Progress in Oceanography 96 (2012) 77-92
PROGRESS IN Contents lists available at SciVerse ScienceDirect OCEANOGRAPHY
Progress in Oceanography
ELSEVIER journal homepage: www.elsevier.com/locate/pocean
Cold seep and oxygen minimum zone associated sources of margin heterogeneity affect benthic assemblages, diversity and nutrition at the Cascadian margin (NE Pacific Ocean)
Katja Guilinia *, Lisa A. Levin b, Ann Vanreusela aBiology Department, Marine Biology Section, Ghent University, Krijgslaan 281/Sterre S8, 9000 Gent, Belgium blntegrative Oceanography Division, Scripps Institution of Oceanography, 9500 Gilman Drive, La Jolla, CA 92093-0218, USA
ARTICLE INFO ABSTRACT
Article history: Hydrate Ridge (HR), located on the northeastern Pacific margin off Oregon, is characterized by the Received 3 June 2011 presence of outcropping hydrates and active methane seepage. Additionally, permanent low oxygen con Received in revised form 19 October 2011 ditions overlay the benthic realm. This study evaluated the relative influence of both seepage and oxygen Accepted 19 October 2011 minima as sources of habitat heterogeneity and potential stress-inducing features on the bathyal Available online 28 October 2011 metazoan benthos (primarily nematodes) at three different seep and non-seep HR locations, exposed to decreasing bottom-water oxygen concentrations with increasing water depth. The nematode seep communities at HR exhibited low diversity with dominance of only one or two genera (D aptonem a a n d Metadesmolaimus), elevated average individual biomass and 513C evidence for strong dependance on chemosynthesis-derived carbon, resembling deep-sea seeps worldwide. Although the HR seep habitats harbored a distinct nematode community like in other known seep communities, they differed from deep-sea seeps in well-oxygenated waters based on that they shared the dominant genera with the surrounding non-seep sediments overlain by oxygen-deficient bottom water. The homogenizing effect of the oxygen minimum zone on the seep nematode assemblages and surrounding sediments was constant with increasing water depth and concomitant greater oxygen-deficiency, resulting in a loss of habitat heterogeneity. © 2011 Elsevier Ltd. All rights reserved.
1. Introduction or to tolerate these physico-chemical conditions (Levin et al., 2003). Several megafaunal seep-specialists are adapted to survive Cold-seeps are geochemical features, found worldwide on in these reduced habitats by carrying endosymbionts and/or by tectonically active and passive continental margins (Sibuet and having sulfide-binding or -oxidizing capabilities (e.g. siboglinid Olu, 1998). They are formed where subsurface interstitial fluids, polychaetes, cladorhizid sponges, bathymodiolin mussels, and ves- rich in reduced chemicals from biogenic or thermogenic origin icomyid, lucinid, solemyid, and thyasirid clams). This successful (mostly methane and sulfide), are released at the sediment surface exploitation of chemosynthetic derived energy is often evidenced as a result of a variety of processes such as tectonic activity, differ by highly increased levels of density and biomass of these few ses ential compaction of organic-rich sediments, gas hydrate dissocia sile or sedentary taxa. Their presence also adds structural complex tion and subsurface salt migration (Cordes et al., 2010). Therefore ity to the seep sediments which increases habitat heterogeneity cold-seep ecosystems appear in a variety of geomorphic and bio and supports many smaller fauna living associated with them logical forms on the seafloor and create a diverse suite of habitats (Levin, 2005). Overall, the aggregations of symbiotic foundation for both endemic seep organisms and more opportunistic colonists species and associated taxa at cold seeps are considered as excep (Levin et al., 2003; Cordes et al., 2010). Together with water depth tions on the general trends of high local metazoan diversity and and age of the geologic features, habitat heterogeneity created by evenness, and low metazoan density and biomass that are found the intensity and volume of fluid flow, and methane and sulfide in most of the deep sea (Levin, 2005). concentrations and fluxes, determines species richness and density In addition to cold seeps, oxygen minimum layers present at of the inhabiting fauna (Cordes et al., 2010). While chemo- intermediate water depths (ca. 100-1000 m), contribute to habitat autotrophic bacteria that thrive on hydrogen sulfide form mats at heterogeneity along continental margins worldwide. This hydro- sites of active seepage, relatively few taxa have the ability to use graphic feature persists over geological time scales and occurs where upwelling leads to high surface productivity, oxygen- depleted source waters are present, and water stability is imposed * Corresponding author. Tel: +32 9 264 85 31; fax: +32 9 264 85 98. by a strong pycnocline that induces accumulation of settling E-mail address: [email protected] (K. Guilini).
0079-6611 /$ - see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.pocean.2011.10.003 78 K. Guilini et al./Progress in Oceanography 96 (2012) 77-92 organic detritus (Levin, 2003; Gooday et al„ 2010). Typically, 02 to the autochthonous chemosynthetic-derived carbon input. When concentrations drop below 0.5 ml I-1. As reduced bottom-water seep-associated fauna depend largely on methane-derived carbon, oxygen concentrations enhance preservation of organic matter either by carrying methane- or sulfide-oxidizing endosymbionts from the surface (Cowie and Hedges, 1992), increased concentra or by feeding direct- or indirectly on chemoautotrophic methane tions of labile organic matter and phytodetrital food are found in oxidizers (aerobic bacteria and anaerobic Achaea with associated sediments where the OMZ intersects with the seafloor (often sulfate-reducing bacteria), their tissues will reflect the thermo- or >3%, up to 20.5% organic carbon; Levin, 2003; Gooday et al., biogenic 513C methane signal (513C o f-5 0 %o to -1 1 0 %a, e.g. Whiti- 2010). Strong zonation of benthic communities occurs in conjunc car et al., 1986). However, as most seep-associated fauna probably tion with varying bottom- and pore-water oxygen concentrations use multiple food sources, the interpretation of isotopic signatures and associated gradient in organic-matter input that change with is complicated. To yield at least upper and lower estimates of meth water depth (e.g. Levin et al., 1991, 2000, 2003, 2009; Wishner ane-derived carbon in organisms, two-end member linear mixing et al., 1990,1995; Levin and Gage, 1998; Neira et al., 2001). Macro- models with single isotope measures can be used (e.g. Levin and and megafauna communities inhabiting OMZ sediments are Michener, 2002; MacAvoyet al., 2002; C arlieret al., 2010; Demopo- generally distinct from the well-oxygenated surrounding margins ulos et al., 2010; Thurber et al., 2010). in having severely reduced densities, biomass and diversity (Wish Our study evaluates the relative influence of both seepage and ner et al., 1990; Levin and Gage, 1998; Smith et al., 2000; Levin oxygen minima as sources of habitat heterogeneity and potential et al., 2001; reviewed in Levin, 2003; Sellanes et al., 2010). stress-inducing features on the bathyal benthic community. We Although total meiobenthic biomass and abundance often follow focus hereby on the nematode community composition, diversity, the same pattern as mega- and macrofauna, in general both proto morphometry, biomass, density and nutrition and compare the lat zoan and metazoan meiofauna are less affected by low oxygen con ter with macrofauna data. Therefore three habitat types (clam beds, ditions (Levin et al., 1991; Bernhard et al., 2000; Cook et al., 2000; microbial mats and nearby non-seep margin sediments) were stud Gooday et al., 2000, 2009; Neira et al., 2001; Veit-Köhler et al., ied at three HR locations categorized according to increasing depth 2009). Among the metazoan meiofauna, deep-sea nematodes seem and decreasing bottom-water oxygen concentrations (500-600 m, very tolerant to low oxygen concentrations (Levin et al., 1991; 0.7 02 m ir1; 700-800 m, 0.2-0.3 02m ir1; 800-900 m, ^0.2 Moodley et al., 1997; Cook et al., 2000) and typically occur at high 02 ml I-1; Levin et al., 2010). This study sought to answer the fol er densities where 02 concentrations are lowest in OMZs compared lowing questions: (1) do the seepage habitats at HR harbor a dis to other taxa that are unable to tolerate low oxygen concentrations tinct nematode community compared to the non-seep margin such as harpacticoid copepods and nauplii (Hicks and Coull, 1983; sites and other known seepage sites worldwide? (2) Does the in Murrell and Fleeger, 1989; Levin et al., 1991; Giere, 2009; Neira creased habitat complexity created by clam beds affect the nema et al., 2001; Levin et al., 2009). If their elongated body-length tode community distribution in the sediment, resulting in an and therefore higher surface area-to-volume ratio could be an increase in biodiversity compared to the microbial mats? (3) Does adaptation to low oxygen partial pressure as suggested before the OMZ homogenize the nem atode com position in such a way that (Jensen, 1986, 1987; Schiemer et al., 1990; Soetaert et al., 2002), seep-related heterogeneity is more pronounced at more oxygen morphometric changes might occur along an oxygen-deficiency ated water depths? (4) Do water depth and co-varying 02 bottom gradient. water concentrations create a bathymetric faunal zonation at the Given the widespread occurrence of both seeps and OMZs on non-seep sites? (5) What is the importance of methane-derived car East Pacific margins it is not surprising to find a considerable num bon for nematodes and how variable is trophic specialization ber of seeps that coincide with OMZs (Levin et al., 2010; Sellanes among the studied locations and habitats? (6) How do the nema et al., 2010). Hydrate Ridge (HR) is such a cold seep ecosystem on tode nutritional trends compare to those observed for macrofauna the Northeastern Pacific margin off Oregon where permanent low at these sites? oxygen conditions overlay the benthic realm. The OMZ extends from about 650 to 1100 m water depth, with a minimum around 800 m (<0.2 ml I-1) (Helly and Levin, 2004) whereas high seepage 2. Material and methods occurs between 600 and 900 m water depth. Key seep habitats at HR include mats of the filamentous sulfide oxidizing Beggiatoa bac 2.Î. Site characteristics teria, Calyptogena vesicomyid clam beds and surrounding Acharax solemyid bivalve beds (Sahling et al., 2002). Levin et al. (2010) Samples for this study were collected at a seepage location in the hypothesized that the effect of oxygen minimum stress would act NE Pacific Ocean, 90 km offshore Oregon (referred to as Hydrate to minimize seep-induced heterogeneity in macrofauna assem Ridge) and on the continental margin within 30 km of the seepage blages at a southern HR location (770 m water depth). In contrast location, at comparable depths (Fig. 1 ). Hydrate Ridge (HR) is one of to what was hypothesized, the species overlap between the HR seep several ridges along an accretionary prism of the Cascadian subduc habitats and the non-seep sediments was less than in more oxygen tion zone, where the Juan de Fuca plate is subducting under the ated realms overlaying the 500-m Eel River (ER) seeps off California North American Plate. The ridge, characterized by the presence of (Levin et al., 2010). However, stronger fluid fluxes and higher sul outcropping gas hydrates and methane venting cold seeps, is about fide levels at HR seeps compared to the ER seeps may have induced 25 km long and 15 km wide, and has a northern and southern peak. these results (Levin et al., 2010). Here we investigate the structur Three regions are distinguished due to their extensive methane ing effect of the OMZ interacting with seep effects on benthic seepage: the southern summit around 770 m, the northern summit assemblages along a bottom-water oxygen gradient at HR only. at around 600 m and the eastern deepest site at around 880 m On the margin along HR, increasing sediment carbon concentra water depth. Gas hydrate dissociation is the main source of the tions were found with increasing depth within the OMZ (500 m: methane enrichment in the seeping fluids. The high sulfide fluxes 0.9% C, 770 m: 1.99% C, 800 m: 2.14% C; Levin et al., 2010; Valentine produced by the anaerobic oxidation of seeping methane (e.g. Boe- et al., 2005). This is likely the result of reduced detrital reminerali tius et al., 2000; Knittel et al., 2003; Treude et al., 2003), and the zation due to lower oxygen concentrations with increasing depth. presence of bacterial and macrofaunal chemosynthetic communi The OMZ might therefore also affect the functional properties of ties that form key habitat patches at HR have been well docu the seep-associated fauna in the sense that with depth the contri mented (e.g. Kuim et al., 1986; Juhl and Taghon, 1993; Sahling bution of allochtonous organic carbon might increase in proportion et al., 2002; Sommer et al., 2002; Levin et al., 2010). Studies on IC Guilini et al./Progress in Oceanography 96 (2012) 77-92 79
ishington
Oregon
California N evada
Fig. 1. Bathymetric map of the Hydrate Ridge study location with indication of the three sampled seep locations, each including clam beds and microbial mats, and the three sampled non-seep reference locations; with indication of water depth.
the associated meiofaunal communities were restricted to the level and in July and October 2006 aboard the RV Atlantis (with submer of phyla (Sommer et al., 2002, 2003, 2007). The habitat types of sible ALVIN). interest for this study are the microbial mats that are primarily composed of white, filamentous sulfide-oxidizing bacteria such as Beggiatoa sp. and the clam beds formed by the vesicomyids Calypto 2.3. Sample processing and analytical procedures gena pacifica and C. kilmeri that harbor sulfide-oxidizing symbionts. Both habitat types occur in patches at scales of meters to tens of The samples destined for meiofaunal identification were verti m eters covering the sea floor (Sahling et al., 2002; Levin et al., cally sectioned at 1 or 2 cm intervals to 5 cm, and from 5 to 7 2003). In the HR microbial mats, H2S concentrations reach over and 7 to 10 cm depth on board. All sample fractions were unsieved 28 mM within the top 5 cm of the sediment, while in clam beds, prior to preservation except for the multicorer samples where in net fluid flux rates are lower and H2S concentrations often only some cases the 5-10 cm fraction was sieved on a 0.3 mm mesh. reach the same (or lower) levels below 5-10 cm down the sediment Samples were preserved in 8% buffered formalin. In the laboratory (Sahling et al., 2002; Valentine et al., 2005; Sommer et al., 2007). of the SCRIPPS Institution of Oceanography part of the samples were sieved on nested 0.3 mm and 42-pm mesh in order to extract the macrofauna from the 0.3 mm fraction. The residue (including 2.2. Sampling meiofauna) was recombined with the finer fraction and stored in 8% buffered formalin again. In the laboratory of the Marine Biology Samples to study the meiobenthos were collected in July- research group at the University of Ghent all meiofaunal samples October 2006 and August 2010 with the submersible ALVIN aboard were sieved on a 1-mm and 32-pm mesh. The fractions retained the RV Atlantis. Seep samples were taken with push cores (6.4 and on the 32-pm mesh sieve were centrifuged three times with the 8.3 cm internal diameter), while sampling of margin sediments, ca. colloidal silica polymer LUDOX 40 (Heip et al., 1985) and rinsed 14-30 km further off-seep site, was conducted by multicorer with tap water. The extracted fraction was preserved in 4% buf (9.6 cm internal tube diameter). We will further refer to the differ fered formalin and stained with Rose Bengal. All metazoan meio- ent types of sampled habitats as ‘clam bed’, ‘microbial mat’ and benthic organisms were classified at higher taxon level and ‘non-seep’, and classify them according to depth zones to facilitate counted under a stereoscopic microscope. Where possible a mini comparison along an increasing stress gradient of oxygen depletion mum of 100 nematodes per sediment layer were randomly hand- in the water column (500-600 m, 0.7 02mll_1; 700-800 m, picked with a fine needle. These nematodes were transferred to 0.2-0.3 02 ml r 1; 800-900 m, ^0.2 02m ir1; Levin et al., 2010) glycerine (De Grisse I, II and III; Seinhorst, 1959), mounted on glass (Table 1). To obtain a more complete view on feeding habits of slides and identified to genus level. Based on mouth morphology, the benthic community, we included 513C isotope data of macrofa all identified individuals were grouped into four groups of feeding una obtained from samples collected with push cores (6.4 and types according to Wieser (1953): selective deposit feeders (1A), 8.3 cm internal diameter) and occasionally a slurp, suction or scoop non-selective deposit feeders (IB), epistratum feeders (2A) and device in clam beds and microbial mats and by multicorer (9.6 cm predators/scavengers (2B). By using a Leica DMR compound micro internal diameter) for margin sediments off seeps (Table 2). Macro scope and Leica LAS 3.3 imaging software nematode length (L, fili fauna samples were collected from the same areas as the meiofa- form tail excluded) and maximal width (W) were measured for the una in August 2005 aboard RV Western Flyer (with ROV Tiburon) first cm fraction in one replicate of each habitat at every location. 80 IC Guilini et al./Progress in Oceanography 96 (2012) 77-92
Table 1 Study site location, habitat type, core code, date and depth of sampling, coordinates of exact sampling location, sampling device, surface area of meiofaunal samples and the lower mesh used for sieving.
Location Habitat Core code Date Depth (m) Latitude N Longitude W Sampling device (surface area) Lower m esh 500-600 m HR North Microbial mat TC 16 4 August 2010 600 44°40.026' 125°5.920' Push core (32.2 cm2) 32 pm HR North Microbial mat TC 23 4 August 2010 600 44°40.026' 125°5.995' Push core (32.2 cm2) 32 pm HR North Clam bed TC 6 21 July 2006 602 44°40.169' 125°5.919' Push core (32.2 cm2) 32 pm HR North Clam bed TC 53 21 July 2006 602 44°40.158' 125°5.855' Push core (54.1 cm2) 42 pm Margin Non-seep MC 3C July 2006 491 44°34.163' 124°50.864' Multicorer (72.4 cm2) 42 pm a Margin Non-seep MC 2C July 2006 494 44°34.163' 124°50.865' Multicorer (72.4 cm2) 42 pm a 700-800 m HR South Microbial mat TC 23 17 July 2006 772 44°34.204' 125°8.841' Push core (32.2 cm2) 32 pm HR South Microbial mat TC 51 18 July 2006 774 44°34.225 125°8.842' Push core (54.1 cm2) 42 pm HR South Clam bed TC 4 20 July 2006 772 44°34.234' 125°8.853' Push core (32.2 cm2) 32 pm HR South Clam bed TC 20 27 September 2006 774 44°34.188' 125°8.849' Push core (32.2 cm2) 32 pm Margin Non-seep MC 2A 19 July 2006 797 44°34.201' 124°57.150' Multicorer (72.4 cm2) 42 pm a Margin Non-seep MC 2C 19 July 2006 797 4 4 °34.20r 124°57.150' Multicorer (72.4 cm2) 42 pm 800-900 m HR East Microbial mat TC 53 23 July 2006 880 44°34.257' 124°59.883' Push core (54.1 cm2) 32 pm HR East Clam bed TC 9 23 July 2006 880 44°34.279' 124°59.921' Push core (32.2 cm2) 32 pm Margin Non-seep MC 2A 28 September 2006 887 44°36.402' 125°07.50r Multicorer (72.4 cm2) 42 pm Margin Non-seep MC4A 30 September 2006 886 44°36.402' 125°07.50r Multicorer (72.4 cm2) 42 pm
a Samples where the 5-10 cm sediment layer was sieved on a 300 pm mesh.
Table 2 Study site location, habitat type, depth of sampling, number of samples for associated macrofaunal 513C measurements, and number of 513C measurements per taxon. The number of samples are given per sampling campaign (A: August 2005, B: July 2006, C: October 2006).
Location Habitat Depth (m) Samples for ô13C (n) ô13C measurements per taxon (n) A B C A B C 500-600 m HR North Microbial mat 588-602 3 3 0 Gastropoda 2 3 0 Polychaeta 3 9 0 HR North Clam bed 590-602 6 7 0 Bivalvia 0 2 0 Crustacea 1 13 0 Gastropoda 0 3 0 Nemertea 0 1 0 Olygochaeta 5 0 0 Ophiuroidea 1 0 0 Polychaeta 15 21 0 Margin Non-seep 500-524 0 1 3 Aplacophora 0 1 0 Bivalvia 0 2 0 Crustacea 0 0 1 Ophiuroidea 0 1 0 Polychaeta 0 2 12 700-800 m HR South Microbial mat 770-800 4 3 6 Bivalvia 0 0 3 Gastropoda 1 3 8 Polychaeta 6 7 17 HR South Clam bed 770-800 3 9 9 Bivalvia 0 3 4 Crustacea 0 2 1 Gastropoda 0 3 2 Oligochaeta 0 1 8 Polychaeta 9 33 38 Margin Non-seep 800 0 0 4 Aplacophora 0 0 2 Cnidaria 0 0 1 Crustacea 0 0 2 Nemertea 0 0 1 Ophiuroidea 0 0 1 Polychaeta 0 0 7 Sipuncula 0 0 1 800-900 m HR East Microbial mat 872-890 4 2 0 Gastropoda 2 0 0 Polychaeta 3 2 0 HR East Clam bed 872-880 2 1 0 Gastropoda 0 1 0 Polychaeta 6 12 0 Margin Non-seep 800 0 2 0 Bivalvia 0 2 0 Cumacea 0 2 0 Ophiuroidea 0 1 0 Polychaeta 0 4 0 IC Guilini et al./Progress in Oceanography 96 (2012) 77-92 81
Nematode biomass was then calculated with Andrassyi formula MDC was calculated based on the use of respectively the ôPOc or 5 SOb (And ras sy, 1956): term in the model. Fm was calculated for each macrofaunal individ ual and bulk nematodes sample, and then averaged within a higher W et weight ((.ig) = L (pm ) x W 2 (pm)/1.6 x IO6 taxon group to generate values for each habitat and location (±stan- dard deviation). Similar to Levin and Michener (2002) we used and a dry-to-wet-weight ratio of 0.25 was assumed (Heip et al., methane S13C values of -65%?, measured by Suess and Whiticar 1985). Nematodes were pooled into biomass and morphometric (1989). Knowledge on potential spatial and temporal variability of classes (length and width) on untransformed geometric scales. Bio methane S13C values is lacking and therefore impedes the use of a mass size spectra were created by plotting nematode cumulative range of values. The S13C of SOB was based on the average value relative abundance versus the biomass classes, while for creating of microbial mat bacteria measured in this study and the study of the morphometric classes, nematode relative abundance were plot Levin and M ichener (2002) (i.e. -3 3 .2 ± 5.6%?). The S13C of POC ted against geometric classes of length (pm) and width (pm). Addi was based on the average value for non-seep animals in each loca tionally body length, body width, oesophagus length and tail length tion (500-600 m: -19.1%^, 700-800 m: -20.2%?, 800-900 m: were measured of two specific dominant genera (five male individ -18.7%?). uals each) to determine the De Man ratios (a = body length/body width, b = body length/oesophagus length, c = body length/tail length; De Man, 1880). 2.4. Data analysis After extracting nematodes for identification, a minimum of 75 nematodes were handpicked for stable isotope analyses with a fine Statistical tests on nematode data were performed on a maxi needle from the top two centimeter fraction and rinsed in MilliQ mum of two replicate samples per habitat type per location. Prior water to remove adhering particles, before being transferred to a to the analyses nematode data were grouped into three sediment drop of MilliQ water in 2.5 x 6 mm aluminium (Al) cups. Nema depth layers (0-1, 1-2 and 2-5 cm), taking into account the indi todes specimens were picked in bulk or when present in sufficient vidual counts per depth layer. To test if the nematode community num bers, genus specific, i.e. Daptonema, and Metalinhomoeus to composition was significantly different between samples with gether with Linhomoeus since they could not be distinguished un different attributes, a non-parametric permutational ANOVA der the stereoscopic microscope. The Al cups were first preheated (PERMANOVA) was performed. Bray-Curtis similarity was used at 550 °C in order to remove all exogenous organic carbon. When as resemblance measure on standardized and square root trans present in the samples, large filamentous bacteria were also iso formed genus abundance data. The model had an unbalanced fully lated in Al cups. The cups with nematodes and bacterial filaments crossed four factor design and included the fixed categorical fac were then oven-dried at 60 °C, pinched closed and stored in Multi tors ‘location’ (Lo), ‘habitat’ (Ha) and ‘sediment depth’ (De), with well Microtitre plates under dry atmospheric conditions until anal the random ‘replicate’ (Re) factor nested in ‘Há’, since data from ysis. A PDZ Europa ANCA-GSL elemental analyzer interfaced to a different depth layers from a single replicate core are not fully PDZ Europa 20-20 isotope ratio mass spectrometer (UC Davis Sta independent, and all interaction terms. Calculation of the ble Isotope Facility, California) was used to measure the carbon sta Pseudo-F ratio and P value (a = 0.05) was based on 9999 permuta ble isotope ratios and carbon content. During this procedure a tions of the residuals under a reduced model. A restricted set of minimal He dilution was applied for the low biomass nematode appropriate a posteriori pair-wise tests was conducted where the samples. Immediate separation of bulk macrofaunal samples and interaction terms were found to be significant and where a PERM- further preparation for carbon stable isotope analysis was feasible DISP test confirmed the homogeneity of multivariate sample on board and performed as described in detail by Levin and dispersions. Where only a restricted number of unique permuta Michener (2002). Analyses were conducted on a Finnigan Conflow tions was possible (mainly in pairwise tests), p-values were ob 2 continuous flow system and a Fisons NA 1500 elemental analyzer tained from Monte Carlo samplings (Anderson and Robinson, coupled to a Finnegan Delta S isotope ratio mass spectrometer at 2003). To visualize the PERMANOVA results a non-metric multidi Boston University or on a continuous flow PDZ Europa 20/20 iso mensional scaling (MDS) plot was constructed based on the tope ratio mass spectrometer at UC Davis.While colloidal silica distance between centroids measure of nematode abundance data gel Ludox did not affect the 513C signal of shallow water nematodes per cm sediment depth for the three different habitats at each (Moens, unpubl. data) and Bengal Rose had no significant effect on location. The global R value of a Two-Way crossed ANOSIM test 513C values of shallow water foraminifera (Serrano et al., 2008), for differences between habitats across all locations indicates the buffered 10% formalin preservation resulted in a mean shift in car strength of the difference between the seepage habitats and the bon isotopes of sg-2.0%c relative to freezing as a control for several non-seep habitats across the locations. To reveal the variability macrofaunal invertebrate taxa (Fanelli et al., 2010). among and between sample groups a SIMPER analysis was per Isotope ratios are expressed as 513C in units per mil (%?), accord formed on standardized and square root transformed abundance ing to the standard formula 513C = [kSampie/^vpDB — 1] x IO3, where data. Additional m ultivariate and univariate PERMANOVA analyses R is the ratio of 13C/12C, RVpdb is Vienna Pee Dee Belemnite as ref (based on Bray-Curtis and Euclidean distance resemblance mea erence standard, and R sam Pie = [(ô13Csampie/1000) + 1] x RVpdb- A sure, respectively) with variations on the number of factors in more positive 5 value is isotopically enriched, meaning that pro the design as described above were used to test for differences in portionally more of the heavy stable isotope is present. The contri nematode trophic composition and nematode densities, meiofaun bution of methane-derived carbon (MDC) in the macrofaunal and al densities, meiofaunal taxon richness and nematode dominance, nematode carbon pool of each location and habitat was estimated respectively. A univariate, two fixed factor design for each location with a single isotope, linear mixing model using two sources (Fry was used to test for difference in 513C values between higher taxa and Sherr, 1984). The formula used was: at the different habitats (factors ‘Ha’ and ’Ta’ for Taxon; i.e. bacte ria, macrofauna and nematodes). A combined analysis with as ex Fm = (<5/ — <5poc or SOB ) f ~ ¿POc) tra factor ‘Lo’ would have been biased by the strongly unbalanced w here Fm is the fraction of MDC, and 8 j, ôm, ôPOc and ôSOb are the car design provoked by the low number of taxa sampled at the deepest bon isotope signatures of fauna, methane, particulate organic car site. Only for nematode, polychaete and gastropod data was bon (POC) and sulfide-oxidizing bacteria (SOB), respectively. For replication sufficient at each location to test for differences in each analyzed sample a maximum and minimum estimate of 513C signatures and MDC with water depth (factor ‘Lo’). 82
Table 3 Synthetic table with meiofauna densities (ind./10 cm2), nematode dominance (%), nematode densities (ind./10 cm2), nematode biomass (jig dwt/10cm2), Hill’s diversity indices (H0, Hi, H2, H¡nf), the rarefaction diversity index EG(51), mean relative abundance of feeding types according to Wieser (1953) (1A: selective deposit feeders, IB: non-selective deposit feeders, 2A: epigrowth feeders, 2B: predators/scavengers), and mean trophic diversity (® _1). Data are presented as averages with indication of the standard deviation, per location, per habitat. T3 g H« x u UC n o n O in ro m (N 05 m m m n i a oo m m oo o ui h ‘I oi c*i o oi in (N 05 O ■sf ÍN 00 ÍN ÍN 1—1 I-1 LO r-< ro ÍN n o 05 T-1 05 ro ro 05 LO p 00 o O ÍN o ÍN < m r< +i +1 +1 +1 +1 +1 +1 +1 +i +i +i +i ÍN K o ro 00 ro ÍN in n i-1 ro i—; O 00 O ÍN 1-1 ro 00 in 05 in T—i ro ÍN 05 r-> p ÍN ÍN ro o m C rn in +i +1 +1 +1 +1 +1 +i +1 +i +1 ra e ; d O 05 ÍN K ro r-> 05 05 in o O ro 1-1 in in ro 00 05 ro ro i—i ÍN ÍN in 00 in ro ÍN o N" in ÍN +i +1 +i +i +1 +1 +1 +1 +i +i m n n ^ ^ o o n to ro oo ro in 05 O 05 (N n to (Nn (N ro o co oo ÍN in 05 i-1 ÍN 05 o 00 o in p ÍN ÍN p O 05 i—i r-> in 00 to o ÍN ÍN cn O o ÍN r-H in +i +1 +1 +1 +i +1 +1 +i +i +1 1 +1 +1 n ro N5 O to ÍN ÍN p i-1 p o o K 00 TH 05 P rH in ÍN o p o 1-H ro - r P rn +i +1 +1 +i +1 +1 +1 +1 +i +i ; oo oq a í io io í a oq oo in N" 00 1-1 K p ÍN 05 p i-1 5 N 1-H 1-1 ÍN 05 O ro ÍN p ro N" o r-> o ÍN ro ÍN ro in ÍN ro 1-H o 1-H O O +i +1 +1 +1 +1 +1 +1 +1 +1 +1 K. Guilini et al./Progress in Oceanography 96 (2012) 77-92 (2012) 96 Oceanography in al./Progress et Guilini K. ÍN rn 05 n i p ro p 05 p 1-1 o IN r-n n i 1-1 ro p 00 p o ÍN 5 N ÍN n p to 1 - i 1-1 o 05 00 ro CO 5 N to r-> C ÍN in o ra e ; in r-. ro o ro 05 O 00 ÍN o t p r-> o p ro 1-1 n i r-. ÍN ro p O ro 05 ÍN ro 00 05 p ÍN ro p 1-H o +i +1 +1 +i +1 +1 +1 +1 +1 +1 posteriori. 3.2. Nematode assemblages Nematode 3.2. densities Meiofaunal 3.1. Results 3. tests pairwise ith w applied ere w analyses PERMANOVA variate Hin[; They w ere followed in decreasing order of appearance by adult adult by appearance of order decreasing in followed ere w They cribed to the type of habitat, location and interaction betw een both both een as betw differences interaction and Density found. location not 0.04). ere P(MC): habitat, w of type the non-seeps to microbial - cribed 0.02; P(MC): beds ats m clam test, microbial 5). - Table beds 0.042; clam test, P(perm): test, (Pairwise Main of (‘HaxDe’; 8.5%± 66 PERMANOVA depth The to ± ent 46% depth. 34 ent sedim from by sedim cm 10 ranged over cm first abundance total top the the in dance signif differed beds clam in classes etric STATISTICA orphom m The in 2008). atodes al., nem et of Anderson 2006; Gorley, and was ate (Clarke estim genera of an ber num Additionally ated replicates. estim the um of one axim for m the reached here was w off cut and data ent sedim per densities the al., account et into Heip _1; 0 taken ( data, diversity abundance trophic and 1968) Sanders, (EG(51); indices Hill’s The generated. ere w beds (Pairwise test, P(perm): 0.003) and microbial m ats and non- non- and densities. ats m est low microbial the and bored 0.003) P(perm): test, locations (Pairwise or beds detected was habitats effect did een A habitat and 4). betw Table 0.29; strongly > significantly varied P(perm) Meiofaunal inance A). differ tests, taxa dom (Main not (Appendix 1% atode other nem than and and therefore 2.3%) less richness (max. abundances taxon Oligochaeta relative and ith 4.8%) w A). (max. Appendix 3, chaeta (Table m 700-800 at sites non-seep in ±0.2% 98 deviations. metad rdal erae t et Fg 2. eaie abun Relative 2). (Fig. depth ith w decreased gradually and ent im abundances, genera relative on dependant variably are at th indices EMNV aayi o nmaoe eea eeld significant revealed genera atode nem of analysis PERMANOVA to m taxon 500-600 at inant beds dom clam 17%±in the 80 to from prised m com ranging A). 800-900 Nematodes habitats, at 3). all in (Appendix ranged ple (Table cm sam 0 ple at -1 depth m sam 0 m 800 from per microbial a layer in 2 ent taxa cm sedim 10 11 ind. the 44 to in from 3 densities from eiofaunal M varied richness mats - non-seeps P(MC): 0.03) and the 5-1 0 cm layer (Pairwise (Pairwise layer cm 0 5-1 the layer and cm 0.03) -2 1 P(MC): the to non-seeps - restricted mats ever how ere w habitat factor differences interaction Density the of effect significant a indicated results chi-square a distributions ith w sites frequency the non-seep the from hether w icantly test to used software was per add-on PERMANOVA+ software ere w 7 ith w analyses All v6 PRIMER families. atode ithin w nem ed of ber formed num untransform the for from made generated Expected the of were (EG(n)) curves Genera resemblance rarefaction of based ber Additional num used. distance ere w ‘Lo’. Euclidean matrices and ‘Ha’ factors uni fixed biomass and diversity atode nem in differences detect To layer. ep ape (arie et (em: .1) mcoil t har ats m microbial clam 0.015); and P(perm): ats m test, (Pairwise microbial ith w samples een 4), betw seep Table 0.009; found P(perm): test, differences (Main significant densities eiofauna m for Poly sized eiofaunal 6%), m (max. Nauplii 11.2%), (max. Copepoda standard by panied accom are means as expressed Results test. fit 1988) w ere calculated on sum m ed (0 -5 cm) and standardized standardized and cm) -5 (0 ed m sum on calculated ere w 1988) 1237 ± 568 ind. l O c n r 2 in non-seep margin samples at 700- 700- at samples margin non-seep in 2 r n c O l ind. 568 ± 1237 To asses nem atode structural and functional diversity, several several diversity, functional and structural atode nem asses To A total of 129 genera belonging to 31 families w ere identified. A identified. ere w families 31 to belonging genera 129 of A total taxon while observed, ere w taxa eiofauna m higher 19 of A total eaoedniis r hgeti h o woc fte sed the of cm o tw top the in highest ere w densities Nematode Hill, 1973), Expected num ber of Genera for 51 individuals individuals 51 for Genera of ber num Expected 1973), Hill, These data required a crossed, tw o factor design w ith ith w design factor o tw crossed, a required data These (H0 = num ber of genera, genera, of ber =num y2) (y goodness-of- goodness-of- Hi, H2, Hi, a Table 4 Main test results from univariate PERMANOVA analyses testing for differences in meiofauna density and taxon richness, and nematode dominance, biomass and diversity between different locations (Lo: 500-600 m, 700-800 m, 800- 900 m), habitats (Ha: non-seep, clam bed, microbial mat), and the interaction term (LoxHa). Significant results are indicated in bold. N O N - f i-i IN O CO (N n ÍO ’vf t CO IN o r s r IN s s s n m in rs rs rs IN OM V ÍN LO IN s o S’ m o mrs IN OO M in r to rv o m d ^ N O ÍN .Giiie l/rgesi caorpy9 21) 77-92 (2012) 96 Oceanography in al./Progress et K.Guilini : 34±., : . 10 c 95 20 ae rsnl undescribed. presently are ±2.0) 9.5 c: ±1.0, 7.5 b: ±7.6, 43.4 a: the and ±0.7) 9.2 abun relative variable highly genera This by 6). inant dom created Table the of partly 71-74%; is dances sites: habitats, age non-seep 42-69%, sites: (seepage and =0.02). habitats p = (R0.954, seepage locations the all een across betw habitats plot difference MDS non-seep the of the een an betw in strength the differences for visualized test ANOSIM ere w crossed ay Two-W A depth results 3). (Fig. ater w These thus and >0.05). (P(MC) location the did nor mats, microbials microbial and and beds beds clam een betw profile com ent Pair-wise sedim fac plete interpreted. com interaction the properly be locations, not three could the of 'LoxRe(Ha)’ each at tor habitats different acs agn fo 728.% Tbe ) Bt gnr ae repre are abun genera relative Both ith w 6). (Table depths 7.2-85.2% ater w from three all ranging at ats dances m microbial and (‘HaxDe’), depth ent sedim by habitats of interaction m). the of 800-900 effects and m 700-800 per m, (500-600 (individuals zones densities depth three meiofauna the for and atode deviations nem absolute average The 2. Fig. p (ae L 19. 40p, : 67±., : . 03 c: ±0.3, 5.8 b: ±2.8, 36.7 a: the pm, Both 74.0 ± sites. 1294.4 seep L: the at (male; species one sp. beds only by clam in sented genera abundant ost m five top the to seep een belonged betw Both as sites well as ithin non-seep w ithin each w variability at than assemblage habitats lower atode nem and seepage ithin variable w more is assemblages location atode nem of indicated larity additionally groups location all across groups sites habitat non-seep clam the at of assemblages position com 700- atode nem were and the on assemblage m effect differences any atode have 500-600 nem the significant the No both een betw for <0.05). (P(MC) found sites, the location over m non-seep 800 assemblages versus atode at nem m in differences identified the of parisons in replication nested ited lim to replicates Due by 5). (Table location ('LoxRe(Ha)’) and (‘LoxHa’) habitats habitat by location 10 cm horizontal bars and dots, respectively) w ith indication of the standard standard the of indication ith w respectively) dots, and bars horizontal 10 cm eut fo h SME nlss e ntae t t h simi the at th onstrated dem analysis SIMPER the from Results 800 - 900 m 900 - 800 0 0 m 800 - 700 0 -60 m 600 - 500 5-10 cm 5-10 cm 2-5 cm 0-1 cm0-1 1-2 cm 0-1 cm 1-2 cm Metadesmolaimus r e 5 30 5 60 5 900 750 600 450 300 150 niiul e 0c 2 cm 10 per Individuals Daptonema D--0- sp. (male; L: 761.2 ± 88.6 pm, pm, 88.6 ± 761.2 L: (male; sp. •i and and i i Clam bed Clam i i ^ Microbial mat mat Microbial ^ Metadesmolaimus. Non-seep Daptonema 83 IC Guilini et al./Progress in Oceanography 96 (2012) 77-92 Table 5 Main test results from multivariate PERMANOVA analyses testing for differences in nematode density, and genus and trophic composition between different locations (Lo: 500- 600 m, 700-800 m, 800-900 m), habitats (Ha: non-seep, clam bed, microbial mat), sediment depths (De: 0-1 cm, 1-2 cm, 2-5 cm), replicates nested within habitats (Re(Ha): A, B), and interaction terms. Significant results are indicated in bold. The significance of the main effect of LoxRe(Ha) on the nematode genus composition is not discussed in the text because low replicate numbers did not allow to test for homogeneity of dispersions (PERMDISP). Factors Nematode density Nematode genus Nematode trophic composition composition Location df 2 2 2 MS 40,952 3912 366 Pseudo-F 2.63 3.05a 2.15 Habitat df 2 2 2 MS 59,748 22,794 8643 Pseudo-F 5.77 10.70a 19.56a Depth df 3 2 2 MS 107,590 2280 76 Pseudo-F 27.93b 3.30a 0.53 Re(Ha) df 3 3 3 MS 10,530 2161 448 Pseudo-F 1.49 3.19b 2.89 Lo x Ha df 4 4 4 MS 37,433 3262 594 Pseudo-F 2.3 2.54a 3.49 Lo x De df 6 4 4 MS 7711 853 68 Pseudo-F 1.09 1.26 0.44 Ha x De df 6 4 4 MS 14,365 1176 298 Pseudo-F 3.67a 1.70a 2.10 Lo x Re(Ha)c df 4 4 4 MS 15,600 1283 170 Pseudo-F 2.21 1.89a 1.10 Re(Ha) x De df 9 6 6 MS 3772 691 142 Pseudo-F 0.53 1.02 0.91 Lo x Ha x De df 11 8 8 MS 3652 791 98 Pseudo-F 0.52 1.17 0.63 Residuals df 9 8 8 MS 7058 678 155 Total df 59 47 47 a 0.001 < p< 0.05. b p < 0.001. c Term has one or more empty cells. Global R: 0.954 seepage site are Leptolaimus, Morlaixia, Thalassomonhystera and p. 0.02 * an undescribed genus spp. of the Chromadoridae family. Common dominant genera (>3%) at all three non-seep sites were Daptonema, Acantholaimus, Tricoma, Monhystrella and Halalaimus. A total of 59 O ° / O . genera (50.4% of all genera, max. 18.8% of total density) were O 1700 - 800 m I exclusively found at non-seep locations. Most of these genera were / ° ♦ / o 1800 - 900 m I rare as only eight occured with >1% mean relative abundance at A Clam bed one of the three locations. At the seepage sites, 11 genera (15.7% Microbial mat of all genera, max. 11.5% of total density) occurred that were not Stress: 0.07 3 Non-seep found at the non-seep sites, with three genera being present with >1% mean relative abundance in one of the habitats at a certain Fig. 3. Non-metric m ultidim ensional scaling (MDS) plot based on distance betw een location ( Cephalochaetosoma, Metacylicolaimus, gen. 1 spp.l (Chro centroids of nematode abundance data per cm sediment depth for the three different habitats at each depth zone. The global R value of the Two-Way crossed madoridae)). With 13 genera exclusively found in microbial mats ANOSIM test for differences between habitat groups across all location groups and 15 genera exclusively found in clam beds, none of the habi indicates the strength of the difference between the seepage habitats and the non- tat-specific seepage assemblages is a subset of the other, at any seep habitats across the depth zones. of the studied locations. A multivariate PERMANOVA test based on feeding type relative abundances revealed a significant effect of the habitat type (‘Ha’; These two species were also found at the three non-seep sites, with Main test, P(perm): 0.023, Table 5). Significant differences in the the dominant seepage Daptonema sp. representing 48-90% of all composition of feeding groups occurred between microbial mats Daptonema’s found there, while only one and the same Metadesmo and clam beds versus non-seep sites (Pairwise test, P(MC): 0.006 laimus sp. occurred at seep and non-seep sites. Other dominant and P(MC): 0.04, respectively), but not between microbial mats genera occurring with >5% mean relative abundance at a certain and clam beds (Pairwise test, P(MC): 0.26). The microbial mats IC Guilini et al./Progress in Oceanography 96 (2012) 77-92 85 Table 6 Mean relative abundance of dominant (>3%) genera per habitat, per location, over 5 cm sediment depth and results from the analysis of similarities and genera contributions (SIMPER) indicating percentages of (dis)similarities. Non-seep % Clam beds % Microbial mats % 500-600 m Leptolaimusa 9.08 Daptonemaa 35.44 Metadesmolaimusa 85.20 Daptonemaa 9.02 Linhomoeusa 18.90 Daptonemaa 8.63 Molgolaimusa 8.06 Leptolaimusa 12.00 Linhomoeus 3.10 Tricomaa 7.00 Morlaixiaa 8.18 Rest 3.06 Acantholaimusa 6.44 Metadesmolaimusa 7.23 Monhystrella 4.72 Neochromadora 3.14 Campylaimus 4.33 Chromadorella 3.09 Halalaimus 4.09 Rest 12.02 Sabatieria 3.88 Thalassomonhystera 3.88 Desmodora 3.85 Rest 35.66 N on-seep Sim. 70.85 Clam beds Dissim. 75.09 Sim. 49.49 Microbial mats Dissim. 89.28 Dissim. 64.12 Sim. 49.97 700-800 m Daptonemaa 24.51 Daptonemaa 60.13 Daptonemaa 51.43 Acantholaimusa 11.53 Metadesmolaimusa 7.55 Metadesmolaimusa 23.39 Leptolaimusa 9.56 Intasiaa 4.00 Thalassomonhystera 6.92 Elzaliaa 7.07 Chromadorinaa 3.07 Rest 18.26 Halalaimusa 6.03 Terschellingiaa 3.06 Tricoma 4.43 Rest 22.20 Microlaimus 3.90 Terschellingia 3.26 Monhystrella 3.10 Rest 26.60 N on-seep Sim. 73.50 Clam beds Dissim. 59.01 Sim. 68.63 Microbial mats Dissim. 70.22 Dissim. 46.18 Sim. 41.86 800-900 m Daptonemaa 10.01 Daptonema 70.38 Metadesmolaimus 31.67 Acantholaimus 7.08 Leptolaimus 7.33 Daptonema 28.54 Monhystrella 6.15 Metadesmolaimus 5.19 gen. 1 spp. 1 (Chromadoridae) 6.68 Tricoma 6.14 Anticoma 4.78 Morlaixia 5.05 Elzalia 4.29 Rest 12.31 Metacylicolaimus 4.80 Microlaimus 4.26 Promonhystera 4.23 Desmoscolex 3.74 Leptolaimus 3.26 Dichromadora 3.69 Rest 15.77 Actinonema 3.53 Halalaimus 3.14 Rest 47.98 N on-seep Sim. 72.01 Clam beds Dissim. 75.09 Sim. N/A Microbial mats Dissim. 73.06 Dissim. 53.83 Sim. N/A a Genera contributing to >5% of the similarity within a habitat. Non - seep (800 - 900 m) Non - see p (500 - 600 m) Non - see p (700 - 800 m) Microbial mat (800 - 900 m) - O Clam bed (700 - 800 m) -O - Clam bed (500 - 600 m) Microbial mat (700 - 800 m) - o - Clam bed (800 - 900 m) - o Microbial mat (500 - 600 m) -<>- 0 , 0 U------1 . i i i i------20° 100 200 300 400 600 600 700 800 900 1000 Number of individuals Fig. 4. EF/G(n) rarefaction curves of the expected num ber of nem atode families (EF) and genera (EG) for a given num ber of sampled individuals (n). Means of replicates were plotted for each type of habitat at every location. and clam beds were characterized by a high dominance of non- tive deposit feeders and the epigrow th feeders (1A, IB, 2A; relative selective deposit feeders (IB; relative abundances range from abundances range from 22.2 ±3.8% to 40.7 ±0.1%; Table 3). 59.5 ±13% to 93.3 ±8.9%; Table 3), while at the non-seep sites Not any of the nematodes showed evidence of chemosynthetic three feeding groups were common: the selective and non-selec- symbionts. IC Guilini et al./Progress in Oceanography 96 (2012) 77-92 were plotted. The same trend was also found at the family level — Non- Seep — Clam bed and visualized as EF(200) (Fig. 4). Although no significant differ — Microbial mat ence was found due to a high variability, structural diversity corre sponded to trophic diversity at the non-seep sites, with highest average index values at 800-900 m water depth, followed by the 500-600 m and 700-800 m sites (Table 3). Structural and func tional diversity did not correspond at the seepage sites. 3.4. Nematode biomass and morphometry Length class (|jm) It is legitimate to interpret the biomass and morphometry re sults of the first sediment centimeter as a proxy for the total com B — Non- Seep munity as the PERMANOVA results showed before that there was — Clam bed no difference in nematode assemblages with sediment depth with — Microbial mat in any of the habitats (‘LoxHaxDe’; Main test, P(perm): 0.268; Table 5; Pairwise test for pairs of levels of ‘De’, within ‘Lo’ and within ‘Há’, P(MC) > 0.22). Generalisations about total biomass can how ever, not be made as densities in the top first centimeter are not representative for the complete sediment profile. Nevertheless, to tal nematode biomass was lowest in the top first centimeter of 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 microbial mats (4.2-16.8 pg dry weight per 10 cm-2), intermediate Width class (|jm) at the non-seep sites (20.6-37.8 pg dry weight per 10 cm-2), and highest at the clam beds (44.4-55.9 pg dry weight per 10 cm-2) Fig. 5. Relative abundance (%) of nematode individuals in (A) length geometric (Table 3). Significant differences were only found between the classes and (B) width geometric classes, from the three habitat types: non-seep, microbial mats and clam beds (‘Ha’; Main test, P(perm): 0.013; Ta Clam bed and microbial mat; based on the average of one replicate per location. ble 4; Pairwise test, P(MC): 0.001). These total biomass values do not reflect the nematode densities alone but are clearly influenced by nematode individual biomasses (i.e. body size). Averaged over A 5-16 — 500-600 m the three depths, highest biomass in the clam beds was linked to ^ 14 — 700-800 m a fraction of nematodes with higher individual length (mean: § 12 — 800-900 m 994 ±395 pm) and width (mean: 31.7 ±10.3 pm) compared to I 10 the non-seep nematodes (mean length: 707 ±518 pm, mean 8 width: 26.0 ± 12.2 pm; Fig. 5). Corresponding length frequency dis 6 tributions differed significantly (x2 test, p = 0.015), but width fre 4 quency distributions did not (x2 test, p = 0.28). The clam bed 2 biomass mainly increased due to the occurrence of nematodes 0 with lengths from 1000 to 1500 pm and width from 30 to 55 pm 1 # #& represented by 35.8 ± 6.5% of the total clam bed nematode assem Length class (gm) blages. This fraction of nematodes was mainly represented by 18 B Daptonema (mean: 93.9 ± 3.2%). Nematodes in the microbial mats 16 — 500-600 m — 700-800 m were also relatively long on average (mean: 1014 ± 1052 pm), 14 — 800-900 m 12 but had the lowest mean individual width (mean: 23.6 ± 10 14.2 pm). Mean individual length and width in the microbial mats 8 was however strongly influenced by the sample from 500-600 m 6 water depth where Metadesmolaimus was most numerous (186 03 4 versus 161 and 64 individuals, respectively), and was much thinner PERMANOVA pairwise tests found th at ô13C values differed signif Table 7 icantly between seepage habitats (clam bed and microbial mat) Main test results from univariate PERMANOVA analyses testing for differences in 613C values between habitats (Ha: non-seep, dam bed, microbial mat), taxa (Ta: and non-seep sites (Pairwise tests, P(MC) sg 0.024), but not be Polychaeta, Bivalvia, Ophiuroidea, Gastropoda, Crustacea, Sipuncula, Nemertea, tween the two seepage habitats (Pairwise tests, P(MC) > 0.138). Oligochaeta, Aplacophora, Cnidaria, Nematodea, bacterial filaments), and the inter Because the Permdisp analysis gave a significant difference in data action term (HaxTa); separately tested for each location (500-600 m, 700-800 m, deviation from the centroids among groups prior to the pairwise 800-900 m). Significant results are indicated in bold. ‘Ha’ analysis for the 700-800 m site, it is unclear if the pairwise test Factors 500-600 m 700-800 m 800-900 m results are based on data dispersion, location or most likely as seen H abitat df 2 2 2 in Fig. 7, a com bination of both. MS 899 1420 82 At the non-seep sites, all faunal 513C values averaged per higher Pseudo-F 10.54b 15.48b 1.38 taxon level and per sampling date, fell more or less well within the Taxa df 2 2 2 range of —15 %o to - 2 5 %o (Fig. 7), indicating th at the assim ilated MS 18 7 92 carbon in the nematodes and macrofauna was mainly of phyto- Pseudo-F 0.21 0.08 1.54 planktonic origin. Taxon mean 513C values at the non-seep loca Ha x Tac df 3 2 0 tions ranged from -16.2%c to -25.9 %B, while mean 513C values MS 14 74 for non-seep nematodes and macrofauna over all locations were Pseudo-F 0.16 0.80 No test -20.12 ±0.46%o and -19.45 ±3.25%c, respectively. The maximum Residuals df 100 202 33 MDC percentage for single non-seep taxa measurements ranged MS 85 92 60 from 0% to 13%, with positive values measured in ophiuroids, an Total df 107 208 37 aplacophoran, an amphipod, and a cnidarian. a 0.001 < p < 0.05. In the clam beds across the seepage sites, nematode 513C values b p < 0.001. differed significantly between the two shallowest and the deepest c Term has one or more empty cells. location (‘Lo’; Pairwise test, P(perm) sg 0.001), with nematodes from the 800-900 m location depending significantly less on MDC than the nematodes from the 500-600 m and 700-800 m of relative abundance, copepods and nauplii, and polychaetes sites (MDC maxima: 18.6 ±1.5% versus 25.4 ±0.7% and followed in dominance and nematodes showed distinct surface 26.1 ± 5.4%, respectively; Pairwise test, P(perm) sg 0.0004). The (0-2 cm) maxima in both studies. According to Giere (2009), the polychaete and gastropod data showed no pattern as such (‘Lo’, colonization of anoxic, sulphidic sediments at HR by the rotifer Main test, P(perm) > 0.55). Most taxa from the seepage sites species is amazing, considering their preference for well-oxygen showed a wide range of carbon isotopic signatures (Fig. 7), reflect ated sediments. We therefore suggest that based on (i) the more ing varying nutritional sources across taxa within a site and across or less equal vertical distribution of rotifers in the sediment in each sites within a taxon. Mean 513C values for higher taxa ranged from sediment core, (ii) the non-significantly differing densities in each -16.2%c to -48.1%c, indicating that 0% to maximum 59 ± 11% of sediment layer between all studied seep habitats and non-seep their carbon was derived from methane. The majority of the seep sites, and (iii) the occurrence of only one species, provisionally age nematodes and macrofauna had mean 513C values between identified as the cosmopolitan fresh and brackish water species Le -27%c and -40 %o (i.e. m inim um 0% to m axim um 46% MDC). 513C cane closterocerca (Shmarda, 1859) (Dr. H. Segers, personal com values of polychaetes, gastropods, and bivalves at several seepage munication) in the study of Sommer et al. (2003, 2007), in sites, however, also showed overlap with carbon isotope signatures addition to (iv) the general lack or very low abundances of rotifers indicative for phytoplankton derived carbon (513C > -2 5 %v\ Fig. 7), in deep-sea marine environments with avoidance of hypoxic con emphasizing a mixed diet observed when combining species into ditions (Neira et al., 2001 ), that the occurrence of this rotifer as de higher taxa. Measurements on a single ophiuroid and nemertean scribed by Sommer et al. (2003, 2007) might possibly result from specimen that were occasionally found in a clam bed at 500- contamination. Rotifera were identified before as the dominant 600 m water depth had fairly high 513C values, indicating the incor taxon in fresh tap water, reaching up to 1500 individuals per m3 poration of mainly phytoplankton derived carbon (513C: -16.4%c of water (Schellart, 1988; Jeurissen and Vermaercke, 1990; Funch and -21.7%c, respectively). Isotopic signatures of large, filamen et al., 1996). tous bacteria forming visible white and orange mats at the two shallowest seepage sites, were indicative for chemosynthetic car 4.2. Seep contribution to nematode community heterogeneity bon fixation via sulfide oxidation, with averages ranging from —28.4 ± 5.6%c to —37.8 ± 0.9 %o. While seep locations with similar habitat types as found at HR exist, geographic heterogeneity manifests through variations in water depth, intensity of fluid flow, long-term stability, availability 4. Discussion of nutrients, the variety of substrates present, CH4 and H2S concen trations and the presence or absence of an OMZ affecting both the 4.1. Metazoan meiofauna in Hydrate Ridge sediments seep and its surrounding sediments (Cordes et al., 2010). The com bination of all these factors may have contributed to some extent Sommer et al. (2007) presents the only study to date on meta to the establishment of a unique nematode community at HR. zoan meiofaunal standing stocks at HR. The surprisingly high dom Average genus richness at HR seep habitats was relatively high, inance of rotifers at clam beds, microbial mats and control while evenness corresponded well with the average Shannon sediments often exceeded nematode dominance (rotifers: 4-85%, diversity index and Pielou’s evenness for known other deep-sea nematodes: 13-90%), and differs from our results (rotifers: 0%, seeps (Vanreusel et al., 2010a). This reflects the fact that the HR nematodes: 80-98%; Appendix A). Despite the discrepancy in nematode assemblages exhibit a typical trend of high dominance occurrence of rotifers, the mean nematode density (individuals by one or two species representing 50-90% of the total community per 10 cm 2) reported by Sommer et al. (2007) for each habitat (Vanreusel et al., 2010a). At HR, Daptonema and Metadesmolaimus type corresponds well with the means observed in this study, irre together represented 43-94% of the total community, whereas a spective of water depth (microbial mats: 162.7 versus 206.4, clam significantly higher nematode diversity was observed, as expected, beds: 546.5 versus 578.6, 921.7 versus 685.5). In decreasing order in the non-seep sediments at all three sampled water depths, with K. Guilini et al./Progress in Oceanography 96 (2012) 77-92 similar trends observed at the genus and family level. Since on mortenseni; reviewed in Vanreusel et al., 2010a). It seems feasible average slightly lower nematode individuals were identified in that these species might tolerate high sulfide and methane concen the non-seep compared to the seep samples (414 ±121 versus trations in shallow marine environments where high loads of or 551 ±263, respectively), diversity in the non-seep reference sites ganic matter can rapidly deplete the sediment of oxygen and might even be underestimated. Seep and non-seep sediments induce biogenic production of reduced elements, while they be shared only 50% of their nematode genera, which is similar to come largely outcompeted in the relatively food-poor, well-oxy the result found for the macrofauna (Levin et al., 2010) and empha genated deep-sea sediments. Deep-sea cold seeps located near sizes the significance of the HR seeps to regional diversity. shore on margins are exceptions that could be accessed by drifts At the smallest scale, varying fluid fluxes and availability of of loose seaweed (e.g. Fucus sp.), as assumed for the HMMV (Van chemicals largely determine the habitat heterogeneity at spatial Gaever et al., 2009c). scales relevant for microbiota but also for the symbiont-bearing Metazoan standing stock (densities and biomass) along the reg mega- and macrofauna such as tubeworms, mussels and clams ular slope generally decreases with water depth and surface pri (MacDonald et al., 1989; Cordes et al., 2010). The higher structural mary productivity (Rex et al., 2006). It is expected that both complexity created by the presence of habitat-forming clam beds endosymbiont-carrying and heterotrophic metazoans at deep-sea at HR resulted in higher diversity of the associated macrofauna seeps profit from the excess supply of autochthonous organic car (Sahling et al., 2002; Levin et al., 2003). Our data, in contrast, indi bon in addition to the carbon input through the water column, cated that nematode communities at HR do not consistently profit which may lead to elevated densities and/or biomass compared from clam bed increased habitat heterogeneity in terms of diver to adjacent non-seep sediments. Nematode densities at seeps vary sity. Beside the pumping and oxygenating effect on the sediment from around 10 to 11,100 ind. lO cnr2 (reviewed in Vanreusel down to 5 cm depth, which may both influence the macro- and et al., 2010a), and often but not always exceed densities in the meiofauna, Calyptogena clams induce an increased habitat com background sediments. Densities at the HR seepage habitats varied plexity to the sediment surface where they visually form extended from 40 to 869 ind. 10 c n r2 and only exceeded the densities in the beds. Macrofaunal epizoonts living on the shells (gastropods, actin- surrounding margin sediments at 800-900 m water depth. It ians, dorvilleids and scale polychaetes; Levin, 2005) or crawling should, however, be noted that total nematode abundances at and swimming macrofaunal species may additionally be using the non-seep reference sites could be underestimated due to the the shells as e.g. refuge or settling and feeding grounds. Addition use of a 42 pm mesh sieve compared to the 32 pm mesh sieve ally, in contrast to the macrofauna at HR where the microbial mat mesh that was generally used at the seep sites (Leduc et al., fauna differed from the clam bed fauna (Sahling et al., 2002) or ap 2010). The clam beds that are characterized by sulfide concentra peared largely to be a stress-tolerant subset of the clam bed fauna tions mostly below 2 mM in the top five centimeters of the sedi (Levin et al., 2010), the nematode communities could not be distin ment generally attained higher meiofaunal and nematode guished based on genus composition. The within-habitat heteroge densities than the bacterial mats where sulfide concentrations neity (at the scale of tens of centimeters to a meter between tube can reach over 28 mM (Sahling et al., 2002; Valentine et al., cores) was on average as high as the between-habitat heterogene 2005; Sommer et al., 2007). The fact that the proportional vertical ity for clam beds and microbial mats, indicating a low genus turn distribution of the nematodes in the HR sediments did not differ over rate between the different seepage habitat types. The between clam beds and microbial mats, illustrates however that turnover both within and between the seep habitats mainly was sulfide concentrations might not necessarily hamper nematode manifested as a shift in dominance at the nematode genus level, densities. Oxygen penetration in microbial mats at HR is less than with the genera Daptonema and Metadesmolaimus as key players. a few millimeters (Knittel et al., 2003), but similar to sulfide, it did Our study reports the first dominant Metadesmolaimus seep- not seem to influence the nematode abundances and distribution associated species, whereas Shirayama and Ohta (1990) also re in the sediment. Neither did the overall quantity of organic carbon ported two Daptonema species dominating with 33% relative abun that was estimated to be almost twice the amount in microbial dance underneath Calyptogena soyoae clam beds at the Hatsushima mats compared to clam beds (Valentine et al., 2005). Therefore, seep site, Sagami Bay (Northwest Pacific Ocean). There is however we expect that it is rather the combination of the transient nature no information available to verify whether one of them is the same of seepage, seep intensity and interactions with the habitat-struc Daptonema species that co-dominates with Metadesmolaimus at the turing and/or associated benthic taxa that determine nematodes HR seep sites. Nevertheless, what distinguishes both the HR and success. Hatsushima seep from other seeps for which nematode data are In terms of biomass, nematode assemblages at the seep habitats available is their occurrence within low oxygen realms. In the at HR confirm the suggested proportionally increased biomass Sagami Bay intense mineralization in the water column results in compared to non-seep assemblages. A higher biomass, induced a relatively stable oxygen-deficient layer reaching minimum val by a fraction of nematodes with relatively higher length and width, ues of ca. 55-60 pmol I-1 at 1200-1400 m water depth (Glud indicates at least a fraction (mainly represented by Daptonema) of et al., 2009), while the seep is found at 1170 m water depth (Shi the nematodes at the clam beds may profit from the high local che- rayama and Ohta, 1990). At both seep locations (HR and Sagami mosynthetic production, despite other factors interfering in deter bay) the dominant seep-associated Daptonema and some other less mining density. Van Gaever et al. (2009a) suggested the same for S. dominant genera (relative abundances >3%) also occurred as mortenseni, the dominant nematode genus present in mussel and (sub)dominant genera in the surrounding non-seep sediments. clam habitats at the REGAB pockmark (Gulf of Guinea, South-East These findings favor the idea that oxygen stress creates a pool of Atlantic). The same seems true for the nematodes in the HR micro species that are tolerant to low-oxygen levels; which may colonize bial mats, except for one microbial mat at 500-600 m water depth, the seeps and further adapt to sulphidic conditions to increase w here Metadesmolaimus dominated, representing 93% of the total their local success. This is in contrast to what has been found pre meiofaunal community, whereas in other measured microbial viously at deep-sea seeps located in well-oxygenated waters, mat samples its relative abundance was at most 52%, and a maxi where generally the dominant nematode species were not or only mum of 5% in clam bed samples. As the microbial mat and clam rarely encountered in background sediments. Instead, these domi bed nematode assemblages were indistinguishable but entailed a nant species were morphologically similar to species that are com high within- and between-habitat variability, by selecting only mon at intertidal habitats over a broad geographical range (e.g. one replicate per location for body measurements, we might have Halomonhystera disjuncta, Terschellingia longicaudata and Sabatieria seen genus-dependant patterns rather than habitat-type induced K. Guilini et al./Progress in Oceanography 96 (2012) 77-92 Non-Seep i Clam bed ! Microbial mat j Non-Seep ! Clam bed I Microbial mai j Non-S&ep I Clam bed I Microbial mat 0 -10 -20 ‘Hr ¿ -30 •tj. O * -40 r H l -50 -60 -70 • Polychaeta ♦ Crustacea + Nemertea o Bacterial filaments o Bulk Nematoda a Bivalvia * Sipuncula • Oligochaeta * Cnidaria A Daptonema spp. ▼ Ophiuroidea «Gastropoda * Aplacophora □ (Meta-)Linhomoeus spp. Fig. 7. ô13C (%o) isotope values (mean ± standard deviation where n > 2, per taxon per sampling campaign) of bacterial filaments, nematodes and macrofaunal taxa collected at the different habitats (non-seep, clam bed, microbial mat) at all three locations (500-600 m, 700-800 m, 800-900 m). The gray area from -15 % o to - 2 5 % o indicates the range of ô13C values obtained for benthic consumers that mainly depend on phytoplanktonic carbon. The gray area below -50 % o indicates the assimilation of mainly methanotrophically derived carbon. differences in geomorphometric characteristics. Our results how quantity, the composition or the quality of the sedimentary organic ever, might illustrate that where the survival of other nematode matter may have influenced nematode diversity. genera is ham pered, Metadesmolaimus might have an advantage The nematode assemblages from the non-seep margin sedi form elongated body-length and therefore higher surface area-to- ments within the OMZ consist of several genera that are known volume ratio; a characteristic that was suggested to be an adapta for their widespread distribution in the deep sea ( Acantholaimus, tion to low oxygen partial pressure, epidermal uptake of DOM and Desmodora, Desmoscolex and Halalaimus) (Vanreusel et al., a high mobility for nematodes living in deeper suboxic or anoxic 2010b). Together with the genera Daptonema, Microlaimus, Sabatie conditions (Jensen, 1986, 1987; Schiemer et al., 1990; Soetaert ria and Thalassomonhystera, which each contribute more than 2.5% et al., 2002). average relative abundance in slope habitats worldwide (Vanreusel et al., 2010b), these genera contributed each more than 3% to the 4.3. OMZ effects on nematode communities similarity between replicates of at least one of the sampled Casca- dian non-seep margin locations. The most remarkable difference in Most diversity research within OMZs has focused on Foraminif relative abundances between the assemblages found in this study era, macrofauna, or megafauna. These groups are characterized by and the averages for genera found in slope sediments worldwide a low number of taxa, low species richness and diversity, and high is the higher relative abundance of Daptonema and the lower rela dominance of a few species when compared with more oxygenated tive abundances of Sabatieria and Thalassomonhystera in our study surrounding habitats (Levin et al., 2001 ; Levin, 2003; Gooday et al., (Daptonema: 14.8 ± 8.6%, Sabatieria: 2.1 ± 1.7%, Thalassomonhyster- 2009, 2010). The only study on metazoan meiofauna diversity a: 2.3 ± 1.6, versus 5.51%, 8.71% and 9.65%, respectively). W hether within and outside an OMZ reports a considerable drop in number this variability is the result of the oxygen minima found on the Cas- of taxa with decreasing oxygen concentrations and increasing or cadian margin or other biotic or abiotic variables, is a question that ganic carbon due to the absence of taxa other than nematodes cannot be answered at this time. Including nearby margin sedi (Veit-Köhler et al., 2009). The non-seep study locations at HR how m ents outside the OMZ will be necessary to find out if the OMZ ever, are all situated close to the OMZ core. Therefore the range of sediments favor the success of nematode species that are tolerant bottom-water oxygen concentrations at the different sampled (colonists) or adapted (endemics) to low oxygen levels and which water depths (±0.7-0.2 0 2 ml I-1) might have been too small to ob species additionally (colonists) or alternatively (endemics) occur at serve any clear elimination effect on meiofaunal taxa that are less seepage sites due to lack of tolerance of most other metazoan life tolerant to oxygen stress. Comparing the rarefaction index for or a competitive successful adaptation, respectively. In the study nematode genus diversity (EG(51)) between the OMZ sediments of Levin et al. (2010), where the macrofaunal assemblages were at HR and average global slope sediments (Vanreusel et al., studied at seep and non-seep sediments at depths of 800 m and 2010b) indicates that nematode genus diversity within the OMZ 500 m on the Oregon and Californian margins, only four species at HR is nevertheless relatively low (14.7 ± 1.6 versus 19.0). On a were confined to the OMZ setting outside of seeps. This suggests regional scale, average nematode diversity was highest at the that the relatively weak NE Pacific OMZ contributes only minimal deepest non-seep location with lowest oxygen concentrations to regional macrofaunal diversity (Gooday et al., 2010). However, (800-900 m water depth; Table 1). Although diversity usually in as nematodes have a different life style, being smaller, less mobile creases with increasing water depth along the upper continental organisms with no pelagic larval stage, it is reasonable to speculate slope (Rex, 1983), this was not the case at HR, where average diver that OMZs have isolated hypoxia-tolerant nematode species more sity was lowest at intermediate water depth. Carbon and nitrogen easily e.g. as a result of expanding or contracted OMZs, resulting concentrations in the sediment increase with increasing water from shifts in global temperatures. Once isolated, spéciation may depth at the non-seep margin locations off Oregon (500 m: 0.9% be further promoted (Rogers, 2000). C, 0.1%N; 800 m: 2.14% C, 0.28% N (Levin et al., 2010); 770 m: Levin et al. (2010) hypothesized that seep-related heterogeneity 1.99% C, 0.25% N (Valentine et al., 2005), but also do not affect would be homogenized where bottom-water oxygen concentra nematode diversity. However, as structural (based on the nema tions are lowest. A comparison of macrofaunal seep and non-seep tode community composition) and functional trophic (based on assemblages from HR and Eel River did not support this hypothesis feeding types according to Wieser, 1953) diversity show the same (Levin et al., 2010). Nor do our nematode assemblages along a trend with water depth, we might assume that rather than water depth and bottom-water oxygen gradient at HR. The 90 K. Guilini et al./Progress in Oceanography 96 (2012) 77-92 dissimilarity between assemblages in surrounding non-seep sedi sulfide-oxidizing bacteria that are common at OMZs (e.g. Thioploca, ments and seepage habitats at HR decreased with decreasing 02 Beggiatoa; Levin, 2003). While Levin and Michener (2002) found concentrations from 500-600 m to 700-800 m water depth distinct macrofaunal 513C values for microbial mats (75.09 and 89.28 versus 59.01 and 70.22, respectively), but in (-4 3 .8 0 ± 3.63%c), clam beds (-3 3 .3 8 ± 2.17%c) and non-seep sedi creased again towards 800-900 m water depth (75.09 and 73.06), ments (-20.78 ± 0.92 %v), our study only found seep versus non- where the 02 concentrations were lowest. seep related differences for all analyzed benthic taxa. High variance Metazoan meiofauna in total and more specific nematode den between and within the seep-associated higher taxa was responsi sities often reach maximum values at lowest oxygen concentra ble for this, and indicates species consume a variety of food sources. tions within OMZs (reviewed in Levin, 2003). The data from the There was no greater incorporation of carbon derived from meth present study and the study by Veit-Köhler et al. (2009) illustrate ane or sulfide oxidation in microbial mats, as found by Levin and that this is not always true. Lowest nematode densities at HR were Michener (2002), who sampled HR 7 years earlier. Mean macrofa reached at the deepest location (886-887 m) while highest densi unal and nematode 513C values ranged from -16.2%c to -48.1%c, ties occurred at the intermediate depth (797 m) (Table 1 ). Macro which corresponds to 0% to maximum 59 ± 11% MDC, and implies benthos densities in margin sediments at HR followed the same that they additionally incorporated photosynthetically produced pattern, with annelids dominating the macrobenthos at each depth carbon. A more detailed interpretation of the stable isotope signa with relative abundances ranging from 67.5 to 76.2% (Levin et al., tures is however complex due to the variety of potential food 2010). In the OMZ of the Arabian Sea, nematode densities were also sources. On the one hand there is MDC w hich can be obtained positively correlated with total macrobenthos and annelid abun through endosymbiotic methanotrophs, or through direct or indi dances, in particular with the dominant tube-building family Spi rect consumption of free-living anaerobic (archaea) or aerobic onidae (Cook et al., 2000). Possibly the effect of different annelid methane-oxidizing bacteria, or sulfide-oxidizing or sulfate-reduc or other macrofaunal species that burrow, bioturbate and/or ing bacteria, which also can take up MDC after it passes into the dis bio-irrigate deep-sea sediments might induce higher nematode solved inorganic carbon (DIC) pool (Thurber et al., 2010). On the survival rates in a similar way as has been shown for different other hand at HR there is an unlimited DOC-supply with a 513C iso tube-building polychaetes in intertidal and shallow subtidal sedi topic composition of -36.5%c that is assumed to contribute to 60% ments (Reise, 1981; Tita et al., 2000; Pinto et al., 2006; Braeckman of the carbon in seep sediments (Valentine et al., 2005). This seep et al., 2010). ing DOC may be taken up directly through uptake of sediment par Reduced body size and flattened tests, which both lead to an in ticles to w hich DOC is absorbed (Sposito et al., 1999; Kennedy et al., crease in the surface area-to-volume ratio, are typical feature of 2002) or indirectly after being accumulated in intermediate con some OMZ metazoan macrofauna and hypoxia-tolerant foraminif- sumers such as heterotrophic sulfate-reducing prokaryotes (Knittel eral species, respectively (Levin, 2003; Gooday et al., 2010). We et al., 2003). A third potential carbon source is phytoplankton-de- speculated that a body elongation and subsequent increase in sur- rived carbon, which can also be taken up directly or indirectly. A face-volume ratio might also emerge in the upper sediment layer mixture of several of these carbon sources may yield faunal 513C along the investigated depth and oxygen depletion gradient, as signatures that were observed for the majority of the macrofauna was suggested before for nematodes living in and being adapted and nematodes (i.e. between -27%o and -40 %v), and which are to deeper suboxic or anoxic conditions (Jensen, 1986, 1987; Schi- indistinguishable from those associated with form I Rubisco sulfide emer et al., 1990; Soetaert et al., 2002). The minimum bottom- oxidation (i.e. -27%o to -35 %v\ Robinson and Cavanaugh, 1995; water oxygen concentrations at 800-900 m water depth seemed Cavanaugh and Robinson, 1996). Moreover, a combined consump to have this body elongation effect on part of the community. How tion of sulfide oxidizers having form I and II Rubisco (i.e. -27%o to ever as it co-occurred with an increase in body width, the resulting -35%c and -9 %o to -16%c, respectively; Robinson and Cavanaugh, increased biomass seems more likely the result of higher concen 1995; Cavanaugh and Robinson, 1996) may yield tissues with trations of food in the sediment. phytoplankton-like 513C signatures (Levin and Michener, 2002). Hitherto none of the seep nematode species in the deep sea 4.4. Importance of methane-derived carbon for macrofaunal and have shown evidence of symbionts. Instead, most nematodes from nematode communities seeps are classified as deposit feeders, based on their small buccal cavity and the absence of teeth (Vanreusel et al., 2010a). At HR, Cold seeps are isolated areas that afford opportunities for higher especially the non-selective deposit feeders (group IB according energy exploitation in the deep sea that is often considered food to Wieser, 1953) dominated at all seepage sites with average rela limited. At HR, methane released from decomposing gas hydrates tive abundances ranging from 59.5 ± 1 37,, to 93.3 ± 8.9'A. Com induces microbial autotrophic processes which provide the benthic bined analyses of fatty acids and stable isotopic signatures of the food web with additional organic carbon. Based on 14C02 fixation non-selective deposit feeding nematode H. disjuncta from the Nor experiments, a 24-fold increase in autochthonous production of or dic margin indicated that this species proliferates on the highly ganic carbon was measured at the southern HR seep sites (700- abundant sulfide-oxidizing bacteria, and is therefore trophically 800 m) in comparison to 5.7 mg C n r 2 d~3 at control sites (Sommer adapted to conditions within the Hákon Mosby Mud Volcano bac et al., 2002). The presence of distinct communities of macrofaunal terial mats (Van Gaever et al., 2009b). Besides evidence of a mixed and nematode species that feed on this chemosynthetic derived diet on autochthonous and allochtonous organic carbon, 513C sig carbon is evidenced by both the results from the macrofaunal and natures of seep-associated nematodes at HR reveal a significantly nematode community composition and 13C stable isotope analyses lower MDC% or potentially higher portion of phytoplankton de in this present study and the study of Levin and Michener (2002) rived carbon in the nematodes diet at the deepest site, compared and Levin and Mendoza (2007). At the non-seep margin locations to the shallower sites. This might reflect an increased availability average 513C values were -20.12 ± 0.46%o for nematodes and of phytodetritus at that depth as suggested by the highest concen -1 9 .4 5 ± 3.25%e for macrofauna, the latter corresponding well with trations of carbon in the sediment at the deepest non-seep site. To the results from Levin and Michener (2002) who found elucidate however, the potential occurrence of chemoautotrophic -20.78 ± 0.92 %o at the southern non-seep Oregon margin site endosymbionts, trophic level, carbon and nitrogen sources and fix (770 m). These signatures result from the incorporation of mainly ation pathways in the nematodes at HR, it is thus highly recom phytoplankton derived carbon and perhaps to some extent from mended to use 51SN in conjunction with 513C analysis, preferably an association with endosymbiotic or consumption of free-living measured on separate fatty acids of both nematodes and bacteria. IC Guilini et al./Progress in Oceanography 96 (2012) 77-92 91 5. Conclusions Andrassy, I., 1956. The determination of volume and weight of nematodes. Acta Zoologica Academiae Scientiarum Hungaricae 2,1-15. Bernhard, J.M., Buck, K.R., Farmer, M.A., Bowser, S.S., 2000. The Santa Barbara Basin Knowledge of trends in nematode diversity and assemblages at is a symbiosis oasis. Nature 403, 77-80. cold seeps and OMZs has steadily grown during the last decade. Boetius, A., Ravenschlag, K., Schubert, C.J., Rickeri, D., Widdel, F., Gieseke, A., Amann, Our efforts reveal that both seep and OMZ sources of margin het R., Jorgensen, B.B., Witte, U., Pfannkuche, O., 2000. A marine microbial consortium apparently mediating anaerobic oxidation of methane. Nature erogeneity generate distinct and unique nematode assemblages, 407, 623-626. although they share certain environmental and faunal characteris Braeckman, U., Van Colen, C., Soetaert, K., Vincx, M., Vanaverbeke, J., 2010. tics w ith seeps and OMZs worldwide. The OMZ homogenizes the Contrasting macrofaunal activities differentially affect nematode density and diversity in a shallow subtidal marine sediment. Marine Ecology Progress Series habitat-related heterogeneity within the HR seep, but the increas 422, 179-191. ing homogenizing effect on nematode assemblages from seep and Carlier, A, Ritt, B., Rodrigues, C.F., Sarrazin, J., Olu, K., Graii, J., Clavier, J., 2010. non-seep sediments did not increase along the relatively narrow Heterogeneous energetic pathways and carbon sources on deep eastern Mediterranean cold seep communities. Marine Biology 157, 2545-2565. bathymetric and bottom-water oxygen gradient. Relative to non- Cavanaugh, C.M., Robinson, J.J., 1996. C02 fixation in chemoautotrophic- seep assemblages the HR nematode seep assemblages exhibited invertebrate symbioses: Expression of form I and form II RubisCO. In: low diversity with dominance of only one or two genera (in this Lidstrom, M.E., Tabita, F.R. (Eds.), Microbial Growth on Q Compounds. Kluwer, Dordrecht, The Netherlands, pp. 285-292. case Daptonema and Metadesmolaimus), higher average nematode Clarke, K., Gorley, R.N., 2006. Prim er v6: User Manual/Tutorial. PRIMER-E, individual biomass, and 513C evidence for strong dependance on Plymouth. chemosynthesis-derived carbon. These attributes are shared with Cook, A.A, Lambshead, P.J.D., Hawkins, L.E., Mitchell, N., Levin, L.A., 2000. Nematode abundance at the oxygen minimum zone in the Arabian Sea. Deep-Sea Research deep-sea seeps worldwide. This study is the first however, to re Part II 47, 75-85. veal that nematode communities at seeps under the influence of Cordes, E.E., Cunha, M.R., Galéron, J., Mora, C., Olu-Le Roy, K., Sibuet, M., Van Gaever, an OMZ may differ from those at deep-sea seeps in well-oxygen 5., Vanreusel, A, Levin, L.A., 2010. The influence of geological, geochemical, and ated water. In the latter the dominant genera are rarely encoun biogenic habitat heterogeneity on seep biodiversity. Marine Ecology 31, 51-65. Cowie, G.L., Hedges, J.I., 1992. The role of anoxia in organic matter preservation in tered in the oxygentated background sediments and presumably coastal sediments: relative stabilities of the major biochemicals under oxic and originate from sulfide-rich, shallow areas. In contrast an OMZ anoxic depositional conditions. Organic Geochemistry 19, 229-234. may host a potential source of infaunal species adapted to low De Man, J.G., 1880. Die einheimischen, frei in der reinen Erde und im süßen W asser lebende Nematoden monographisch bearbeitet. Vorläufiger Bericht und oxygen levels, which can then colonize and further adapt to the descriptiv-systematischer Theil. Tijdschrift der Nederlandse Dierkundige prevailing reducing seep conditions. Future taxonomic work that Vereeniging 5, 1-104. combines morphological and molecular characterizations of Demopoulos, A.W.J., Gualtieri, D., Ko vacs, K., 2010. Food-web structure of seep sediment macrobenthos from the Gulf of Mexico. Deep-Sea Research Part II 57, nematode species from HR sediments, adjacent oxygenated slope 1972-1981. sediments, shallow-water sediments and nearby cold seeps is Fanelli, E., Cartes, J.E., Papiol, V., Rumolo, P., Sprovieri, M., 2010. Effects of however, required to resolve the degree of endemicity at the HR preservation on the ô13C and 815N values of deep sea macrofauna. Journal of Experimental Marine Biology and Ecology 395, 93-97. seep and OMZ. Fry, B., Sherr, E.B., 1984. A13C m easurem ents as indicators of carbon flow in m arine and freshwater systems. Contributions in Marine Science 27,13-46. Funch, P., Segers, H., Dumont, H.J., 1996. Rotifera in tap water in Gent, Belgium. Acknowledgements Biologisch Jaarboek Dodonaea 63, 53-57. Giere, O., 2009. Meiobenthology. The Microscopic Motile Fauna of Aquatic We thank Dr. Wilhelm Weinrebe for providing bathymetric Sediments. Springer, Berlin. data on the Hydrate Ridge and the Flanders Marine Institute (VLIZ), Glud, R.N., Thamdrup, B., Stahl, H., W enzhoefer, F., Glud, A., Nomaki, H., Oguri, K., Revsbech, N.P., Kitazato, H., 2009. Nitrogen cycling in a deep ocean margin in particular Nathalie De Hauwere, for generating the bathymetric sediment (Sagami Bay, Japan). Limnology and Oceanography 54 (3), 723-734. map. We are grateful to Guillermo Mendoza for his logistic assis Gooday, A.J., Bernhard, J.M., Levin, L.A., Suhr, S., 2000. Foraminifera in the Arabian tance in providing the samples and accompanying information, Sea oxygen minimum zone and other oxygen deficient settings: taxonomic composition, diversity and relation to metazoan faunas. Deep-Sea Research Part Niels Viaene for nematode measurements, Jennifer Gonzalez for II 47, 25-54. assisting with macrofaunal isotope sample preparation, Robert Gooday, A.J., Levin, L.A, Aranda da Silva, A., Bett, B.J., Cowie, G.L., Dissard, D., Gage, Michener and D. Harris for analyzing these samples, Annick Van J.D., Hughes, D.J., Jeffreys, R., Lamont, P.A., Larkin, K.E., Murty, S.J., Schumacher, 5., Whitcraft, C., Woulds, C., 2009. Faunal responses to oxygen gradients on the Kenhove assistance with the preparation of glass slides, Guy De Pakistan margin: a comparison of foraminiferans, macrofauna and megafauna. Smet for help with the sorting of meiofauna taxa. Sample collection Deep-Sea Research Part II 56, 488-502. at HR was supported by NSF Grants OCE 04-35217 and OCE OS- Gooday, A.J., Bett, B.J., Escobar, E., Ingole, B., Levin, L.A., Neira, C., Raman, A.V., Sellanes, J., 2010. Habitat heterogeneity and its influence on benthic 26254 from the US National Science Foundation and by UAF- biodiversity in oxygen minimum zones. Marine Ecology 31,125-147. 050141 from the West Coast National Undersea Research Center, Heip, C.H.R., Vincx, M., Vranken, G., 1985. The ecology of marine nematodes. and NA17RJ1231 from the US NOAA Office of Ocean Exploration. Oceanography and Marine Biology 23, 399-489. Heip, C.H.R., Herman, P.M.J., Soetaert, K., 1988. Data processing, evaluation, and The research leading to these results has received funding from analysis. In: Higgins, R.P., Thiel, H. (Eds.), Introduction to the Study of the Flanders Fund for Scientific Research (FWO, project number Meiofauna. Smithsonsian Institution Press, Washington, pp. 197-232. 3G0346) and Special Research Fund (BOF, the relation between Helly, J., Levin, L.A., 2004. Global distribution of naturally occurring m arine hypoxia function and biodiversity of nematoda in the deep-sea [FUNDEEP], on continental margins. Deep-Sea Research 51,1159-1168. Hicks, G.F.R., Coull, B.C., 1983. The ecology of marine meiobenthic harpacticoid project number 01J14909). We also thank the two anonymous ref copepods. Oceanography and Marine Biology 21, 67-175. erees for their detailed and constructive feedback. Jensen, P., 1986. Nematode fauna in the sulphide-rich brine seep and adjacent bottoms of the East Flower Garden, NW Gulf of Mexico. IV. Ecological aspects. Marine Biology 92, 489-503. Appendix A. Supplementary material Jensen, P., 1987. Differences in micro habitat, abundance, biomass and body size between oxybiotic and thiobiotic free-living marine nematodes. Oecologia 71, 564-567. Supplementary data associated with this article can be found, in Jeurissen, B., Vermaercke, H., 1990. De biotische leidingnet-index: een methode the online version, at doi:10.1016/j.pocean.2011.10.003. voor de kwaliteitsbeoordeling van leidingwater. Water 55, 264-271. Juhl, A., Taghon, G., 1993. Biology of an active methane seep on the Oregon continental shelf. Marine Ecology Progress Series 102, 287-294. References Kennedy, M., Pevear, D., Hill, R., 2002. Mineral surface control of organic carbon in black shale. Science 295, 657-660. Knittel, K., Boetius, A., Lemke, A., Eilers, H., Lochte, K., Pfannkuche, O., Linke, P., Anderson, M.J., Robinson, J., 2003. Generalized discriminant analysis based on Amann, R., 2003. Activity, distribution, and diversity of sulphate reducers and distances. Australian & New Zealand Journal of Statistics 45, 301-318. other bacteria in sediments above gas hydrate (Cascadia margin, Oregon). Anderson, M.J., Gorley, R.N., Clarke, K.R., 2008. PERMANOVA+ for PRIMER: Guide to Geomicrobiology Journal 20, 269-294. Software and Statistical Methods. PRIMER-E, Plymouth. 92 IC Guilini et al./Progress in Oceanography 96 (2012) 77-92 Kuim, L.D., Suess, E., Moore, J.C., Carson, B., Lewis, B.T., Ritger, S.D., Kadlco, D.C., Sellanes, J., Neira, C., Quiroga, E., Teixido, N., 2010. Diversity patterns along and Thornburg, T.M., Embley, R.W., Rugh, W.D., M assoth, G.J., Langseth, M.G., across the Chilean margin: a continental slope encompassing oxygen gradients Cochrane, G.R., Scamman, RL., 1986. Oregon subduction zone: venting, fauna, and methane seep benthic habitat. Marine Ecology 31,111-124. and carbonates. Science 231, 561-566. Serrano, 0., Serrano, L., Matea, M.A., 2008. Effects of simple pre-treatment on the Leduc, D., Probert, P.K., Berkenbusch, K., Nodder, S.D., Pilditch, C.A., 2010. 813C and ô180 values of living benthic foraminifera. Chemical Geology 257,218- Abundance of small individuals influences the effectiveness of processing 220. techniques for deep-sea nematodes. Deep-Sea Research Part I 57,1363-1371. Shirayama, Y., Ohta, S., 1990. Meiofauna in a cold seep community off Hatsushima, Levin, LA., 2003. Oxygen minimum zone benthos: adaptation and community Central Japan. Journal of the Oceanographical Society of Japan 46,118-124. response to hypoxia. Oceanography and Marine Biology 41,1-45. Sibuet, M., Olu, K., 1998. Biogeography, biodiversity and fluid dependence of deep- Levin, LA., 2005. Ecology of cold seep sedim ents: interactions of fauna w ith flow, sea cold-seep communities at active and passive margins. Deep-Sea Research chemistry and microbes. Oceanography and Marine Biology 43,1-46. Part II 45, 517-567. Levin, LA., Gage, J.D., 1998. Relationships betw een oxygen, organic m atter and the Smith, C.R., Levin, LA., Hoover, D.J., McMurtry, G., Gage, J.D., 2000. Variations in diversity of bathyal macrofauna. Deep-Sea Research Part II 45,129-163. bioturbation across the oxygen minimum zone in the northwest Arabian Sea. Levin, LA., Mendoza, G.F., 2007. Community structure and nutrition of deep Deep-Sea Research Part II 47, 227-257. methane-seep macrobenthos from the North Pacific (Aleutian) Margin and the Soetaert, K., Muthumbi, A., Heip, C.H.R., 2002. Size and shape of ocean margin Gulf of Mexico (Florida Escarpment). Marine Ecology 28,131-151. nematodes: morphological diversity and depth-related patterns. Marine Levin, L.A., Michener, R.H., 2002. Isotopic evidence for chemosynthesis-based Ecology Progress Series 242,179-193. nutrition of macrobenthos: the lightness of being at Pacific methane seeps. Sommer, S., Pfannkuche, 0., Rickeri, D., Kahler, A., 2002. Ecological implications of Limnology and Oceanography 47,1336-1345. surficial marine gas hydrates for the associated small-sized benthic biota at the Levin, LA., Huggett, C.L., Wishner, K.F., 1991. Control of deep-sea benthic Hydrate Ridge (Cascadia Convergent Margin, NE Pacific). Marine Ecology community structure by oxygen and organic-matter gradients in the eastern Progress Series 243, 25-38. Pacific Ocean. Journal of Marine Research 49, 763-800. Sommer, S., Gutzmann, E., Ahlrichs, W., Pfannkuche, 0., 2003. Rotifers colonising Levin, L.A., Gage, J.D., M artin, C., Lamont, P.A., 2000. M acrobenthic com m unity sediments with shallow gas hydrates. Naturwissenschaften 90, 273-276. structure within and beneath the oxygen minimum zone, NW Arabian Se. Deep- Sommer, S., Gutzmann, E., Pfannkuche, 0., 2007. Sediments hosting gas hydrates: Sea Research Part II 47,189-226. oases of metazoan meiofauna. Marine Ecology Progress Series 337, 27-37. Levin, LA., Etter, R.J., Rex, M.A, Gooday, A.J., Smith, C.R., Pineda, J., Stuart, C.T., Sposito, G., Skipper, N., Sutton, R, Park, S., Soper, A., Greathouse, J., 1999. Surface Hessler, RR., Pawson, D., 2001. Environmental influences on regional deep-sea geochemistry of the clay minerals. Proceedings of the National Academy of species diversity. Annual Review of Ecology and Systematics 132, 51-93. Sciences of the United States of America 96, 3358-3364. Levin, L.A., Ziebis, W., Mendoza, G.F., Growney, V.A., Tryon, M.D., Brown, K.M., Mahn, Suess, E., Whiticar, M.J., 1989. Methane-derived C02 in pore water fluids expelled C., Gieskes, J.M., Rathburn, A.E., 2003. Spatial heterogeneity of macrofauna at from the Oregon subduction zone. Palaeogeography Palaeoclimatology northern California methane seeps: influence of sulfide concentration and fluid Palaeoecology 71,119-136. flow. Marine Ecology Progress Series 265,123-139. Thurber, A.R., Kroger, K., Neira, C., Wiklund, H., Levin, L.A., 2010. Stable isotope Levin, LA., Ekau, W., Gooday, A.J., Jorissen, R, Middelburg, J.J., Naqvi, S.W.A., Neira, signatures and methane use by New Zealand cold seep benthos. Marine Geology C., Rabalais, N.N., Zhang, J., 2009. Effects o f natural and hum an-induced hypoxia 272, 260-269. on coastal benthos. Biogeosciences 6, 2063-2098. Tita, G., Desrosiers, G., Vincx, M., Nozáis, C., 2000. Predation and sedim ent Levin, LA., Mendoza, G.F., Gonzalez, J.P., Thurber, A.R, Cordes, E.E., 2010. Diversity disturbance effects of the intertidal polychaetes Nereis virens (Sars) on of bathyal macrofauna on the northeastern Pacific margin: the influence of associated meiofaunal assemblages. Journal of Experimental Marine Biology methane seeps and oxygen minimum zones. Marine Ecology 31, 94-110. and Ecology 243, 261-282. MacAvoy, S.E., Carney, R.S., Fisher, C.R., Macko, S.A., 2002. Use of chemosynthetic Treude, T., Boetius, A., Knittel, K., Wallmann, K., Jorgensen, B.B., 2003. Anaerobic biomass by large, mobile, benthic predators in the Gulf of Mexico. Marine oxidation of methane above gas hydrates at Hydrate Ridge, NE Pacific Ocean. Ecology Progress Series 225, 65-78. Marine Ecology Progress Series 264,1-14. MacDonald, I.R, Boland, G.S., Baker, J.S., Brooks, J.M., Kennicutt, M.C., Bidigare, Valentine, D.L, Kästner, M., W ardlaw, G.D., Wang, X., Purdy, A., Bartlett, D.H., 2005. R.R., 1989. Gulf of Mexico hydrocarbon seep communities. II. Spatial Biogeochemical investigations of marine methane seeps, Hydrate Ridge, distribution of seep organisms and hydrocarbons at Bush Hill. Marine Oregon. Journal of Geophysical Research 110 (G2), G02005. doi:10.1029/ Biology 101, 235-247. 2005JG000025. Moodley, L, van der Zwaan, G.L, Herman, P.M.J., Van Breugel, P., 1997. Differential Van Gaever, S., Galéron, J., Sibuet, M., Vanreusel, A., 2009a. Deep-sea habitat response of benthic meiofauna to anoxia with special reference to foraminifera heterogeneity influence on meiofaunal communities in the Gulf of Guinea. (Protista: Sarcodina). Marine Ecology Progress Series 158, 151-163. Deep-Sea Research Part II 56, 2259-2269. Murrell, M.C., Fleeger, J.W., 1989. Meiofauna abundance on the Gulf of Mexico Van Gaever, S., Moodley, L., Pasotti, F., Houtekamer, M., Middelburg, J.J., Danovaro, continental shelf affected by hypoxia. Continental Shelf Research 9, 1049- R., Vanreusel, A., 2009b. Trophic specialisation of metazoan meiofauna at the 1062. Hákon Mosby Mud Volcano: fatty acid biomarker isotope evidence. Marine Neira, C., Sellanes, J., Levin, L.A, Arntz, W.E., 2001. Meiofaunal distributions on the Biology 156,1289-1296. Peru margin: relationship to oxygen and organic matter availability. Deep-Sea Van Gaever, S., Olu, K., Derycke, S., Vanreusel, A., 2009c. Metazoan meiofaunal Research Part I 48, 2453-2472. communities at cold seeps along the Norwegian margin: influence of habitat Pinto, T.K., Austen, M.C., Bemvenuti, C.E., 2006. Effects of macroinfauna sediment heterogeneity and evidence for connection with shallow-water habitats. Deep- disturbance on nematode vertical distribution. Journal of the Marine Biological Sea Research Part I 56, 772-785. Association of the United Kingdom 86, 227-233. Vanreusel, A., De Groote, A, Gollner, S., Bright, M., 2010a. Ecology and biogeography Reise, K., 1981. High abundance of small zoobenthos around biogenic structures in of free-living nematodes associated with chemosynthetic environments in the tidal sediments of the Wadden Sea. Helgoländer Meeresuntersuchungen 34, deep sea: a review. PLoS ONE 5, e l2449. doi:10.1371/journal.pone.0012449. 413-425. Vanreusel, A., Fonseca, G., Danovaro, R., Cristina da Silva, M., Esteves, A.M., Ferrero, Rex, M.A, 1983. Geographic patterns of species diversity in deep-sea benthos. In: T., Gad, G., Galtsova, V., Gambi, M.C., da Fonsêca-Genevois, V., Ingels, J., Ingole, Rowe, G.T. (Ed.), The Sea, vol. 8. John W iley and Sons, Inc New York, pp. 4 53- B., Lampadariou, N., Merckx, B., Miljutin, D., Miljutina, M.A, Muthumbi, A., 472. Netto, S.A., Portnova, D., Radziejewska, T., Raes, M., Tchesunov, A.V., Rex, M.A, Etter, R.J., Morris, J.S., Crouse, J., McClain, C.R., Johnson, N.A., Stuart, C.T., Vanaverbeke, J., Van Gaever, S., Venekey, V., Bezerra, T.N., Flint, H.C., Copley, Deming, J.W., Thies, R., Avery, R, 2006. Global bathymetric patterns of standing J., Pape, E., Zeppilli, D., M artinez Arbizu, P., Galerón, J., 2010b. The contribution stock and body size in the deep-sea benthos. Marine Ecology Progress Series of deep-sea macrohabitat heterogeneity to global nematode diversity. Marine 3 1 7 ,1 -8 . Ecology 31, 6-20. Robinson, J.J., Cavanaugh, C.M., 1995. Expression of form I and from II Rubisco in Veit-Köhler, G., Gerdes, D., Quiroga, E., Hebbeln, D., Sellanes, J., 2009. M etazoan chemoautotrophic symbiosis: Implications for the interpretation of stable meiofauna within the oxygen-minimum zone off Chile: results of the 2001- isotope values. Limnology and Oceanography 40,1496-1502. PUCK expedition. Deep-Sea Research Part II 56,1105-1111. Rogers, A.D., 2000. The role of the oceanic oxygen m inim a in generating biodiversity Whiticar, M.J., Faber, E., Schoell, M., 1986. Biogenic methane formation in marine in the deep sea. Deep-Sea Research Part II 47,119-148. and freshwater environments. C02 reduction vs. acetate fermentation-isotope Sahling, H., Rickeri, D., Lee, RW., Linke, P., Suess, E., 2002. M acrofaunal com m unity evidence. Geochimica Cosmochimica Acta 50, 693-709. structure and sulfide flux at gas hydrate deposits from the Cascadia convergent Wieser, W., 1953. Die Beziehung zwischen Mundhöhlengestalt, Ernährungsweise margin, NE Pacific. Marine Ecology Progress Series 231,121-138. und Vorkommen bei freilebenden marinen Nematoden. Eine ökologische- Schellart, J.A., 1988. Benefits of covered slow sand filtration. In: Graham, N.J.D. (Ed.), morphologische Studie. Arkiv für Zoologi 2, 439-483. Slow Sand Filtration: Recent Developments in Water Treatment Technology. Wishner, K., Levin, L.A., Gowing, M., Mullineaux, L., 1990. Involvement of the oxygen Ellis Horwood, Chichester, pp. 253-264. minimum in benthic zonation on a deep seamount. Nature 346, 57-59. Schiemer, F., Novak, R, Ott, J., 1990. Metabolic studies on thiobiotic free-living W ishner, K.F., Ashjian, J., Gelfman, C., Gowing, M.M., Kann, L., Levin, L.A, nematodes and their symbiotic microorganisms. Marine Biology 106,129-137. Mullineaux, L., Saltzman, J., 1995. Pelagic and benthic ecology of the lower Seinhorst, J.W., 1959. A rapid method for the transfer of nematodes from fixative to interface of the Eastern Tropical Pacific oxygen minimum zone. Deep-Sea anhydrous glycerine. Nematologica 4, 67-69. Research 42, 93-115.