosynthetic (Luria et al., 2014). These alveolates may be related to the syndinian parasite groups which we report here, suggesting that potentially the importance of the role of parasites may be temporally variable over the year (Luria et al., 2014). Parasites can also have effects on the population dynamics of hosts and the diversity of the ecosystem more broadly. Parasitic infections among the plankton can be widespread, with observations of over 80% of host Gymnodinium dinoflagellates infected with Amoebophrya, and 29% of host Paracalanus copepods infected with Syndinium (Coats, 1999). Parasites can cause mortality among microplanktonic hosts rivaling that caused by grazing zooplankton (Coats, 1999). Syndinium (MALV IV) infestation is estimated to cause copepod mortality comparable with predation mortality with rates as high as 42% j NUMBER j PAGES – j mortality per day reported (Konovalova, 2008), Perkinsidae or Haplosporidium infections can wipe out 90% of oyster beds (Sousa, 1991) and Amoebophrya spp. (MALV II) have even been suggested as a biocontrol on harmful algal blooms as they can remove over 50% of their hosts daily (Coats and Park, 2002; Skovgaard, 2014). Many parasites are highly host-specific, parasitizing only a single species (Tillmann et al., 1999; Hudson et al., 2006; Skovgaard, 2014). Such high-specificity parasites may play a role in maintaining diversity and structuring seasonal succession within Antarctic marine ecosystems; by causing mortality or reducing fitness of an abundant species, parasites may create opportunities for other species within the ecosystem. Parasitism has traditionally been considered to be more important in warmer ecosystems (Rhode, 1984), and thus as temperatures continue to increase, parasitism may play a role in the increasing importance of the microbial loop. The prevalence of protozoan marine parasites within a single geographic area over interannual variations has also been linked to environmental temperatures, with increasing infections by Perkinsus spp. in the Gulf of Mexico under warmer La Niña waters (Harvell et al., 2002). The role of parasites in the WAP may therefore be increasing as this region continues to experience ongoing anthropogenic climate change and some of the most rapid warming observed globally (Ducklow et al., 2007). Thus, incorporating parasites and their roles in regulating plankton populations into ecosystem models may allow for better understanding and predictions of the trends in plankton species dynamics in the WAP. Parasitism may be more important than commonly considered in marine ecosystems more broadly. When considering metazoan parasites on larger organisms such as fish, the Antarctic has been found to have lower parasite loads than other parts of the world ocean (Rhode, 1984). Here, we present data on protistan parasites for a limited area from the WAP. DNA sequencing technologies have been rapidly improving, and with the public availability of reference databases for eukaryotes (Quast et al., 2013), it has only very recently become feasible to conduct and analyze broad-scale surveys of eukaryote communities. Data from 18S NGS and clone libraries in other regions of the world ocean suggest the unexpectedly high prevalence of parasite sequences observed here may be a wider phenomenon. The largest survey available to date, throughout the temperate and tropical euphotic zone, also found high abundances of parasite-associated sequences, with these sequence reads accounting for 53% of the heterotrophic protists in their smallest size fraction, with lower per cent in the larger size fractions (de Vargas et al., 2015). That abundance is still lower than the 52% of all protists (of all sizes combined) observed in deep waters 414 A. C. CLEARY AND E. G. DURBIN j ABUNDANT PARASITE 18S IN ANTARCTIC WINTER here, but indicates these parasites may be abundant globally. Sequence assemblages from the English Channel included 18% parasite sequences (Genitsaris et al., 2015). Meta-analyses of clone libraries suggest Syndiniales make up over half of the dinoflagellate sequences observed in marine samples (Guillou et al., 2008). Clone library sequences of Antarctic deep water samples north of our sampling region in austral summer consisted of 65–76% unclassified Alveolata, many of which are related to the Syndiniales parasite groups (López-Garcı́a et al., 2001). Clone libraries in the high arctic also contained high abundances of Syndiniales, with various groups making up 44% of the clones (Sørensen et al., 2012). Radiolarians sampled in the Arctic were associated with alveolates, as observed here in the Antarctic (Fig. 7) (Bråte et al., 2012). Similar syndinian-like alveolates have also been observed near hydrothermal vent systems in both the Atlantic and Pacific along with more well-known parasites such as Perkinsus spp. and Pirsonia spp. (Edgcomb et al., 2002; López-Garcı́a et al., 2003; Moreira and López-Garcı́a, 2003; Takishita et al., 2005). Understanding the relationships and biogeography of these global distributed yet poorly known parasite groups, such as the Syndiniales, may prove an interesting area for future research. The much higher abundances of parasites observed in winter, when compared with the better studied summer season, underscore the importance of year-round observations, ideally with fine temporal resolution, in understanding plankton communities. As new observations over diverse areas of the world ocean and over the seasonal cycle become available, it will be interesting to see how the importance of these parasite-associated sequence groups varies globally and temporally, and begin to understand the magnitude of their ecological roles more widely. DATA A RC H I V I N G DNA sequences of parasite OTUs are available from GenBank under accession numbers KR864908 – KR865949. AC K N OW L E D G E M E N T S Many thanks to Kerry Whitaker for assistance in winter seawater sampling, and to Rebecca Robinson for assistance in sediment sampling. Thanks to Maria Casas, Iain McCoy, Michelle Dennis, Jessica Perreault and David Gleeson for assistance with summer water sampling. Thanks to John Kirkpatrick for advice on molecular and analytical techniques. Thanks also to the science party, command, and crew of RVIB Nathaniel B Palmer cruises 1304 and 1410. This manuscript was improved by comments from four anonymous reviewers, who we thank. FUNDING This research is based in part upon work conducted using the Rhode Island Genomics and Sequencing Center, URI, which is supported in part by the National Science Foundation under EPSCoR Grants Nos. 0554548 and EPS-1004057, and based in part using resources and services at the Center for Computation and Visualization, Brown University. This research was supported by National Science Foundation Office of Polar Programs grant #ANT-1142107 to E.G.D. REFERENCES Altschul, S. F., Gih, W., Miller, W., Myers, E. W. and Lipman, D. J. (1990) Basic local alignment search tool. J. Mol. Biol., 215, 403–410. Arias-Gonzáles, J. and Morand, S. (2006) Trophic functioning with parasites: a new insight for ecosystem analysis. Mar. Ecol. Prog. Ser., 320, 43–53. Bråte, J., Krabberød, A. K., Dolven, J. K., Ose, R. F., Kristensen, T., Bjørklund, K. R. and Shalchian-Tabrizi, K. (2012) Radiolaria associated with large diversity of marine alveolates. Protist, 163, 767 –777. Bray, J. R. and Curtis, J. T. (1957) An ordination of upland forest communities of southern Wisconsin. Ecol. Monographs, 27, 325– 349. Cachon, J. (1964) Contribution a l’étude des Péridiniens parasites. Cytologie. Cycles évolutifs. Ann. Sci. Nat. Zool., 12, 1– 158. Caporaso, J. G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F. D., Costello, E. K., Fierer, N., Gonzalez-Pena, A. et al. (2010) QIIME allows analysis of high-throughput community sequencing data. Nat. Methods, 7, 335– 336. Chambouvet, A. (2009) Les amoebophryidae (Syndiniales) parasitoı̈des de dinoflagellés: cycle de vie, dynamique et spécificité in situ. PhD thesis, University Pierre and Marie Curie, Paris, France. Chambouvet, A., Alvez-de-Souza, C., Cueffa, V., Mariea, D., Karpova, S. and Guillou, L. (2011) Interplay between the parasite Amoebophrya sp. (Alveolata) and the cyst formation of the red tide dinoflagellate Scripsiella trochoidea. Protist, 162, 637– 649. Chambouvet, A., Berney, C., Romac, S., Audic, S., Maguire, F., De Vargas, C. and Richards, T. A. (2014) Diverse molecular signatures for ribosomally “active” Perkinsea in marine sediments. BMC Microbiol., 14, 110. Cleary, A. C. (2015) Distributions and interactions in three groups of polar marine plankton. PhD dissertation, University of Rhode Island, Graduate School of Oceanography. Cleary, A. C., Durbin, E. G., Casas, M. C. and Zhou, M. (in review) Winter distribution and size structure of Antarctic krill, Euphausia superba, populations inshore in the West Antarctic Peninsula. Mar. Ecol. Prog. Ser. Cleary, A. C., Durbin, E. G., Rynearson, T. A. and Bailey, J. (2015) Feeding by Pseudocalanus copepods in the Bering Sea: trophic linkages and a potential mechanism of niche partitioning. Deep-Sea Res, II, doi:10.1016/j.dsr2.2015.04.001. 415 JOURNAL OF PLANKTON RESEARCH j VOLUME Coats, D. W. (1999) Parasitic life styles of marine dinoflagellates. J. Eukaryot. Microbiol., 46, 402– 409. Coats, D. W. and Park, M. G. (2002) Parasitism of photosynthetic dinoflagellates by three strains of Amoebophrya (dinophyta): parasite survival, infectivity, generation time, and host specificity. J. Phycol., 38, 520–528. De Vargas, C., Audic, S., Henry, N., Decelle, J., Mahé, F., Logares, R., Lara, E., Berney, C. et al. (2015) Eukaryotic plankton diversity in the sunlit ocean. Science, 348, 1261605. Diez, B., Massana, R., Estrada, M. and Pedrós-Alió, C. (2004) Distribution of eukaryotic picoplankton assemblages across hydrographic fronts in the Southern Ocean, studied by denaturing gradient gel electrophoresis. Limnol. Oceanogr., 49, 1022–1034. Ducklow, H. W., Baker, K., Martinson, D. G., Quetin, L. B., Ross, R. M., Smith, R. C., Stammerjohn, S. E., Vernet, M. et al. (2007) Marine pelagic ecosystems: the west Antarctic Peninsula. Phil. Trans. Roy. Soc. B: Biol. Sci., 362, 67–94. j NUMBER j PAGES – j Jespersen, A. M. and Christoffersen, K. (1987) Measurements of chlorophyll-a from phytoplankton using ethanol as extractions solvent. Arch. Hydrobiol., 9, 445–454. Jones, D. L. (2014) Fathom Toolbox for Matlab: Software for Multivariate Ecological and Oceanographic Data Analysis. College of Marine Science, University of South Florida, St Petersburg, FL. http://www.marine .usf.edu/user/djones/. Jumpstart Consortium Human Microbiome Project Data Generation Working Group Evaluation of 16S rDNA-Based Community Profiling for Human Microbiome Research. (2012) PLoS One, 7, 39– 315. Konovalova, G. V. (2008) Parasitic dinoflagellates and ellobiopsids (Ellobiopsidae) of the coastal waters of the Sea of Japan. Russ. J. Protistol., 34, 28– 37. Edgar, R. C. (2010) Search and clustering orders of magnitude faster than BLAST. Bioinformatics, 26, 2460– 2461. Kühn, S., Medlin, L. and Eller, G. (2004) Phylogenetic position of the parasitoid nanoflagellate Pirsonia inferred from nuclearencoded small subunit ribosomal DNA and a description of Pseudopirsonia n. gen. and Pseudopirsonia mucosa (Drebes) comb. nov. Protist, 155, 143– 156. Edgcomb, V. P., Kysela, D. T., Teske, A., de Vera Gomez, A. and Sogin, M. L. (2002) Benthic eukaryotic diversity in the Guaymas Basin hydrothermal vent environment. Proc. Natl Acad. Sci. USA, 99, 7658– 7662. Kuris, A. M., Hechinger, R. F., Shaw, J. C., Whitney, K. L., Aguirre-Macedo, L., Boch, C. A., Dobson, A. P., Dunham, A. J. et al. (2008) Ecosystem energetic implication of parasite and free-living biomass in three estuaries. Nature, 454, 515 –518. Egge, E., Bittner, L., Andersen, T., Audic, S., de Vargas, C. and Edvardsen, B. (2013) 454 Pyrosequencing to describe microbial eukaryotic community composition, diversity and relative abundance: a test for marine haptophytes. PLoS One, doi:10.1371/journal.pone.0074371. Leander, B. S. (2007) Marine gregarines: evolutionary prelude to the apicomplexan radiation? Trends Parasitol., doi:10.1016/ j.pt.2007.11.005. Garrison, D. L. (1991) An overview of the abundance and role of protozooplankton in Antarctic waters. J. Mar. Syst., 2, 317– 331. Gast, R. J., Dennett, M. R. and Caron, D. A. (2004) Characterization of protistan assemblages in the Ross Sea, Antarctica, by denaturing gradient gel electrophoresis. Appl. Environ. Microbiol., 70, 2028–2037. Genitsaris, S., Monchy, S., Viscogliosi, E., Sime-Ngando, T., Ferreira, S. and Christaki, U. (2015) Seasonal variations of marine protist community structure based on taxon-specific traits using the eastern English Channel as a model coastal ecosystem. FEMS Microbiol. Ecol., 91 (doi: 10.1093/femsec/fiv034). Glockling, S. L., Marshall, W. L. and Gleason, F. H. (2013) Phylogenetic interpretations and ecological potentials of the Mesomycetozoea (Ichthyosporea). Fungal Ecol., 6, 237–247. Guillou, L., Bachar, D., Audic, S., Bass, D., Berney, C., Bittner, L., Boutte, C., Burgaud, G. et al. (2012) The Protist Ribosomal Reference database (PR2): a catalog of unicellular eukaryote small sub-unit rRNA sequences with curated taxonomy. Nucleic Acids Res., 41, 597–604. Guillou, L., Viprey, M., Chambouvet, A., Welsh, R. M., Kirkham, A. R., Massana, R., Scanlan, D. J. and Worden, A. Z. (2008) Widespread occurrence and genetic diversity of marine parasitoids belonging to Syndiniales (Alveolata). Environ. Microbiol., 10, 3349– 3365. Harvell, C. D., Mitchell, C. E., Ward, J. R., Altizer, S., Dobson, A. P., Ostfeld, R. S. and Samuel, M. D. (2002) Climate warming and disease risks for terrestrial and marine biota. Science, 296, 2158–2162. Leander, B. S., Saldarriaga, J. F. and Keeling, P. J. (2002) Surface morphology of the marine parasite Haplozoon axiothellae Siebert (Dinoflagellata). Eur. J. Protistol., 38, 287–297. Lima-Mendez, G., Faust, K., Henry, N., Decelle, J., Colin, S., Carcillo, F., Chaffron, S., Ignacio-Espinosa, J. et al. (2015) Determinants of community structure in the global plankton interactome. Science, 348, 1262073. López-Garcı́a, P., Phillippe, H., Gail, F. and Moreira, D. (2003) Autochthonous eukaryotic diversity in hydrothermal sediment and experimental microcolonizers at the Mid-Atlantic Ridge. Proc. Natl Acad. Sci. USA, 100, 697–702. López-Garcı́a, P., Rodrı́guez-Valera, F., Pedr?s-Alló, C. and Moreira, D. (2001) Unexpected diversity of small eukaryotes in deep-sea Antarctic plankton. Nature, 409, 603– 607. Luria, C. M., Ducklow, H. W. and Amaral-Zettler, L. A. (2014) Marine bacterial, archaeal, and eukaryotic diversity and community structure on the continental shelf of the western Antarctic Peninsula. Aquat. Microb. Ecol., 73, 107–121. Melbourne-Thomas, J., Constable, A., Wotherspoon, S. and Raymond, B. (2013) Testing paradigms of ecosystem change under climate warming in Antarctica. PLoS One, doi:10.1371/journal.pone.0055093. Moreira, D. and López-Garcı́a, P. (2003) Are hydrothermal vents oases for parasitic protists? Trends Parasitol., 19, 556–558. Morgulis, A., Coulouris, G., Raytselis, Y., Madden, T. L., Agarwala, R. and Schäffer, A. A. (2008) Database indexing for production MegaBLAST searches. Bioinformatics, 15, 1757– 1764. Hudson, P. J., Dobson, A. P. and Lafferty, K. D. (2006) Is a healthy ecosystem one that is rich in parasites? Trends Ecol. Evol., 21, 381–385. Neuhauser, S., Kirchmair, M. and Gleason, F. H. (2011) Ecological roles of the parasitic phytomyxids ( plasmodiophorids) in marine ecosystems—a review. Mar. Freshwater Res., 62, 365–371. Huntley, M. E., Lopez, M. D. G. and Karl, D. M. (1991) Top predators in the Southern Ocean: a major leak in the biological carbon pump. Science, 253, 64–66. Not, F., del Campo, J., Balagué, V., de Vargas, C. and Massana, R. (2009) New insights into the diversity of marine picoeukaryotes. PLoS One, 4, e7143. 416 A. C. CLEARY AND E. G. DURBIN j ABUNDANT PARASITE 18S IN ANTARCTIC WINTER Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T., Yarza, P., Peplies, J. and Glöckner, F. O. (2013) The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids. Res., 41, 590– 596. Takahashi, K. T., Kawaguchi, S., Kobayashi, M., Toda, T., Tanimura, A., Fukuchi, M. and Odate, T. (2011) Eugregarine infection within the digestive tract of larval Antarctic krill, Euphausia superba. Polar Biol., 34, 1167–1174. Rhode, K. (1984) Zoogeography of marine parasites. Helgol. Meeresunters, 37, 35–52. Takishita, K., Miyake, H., Kawato, M. and Maruyama, T. (2005) Genetic diversity of microbial eukaryotes in anoxic sediment around fumaroles on a submarine caldera floor based on the small-subunit rDNA phylogeny. Extremophiles, 9, 185– 196. Sailley, S. F., Ducklow, H. W., Moeller, H. V., Fraser, W. R., Schofield, O. M., Steinberg, D. K., Garzio, L. M. and Doney, S. C. (2013) Carbon fluxes and pelagic ecosystem dynamics near two western Antarctic Peninsula Adélie penguin colonies: an inverse model approach. Mar. Ecol. Prog. Ser., 492, 253– 272. Skovgaard, A. (2014) Dirty tricks in the plankton: diversity and role of marine parasitic protists. Acta Protozool., 53, 51–62. Skovgaard, A. and Daugbjerg, N. (2008) Identity and systematic position of Paradinium poucheti and other Paradinium-like parasites of marine copepods based on morphology and nuclear-encoded SSU rDNA. Protist, 159, 401– 413. Sogin, M. L., Morrison, H. G., Huber, J. A., Welch, M. D., Huse, S. M., Neal, P. R., Arrieta, J. M. and Herndl, G. J. (2006) Microbial diversity in the deep sea and the underexplored “rare biosphere”. Proc. Natl Acad. Sci. USA, 103, 12115–12120. Sørensen, N., Daugbjerg, N. and Gabrielsen, T. M. (2012) Molecular diversity and temporal variation of picoeukaryotes in two Arctic fjords, Svalbard. Polar Biol., 35, 519–533. Sousa, W. P. (1991) Can models of soft-sediment community structure be complete without parasites? Am. Zool., 31, 821–830. Thomson, S. (2014) Seasonal abundance of parasitic Marine Alveolate group II (MALV II) in an Arctic fjord, Svalbard. MS thesis, University of Tromsø, Department of Arctic and Marine Biology. Tillmann, U., Hesse, K. J. and Tillmann, A. (1999) Large-scale parasitic infection of diatoms in the Northfrisian Wadden Sea. J. Sea Res., 42, 255 –261. Toft, C. A. (1986) Communities of species with parasitic life-styles. In Diamond, J. and Case, T. J. (eds), Community Ecology. Harper & Row, New York, pp. 445 –463. Wang, Q., Garrity, G. M., Tiedje, J. M. and Cole, J. R. (2007) Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl. Environ. Microb., 73, 5261–5267. Wright, A. D. G. and Colorni, A. (2002) Taxonomic re-assignment of Crytocaryon irritans, a marine fish parasite. Eur. J. Protistol., 37, 375 –378. Zhu, F., Massana, R., Not, F., Marie, D. and Vaulot, D. (2005) Mapping of picoeukaryotes in marine ecosystems with quantitative PCR of the 18S rRNA gene. FEMS Microbiol., 52, 79–92. 417