The outer western Crimean shelf of the Black Sea is a natural laboratory to
investigate effects of stable oxic versus varying hypoxic conditions on
seafloor biogeochemical processes and benthic community structure.
Bottom-water oxygen concentrations ranged from normoxic (175
Hypoxia describes a state of aquatic ecosystems in which low oxygen concentrations affect the physiology, composition and abundance of fauna, consequently altering ecosystem functions including biogeochemical processes and sediment–water exchange rates (Middelburg and Levin, 2009). Low faunal bioturbation rates in hypoxic zones limit sediment ventilation (Glud, 2008), decreasing oxygen availability for aerobic respiration. Hence, sediments underlying a low-oxygen water column often show oxygen penetration depths of only a few millimeters (Archer and Devol, 1992; Glud et al., 2003; Rasmussen and Jørgensen, 1992). This increases the contribution of anaerobic microbial metabolism to organic matter remineralization at the expense of aerobic degradation by microbes and fauna, as reported from the Romanian shelf area of the Black Sea (Thamdrup et al., 2000; Weber et al., 2001), the Neuse River estuary (Baird et al., 2004) and the Kattegat (Pearson and Rosenberg, 1992). Consequently, oxygen is channeled into the reoxidation of reduced substances produced during anaerobic degradation of organic matter and lost for direct aerobic respiration. Even temporarily reduced bottom-water oxygen concentrations can repress seafloor oxygen uptake that should become enhanced by algae blooms and temperature increases (Rasmussen and Jørgensen, 1992). However, depending on frequency and duration of oxygen oscillations, oxygen consumption following an anoxic event can also be significantly increased (Abril et al., 2010). Hence, these and other studies have indicated, that not only the degree of oxygenation plays an important role in oxygen uptake, but also the frequency and persistency of the low oxygen conditions can shape faunal activity, biogeochemical processes and the functioning of the ecosystem as a whole (Boesch and Rabalais, 1991; Diaz, 2001; Friedrich et al., 2014).
The outer western Crimean shelf of the Black Sea is a natural laboratory
where long-term effects of different, and locally fluctuating oxygen
concentrations on benthic oxygen consumption and biogeochemical processes
can be investigated, which was the main aim of this study. In the Black Sea,
the depth of the oxic–anoxic interface changes from about 70 to 100 m in open
waters (Friedrich et al., 2014) to depths of > 150 m above the
shelf break (Stanev et al., 2013). This interface is stabilized by a
halocline that separates the upper layer of brackish, oxic water (salinity
< 17) from the saline, anoxic and sulfidic deep waters below
(Tolmazin, 1985). Due to mixing processes by internal waves and eddies, the
location of this interface zone is more dynamic along the margins of the
Black Sea compared to the open sea. In the shelf region, hypoxic waters with
oxygen concentrations < 63
Here we investigated biogeochemical processes on the outer western Crimean shelf to assess how different ranges of oxygen availability, and also of fluctuations in bottom-water oxygen concentrations, influence respiration, organic matter remineralization and the distribution of benthic organisms. The questions addressed are to what extent the variability in oxygen concentration has an effect on (1) the remineralization rates, (2) the proportion of microbial vs. fauna-mediated respiration, (3) the community structure and (4) the share of anaerobic vs. aerobic microbial respiration pathways.
Investigations of bottom-water oxygen concentrations and biogeochemistry of
the underlying seafloor of the outer western Crimean shelf were carried out
over a time period of 2 weeks (20 April–7 May 2010) during
leg MSM 15/1 of R/V
Measurements and samples (including PANGAEA event labels) taken in zones with different oxygen regime. PUC represents JAGO pushcores, MOVE represents the benthic crawler move (in situ microsensor measurements and/or benthic chamber deployment), TVMUC represents the video-guided multicorer and KAMM represents the lander (in situ microsensor measurements and/or benthic chamber deployment).
Bottom-water oxygen concentrations were recorded repeatedly between 95 and
218 m water depth at different spatial and temporal scales with various
sensors, which were all calibrated by Winkler titration (Winkler, 1888). A
total of 26 casts were performed with a CTD rosette equipped with a SBE 43
oxygen sensor (Seabird Electronics, Bellevue, WA, USA). A mooring was
deployed at 135 m water depth 1.5 m above the sediment, equipped with a
Seaguard current meter with CTD and a type 4330 oxygen optode (Aanderaa Data
Instruments, Bergen, Norway) recording at 60 s intervals at a distance
of 1.5 m above the sediment from 30 April to 7 May
2010. A second mooring was deployed for the same time period at 100 m water
depth, with a CTD attached at 1.5 m above the sediment (type SBE 16,
Seabird Electronics) to record density, salinity and temperature. CTD
water-column casts and the mooring at 135 m showed that oxygen
concentrations strongly correlate with density (
Sediment sampling locations (TVMUC represents the video-guided multicorer, PUC represents the JAGO pushcores) and deployment sites of benthic chamber and microprofiler with MOVE and lander (KAMM) along the transect from shallower (101 m) to deeper (207 m) water depth. Inset: working area on the outer western Crimean shelf (red square) in the Black Sea.
To observe organisms, their traces of life, and the resulting
microtopography at the surface of the different seafloor habitats, a laser
scanning device (LS) and the high-resolution camera MEGACAM were used on the
benthic crawler MOVE (MARUM, Bremen). The LS consisted of a linear drive
that moved a downward-facing line laser together with a monochrome digital
camera horizontally along a 700 mm long stretch of the seafloor. The
position of the approx. 200 mm wide laser line was recorded by the camera
from an angle of 45
The downward-facing MEGACAM (Canon EOS T1i with 15 megapixel imager and 20 mm wide-angle lens) was either attached directly to MOVE or added to the horizontal drive of the LS; the latter configuration facilitating imaging of larger sediment stretches by photo-mosaicking. In addition, visual seafloor observations were carried out before pushcore sampling by JAGO. Dive videos were recorded with a type HVR-V1E HDV camcorder (SONY, Tokyo, Japan) mounted in the center of JAGO's large front viewport during 19 dives. During each dive, video still images were captured by the video-grabber from the running camera.
Meiofauna organisms were studied in the upper 5 cm sediment horizons of
two–four cores per station, with each core covering an area of 70.9 cm
Vertical solute distributions were measured in situ at high resolution in
sediment porewaters and the overlying waters with microsensors mounted on
microprofiler units (Boetius and Wenzhöfer, 2009). In particular,
Clark-type O
Profiler units were mounted either on the benthic crawler MOVE (Waldmann and
Bergenthal, 2010) or on a benthic lander (Wenzhöfer and Glud, 2002). The
MOVE vehicle was connected to the ship via a fiber optic cable that allowed
continuous access to video and sensor data. The maneuverability of the
vehicle allowed targeting spots of interest on the seafloor in the centimeter range.
The profiler units were equipped with three–four O
From the obtained oxygen profiles the diffusive oxygen uptake (DOU) was
calculated based on the gradients in the diffusive boundary layer (DBL)
according to Fick's first law of diffusion:
Total oxygen uptake (TOU) of sediments was measured by in situ benthic
chamber incubations using two platforms: (1) two benthic chambers, each
integrating an area of 0.2
Sediments for geochemical analyses were sampled with a video-guided
multicorer (TVMUC) at four stations between 104 and 207 m (Table 1). Porewater
was extracted from sediment cores within 3 h after retrieval in 1 cm (upper
5 cm) or 2 cm (> 5 cm) intervals with Rhizons (type: CSS,
Rhizosphere Research Products, pore size < 0.2
Porewater constituents were analyzed by the following procedures: dissolved
Mn (II) and Fe (II) were measured with a Perkin Elmer 3110 flame atomic
absorption spectrophotometer (AAS) with a detection limit of 5
Total zero-valent sulfur in sediments was extracted with methanol from
sediment preserved in ZnAc (Zopfi et al., 2004) and analyzed by HPLC.
Concentrations of acid volatile sulfide (AVS; Fe
Sulfate reduction rates were determined with the whole core incubation
method described in Jørgensen (1978). On board, 10
Porosity and solid-phase density were determined by drying a wet sediment
aliquot of known volume at 105
Sediment accumulation rates were determined from excess
Recordings of bottom-water oxygen concentrations (
The “oxic zone” at water depths of 95 to 130 m had oxygen concentrations
of on average 116
Synthesis of oxygen concentrations in bottom water (circles)
measured during the 2 weeks of the cruise (
In the “oxic–hypoxic zone” at water depths between 130 and 142 m, average
bottom-water oxygen concentrations were 94
The “hypoxic–anoxic” zone between 142 and 167 m water depth sediments
showed fluctuating hypoxic conditions between 0 and 63
Below 167 m, the bottom water was permanently anoxic during the time period
of our campaign. Below 180 m sulfide was constantly present in the bottom
water, with concentrations ranging between 5 and 23
Abundance and composition of meiobenthos as retrieved from the
top 5 cm of pooled core samples were compared across the different zones of
oxygen availability indicated in Fig. 2 (Table S2 in the
Supplement). The macrobenthos abundances and taxonomic composition presented
here are not quantitative, nor statistically significant, for the entire
size class due to the limited sample size available; they might represent
mostly small types and juvenile stages (Table S1). These
decreased by more than 1 order of magnitude from the oxic zone (21
Meiobenthos was composed of similar groups and abundances in the oxic and
oxic–hypoxic zone with densities of around 200
Abundance of meiobenthos in the upper 5 cm of the sediment under different oxygen regimes. The middle line in each box depicts the median, while both whiskers and outliers indicate the distribution of remaining data points.
A total of 33 oxygen microprofiles were measured during seven deployments of
the benthic crawler MOVE and the lander at water depths between 104 and 155 m. Oxygen penetration depths and dissolved oxygen uptake rates are
summarized in Table 2. The shape of the profiles and the differences in
oxygen penetration depth as shown in Fig. 5 reflect the spatial
variations of oxygen bottom-water concentrations and oxygen consumption
rates. In the shallowest depth of the oxic zone (104 m), clear signs of bioturbation were
visible from the irregular shape of about 25 % of the profiles,
occasionally increasing the oxygen penetration depth up to approximately 10 mm. Bioturbation activity was in accordance with a significant bioturbated
surface layer and more pronounced roughness elements at the sediment surface
at the shallowest site as compared to deeper waters (see Sect. 3.5). In
contrast, the shape of the oxygen profiles obtained in the oxic–hypoxic and
the hypoxic–anoxic zone showed no signs of bioturbation. Small-scale spatial
heterogeneity was low between parallel sensor measurements and within one
deployment (area of 176 cm
Diffusive oxygen uptake (DOU) rates, total oxygen uptake
(TOU) rates and oxygen penetration depth under different oxygen regimes at
the outer western Crimean shelf. Chamber measurements in the hypoxic–anoxic
zone represent potential rates, scaled to a bottom-water oxygen
concentration of 20
Diffusive oxygen uptake (DOU) ranged within an order of magnitude between
all zones (Table 2). The highest DOU of 8.1 mmol m
Cluster dendrogram of meiofauna abundances for different station depths based on the inverse of Bray–Curtis dissimilarity.
In bottom waters of the hypoxic–anoxic zone, high-resolution measurements of
pH indicated a pH
Examples of high-resolution oxygen profiles under different oxygen regimes. Differences in bottom-water oxygen concentrations (reflected in profile shape and oxygen penetration depth) are clearly visible between sites and deployments.
Total oxygen uptake (TOU) including the faunal respiration, was generally
higher than DOU (Table 2). Individual measurements varied from 20.6 to 3.2 mmol m
Trapping of oxygen-enriched waters in the chambers during deployments
carried out at the hypoxic–anoxic zone led to higher initial oxygen
concentrations in the enclosed water as compared to ambient bottom waters.
Therefore, we could only obtain potential TOU rates at elevated bottom-water
oxygen concentrations of 70
Examples of individual oxygen profiles measured in the sediment (white circles) and modeled with PROFILER (black lines). Volumetric rates are combined in discrete layers (dashed lines) and exhibit different depths and degrees of oxygen consumption rates in different zones and under different bottom-water oxygenation.
Cores from all sites had the typical vertical zonation of modern Black Sea
sediments with a brown/black fluffy layer (oxic and hypoxic zones, Fig. S2d), or dark/grey fluffy layer (anoxic–sulfidic zone), covering beige–grey,
homogenous, fine-grained mud. Substantial differences in the concentration
profiles and volumetric production and consumption rates of dissolved iron,
dissolved manganese, sulfide and ammonium were found in porewaters from
surface sediments sampled from the four different oxygen regimes (Fig. 7).
In the oxic zone, dissolved iron and manganese were present in the porewater with maximal concentrations of 217
Distribution of reduced porewater species and oxidized and reduced solid-phase iron and sulfur species along the depth transect in the upper 30 cm of the sediment (symbols with dotted lines). Solid lines are the model results and dashed lines represent production and consumption rates.
In solid-phase extractions, reactive iron was elevated in the 0-1 cm interval of the oxic zone and iron oxides were present throughout the upper 30 cm of surface sediments (Fig. 7e). In contrast, concentrations of iron-oxides in the upper 10 cm of the oxic–hypoxic zone were clearly reduced and dropped to background concentrations below 10 cm. The same trend was observed in sediments of the hypoxic–anoxic and the anoxic–sulfidic zone (Fig. 7l, s, z). Solid-phase manganese concentration was only clearly elevated in the 0–1 cm interval of the oxic zone (Fig. 7f) and at, or close to, background concentration below 1 cm, as in all other zones (Fig. 7m, t, aa).
Although porewater concentrations of sulfide were below the detection limit in
the oxic to hypoxic–anoxic zones, the presence of reduced solid sulfide
phases (AVS, CRS and S, Fig. 7g, n, u, ab) and measured
sulfate reduction rates indicate that some sulfate reduction took place
below the oxygenated sediment. Sulfate reduction rates, integrated over the
upper 10 cm of the sediment, represent gross sulfide production and compare
well to net sulfide fluxes calculated from the porewater profiles in Table 3.
Altogether, seafloor sulfate reduction rates were increasing nearly
40-fold from < 0.1 mmol m
Diffusive oxygen uptake compared to fluxes of reduced species, calculated from the modeled profiles (Fig. 7) or measured directly (SRR represents sulfate reduction rates). The sum of oxygen equivalents is calculated from the stoichiometry of the oxidation processes (respective formulas are displayed at the lower end of the table), and oxygen available for direct aerobic respiration is calculated by subtracting the potential oxygen demand from the available oxygen flux.
Note: negative numbers denote downward flux, positive numbers upward flux.
Sediment porosity was similar across all sites with 0.9
Rates of benthic oxygen consumption are governed by a variety of factors
including primary production, particle export, quality of organic matter,
bottom-water oxygen concentrations and faunal biomass (Jahnke et al., 1990;
Middelburg and Levin, 2009; Wenzhöfer and Glud, 2002). Here we
investigated the effects of variable hypoxic conditions, with bottom-water
oxygen concentrations ranging from 180 to 0
Oxygen consumption in the sediment is usually directly proportional to the
total carbon oxidation rate, i.e., carbon oxidized by both aerobic and
anaerobic pathways. An imbalance could be the result of denitrification
processes, where the reduced product is N
Oxygen consumption in hypoxic areas of the Black Sea, n.d. depicts values not determined.
To evaluate the contribution of chemical reoxidation to TOU at the outer western Crimean shelf, we fitted measured porewater profiles of dissolved manganese, iron, ammonium and sulfide with 1-D models to quantify upward-directed fluxes (Berg et al., 1998, Table 3, Fig. 7). Taking the stoichiometries of the reaction of oxygen with the reduced species into account, the maximal oxygen demand for the reoxidation of reduced porewater species was less than 8 % (Table 3). This is less than in other studies in eutrophic shelf sediments, where the chemical and microbial reoxidation of reduced compounds, such as sulfide, dominated and the heterotrophic respiration by fauna contributed around 25 % to total oxygen consumption (Glud, 2008; Heip et al., 1995; Jørgensen, 1982; Konovalov et al., 2007; Soetaert et al., 1996).
Comparing total remineralization rates across all zones, including the oxygen demand by anaerobic microbial processes (Table 3), the capacity of the benthic communities to remineralize the incoming particle flux decreased from the oxic zone, to the oxic–hypoxic, hypoxic–anoxic and the anoxic zone. Total remineralization rates were similar in the hypoxic–anoxic and stable anoxic zone, but only the latter, anaerobic processes dominated, most likely due to the persistent absence of oxygen, allowing anaerobic microbial communities to thrive.
Total oxygen uptake (TOU), as measured in situ with benthic chambers,
represents an integrated measure of diffusive microbial respiration, as well
as oxygen uptake by benthic fauna. The diffusive oxygen uptake (DOU), as
calculated from microsensor profiles, represents mainly aerobic respiration
of microorganisms or – although not relevant in our area (see above) –
chemical reoxidation (Glud, 2008). In general, the DOU of the outer western
Crimean shelf sediments was lower than in other shelf zones with
seasonally hypoxic water columns (e.g., Glud et al., 2003), but in the same
range as fluxes reported in other Black Sea studies (Table 4). Average DOU
was similar in the oxic and oxic–hypoxic zone and only clearly reduced when
oxygen concentrations were close to zero (20
Despite a relatively uniform sediment accumulation rate, TOU at the
oxic–hypoxic zone was substantially lower as compared to the oxic zone
despite bottom-water oxygen concentrations which remained mostly above the common
threshold for hypoxia of 63
It has previously been shown that sediment–water exchange rates can be
altered due to changes in fauna composition in response to different
bottom-water oxygenation (Dale et al., 2013; Rossi et al., 2008). Coastal
hypoxic zones often show reduced faunal abundances, biodiversity and loss
of habitat diversity below a threshold of 63
The restriction of bivalves and gastropods to the upper oxic–hypoxic zone is
surprising, as representatives of these groups are known to be able to
maintain their respiration rate at hypoxic oxygen concentrations (Bayne,
1971; Taylor and Brand, 1975). Oxygen concentrations on the outer western
Crimean shelf (Fig. 2) were mostly well above reported oxygen thresholds,
e.g., 50
The overall role of meiobenthos in oxygen consumption is difficult to assess as it can add to both BMU and DOU by bio-irrigating the sediment as well as enhancing diffusional fluxes (Aller and Aller, 1992; Berg et al., 2001; Rysgaard et al., 2000; Wenzhöfer et al., 2002). Altogether, different distribution patterns were found for meiofauna as compared to macrofauna. Meiobenthos abundances were similar in the oxic and oxic–hypoxic zone, and only sharply decreased in the hypoxic–anoxic zone. As shown previously (Levin et al., 2009) nematodes and foraminifera dominate meiofauna in hypoxic zones due to their ability to adapt to low oxygen concentrations. In particular, nematodes are known to tolerate hypoxic, suboxic, anoxic or even sulfidic conditions (Sergeeva et al., 2012; Sergeeva and Zaika, 2013; Steyaert et al., 2007; Van Gaever et al., 2006). Some meiobenthos species are known to occur under hypoxic conditions (Sergeeva and Anikeeva, 2014; Sergeeva et al., 2013). The relatively high abundance of apparently living foraminifera in the hypoxic zone might be related to the ability of some species to respire nitrate under anoxic conditions (Risgaard-Petersen et al., 2006).
Regarding the validation of the traditionally used hypoxia threshold for
impact on fauna (63
This study assesses the effect of different ranges of bottom-water
oxygenation and its local fluctuation on carbon remineralization rates, the
proportion of microbial vs. fauna-mediated respiration, the benthic community
structure and the share of anaerobic vs. aerobic microbial respiration
pathways. We could show that fauna-mediated oxygen uptake and biogeochemical
fluxes can be strongly reduced already at periodically hypoxic conditions
around 63
We thank the Captain and shipboard crew of the RV