Old Dominion University
ODU Digital Commons
"'):29> :(2/)'9/548 )+'4'79.93586.+7/)")/+4)+8
Elevated Trace Metal Content of Prokaryotic
Communities Associated with Marine Oxygen
De)cient Zones
Daniel C. Ohnemus
Sara Rauschenberg
Gregory A. Cu*er
Old Dominion University-):B+75*:+*:
Jessica N. Fitzsimmons
Robert M. Sherrell
See next page for additional authors
5225<9./8'4*'**/9/54'2<5718'9 .B68*/-/9'2)5335485*:+*:5+'8&,')&6:(8
'795,9.+ '7/4+/525->533548'4*9.+ )+'45-7'6.>533548
@/879/)2+/8(75:-.995>5:,57,7++'4*56+4'))+88(>9.+)+'4'79.93586.+7/)")/+4)+8'9$/-/9'25335489.'8(++4'))+69+*,57
/4)2:8/54/4"'):29> :(2/)'9/548(>'4':9.57/?+*'*3/4/897'9575,$/-/9'253354857357+/4,573'9/5462+'8+)549')9
*/-/9'2)5335485*:+*:
!+658/957>/9'9/54
.4+3:8'4/+2!':8).+4(+7-"'7':B+77+-57>/9?8/33548+88/)'".+77+22!5(+79'4*#</4/4-+40'3/4
"2+;'9+*#7')++9'2549+495, 751'7>59/)533:4/9/+8885)/'9+*</9.'7/4+=>-+4+A)/+49%54+8 OEAS
Faculty Publications
.B68*/-/9'2)5335485*:+*:5+'8&,')&6:(8
7/-/4'2 :(2/)'9/54/9'9/54
.4+3:8!':8).+4(+7-":B+7/9?8/33548".+77+22!#</4/4-"2+;'9+*97')+3+9'2
)549+495,6751'7>59/))533:4/9/+8'885)/'9+*</9.3'7/4+5=>-+4*+A)/+49?54+8 Limnology and Oceanography 62
*5/245
Elevated trace metal content of prokaryotic communities associated
with marine oxygen deficient zones
Daniel C. Ohnemus,
1
* Sara Rauschenberg,
1
Gregory A. Cutter,
2
Jessica N. Fitzsimmons,
3,4
Robert M. Sherrell,
3,5
Benjamin S. Twining
1
1
Bigelow Laboratory for Ocean Sciences, East Boothbay, Maine
2
Department of Ocean, Earth, and Atmospheric Sciences, Old Dominion University, Norfolk, Virginia
3
Department of Marine and Coastal Sciences, Rutgers University, New Brunswick, New Jersey
4
Now at Department of Oceanography, Texas A&M University, College Station, Texas
5
Department of Earth and Planetary Sciences, Rutgers University, Piscataway, New Jersey
Abstract
Little is known about the trace metal content of marine prokaryotes, in part due to their co-occurrence
with more abundant particulate phases in the upper ocean, such as phytoplankton and biogenic detritus,
lithogenic minerals, and authigenic Mn and Fe oxyhydroxides. We attempt to isolate these biomass signals
in particulate data from the US GEOTRACES Eastern Pacific Zonal Transect (cruise GP16) in the Eastern Trop-
ical South Pacific (ETSP), which exhibited consistent maxima in P and other bioactive trace metals, and min-
ima in particulate Mn, in the oxygen deficient zones (ODZs) of 13 stations. Nitrite maxima and nitrate
deficits indicated the presence of denitrifying prokaryotic biomass within ETSP ODZs, and deep secondary
fluorescence maxima at the upper ODZ boundaries of 10 stations also suggested the presence of low-light,
autotrophic communities. ODZs were observed as far west as 998W, more than 2300 km from the South
American coast, where eolian lithogenic and lateral/resuspended sedimentary inputs were negligible, present-
ing a unique opportunity to examine prokaryotic metal stoichiometries. ODZ particulate P maxima can rival
gyre mixed layer biomass concentrations, are highly sensitive to oxygen, and are in excess of amounts scav-
engable by local Fe oxyhydroxides and acid–volatile sulfides. Even after correction for lithogenic and ferrugi-
nous–scavenged metals, ODZ P-maxima are often enriched in Cd, Co, Cu, Ni, V, and Zn, exhibiting
particulate trace metal ratios to P that exceed mixed layer biomass ratios by factors of 2–9. ODZ prokaryotic
communities may be largely hidden, TM–rich pools involved in the marine cycles of these bioactive trace
metals.
Oxygen deficient zones (ODZs) of the Arabian Sea and the
Eastern Tropical North and South Pacific (ETNP, ETSP) are
sites of diverse microbial productivity that mediate redox
cycling of N, S, and C in the virtual absence of oxygen
(Ulloa et al. 2012). These zones consist of functionally
anoxic cores and associated oxic–suboxic transition zones at
the bottom of the upper oxycline, hosting a complex cas-
cade of aerobic and anaerobic microbially mediated biogeo-
chemical processes (Tiano et al. 2014). Significant attention
has been paid to ODZs in recent decades because of their
importance to the global cycles of C and N, their production
of the greenhouse gas N
2
O, and their ongoing expansion
and shoaling in response to climate change (Whitney et al.
2007; Stramma et al. 2010, 2012; Wright et al. 2012).
Low–oxygen marine environments have long been known
to contain particle maxima that were originally termed
“intermediate nepheloid layers” (Pak et al. 1980; Spinrad
et al. 1989), and several recent studies have highlighted the
likely biotic nature of these maxima within the ETSP ODZ.
Whitmire et al. (2009) used drifter–collected transects of the
Peruvian upwelling region to demonstrate consistent particle
maxima in association with the upper edge of the ODZ and
also noted a low–light, secondary deep chlorophyll maxi-
mum at stations where the ODZ is shallow enough to
impinge on the euphotic zone. Using ultra–low detection
limit oxygen sensors, Tiano et al. (2014) demonstrated that
ODZ–associated secondary fluorescence maxima maintain
oxygen concentrations in the low nanomolar range, suggest-
ing that the autotrophic upper ODZ layer may diurnally
cycle between anoxic and vanishingly oxic conditions, indi-
cating a tight coupling between anaerobic and aerobic
*Correspondence: dan@bigelow.org
3
LIMNOLOGY
and
OCEANOGRAPHY
Limnol. Oceanogr. 62, 2017, 3–25
V
C
2016 Association for the Sciences of Limnology and Oceanography
doi: 10.1002/lno.10363
processes. Flow cytometry and genetic phylogeny conducted
at ODZ sites have identified populations (10
3
to 10
4
cells
mL
21
) of low-light–adapted cyanobacteria in this fluorescent
zone, primarily Prochlorococcus (Goericke et al. 2000; Lavin
et al. 2010). Acoustic backscatter of larger, oxycline–associ-
ated biota (zooplankton and larger eukaryotes) can be a reli-
able means to observe the base of the upper oxycline
(Bertrand et al. 2010) though vertically migrating zooplank-
ton, which feed on settling particulate organic material
(POM) (Wishner et al. 1995; Williams et al. 2014), are typi-
cally excluded from the ODZ itself by the lack of oxygen
(Wishner et al. 2013). Increased cell counts and particle
loads in the anoxic ODZ core, by comparison, are correlated
with the intensity of its denitrifying activity (Naqvi et al.
1993). Taken together, the upper ODZ appears to be a maxi-
mum for autotrophic cyanobacteria and annamox activity
(Gal
an et al. 2009), while heterotrophic denitrification domi-
nates within the ODZ core (Ulloa et al. 2012).
Regardless of the absolute identities of the underlying
communities, ODZs are sites of intense and uniquely pro-
karyotic productivity in the near–micron size fraction (Lavin
et al. 2010; Close et al. 2014). Furthermore, the absence of
an underlying sulfidic zone, as is consistently present
beneath the suboxic zones of the Black Sea and Cariaco
Basin, precludes the intense and persistent diffusive metal
fluxes associated with euxinic environments (Lewis and
Landing 1991, 1992; Ho et al. 2004; Yi
giterhan et al. 2011).
This makes the ETSP ODZ, especially offshore away from
continental shelf influences, a good environment for exam-
ining the elemental composition of prokaryotic communities
in the relative absence of lithogenic particles, eukaryotic
phytoplankton, and detrital mixed layer biomass. These
interfering phases typically mask prokaryotic biomass signa-
tures in the surface mixed layer and near–shore environ-
ments. Particulate interferences unique to the suboxic water
column, such as authigenic Fe and Mn oxyhydroxides that
form and cycle there (Senn et al. 2015), do require special
consideration, however.
Examinations of bioactive particulate trace metals (pTMs)
in prokaryotes have typically focused on trace metal toxicity,
e.g., in human–influenced coastal or sedimentary environ-
ments and wastewater bioreactors. While extremely well
studied in regards to their macronutrient cycling, ODZ–re-
presentative cultured organisms such as denitrifiers, ammonia–
oxidizing archaea and bacteria (AOAs and AOBs), have not
to our knowledge been rigorously examined in regards to
their trace metal content. Vogel and Fisher (2010) investi-
gated uptake and utilization rates of Fe, Mn, Cd, and Zn,
among other elements, in five bacterioplankton species in
surface seawater cultures and found that bacterial accumula-
tion of Fe may be significant in the surface ocean. Other
pTM–prokaryote associations have been reported for Fe
(Tortell et al. 1996; Sunda and Huntsman 1997; Schmidt and
Hutchins 1999), Cu, and Pb (Dixon et al. 2006; Rossi and
Jamet 2008), Cd and Zn (Sunda and Huntsman 1995; Binet
et al. 2003; Iyer et al. 2005), Mn (Francis et al. 2001), and V
and Ni (Shirdam et al. 2006), though only a few examined
prokaryotic strains relevant to the open ocean or ODZs.
With prokaryotic biomass accounting for potentially half of
oceanic POM (Cho and Azam 1988; Fuhrman et al. 1989)
and despite these organisms’ crucial global biogeochemical
roles, there remains a surprising dearth of information on
prokaryotic metal utilization in natural marine
environments.
Redox–active trace metals, many of which are ultimately
delivered to sediments via settling organic matter, also have
importance as sedimentary paleotracers. Bioactive pTMs
have been used to record the past extent of ODZs via sedi-
mentary deposition of Ba, Cu, and Cd (Tribovillard et al.
2006). Sedimentary redox state, which can be influenced by
ODZ extent due to its impingement along the shelf, is
thought to control sedimentary retention of the paleoredox
tracers Mo, Re, U, and V (Nameroff et al. 2004). Tribovillard
et al. (2006) indicated that Ni and Cu may be paleoproxies
for periods of high organic carbon delivery to sediments, as
these pTMs can be preserved within authigenic sedimentary
sulfides. In an extensive sediment core survey beneath the
ETSP ODZ, B
€
oning et al. (2004) indicated a biotic source,
presumed to be the overlying euphotic zone, for Cd, Cu, Ni,
Zn, and other pTMs that have delivery rates dominated by
the downward vertical flux of organic matter. Metal stoichio-
metries and bulk TM associations for local biomass, includ-
ing eukaryotic phytoplankton and ODZ prokaryotes, are thus
relevant to interpretation of sedimentary paleoproxies, in
addition to understanding the communities themselves.
Here we report pTM enrichments found in bulk particles
associated with heterotrophic and autotrophic prokaryotic
communities of the ETSP ODZ, sampled during the US GEO-
TRACES Eastern Pacific Zonal Transect (EPZT) in Oct–Dec
2013 (Fig. 1). The pTM composition of these communities is
often significantly different, and pTM–enriched, compared
to mixed layer communities. Prokaryotic biomass associated
with the ODZ may influence the marine cycles and sedimen-
tary imprint of trace metals in the ETSP and other oxygen–
deficient regions.
Materials and methods
Sample collection, analyses, and data access
All particulate samples were collected on the US GEOTRA-
CES EPZT cruise GP16 from Lima, Peru to Tahiti in Oct–Dec
2013 (Fig. 1). Seawater was collected with the non–contami-
nating GEOTRACES rosette and filtered onto acid–cleaned,
25-mm diameter, 0.45 mm pore–size Supor
TM
(polyethersul-
fone) filters in the GEOTRACES sampling van (Cutter et al.
2010; Planquette and Sherrell 2012). Typical filtration vol-
umes through the analyzed filter halves were 4.4 L 6 1.4 L
seawater (median 6 one SD). Digestion and analytical
Ohnemus et al. Elevated trace metals in ODZ prokaryotes
4
procedures were as previously reported by our laboratory
(Twining et al. 2015). Briefly, in a land–based analytical
clean lab, filters were halved, digested in a mixture of 4 mol
L
21
each HCl/HNO
3
/HF optima–grade acids at 1108Cin
sealed Teflon vials, dried, and brought up in 0.32 mol L
21
HNO
3
for analysis via ICP–MS (Thermo Element2). External
multi–elemental standard curves and internal spikes of In
(for ICP–MS instrumental drift) and Cs (for sample recovery
purposes) were used to standardize and monitor elemental
recoveries. Recoveries of the certified reference materials
(CRMs) PACS-2 (marine sediment) and BCR-414 (plankton)
and multi–elemental standard addition spikes into random
samples were also used to monitor reported values. Recov-
eries of CRMs averaged 95% 6 8% for PACS-2 and 102% 6
7% for BCR-414 across all elements. Full CRM recoveries,
blank information, and methodological information includ-
ing the details of hydrothermal sample analyses and labora-
tory intercalibration within the particle dataset, are available
in the dataset metadata available via the NSF data archive at
the Biological & Chemical Oceanography Data Management
Office (BCO-DMO), dataset #639857.
Samples analyzed for acid–volatile sulfides (AVS) were col-
lected using in situ pumps (McLane) suspended from a trace
metal wire (Hytrel–jacketed Vectran), at the same stations
and target depths as the GEOTRACES carousel, further
described in Ohnemus and Lam (2015). Pumps filtered for
4 h, with the relevant flow path passing through a 51-mm
polyester pre–filter, then through paired 0.8-mm pore–size,
142-mm Supor
TM
polyethersulfone filters. A small portion of
the top Supor filter, representing 2% of the filter area (0.8–
51-mm size–fraction material), was sub–sectioned into poly-
ethylene vials and frozen at 2858C until shipboard AVS anal-
ysis, described below. Filter blanks were held within
identical filter housings and submerged on one of the deeper
pumps on each cast but were not actively filtered through.
Blanks were handled identically to samples during process-
ing, which was conducted in a HEPA–filtered environment
using trace metal clean procedures. The slight pore–size mis-
match between pump–collected AVS samples and bottle–
filtered pTMs (0.8 mm vs. 0.45 mm cutoff, respectively) is not
expected to influence our results significantly, as used here.
AVS data are available through the NSF data archive at BCO-
DMO, dataset #646143.
Acid–volatile sulfide determination
Determinations of acid–volatile sulfide (AVS) were made
onboard ship and used the method of Cutter and Oatts (1987)
that includes acidification, gas stripping, and cryogenic trap-
ping/preconcentration of the generated hydrogen sulfide. In
this case, the filter slice was placed in a glass stripping vessel,
10 mL of deionized water added, and the stripper purged with
He. The cryogenic trap was then immersed in liquid nitrogen,
10 mL of 1 mol L
21
HCl added through a septum in the strip-
ping vessel and the evolved and trapped hydrogen sulfide was
then quantified using the gas chromatography/flame photo-
metric detection method of Radford-Knoery and Cutter
(1993). The system was calibrated with hydrogen sulfide from
a permeation tube (calibrated diffusion rate) and assuming a
filtration volume of at least 5 liter (median sample volume:
7.9 liter), the detection limit was 0.2 pmol/L. It should be
noted that AVS determined by this method includes monosul-
fides of Fe, Zn, and Ni, but not those with Cu or Hg, or mixed
oxidation state sulfides such as greigite or pyrite (Cutter and
Oatts 1987; Radford-Knoery and Cutter 1993).
Profiling CTD data
Reported values for transmissometer voltage (WET Labs,
25-cm path length, 660 nm), fluorescence (Seapoint), and
oxygen (Seabird SBE 43) were collected via the GEOTRACES
carousel Seacat 91 (Seabird) CTD after 1 m binning of
higher–frequency profiling data; upcast traces are used
herein, except when otherwise noted. Station occupations
were often longer than 24 h and consisted of multiple casts
on three sampling systems (GEOTRACES carousel, McLane
pumps, and Oceanographic Data Facility [ODF] rosette) each
140˚W 120˚W 100˚W 80˚W
15˚S
10˚S
5˚S
EQ
6500 m
6000 m
5500 m
5000 m
4500 m
4000 m
3500 m
3000 m
2500 m
2000 m
1500 m
1250 m
1000 m
750 m
500 m
250 m
100 m
50 m
Peru
6
8
10
11 9 7
12
13
14
15
18
20
21
26
30
36
79˚W 78˚W 77˚W 76˚W
13˚S
12˚S
Lima
2345
1
(5.5)
Fig. 1. Key stations from R/V Thompson cruise TT303 (US GEOTRACES Eastern Pacific Zonal Transect GP16; Oct–Dec 2013), including all stations
directly referenced in the text. Inset: “shelf” stations near the Peruvian coast, including Sta. 1 which is plotted as Sta. 5.5 in all other figures for longi-
tudinal consistency. Sta. 6–13 are “ODZ” stations; Sta. 14–36 “gyre” stations. Shading indicates bathymetry.
Ohnemus et al. Elevated trace metals in ODZ prokaryotes
5
![Fig. 6. Scatterplots of select bioactive pTMs (Cd, P, Cu, and V; y-axes) vs. particulate Fe (x-axes) after correction for only lithogenic abundances (“_NoLith”) in shelf ODZ samples (gray asterisks) and samples from the East Pacific Rise hydrothermal plume (Sta. 18–26, 2000–2800 m depth, pFe>1 nM; black diamonds). Robust linear regressions (MATLAB, which iteratively re-weights samples to remove the influence of outliers [Holland and Welsch 1987]) and r2 values for the fits are shown (color coded lines, text). Insets: close-up views of the lower ends of the distributions.](/figures/fig-6-scatterplots-of-select-bioactive-ptms-cd-p-cu-and-v-y-2vgn90cl.webp)
