Marine Oil Snow Sedimentation & Flocculent Accumulation MOSSFA Events—The Rule or the Exception to the Rule?

Comparative Analysis of the Deepwater Horizon & IXTOC-1 Blowouts

David J. Hollander, M.-L. Machain-Castillo, A. Gracia, H.A. Alexander-Valdés, G.R. Brooks, J. Chanton, J., E. Escobar-Briscon, D.W. Hastings, J. Kostka, R.A. Larson, I.C. Romero, I.C., A.C. Ruiz-Fernández, J. Sánchez-Cabeza, P.T. Schwing, S. Lincoln, T. Oldenburg, K. Freeman, P. Montange

Where did the oil go? Sea Grant-NOAA SymposiumMobile, AL April 14, 2016

WHAT factors control the formation sinking of

oil-associated particles (Marine Oil Snow Sedimentation)?

Microbial mucus snow

Aggregates coagulation of particles

Zooplankton Activity OMA: Oil mineral aggregates

Oil, Dispersant

Riverine Influences

Marine biota Marine Oil

Snow“Dirty Blizzard”

Pyrogenic

Petrogenic Dispersant

Algal Bloom

Microbial loop

Mesozooplanktonfood

Salinity Nutrients

Clays

2

StatisticalanalysisCorrelationcoefficientsaswellasdifferencesbetweentwo

averagevaluesgivenastheirstatisticalmean6standarddeviation

weretestedfortheirsignificanceusingtheStudentst-test.Analysis

ofvariance(one-wayANOVA)wasusedforcomparingaverage

valuesofmorethantwogroupsofdata.IfANOVAwas

significant,posthocpairwisecomparisonsofmeanswere

performedusingtheBonferroni-Holmestestofvariability.All

statisticalanalysiswasperformedinExcelHusingthedataanalysis

toolpackaswellasDaniel’sXLtoolbox(bothopensourceadd-

ins).

Results

FormationofoilaggregatesinrollerbottlesRollertableincubationofuncontaminatedseawatercollected

neartheDeepwaterHorizonoilspillsitewithsurfaceoilsampled

inthesamearea(hereafterreferredtoasSW+oil1andSW+oil2

bottles;seeMaterialandmethodsforrollerbottlesset-up)ledto

rapidformationofaggregates(hereafterreferredtoasoil

aggregates)withinoneday(seeFig.S1forclose-upphotosofoil

aggregates).OilaggregatesinbothSW+oilbottlesclumped

togetherafter7days,formingasingleaggregateupto30mm

indiameter,withvisiblyincorporatedoildroplets.Oilaggregate

formationincontrolSW+oil(seawaterthathadbeenfilteredand

autoclavedbeforetheoilwasadded)wasfirstobservedatday10

aftertheappearanceofgelatinousnetworksofparticulatematter

withincorporatedoildroplets(hereafterreferredtoasoilgels).Oil

gelsappearedtobeverystickysurfacesontowhichoilaggregates

attacheduponcollision(Fig.1).

Incontrasttooil-amendedbottles,aggregateformationinrollerbottleswithseawaternotamendedwithoil(SWbottles)was

delayedandreducedinscale(aggregatesthatweremuchmoretransparentthanoilaggregatesfirstappearedafter3days;Fig.

S1E),andaggregateswerelessabundant(2to3perbottle).Theseaggregatesdidnotchangeinsizeandnumberthroughoutthe21

daysofincubation.NoaggregatesformedinthecontrolSWbottlecontainingfilteredandsterilizedseawater.

WetweightsofoilaggregatesinSW+oil1,SW+oil2,andcontrol

SW+oilbottlesafter21dayswere5.5g,7.1g,and9.8g,respectively.Assumingafinalbottlewatervolumeof900mlat

day21(weight

Freshwater Discharge

> 30 meter of water~ 30 miles off shore

3

In Situ Burning

~ 25 to 55 mi off shore

Area of Dispersant

Low SalinityCoastal/Offshore

Mitigation Strategies of Surfacing Oil Did mitigation strategies of surface oil intensify Marine Oil Snow Sedimentation & increase the “footprint” of sedimentary oil deposition?

1. River discharge releases clays & nutrients and freshwater to offshore

2. Dispersant application decreases oil droplet size and facilitates oil binding with clays, algae and bacteia

3. Algae-bacteria exposed to oil dispersant form biopolymers (stress-response” that flocculates and traps clay , oil & plankton

4. Oil burning produces pyrogenic PAHs and soot particles

Two Possible Mechanisms of Sedimentary Oil Deposition

•1-Toxic Bath-Tub Ring:

Plume impinges on the sediment directly; poisoning the benthic ecosystem with BTEX, PAHs.

•2-Flocculent Blizzard:

Rapid flocculation and sinking of clay, algae, bacteria- oiled aggregates (weathered oil pyrogenic PAHs, dispersant) aggregates; rapid pulsed sediment accumulation of surfacing oil-dispersant

2. Flocculent “Dirty” Blizzard: Oil w/particles: lithogenic, orgs.

Surfacing Oil Slick and Sheen

Jet ReleaseOil-Gas RatioPressure GradientOil Composition

1000-1300m

BOP

1. Toxic Bath-Tub Ring: Plume Impingement

Continental Slope Sediments

Statistical analysisCorrelation coefficients as well as differences between two

average valuesgiven astheir statistical mean 6 standard deviation

were tested for their significance using the Students t-test. Analysis

of variance (one-way ANOVA) was used for comparing average

values of more than two groups of data. If ANOVA was

significant, post hoc pairwise comparisons of means were

performed using the Bonferroni-Holmes test of variability. All

statistical analysis wasperformed in ExcelHusing the data analysis

toolpack as well as Daniel’s XL toolbox (both open source add-

ins).

Results

Formation of oil aggregates in roller bottlesRoller table incubation of uncontaminated seawater collected

near the Deepwater Horizon oil spill site with surface oil sampled

in the same area (hereafter referred to as SW+oil1 and SW+oil2

bottles; see Material and methods for roller bottles set-up) led to

rapid formation of aggregates (hereafter referred to as oil

aggregates) within one day (see Fig. S1 for close-up photos of oil

aggregates). Oil aggregates in both SW+oil bottles clumped

together after 7 days, forming a single aggregate up to 30 mm

in diameter, with visibly incorporated oil droplets. Oil aggregate

formation in control SW+oil (seawater that had been filtered and

autoclaved before the oil was added) was first observed at day 10

after the appearance of gelatinous networks of particulate matter

with incorporated oil droplets (hereafter referred to asoil gels). Oil

gels appeared to be very sticky surfaces onto which oil aggregates

attached upon collision (Fig. 1).

In contrast to oil-amended bottles, aggregate formation in rollerbottles with seawater not amended with oil (SW bottles) was

delayed and reduced in scale (aggregates that were much moretransparent than oil aggregates first appeared after 3 days; Fig.

S1E), and aggregates were lessabundant (2 to 3 per bottle). Theseaggregates did not change in size and number throughout the 21

daysof incubation. No aggregates formed in the control SW bottlecontaining filtered and sterilized seawater.

Wet weightsof oil aggregates in SW+oil1, SW+oil2, and control

SW+oil bottles after 21 days were 5.5 g, 7.1 g, and 9.8 g,respectively. Assuming a final bottle water volume of 900 ml at

day 21 (weight< 924.3 g), oil aggregates occupied approximately0.6% (SW+oil1), 0.8% (SW+oil2), and 1.1% (control SW+oil) of

the total roller bottle volume. SW1 and SW2 aggregates were0.19 g and 0.17 g, respectively, and thus 0.02% of total rollerbottle volume.

Microbial cell abundance in ambient watersThe microbial cell counts documented the impact of oil

amendments on the abundance of prokaryotic cells in surface

seawater during the roller table incubations, compared touncontaminated seawater. Uncontaminated ambient water

(SW2) had 0.56 0.46 106 cells mL2 1 at day 0 (Fig. 2A); thisnumber was lower (p, 0.05) but the same order of magnitude as

the cell abundance of uncontaminated water fixed shortly aftersampling (0.86 0.26 106 mL2 1), indicating that storage time and

conditions from the time of sampling until the beginning of theexperiment had little influence on cell numbers in uncontaminated

water (note that a fixed sample of the oil slick was not available).Initial SW2 cell numbers were also lower than the cell numbers

from SW+oil1 (1.36 0.66 106 mL2 1) and SW+oil2(3.56 2.36 106 mL2 1) at day 0 of the experiment (p, 0.01),

suggesting that bacterial cells were introduced into oil-amendedbottles along with the oil sample.

Throughout the incubation, cell numbers in uncontaminated

bottles remained low and were either indistinguishable from oneanother (control SW, p= 0.2; SW1, p= 0.2), or decreased towards

the end of the incubation (SW2, p, 0.05). In contrast, SW+oil1 cellnumbers increased after the start of the incubation (all time points

were significantly higher than day 0, p, 0.001), peaking at day 14

(8.56 1.56 106 mL2 1; Fig. 2B). SW+oil2 cell numbers weresignificantly higher at day 10, 16, and 21 compared with day 0

(p, 0.001), and control SW+oil cells showed significantly highernumbers at day 16 (7.36 2.26 106 mL2 1) than day 7

(4.16 1.16 106 mL2 1, p, 0.001; note that no cell counts areavailable for days 0 and 2 due to high autofluorescence of the

samples).

Microbial cell abundance in oil aggregatesAggregate-associated microbial cells accounted for high pro-

portions of the total cell counts in the oil-amended incubations.Average cell numbers in oil aggregates at day 21 were

116 0.016 108 (mL aggregate)2 1 in SW+oil1 as well as46 0.016 108 (mL aggregate)2 1 in SW+oil2 and control SW+oil.

SW1 and SW2 aggregates had 16.76 0.046 108 cells (mLaggregate)2 1 and 28.26 0.016 108 cells (mL aggregate)2 1, respec-tively (data not shown). Corrected for their approximate volume in

each of the roller bottles (e.g. SW+oil1 aggregates: 0.6% of 900 mlbottle water< 5.4 ml oil aggregates), total aggregate-associated cell

numbers in oil-amended bottles were 60.76 0.056 109 (SW+oil1),28.46 0.056 109 (SW+oil2), and 39.36 0.056 109 (control SW+oil;

Fig. S2). Uncontaminated bottles had fewer cells associated withaggregates compared to oil-amended bottles (p, 0.001), with total

aggregate-associated cell numbers at 3.26 0.016 109 (SW1) and

Figure 1. Photo of an oil aggregate formed in one of the rollerbott les. Oil aggregate attached to surface water oil slick through stickyoil gels. Photo was taken at the end of the 21-day roller table incubationin one of the roller bottles containing seawater and oil (SW+oil1). Scalebar is approximately 10 mm.doi:10.1371/journal.pone.0034816.g001

Microbial Activities in Oil-Contaminated Seawater

PLoS ONE | www.plosone.org 3 April 2012 | Volume 7 | Issue 4 | e34816

•1500 m depth

• 775 million liters

• 87 days (Apr-Jul)

• 70 kms offshore

•56 m depth

• 525 million liters

• 9.7 months (Jun-Mar)

• 70 kms offshore

(2010)

(1979-80)

Deepwater Horizon (2010) – IXTOC (1979-80) ComparisonDeepwater Horizon (2010), not the first submarine blowout It was IXTOC-1 (1979-1980), Bay of Campeche, Offshore MX

Research Approach:

• Comparative Analyses of DWH and IXTOC• Use sediments record of MOSSFA events in

the present & past, recovery rates, predict DWH

Research Questions:• Does traditional oil spill response facilitate

MOSSA events?

• Are sub-marine oil-well blowouts and

MOSSFA events linked?

Oil Spill Response at DwH & IXTOC Blowouts Included:

Dispersants, Oil Burning and River Discharge

IXTOC SW GoMDWH NGoM

Sediment Core Comparison DwH-IXTOCLaminated Facies

2010

1979

Dw

HIn

fluen

ce

IXTO

C-1

In

fluen

ce

InfluencesSeen Above and Below Actual Date of Event Due to Carbon Loading, Redox Changes, and Redeposition

DWH (present)Ixtoc (past) Can we predict DWH recovery (future)?

“The present is the key to the past and the past is a window to

see into the future”

DWHSW-01 1110 m

IXTOCE55 1553 m

Sediment Coring Sites and Analytical Methods

Cores extruded @ high resolution- 2 mm to 20 cm, 5 mm to 60 cm

Methods:

1. Geochronology

(234Th, 210Pb,

MAR-gm/cm2/yr)

2. Sedimentology

(Grain size , clays)

3. Organic Geochemistry

(Org-C, aliphatic, PAH,

polars)

4. Benthic Foraminifera

(mortaility, recovery)

5. Microbial Ecology

(community structure)

6. Redox metal chemistry

(MnO2- oxic, Re- anoxia)

7. Bulk 14C

(Org-C source indicator)

1 cm1047m Sediments

PCB-06 DeSoto Canyon

70 nm ENE of DWH

1115 m Sediments

DSH 08 (N-S line)

20 nm NE of DWH

5

cm

Sediment Cores- August and December, 2010

1000-1200 m. “Plume Depth”

Why no Bioturbation?

UV fluorescing

particles & sheen

Sheen on Surface Sediment

Water DepthM-04 400m

Site ID

P-06 1043mD-08 1140mD-10 1520m

DWHBlowout

Event

0.0

0.5

1

1.5

2

2010 2011 2012 2013

Mas

s A

ccu

mu

lati

on

Rat

e –

MA

R (

g/c

m2/y

r)

Feb.

Dec.

Au

g.

Au

g.

Sept.

2014

Au

g.Sediment Pulse Event During the DWH

• 234Th-Mass Accumulation Rate Define Sediment Event (MOSSFA)Deposition/Sedimentation

Pulse (234Th)

Au

g.

Ap

ril

July

0.1

Pre-Event 210Pb MAR

1900-2000

Recovery ofBioturbation

at DSH-08Increases ThPenetration

Decreasing MOSSFA InputsStill Elevated wrt Pre-Event

(234Th)

234Th – Inventories (input indicator) show continued reduction in 2013-2014 at all sitesAt DSH-08 a return of sediment bioturbation is controlling increasing 234Th MAR

0

50

100

150

200

250

300

J… F… M… A… M… J… J… A… S… O… N… D…

2003-2009

2010

Mis

siss

ipp

i Riv

er D

isch

arge

(m

3s-

1)

DwH

2010 Discharge

2003-2009 Average Discharge

Elevated Discharge Mississippi River & Diversionary Channels To Push Oil Offshore and the Purge Marsh of Oil

23

Grain size changes W and E of DeSoto Canyon2010 2011-2012

2010 DwH Event

• Prior to 2010 all sites exhibit unique distribution of sediment grain-size.

• Since DwH event in 2010, all sediment core sites >100 m water depth show convergence to fine-grained sediments.

11

Interaction of Particles With Oil +/- Dispersant

Oil only

Oil +Dispersant

No oilNo disp.

No

particles

Nettle

tea

Diato

mateo

us

Earth

Kao

linite

Algae

No

particles

Nettle

tea

Diato

mateo

us

Earth

Kao

linite

Algae

Top of Test Tube Bottom of Test Tube

12

0

0.04

0.08

0.12

0.16

050

100150

200

Percent Phytoplankton

Depth (m

m)

Diatom 16S RNA Sequences

DwHEvent

• Diatom 16S rRNA genes - w/Chloroplast genes- only in top 3 cm of core

• Indicates large input of surface particle

•

DWH Surface Produced Microbial Sources and EPS Preserved in Deep-Sea Sediments

0

20

40

60

80

100

120

140

160

0 200 400

Bulk polysaccharides (ng/ml sediment)

DwH Event

Fluorescent lectin-bound carbohydrates showing drape (a) and stringer (b) morphologies. Scale bars = 100 µm. Phytoplankton and EPS

Inputs are evidence for the

MOSSFA PhenomenaOverholt and Kostka,Georgia Tech

Lincoln and Freeman,Penn State

0

5

10

15

20

25

30

0 10000 20000 30000

Do

wn

co

re (

mm

)

PFAR (F cm-2 yr-1)

DwH Event

Planktic Foram.Accum. Rate (F/cm2/yr)

Oil Only

DispersantOnly

Oil + Dispersant

ClearSeawater

Seawater + AlgeaTube top Middle

**

Proteins/Carbohyd.BiopolymerWebbing

Murk , Zeinstra, Koops et al, in review

Algae exposed to Oil/dispersant release EPS Forms ‘sticky-webbing’ that aggregates w/oil & clays

Statistical analysisCorrelation coefficients as well as differences between two

average valuesgiven astheir statistical mean 6 standard deviation

were tested for their significance using the Students t-test. Analysis

of variance (one-way ANOVA) was used for comparing average

values of more than two groups of data. If ANOVA was

significant, post hoc pairwise comparisons of means were

performed using the Bonferroni-Holmes test of variability. All

statistical analysis wasperformed in ExcelHusing the data analysis

toolpack as well as Daniel’s XL toolbox (both open source add-

ins).

Results

Formation of oil aggregates in roller bottlesRoller table incubation of uncontaminated seawater collected

near the Deepwater Horizon oil spill site with surface oil sampled

in the same area (hereafter referred to as SW+oil1 and SW+oil2

bottles; see Material and methods for roller bottles set-up) led to

rapid formation of aggregates (hereafter referred to as oil

aggregates) within one day (see Fig. S1 for close-up photos of oil

aggregates). Oil aggregates in both SW+oil bottles clumped

together after 7 days, forming a single aggregate up to 30 mm

in diameter, with visibly incorporated oil droplets. Oil aggregate

formation in control SW+oil (seawater that had been filtered and

autoclaved before the oil was added) was first observed at day 10

after the appearance of gelatinous networks of particulate matter

with incorporated oil droplets (hereafter referred to asoil gels). Oil

gels appeared to be very sticky surfaces onto which oil aggregates

attached upon collision (Fig. 1).

In contrast to oil-amended bottles, aggregate formation in rollerbottles with seawater not amended with oil (SW bottles) was

delayed and reduced in scale (aggregates that were much moretransparent than oil aggregates first appeared after 3 days; Fig.

S1E), and aggregates were lessabundant (2 to 3 per bottle). Theseaggregates did not change in size and number throughout the 21

daysof incubation. No aggregates formed in the control SW bottlecontaining filtered and sterilized seawater.

Wet weightsof oil aggregates in SW+oil1, SW+oil2, and control

SW+oil bottles after 21 days were 5.5 g, 7.1 g, and 9.8 g,respectively. Assuming a final bottle water volume of 900 ml at

day 21 (weight< 924.3 g), oil aggregates occupied approximately0.6% (SW+oil1), 0.8% (SW+oil2), and 1.1% (control SW+oil) of

the total roller bottle volume. SW1 and SW2 aggregates were0.19 g and 0.17 g, respectively, and thus 0.02% of total rollerbottle volume.

Microbial cell abundance in ambient watersThe microbial cell counts documented the impact of oil

amendments on the abundance of prokaryotic cells in surface

seawater during the roller table incubations, compared touncontaminated seawater. Uncontaminated ambient water

(SW2) had 0.56 0.46 106 cells mL2 1 at day 0 (Fig. 2A); thisnumber was lower (p, 0.05) but the same order of magnitude as

the cell abundance of uncontaminated water fixed shortly aftersampling (0.86 0.26 106 mL2 1), indicating that storage time and

conditions from the time of sampling until the beginning of theexperiment had little influence on cell numbers in uncontaminated

water (note that a fixed sample of the oil slick was not available).Initial SW2 cell numbers were also lower than the cell numbers

from SW+oil1 (1.36 0.66 106 mL2 1) and SW+oil2(3.56 2.36 106 mL2 1) at day 0 of the experiment (p, 0.01),

suggesting that bacterial cells were introduced into oil-amendedbottles along with the oil sample.

Throughout the incubation, cell numbers in uncontaminated

bottles remained low and were either indistinguishable from oneanother (control SW, p= 0.2; SW1, p= 0.2), or decreased towards

the end of the incubation (SW2, p, 0.05). In contrast, SW+oil1 cellnumbers increased after the start of the incubation (all time points

were significantly higher than day 0, p, 0.001), peaking at day 14

(8.56 1.56 106 mL2 1; Fig. 2B). SW+oil2 cell numbers weresignificantly higher at day 10, 16, and 21 compared with day 0

(p, 0.001), and control SW+oil cells showed significantly highernumbers at day 16 (7.36 2.26 106 mL2 1) than day 7

(4.16 1.16 106 mL2 1, p, 0.001; note that no cell counts areavailable for days 0 and 2 due to high autofluorescence of the

samples).

Microbial cell abundance in oil aggregatesAggregate-associated microbial cells accounted for high pro-

portions of the total cell counts in the oil-amended incubations.Average cell numbers in oil aggregates at day 21 were

116 0.016 108 (mL aggregate)2 1 in SW+oil1 as well as46 0.016 108 (mL aggregate)2 1 in SW+oil2 and control SW+oil.

SW1 and SW2 aggregates had 16.76 0.046 108 cells (mLaggregate)2 1 and 28.26 0.016 108 cells (mL aggregate)2 1, respec-tively (data not shown). Corrected for their approximate volume in

each of the roller bottles (e.g. SW+oil1 aggregates: 0.6% of 900 mlbottle water< 5.4 ml oil aggregates), total aggregate-associated cell

numbers in oil-amended bottles were 60.76 0.056 109 (SW+oil1),28.46 0.056 109 (SW+oil2), and 39.36 0.056 109 (control SW+oil;

Fig. S2). Uncontaminated bottles had fewer cells associated withaggregates compared to oil-amended bottles (p, 0.001), with total

aggregate-associated cell numbers at 3.26 0.016 109 (SW1) and

Figure 1. Photo of an oil aggregate formed in one of the rollerbott les. Oil aggregate attached to surface water oil slick through stickyoil gels. Photo was taken at the end of the 21-day roller table incubationin one of the roller bottles containing seawater and oil (SW+oil1). Scalebar is approximately 10 mm.doi:10.1371/journal.pone.0034816.g001

Microbial Activities in Oil-Contaminated Seawater

PLoS ONE | www.plosone.org 3 April 2012 | Volume 7 | Issue 4 | e34816

14

2005

2006

2007

2008

2009

2010

2011

0 50000100000150000

DSH08

PCB06

DSH10 (1043 m)

2005

2006

2007

2008

2009

2010

2011

0 50000100000150000

DSH08

PCB06

DSH10

(1140 m)

2005

2006

2007

2008

2009

2010

2011

0 5 10 15

2005

2006

2007

2008

2009

2010

2011

0 5 10 15

LMW PAH(μg/g OC)

0 5 10

2005

2006

2007

2008

2009

2010

2011

0.0 150.0 300.0

2005

2006

2007

2008

2009

2010

2011

0.00 10.00 20.00

TOC (g m2 month)

Total PAH (μg m2 month)

0 150 3000 10 20

year

15

HMW PAH(μg/g OC)

0 5 10 15

LethalLethal2005

2006

2007

2008

2009

2010

2011

0 60 120

2005

2006

2007

2008

2009

2010

2011

0 120 240

LMW PAH(μg m-2 month-1)

HMW PAH(μg m-2 month-1)

DWH Hydrocarbons, Dispersant in Sediments • High Accumulation Rates of Organic-C & PAHs During DwH• LMW- Petrogenic & HMW- Pyrogenic Sources, DOSS-Dispersant

Romero et al, PlosOne (in review)

2005

2006

2007

2008

2009

2010

2011

0 50000100000150000

DSH08

PCB06

DSH10

(1500 m)

PetrogenicSource

PyrogenicSource

TOXIC and CARCENOGENICCompounds

DWH Event

DOSS, ngFT-ICRMS

DispersantDSH-08

16

Benthic Habitat Changes and Organismal Decline• Oil Sedimentation Causes changes in toxicity, chemical

conditions and widespread mortality of benthic fauna

DWH Event

DWH Event

• Benthic Foram Die-off: What is the cause? How is recovery?

2010 Map of Sediment Flocculent Thickness- nGoM

Statistical analysisCorrelation coefficients as well as differences between two

average valuesgiven astheir statistical mean 6 standard deviation

were tested for their significance using the Students t-test. Analysis

of variance (one-way ANOVA) was used for comparing average

values of more than two groups of data. If ANOVA was

significant, post hoc pairwise comparisons of means were

performed using the Bonferroni-Holmes test of variability. All

statistical analysis wasperformed in ExcelHusing the data analysis

toolpack as well as Daniel’s XL toolbox (both open source add-

ins).

Results

Formation of oil aggregates in roller bottlesRoller table incubation of uncontaminated seawater collected

near the Deepwater Horizon oil spill site with surface oil sampled

in the same area (hereafter referred to as SW+oil1 and SW+oil2

bottles; see Material and methods for roller bottles set-up) led to

rapid formation of aggregates (hereafter referred to as oil

aggregates) within one day (see Fig. S1 for close-up photos of oil

aggregates). Oil aggregates in both SW+oil bottles clumped

together after 7 days, forming a single aggregate up to 30 mm

in diameter, with visibly incorporated oil droplets. Oil aggregate

formation in control SW+oil (seawater that had been filtered and

autoclaved before the oil was added) was first observed at day 10

after the appearance of gelatinous networks of particulate matter

with incorporated oil droplets (hereafter referred to asoil gels). Oil

gels appeared to be very sticky surfaces onto which oil aggregates

attached upon collision (Fig. 1).

In contrast to oil-amended bottles, aggregate formation in rollerbottles with seawater not amended with oil (SW bottles) was

delayed and reduced in scale (aggregates that were much moretransparent than oil aggregates first appeared after 3 days; Fig.

S1E), and aggregates were lessabundant (2 to 3 per bottle). Theseaggregates did not change in size and number throughout the 21

daysof incubation. No aggregates formed in the control SW bottlecontaining filtered and sterilized seawater.

Wet weightsof oil aggregates in SW+oil1, SW+oil2, and control

SW+oil bottles after 21 days were 5.5 g, 7.1 g, and 9.8 g,respectively. Assuming a final bottle water volume of 900 ml at

day 21 (weight< 924.3 g), oil aggregates occupied approximately0.6% (SW+oil1), 0.8% (SW+oil2), and 1.1% (control SW+oil) of

the total roller bottle volume. SW1 and SW2 aggregates were0.19 g and 0.17 g, respectively, and thus 0.02% of total rollerbottle volume.

Microbial cell abundance in ambient watersThe microbial cell counts documented the impact of oil

amendments on the abundance of prokaryotic cells in surface

seawater during the roller table incubations, compared touncontaminated seawater. Uncontaminated ambient water

(SW2) had 0.56 0.46 106 cells mL2 1 at day 0 (Fig. 2A); thisnumber was lower (p, 0.05) but the same order of magnitude as

the cell abundance of uncontaminated water fixed shortly aftersampling (0.86 0.26 106 mL2 1), indicating that storage time and

conditions from the time of sampling until the beginning of theexperiment had little influence on cell numbers in uncontaminated

water (note that a fixed sample of the oil slick was not available).Initial SW2 cell numbers were also lower than the cell numbers

from SW+oil1 (1.36 0.66 106 mL2 1) and SW+oil2(3.56 2.36 106 mL2 1) at day 0 of the experiment (p, 0.01),

suggesting that bacterial cells were introduced into oil-amendedbottles along with the oil sample.

Throughout the incubation, cell numbers in uncontaminated

bottles remained low and were either indistinguishable from oneanother (control SW, p= 0.2; SW1, p= 0.2), or decreased towards

the end of the incubation (SW2, p, 0.05). In contrast, SW+oil1 cellnumbers increased after the start of the incubation (all time points

were significantly higher than day 0, p, 0.001), peaking at day 14

(8.56 1.56 106 mL2 1; Fig. 2B). SW+oil2 cell numbers weresignificantly higher at day 10, 16, and 21 compared with day 0

(p, 0.001), and control SW+oil cells showed significantly highernumbers at day 16 (7.36 2.26 106 mL2 1) than day 7

(4.16 1.16 106 mL2 1, p, 0.001; note that no cell counts areavailable for days 0 and 2 due to high autofluorescence of the

samples).

Microbial cell abundance in oil aggregatesAggregate-associated microbial cells accounted for high pro-

portions of the total cell counts in the oil-amended incubations.Average cell numbers in oil aggregates at day 21 were

116 0.016 108 (mL aggregate)2 1 in SW+oil1 as well as46 0.016 108 (mL aggregate)2 1 in SW+oil2 and control SW+oil.

SW1 and SW2 aggregates had 16.76 0.046 108 cells (mLaggregate)2 1 and 28.26 0.016 108 cells (mL aggregate)2 1, respec-tively (data not shown). Corrected for their approximate volume in

each of the roller bottles (e.g. SW+oil1 aggregates: 0.6% of 900 mlbottle water< 5.4 ml oil aggregates), total aggregate-associated cell

numbers in oil-amended bottles were 60.76 0.056 109 (SW+oil1),28.46 0.056 109 (SW+oil2), and 39.36 0.056 109 (control SW+oil;

Fig. S2). Uncontaminated bottles had fewer cells associated withaggregates compared to oil-amended bottles (p, 0.001), with total

aggregate-associated cell numbers at 3.26 0.016 109 (SW1) and

Figure 1. Photo of an oil aggregate formed in one of the rollerbott les. Oil aggregate attached to surface water oil slick through stickyoil gels. Photo was taken at the end of the 21-day roller table incubationin one of the roller bottles containing seawater and oil (SW+oil1). Scalebar is approximately 10 mm.doi:10.1371/journal.pone.0034816.g001

Microbial Activities in Oil-Contaminated Seawater

PLoS ONE | www.plosone.org 3 April 2012 | Volume 7 | Issue 4 | e34816

Resuspension and Downslope Transport of DWH Oil-Associated Sediment

Spatial & temporal offset between surface water oil coverage And ”foot-print” of sedimentary oil deposition

Significant quantities of oil remain trapped in deep-sea sediments (4-10% of the total oil released to the ocean)

85 Day-Gridded Average Oil-CoverRed = >90%Yellow =

Three Mechanisms of Sedimentary Oil Deposition:

2. Flocculent “Dirty” Blizzard: Oil w/particles: lithogenic, orgs.

Surfacing Oil Slick and Sheen

Jet ReleaseOil-Gas RatioPressure GradientOil Composition

1000-1300m

1. Toxic Bath-Tub Ring: Plume Impingement

Continental Shelf

BOP

3. Cross-Shelf Oil-Snow Transport:Outer Shelf and Slope Deposition

1-Toxic Bath-Tub Ring:

Plume impinges on sediment directly

2-Flocculent Blizzard:

Rapid flocculation and sinking of oil-associated clays, algae and particles

3-Cross-Shelf Transport/Deposition:

Persistent transport of oil-floc from shallow shelf to outer shelf (>100) and slope environments 19

Oil Can Sink… At DwH, oiled-sediment pulse layers

confirmed through multiple lines evidence: sedimentological, biological, organic and

inorganic geochemical, redox metal, micropaleontological, isotopic analyses…

What do we see in sediments from the IXTOC region in the SGoM?

Oil, Dispersant

Riverine Influences

Marine biota Marine Oil

Snow“Dirty Blizzard”

Pyrogenic

Petrogenic Dispersant

Algal Bloom

EPS Formation

MicrobialLoop

Salinity Nutrients

Clays

Freshwater Discharge

> 30 meter of water~ 30 miles off shore

20

In Situ Burning

~ 25 to 55 mi offshore

Area of Dispersant

Low SalinityCoastal/Offshore

Response strategies intensified MOSSFA and increased the

“footprint” of sedimentary oil

deposition?

Environmental factors that control MOSSFA and the

formation and sinking of oil-associated particles

Clays, Nutrients Productivity

Low SalinityCoastal/Offshore

ClaysNutrients

Surface oil footprint and trajectory of the Ixtoc-I oil spillCummulative Days Coverage of IXTOC and DWH oil spill

100 km

IXTOC Cruise: Sediment Coring & Water Sampling Transects

IXTOC-1 Exclusion Zone

IXTOC-1 Exclusion Zone

IXTOCSW GoM

DWHNE GoM

Sediment Core Comparison: DwH vs IXTOCRedox Jump Facies

2010

1979

Dw

HIn

fluen

ce

IXTO

C-1

In

fluen

ce

Influences Below and Above Actual Date of Event Due to Carbon Loading, Redox Changes and Redeposition

“The present is the key to the past and the past is the key to see the future”DWH (present)Ixtoc (past) Can we predict DWH/other

spill (future)???

“The present is the key to the past and the past is a window to see into the future”

DSH-08 1043 mDWH nGoM

E58 1525 mIXTOC sGoM

IXTOC SedimentMass Accumulation Rate

g/cm2/yr

IXTOC Event

IXTOC Sediments: Mass Accumulation Rate Increases

IXTOC Event MAR increase by ~3.5-fold

0.09 to 3.2 g/cm2/yr(DWH Event increases

by 2.5 fold)

MAR increase is synchronous with date of IXTOC blowout, 1979, but elevated MAR extends for years after event ends.

The prolonged IXTOC MAR is similar to that seen after DWH event

Dep

th, c

ms

0

2

4

6

8

10

12

14

16

18

0 1 2 3

(210Pb-derived)

1979

Seasonal Rainfall Pattern and River Discharge of Suspended Solids in the Vera Cruz-IXTOC Region

• Majority of Wet-Season Overlaps with Timing of IXTOC Event, June - March • Rivers adjacent to IXTOC contain abundant Suspended Particles and Nutrients

IXTOC

River Suspended Solids (concentrations)

River Suspended Solids (discharge rate)

June-MarchWet

Dry

IXTOC Sediment Core Locations Surface Sediment PAH Distribution & Concentrations

Gracia et al & Machain et al., Unpublished Data

1979

1979

Correlation of IXTOC Oil and Sediment: HOPANES M/Z 191

Sediment E52 1650m4-5 cm depth

IXTOC Oil Samplenorhopane hopane

S/R homo-hopane

S/R bishomo-hopane S/R trishomo-

hopaneS/R tetrakish-homohopane S/R pentakish-

homohopane

Tm

Ts

norhopane

hopane

S/R homo-hopane

S/R bishomo-hopane

S/R trishomo-hopane

S/R tetrakish-homohopane

S/R pentakish-homohopane

BITUMEN ENCRUSTERD

IXTOC Blowout Interval

IXTOC-1 Sediment CoreSite E63

4-5cm

13-14 cm

0-1 cm

Magnetic Susceptibility

Core 631650 mwd

2010

1980

Foraminifera Abundance (tests/g)

50 200 400

1950

1912

Age210Pb

Sediment > 63 uM

IXTOC Blowout Event Recorded in Sediments-1600 mwdChanges in Sediment type, Foram Abundance, Redox, HC Inputs

Machain-Unpublished

IIx

Hydrocarbon StainedForam Tests

IXTOC EventIXTOC Event

Indicator of Volcanic Minerals

17

19

21

23

25

27

1975 1995 2015

Me

an F

ish

ers

Alp

ha

(S)

Year

Benthic Recovery Rate

11

12

13

14

15

2010 2011 2012 2013

Me

an F

ish

ers

Alp

ha

(S)

Year

Appears that Recovery Takes Longer in SGoM, need more sites to validate?

NGOM: integrated diversity from 8

sites took 3 years to recover following DWH, resembles 234Th evidence of bioturbation (Larson, Brooks, et al.)

SGOM: diversity from 1 site suggests

recovery took ~7 years following Ixtoc

SGoM- IXTOC-1NGoM- DWH

Ixto

c

DW

HBackground Mean

Background Mean

Schwing et al.

Gracia, Unpublished Data

Spatial Offset Between Surface Water & Sediment Oil Coverage MOSSFA processes after IXTOC : Deposition in deepsea

Sediment Oil Coverage

Surface WaterOil Coverage

IXTOC

1979

1979

Significance and Implication of a MOSSFA Event: Oil Spill Response and Long-Term Consequences

• A “new” consideration for real-time oil spill response- Concentrating mechanism of mineral-oil-biota aggregates (MOBAs)- Predict MOBA formation mechanisms: spatial-temporal - Formation, in part, related to traditional response strategies:

Freshwater discharge (clays/nutrients), burning and dispersant application

- Target for real-time collection and cleanup

• Widespread deposition & accumulation MOBAs/Sediments- Long-term persistence in the benthic environment - Important for calculations of the final oil budget- Predict area of benthic ecosystem impact- If transported to slope and buried, may be best alternative to

remove oil from biologically active areas and minimize impact to economically important species

MOSSFA Time in the nGoM, Sunset May 22, 2010

Backup Slides