Content uploaded by Danielle M DeLeo
Author content
All content in this area was uploaded by Danielle M DeLeo on Sep 18, 2018
Content may be subject to copyright.
Response of deep-water corals to oil and chemical dispersant exposure
Danielle M. DeLeo
a
, Dannise V. Ruiz-Ramos
b
, Iliana B. Baums
b
, Erik E. Cordes
a
a
Department of Biology, Temple University, 315 Bio-Life Sciences Bldg, Philadelphia, PA 19122, United States
b
Department of Biology, The Pennsylvania State University, 208 Mueller Lab, University Park, PA 16802, United States
article info
Available online 5 March 2015
Keywords:
Deepwater Horizon
Deep sea
Oil spill
Dispersant
Corexit
Survival
Toxicity tests
Gulf of Mexico
Gorgonian
Octocoral
Black coral
Antipatharian
abstract
Cold-water corals serve as important foundation species by building complex habitat within deep-sea
benthic communities. Little is known about the stress response of these foundation species yet they are
increasingly exposed to anthropogenic disturbance as human industrial presence expands further into the
deep sea. A recent prominent example is the Deepwater Horizon oil-spill disaster and ensuing clean-up
efforts that employed chemical dispersants.Thisstudyexaminedtheeffectsofbulkoil–water mixtures,
water-accommodated oil fractions, the dispersant Corexit 9500A
s
, and the combination of hydrocarbons
and dispersants on three species of corals living near the spill site in the Gulf of Mexico between 500 and
110 0 m dept h s : Paramuricea type B3, Callogorgia delta and Leiopathes glaberrima. Following short-term
toxicological assays (0–96 h), all three coral species examined showed more severe health declines in
response to dispersant alone (2.3–3.4 fold) and the oil–dispersant mixtures (1.1–4.4 fold) than in the oil-only
treatments. Higher concentrations of dispersant alone and the oil–dispersant mixtures resulted in more
severe health declines. C. delta exhibited somewhat less severe health declines than the other two species in
response to oil and oil/dispersant mixture treatments, likely related to its increased abundance near natural
hydrocarbon seeps. These experiments provide direct evidence for the toxicity of both oil and dispersant on
deep-water corals, which should be taken into consideration in the development of strategies for
intervention in future oil spills.
&2015 Elsevier Ltd. All rights reserved.
1. Introduction
The Deepwater Horizon (DWH) oil spill was one of the largest
environmental disasters in history, releasing approximately 5 million
barrels of crude oil at depth in the Gulf of Mexico (GoM) over a three-
month period (Crone and Tolstoy, 2010; Camilli et al., 2011). In
addition, nearly 7 million liters of oil dispersants were applied during
the ensuing cleanup efforts. Dispersants are chemical emulsifiers
that act to increase the rate of oil dispersion thereby increasing the
amount of small oil droplets suspended in the water column,
reducing oil slicks at the surface. Thus, dispersant applications affect
the fate, transport and physical composition of oil. Of the 7 million
liters of oil dispersants used, approximately 3 million liters were
applied at depth for the first time in history (Barron, 2012), without a
comprehensive understanding of how this subsea application might
alter the fate of oil and impact benthic ecosystems (National Research
Council, 2005).
Petroleum hydrocarbons released under high-pressure undergo a
series of interconnected physical and chemical processes that affect
their fate and transport in the deep sea (Camilli et al., 2010; Kessler
et al., 2011; Reddy et al., 2012). Following the direct injection of
disperant (Corexit 9527A and 9500A) to the Macondo well head at a
depth of 1544 meters (m) (Hazen et al., 2010), a large oil plume
persisted for months centered at approximately 1100 m depth,
without substantial biodegradation (Camilli et al., 2010). Oil spewing
from the wellhead encountered turbulent mixing and was emulsi-
fied as a result of its reduced buoyancy at depth and the application
of dispersant (Fodrie and Heck Jr., 2011). Measurements of water-
column samples collected from this deep-water plume (defined by
Camilli et al., 2010) indicated that a significant portion of water-
soluble hydrocarbon components were retained in deep waters,
with unknown portions of insoluble hydrocarbons drifting to the
sea floor (Reddy et al., 2012).Despitesomeemulsification of oil
throughout the water column, surface waters were still polluted with
oil slicks (Fodrie and Heck Jr., 2011). At the surface, some components
of the oil were then transformed into aggregations of marine snow
(and floc) by coagulation with suspended particulates and planktonic
organisms. Although this marine snow disappeared from the surface
layers of the GoM within a month, it is likely that it sunk into the deep
sea as the oil weathered (Passow et al., 2012).
Recent studies have found both lethal and sub-lethal effects of the
DWH blowout on species inhabiting pelagic and coastal environments
(Barron, 2012; Silliman et al., 2012; Whitehead et al., 2012; Dubansky
et al., 2013; Almeda et al., 2013). Prior studies have shown variable
levels of crude oil toxicity on aquatic organisms with some fauna being
more susceptible than others (Anderson et al., 1974; Bonsdorff et al.,
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/dsr2
Deep-Sea Research II
http://dx.doi.org/10.1016/j.dsr2.2015.02.028
0967-0645/&2015 Elsevier Ltd. All rights reserved.
Deep-Sea Research II 129 (2016) 137–147
1990; Coull and Chandler, 1992; Stark et al., 2003). Dispersant addition
to the oil triggers a transient increase in hydrocarbon concentrations
throughout the water-column (Pace et al., 1995), which can then lead
to higher, more toxic exposures of dissolved and dispersed oil
components upon contact with marine life.
Spill-impacted deep-sea coral communities were first discov-
ered at a depth of approximately 1370 m, 11 km southwest of the
Macondo well explosion, at the lease block site Mississippi Canyon
(MC) 294 (White et al., 2012). Various species of coral, primarily
Paramuricea biscaya (Grasshoff, 1977), were found covered with
brown flocculent material (floc), exhibiting characteristic signs of
stress and mortality, including excess mucus production, sclerite
enlargement, and tissue loss. Further analysis of this floc revealed
hydrocarbons from the Macondo well were indeed present (White
et al., 2012). Whether the damage observed to the corals was
induced by sinking oil-filled particulates, dissolved hydrocarbons,
dispersants, or a combination of all of these sources is unknown.
Subsequently, two additional sites were discovered to contain
impacted deep-sea coral communities (Fisher et al., 2014).
Deep-sea corals alter the terrain of the sea floor and produce
complex, heterogeneous habitat, which promotes benthic biodi-
versity (Cordes et al., 2008, 2010). In addition to reef-forming
scleractinian corals, which generally occur at upper-slope depths
(300–1000 m), octocorals and black corals (antipatharians) form
large, tree-like structures from the subtidal to over 3000 m depth.
These corals colonize hard substrata, and can form dense fields
(Roberts et al., 2006). By increasing the complexity of the seafloor,
they provide shelter, feeding areas, and nursery grounds for many
fish and invertebrates.
Because deep-sea corals build the foundation for these com-
munities, damage to them can impact biodiversity and ecosystem
function (Husebo et al., 2002; Freiwald et al., 2004). Their long-
evity and slow growth rates make them particularly vulnerable to
anthropogenic disturbance (Grigg, 1974; Emiliani et al., 1978;
Druffel et al., 1990, 1995; Risk et al., 1998, 2002; Andrews et al.,
2002; Adkins et al., 2004;Roark et al., 2009). As crude oil reserves
are abundant in the GoM, with 1.5 billion barrels of oil extracted
from the sea floor each day (Minerals Management Service, 2009),
it is now a critical time for further examination of deep-sea coral
response to oil and dispersant exposure.
Here, the effects of oil, dispersant and oil–dispersant mixtures
were tested experimentally on three species of deep-sea coral
living near the DWH oil spill site in the Gulf of Mexico, including
Paramuricea type B3 (Doughty et al., 2014), Callogorgia delta (Bayer
et al., 2014) and Leiopathes glaberrima (as re-described in Opresko
and Baron-Szabo, 2001). P. biscaya was the most common of the
corals impacted by the DWH oil spill (White et al., 2012; Fisher et
al., 2014), and Paramuricea type B3 is the sister species to this coral
(Doughty et al., 2014). Paramuricea type B3 was chosen because its
shallower depth distribution (830–1090 m for Paramuricea type
B3 vs. 1370–2600 m for P. biscaya with one individual collected at
850 m, Doughty et al., 2014) results in higher survivorship ship-
board, and to avoid further impact to the relatively small popula-
tions of P. biscaya that have thus far been discovered. C. delta
preferentially occupies habitats near natural oil seeps in the deep
GoM (Quattrini et al., 2013), suggesting that the species may have
evolved a tolerance for hydrocarbon exposure. L. glaberrima is slow
growing and lives to very old ages, making it one of the oldest
skeletal secreting organisms known to date (Roark et al., 2009).
Slow growth rates make this species highly sensitive to natural
and anthropogenic disturbances.
This study examined the effects of exposure to bulk oil–water
mixtures, water-accommodated oil fractions (WAF), dispersants,
and mixtures of hydrocarbons and dispersants using short-term
toxicological assays (r96 h) that monitored phenotypic responses
and survivorship. Specifically, we tested the hypotheses that oil/
dispersant mixtures would be the most toxic to corals, and that C.
delta would have a higher tolerance for hydrocarbons due to its
affinity for natural seep habitats.
2. Methodology
2.1. Sample collection and acclimatization
All samples were collected from two sites in the GoM. C. delta
and L. glaberrima were collected from the Viosca Knoll (VK) 826
site at a depth of approximately 500 m (29 109.5
0
N, 88 101.0
0
W;
Cordes et al., 2008;Davies and Guinotte, 2011). Paramuricea type
B3 colonies were collected from a large population of corals at
approximately 1050 m depth at Atwater Valley (AT) 357 (27 158.6
0
N,
89 170.4
0
W; Doughty et al., 2014). At each site, corals were hapha-
zardly collected with the remotely operated vehicles (ROV) Global
Explorer MK3 or Hercules.
Samples were taken on multiple dives, with 5–6 colonies of
both C. delta and L. glaberrima collected from VK826, and 5–6
colonies of Paramuricea type B3 gathered from AT357. Samples
were collected several meters apart from conspecific colonies to
reduce the likelihood of sampling clones. Corals were visually
identified using live video stream from cameras attached to each
ROV, before being collected with a manipulator arm and secured in
an insulated “bio”box and or sealable collection quivers. When
possible, branches of colonies were sampled to reduce impact.
At the surface, colonies were immediately transferred to con-
tainers with filtered seawater of the species-appropriate tempera-
ture and salinity (35 psu). C. delta and L. glaberrima were jointly
maintained at approximately 8 1C and later, Paramuricea type B3 at
51C (the average in situ temperatures at depth) in a temperature-
controlled room for the duration of the experiment. Temperature
in holding vessels was continuously monitored using temperature
probes (Hobo
s
Data Loggers). Corals were allowed to acclimate for
6–12 h prior to experimentation.
2.2. Preparation of bulk-oil treatments
For the bulk-oil experiment three stock solutions were prepared:
crude oil (MASS oil collected from the Macondo well during the spill),
dispersant (Corexit 9500A), an oil/dispersant mixture, and artificial
seawater controls. All solutions were made with sterile artificial
seawater (ASW, Instant Ocean™)at35psu,theaverageinsitusalinity
for both sites. ASW allowed us to accurately maintain desired salinity
and temperature for large volumes of water without the potential for
introducing contaminants from the ship's seawater system, and to
avoid the unreliability of collecting buckets of seawater from over the
side in variable sea states. We have used ASW to maintain other cold-
water coral species alive in laboratory aquaria for extended periods of
time without adverse affects (Lunden et al., 2014).
A stock bulk-oil solution was prepared at a concentration of 250
parts per million (ppm) by adding 50
μ
LofMASSoilto199.95mL
ASW. The solution was mixed at room temperature for a 24-h
period on an orbital shaker at approximately 500 rpm to achieve
highest possible homogeneity. Oil dilutions were prepared from this
stock solution. The subsequent oil concentrations were chosen in an
attempt to determine the threshold for lethal toxicity, following
preliminary toxicity studies on L. glaberrima. Dispersant concentra-
tions were the same as the oil concentrations so as to examine the
relative toxicity of oil vs. dispersant. The oil/dispersant-mixture
stock solution was prepared with an initial targeted concentration
of 250 ppm each of crude oil and Corexit 9500A by adding 50
μ
Lof
each to 199.90 mL of ASW. The dispersant stock solution was
prepared by adding 50
μ
L Corexit 9500A to 199.95 mL ASW to
achieve an initial concentration of 250 ppm. Serial dilutions were
D.M. DeLeo et al. / Deep-Sea Research II 129 (2016) 137–147138
prepared from each of the three stock solutions to produce three
target concentrations: 25 ppm (High), 7.9 ppm (Medium) and
0.8 ppm (Low).
All solutions were placed into sterile 50 mL glass vials. These
were then incubated at 5 or 8 1C, dependent on species, and mixed
continuously at low speeds for 24 h on an orbital shaker table to
reduce separation and to encourage even oil distribution. Experi-
ments were conducted between 8 and 27 November 2012 onboard
the R/V Falkor.
2.3. Preparation of treatments using water-accommodated oil
fractions (WAF)
For this experiment, stock solutions were prepared using only
the water-accommodated oil fractions (WAF). For the WAF oil
treatment, a higher oil volume (9.5 mL) of surrogate oil was added
to 475 mL of ASW and mixed at high speeds ( 350 rpm) in an
attempt to produce a 1.2 mM WAF oil solution. The WAF was
separated from the insoluble oil layer using a sterile separatory
funnel, and used as a stock solution to produce experimental
treatments with targeted initial total hydrocarbon concentrations
of 250
μ
M (High), 150
μ
M (Medium) and 50
μ
M (Low) WAF. Target
concentrations were chosen to find lethal doses, as none of the
previous bulk-oil (only) concentrations proved to be lethal. This
was done using a standardized WAF protocol (S. Joye, personal
communication) and based on the highest concentrations of oil
detected during the spill (300
μ
M, Joye et al., 2011).
The oil/dispersant mixture treatment was prepared using the
same oil volume, with 950
μ
L of Corexit 9500A added (one-tenth
of the oil concentration) to produce a dispersant enhanced WAF
(DE–WAF; oil/dispersant treatment), also mixed at high speeds
(350 rpm). As the dispersant concentrations in the bulk-oil
exposures were not entirely lethal to C. delta in the short term
and most of the observed health decline was seen towards the end
of the exposures at the highest Corexit 9500A concentration, the
range of dispersant concentrations was progressively increased
from those used in the previous exposures to attempt to reveal the
lethal concentration (LC50). The dispersant stock solution was
made by adding 950
μ
L of Corexit 9500A to 475 mL of ASW,
with an initial dispersant concentration of 848 mg/L (mixed at
200–300 rpm). All stock solutions were mixed at room tempera-
ture for 48–72 h. Experimental solutions were then made from
these two treatments with targeted initial oil concentrations of
250
μ
M (High), 150
μ
M (Medium) and 50
μ
M (Low) and targeted
initial total dispersant concentrations of 176.7 mg/L (High),
106.0 mg/L (Medium) and 35.3 mg/L (Low).
All solutions were placed into sterile 50 mL acid-washed glass
vials prior to experimentation. There was an anticipated and
unavoidable loss of hydrocarbons and dispersant due to the
adhesion of hydrophobic components to the dilution containers
with each sequential transfer, as well as the chemical and coral–
microbial alterations of hydrocarbons and dispersant components
over the course of the treatments. Therefore oil and dispersant
concentrations are reported as conservative, initial targeted values
only, and qualitatively designated as “High”“Medium”and “Low”
in the analysis. Experiments were conducted from 23 June 2013 to
3 July 2013 onboard the R/V Nautilus.
2.4. Fragmentation and exposure experiments
For both bulk-oil and WAF experiments, four to six colonies of
each species (n¼3) were fragmented into similar sized (approxi-
mately 3–6 cm tall), genetically identical replicates, or “nubbins”
(n¼11) and placed into the oil, dispersant, oil/dispersant mixture
and the control (ASW) treatments. Paramuricea type B3 had only
three healthy colonies for the bulk-oil exposures. The number of
polyps per nubbin varied for each species because of the wide
range in polyp sizes and unique branching morphology. Samples
were placed in 50 mL pyrex test tubes, mounted on a shaker table
in a temperature controlled environment, and aerated every 24 h
by bubbling air into the tubes and gently inverting each sample.
Each sample was photographed together with a scale and
monitored for signs of stress at four time points (24, 48, 72 and
96 h) during the bioassay. Each experimental nubbin was assigned
an overall health rating on a scale ranging from 0 to 5. The
percentage of live polyps and tissue-covered skeleton primarily
contributed to this rating: dead fragment (score of 0), ⪡50% (score
of 1–2), 50% (score of 3), ⪢50% (score of 4–5), while the other
stress responses further differentiated between scores. Ratings were
further refined based on the following phenotypic stress responses:
percentage of polyp retraction and or inflation, presence and
persistence of mucus discharge, dead or darkened tissue, sloughing
tissue and exposed skeleton. While polyp mortality, polyp retrac-
tion, mucus release, loose tissue, and exposed skeleton were
observed in all three species, swollen polyps were only observed
in L. glaberrima, while darkened tissue was specifictoParamuricea.
Tissue discoloration and whitening was only observed in C. delta.
Furthermore, C. delta displayed a distinctive polyp coiling, ulti-
mately forming node-like structures that eventually disintegrated,
leaving behind exposed skeleton. Samples and treatments were
randomized in an attempt to reduce health-scoring bias.
2.5. Survival analysis
Health rankings were averaged for replicate coral fragments in
each experimental concentration and plotted over time to inves-
tigate health decline. This was done discretely for each round of
experiments (bulk-oil or WAF), type of treatment (oil, dispersant
and oil/dispersant) and species to determine the effect of concen-
tration on fragment health over time. Health differences within
the different treatments at the 96-h end-point were tested using a
non-parametric Kruskal–Wallis test, and if applicable (po0.05),
non-parametric post-hoc, pair-wise comparisons were performed
using the Wilcoxon method (using JMP
s
Pro 10.0.2).
To investigate fragment survival over time, a Kaplan–Meier (K–M)
“time to event”survival analysis was performed separately for each
experimental series (IBM
s
SPSS
s
Statistics v22, Kaplan and Meier,
1958). This test measures the fraction of fragments declining to a
health status of 3 or below at each time point and generates a
survival curve. To quantify differences amongst the survival curves
for a given species and treatment, a Mantel–Cox log-rank test was
used to evaluate statistical significance (
α
¼0.05); if significant,
pair-wise comparisons were made, again using a Mantel–Cox log-
rank test.
An additional K–M analysis was performed to compare survival
across species in each treatment. Only “event”occurrences con-
tribute to survival estimates; the remaining data becomes censored
in the analysis. For this reason the ASW control treatments, in
which all fragments maintained health ratings 43, were excluded
from survival-estimate statistics during species comparisons. A
similar percentage of censored cases were present in the oil,
dispersant and oil–dispersant treatments for each species, and the
pattern of censoring was similar.
Additionally, Cox regressions were performed to quantify the
hazard (i.e. a decline in health) associated with (a) treatment
(water, oil, dispersant/oil and dispersant), (b) concentration (High,
Medium, Low, Zero), and (c) species (C. delta,L. glaberrima,
Paramuricea type B3) for the two sets of experiments (bulk-oil
and WAF). The “event”in the time-to-event analysis was reaching
a health rating of 3 or below (3, 1, 2 or 0), as mortality was not
observed in every treatment and concentration during the expo-
sure. The hazard ratios were calculated for each factor with respect
D.M. DeLeo et al. / Deep-Sea Research II 129 (2016) 137–147 13 9
to control treatment (a), the zero concentration (b) and C. delta (c),
as we had hypothesized this to be the species most likely adapted
to oil exposure. Cox regression was performed in IBM
s
SPSS
s
Statistics v22.
3. Results
3.1. Exposure effects on Paramuricea type B3
3.1.1. Oil treatment
Complete fragment mortality was not observed for Paramuricea
type B3 in the control, bulk-oil or oil–WAF trea tme nts ( Figs. 1Aand2A).
In examining the effect of concentration on fragment condition at the
end of the bulk-oil and WAF exposures, the Kruskal–Wallis test showed
no significant differences among the 96-h health ratings across all oil
concentrations and controls (p40.05).
3.1.2. Dispersant treatment
Whole fragment mortality was observed in Paramuricea type B3
nubbins exposed to the High dispersant treatment (Fig. 1D). This
decline in health originated in the dispersant mixture within 48–72 h,
with two of three colonies exhibiting complete fragment mortality at
theendoftheexposureperiod.TheKruskal–Wallis test revealed
significant differences (po0.05)inhealthrankingsforParamuricea
type B3 at the end of the exposure; pair-wise comparisons revealed
significant differences between nubbins in the High dispersant relative
to the control samples (po0.05).
High coral fragment mortality was observed in the dispersant
treatment across all concentrations tested in the WAF experiments.
One of six Paramuricea type B3 replicates died in the Low dispersant
solution, with complete mortality observed in four of six replicates in
the Medium dispersant treatment by 96 h. At High dispersant
concentrations, four of six replicates were dead after only 48 h, with
complete mortality of all fragments after 96 h (Fig. 2D). The Kruskal–
Wallis test and pair-wise comparisons revealed significantly higher
health ratings among the control Paramuricea type B3 nubbins relative
to all levels of dispersant (Low, Medium and High; po0.005) as well
as in the Low vs. High dispersant concentrations (po0.005).
3.1.3. Oil/dispersant treatment
Whole fragment mortality was observed in Paramuricea type B3
nubbins exposed to the High oil/dispersant treatment (Fig. 1G), with
complete mortality in two of three fragments by 96 h. There were
significant health differences among concentrations (Kruskal–Wallis,
po0.05), and subsequent pair-wise comparisons revealed significant
differences between fragments in the High oil/dispersant relative to
the control samples (po0.05).
During the WAF exposures, complete mortality was observed in
the oil/dispersant mixture (DE–WAF), for one of six Paramuricea type
B3 samples in both the Low and High concentrations (Fig. 2G). The
Kruskal–Wallis and post-hoc tests detected significant health differ-
ences in fragments exposed to all concentrations of the mixture
relative to the controls (po0.05).
3.1.4. Comparisons between treatments for Paramuricea type B3
For comparisons made between treatments in the bulk-exposure
series, the log-rank test revealed significant differences among the
K–Msurvivalestimates(
χ
¼7.62, df¼2, p¼0.022); pairwise compar-
isons (Table 1 ) indicated these differences were between the oil and
oil/dispersant treatments (po0.0167). The oil/dispersant treatment
hadthelowestmeanestimatedsurvivaltimeof87.6h,comparedto
the overall mean estimate of 90.2 h (Table 2a, Fig. 3). In the WAF
exposures there were also significant differences among time-to-
event occurrences (
χ
¼57.3, df ¼2, po0.001), and pair-wise compar-
isons affirmed significantly different estimates between all treat-
ments. The lowest time-to-event estimate was 82.5 h in dispersant
comparedto96hinoilandanoverallaverageestimateof90.8h
(Tabl e 2b, Fig. 3).
Fig. 1. Average health ratings over time for coral fragments exposed to various concentrations of bulk-oil mixtures (yellow/ top row), Corexit 9500A dispersant solutions
(blue/ middle row) and oil–dispersant (oil/disp.) combination mixtures (red/ bottom row). Health rating scale 0–5. Bars show standard error.
D.M. DeLeo et al. / Deep-Sea Research II 129 (2016) 137–147140
3.2. Exposure effects on C. delta
3.2.1. Oil treatment
There was no complete fragment mortality in the control or bulk-
oil treatments (Fig. 1B). However, one C. delta replicate in the Low
oil–WAFdiedbytheendoftheexposure(Fig. 2B). The Kruskal–Wallis
test showed no significant differences among the 96-h health ratings
across all concentrations of bulk and WAF oil (p40.05).
3.2.2. Dispersant treatment
C. delta showed a decline in health in the High dispersant
(Fig. 1E), though complete fragment mortality was not observed
during the 96 h assay. The Kruskal–Wallis test revealed significant
differences (po0.05) in health rankings, with the High dispersant
showing a significantly greater decline in health than the Medium
and Low concentrations (po0.05).
During the WAF exposures, 75% of C. delta fragments died in the
Low dispersant, 25% in the Medium and 75% in the High dis-
persant after 96 h (Fig. 2E). Control fragment health was signifi-
cantly higher relative to all concentrations of dispersant (po0.05).
3.2.3. Oil/dispersant treatment
Coral fragments also showed a decline in health within the
High oil/dispersant treatment (Fig. 1H), but again complete frag-
ment mortality was not observed. Significant differences were
detected between nubbins in the High oil/dispersant relative to
the control samples (po0.05).
During the DE–WAF exposures, mortality was observed in one
colony in the Medium concentration and three of the four colonies
in the High concentration (Fig. 2H). A Kruskal–Wallis test revealed
significant health differences among treatments, with the Medium
and High DE–WAF treatments significantly lower than the controls
(po0.05), and the High DE–WAF also significantly lower than the
Low treatment (po0.05).
Fig. 2. Average health ratings over time for coral fragments exposed to various concentrations of water accommodated oil fractions (yellow/ top row), Corexit 9500A
dispersant solutions (blue/ middle row) and water accommodated oil–dispersant (oil/disp.) combination mixtures (red/ bottom row). Health rating scale 0–5. Bars show
standard error.
Table 1
Pair-wise comparisons of K–M survival estimates in oil, dispersant and oil/
dispersant treatments within the bulk-oil and oil–WAF exposure series, using a
Mantel–Cox log-rank analysis. Comparisons were done discretely for each of the
three coral species: C. delta,Paramuricea (type) B3 and L. glaberrima (χ
2
¼chi-
square, α¼0.05). The event was a decline in health rating to 3 or below (bulk) or
1 and below (WAF). Bonferroni adjusted p-values for each within species compar-
ison are po0.0167, with values in bold being significant.
Log-rank (Mantel–Cox) Oil Dispersant Oil /dispersant
χ
2
p-Val χ
2
p-Val χ
2
p-Val
Bulk exposures
C. delta
Oil ––1.766 0.184 0.284 0.594
Dispersant 1.766 0.184 ––3.594 0.058
Oil/dispersant 0.284 0.594 3.594 0.058 ––
Paramuricea B3
Oil ––3.958 0.047 10.634 0.001
Dispersant 3.958 0.047 ––2.401 0.121
Oil/dispersant 10.634 0.001 2.401 0.121 ––
L. glaberrima
Oil ––0.152 0.696 7.364 0.007
Dispersant 0.152 0.696 ––6.919 0.009
Oil/dispersant 7.364 0.007 6.919 0.009 ––
WAF exposures
C. delta
Oil ––14.127 0.0 00 4.788 0.029
Dispersant 14.127 0.000 ––3.651 0.056
Oil/dispersant 4.788 0.029 3.651 0.056 ––
Paramuricea B3
Oil ––46.594 0.000 8.695 0.003
Dispersant 46.594 0.000 ––25.770 0.000
Oil/dispersant 8.695 0.003 25.770 0.000 ––
L. glaberrima
Oil ––65.367 0.000 45.871 0.000
Dispersant 65.367 0.000 ––6.077 0.014
Oil/dispersant 45.871 0.000 6.077 0.014 ––
D.M. DeLeo et al. / Deep-Sea Research II 129 (2016) 137–147 141
3.2.4. Comparisons between treatments for C. delta
No significant differences were detected among K–Mtime-to-
event estimates for C. delta fragments in all treatments within the
bulk-oil series (
χ
¼1.7 2, d f ¼2, p¼0.422), with an overall time-to-
event (health rating of 3 or less) estimate of 91.4 h (Table 2a, Fig. 3).
However, significant differences were detected among treatment
estimates in the WAF series (
χ
¼12.5 , df ¼2, p¼0.002); these
differences were between the oil-only and dispersant-only treat-
ments (Table 1). The lowest estimate was 89.1 h in the dispersant
treatment relative to the 93.5 h in the oil, and an overall average
time-to-event estimate of 91.6 h (Ta bl e 2b).
3.3. Exposure effects on L. glaberrima
3.3.1. Oil treatment
There was no complete fragment mortality for L. glaberrima
nubbins in the control, bulk-oil (Fig. 1C) or oil–WAF tr eat men t s
(Fig. 2C). However, the Kruskal–Wallis test detected significant differ-
ences (po0.05) among fragment health ratings in bulk-oil mixtures at
96 h; this difference was due to lower rankings in the Medium oil
compared to those in the Low oil (po0.05) and the controls (p¼0.01),
although rankings were similar between the Medium and High oil
concentrations.
There was a significant difference among L. glaberrima health
ratings in the oil–WAF exposure (po0.001); pairwise comparisons
revealed that all concentrations of oil had significantly higher
health ratings than control fragments (pr0.005). Ratings in the
Medium oil–WAF were also significantly higher than the Low and
High (po0.05) oil concentrations.
3.3.2. Dispersant treatment
Whole fragment mortality was not observed for L. glaberrima
samples, though there was a decline in health within the High
dispersant treatment (Fig. 1F). The Kruskal–Wallis test also detected
no significant differences among sample health ratings at 96 h
(p40.05).
During the WAF exposures, L. glaberrima samples in the High and
Medium dispersant concentrations were dead by 72 h. By 96 h four
of six fragments were also dead in the Low dispersant treatment
(Fig. 2F). The Kruskal–Wallis test revealed significantly lower health
ratings in all concentrations of dispersant: Low (po0.05), Medium
and High (po0.005) relative to controls.
3.3.3. Oil/dispersant treatment
Whole fragment mortality was not observed in the bulk-oil/
dispersant mixture (Fig. 1I). The Kruskal–Wallis test and pair-wise
comparisons revealed significant health differences between L.
glaberrima samples in the High and Medium oil/dispersant
(po0.05) and between both the High and Medium concentrations
relative to the control samples (pr0.01).
For L. glaberrima samples in the DE–WAF, t here wa s com ple te
sample mortality in the High concentration by 72 h, with two of six
colonies dead in the Medium DE–WAF ( Fig. 2I). Health ratings for
nubbins in the control and Low DE–WAF were significantly higher
than those in the Medium and High concentrations (po0.005).
3.3.4. Comparisons between treatments for L. glaberrima
The K–M analysis and log rank test revealed significantly
different time-to-event estimates for L. glaberrima samples
(
χ
¼7.20, df¼2, p¼0.027). Pairwise comparisons (Table 1) indi-
cated this difference was between both the oil and dispersant
treatments relative to the oil/dispersant mixture, which had the
shortest time-to-event estimate of 91.0 h compared to an overall
time-to-event estimate of 94.0 h (Table 2a, Fig. 3). Significant
differences were also detected among time-to-event estimates in
the WAF exposures (
χ
¼61.7, df¼2, po0.001) across all treat-
ments. The lowest time-to-event estimate was in the dispersant
treatment (77.8 h) with the highest estimate (96 h) in the oil
treatment and an overall estimate of 86.4 h (Table 2b, Fig. 3).
3.4. Overall comparisons between treatments and concentrations
The Cox regression analysis for the bulk-oil series revealed
significant differences (
χ
¼57.8, df¼5, po0.001) among rates in
health decline (to health-rating 3, 50% survival) among treat-
ments (control, oil, dispersant and oil/dispersant) and concen-
trations (zero, low, medium and high). The High concentration
significantly increased the hazard of reaching a health rating of
3 or below by 2.5 fold relative to control concentrations, but the
Medium concentration did not significantly increase the hazard.
Also, relative to controls, samples in the dispersant had an increased
hazard risk of 2.3 fold, however the hazard increase in the bulk-oil
and oil/dispersant mixture treatments were not significantly differ-
ent from the control treatment (Table 3 ).
Similar regression analyses for the WAF exposures also revealed
significant differences (
χ
¼176.470, df¼7, po0.001) among rates
of health decline between treatments and concentrations. Relative
to the controls, dispersant significantly increased the hazard of
reaching a health rating of 3 or below by 3.4 fold, compared to
4.4 fold in the oil/dispersant treatment; being exposed to oil did not
significantly increase the hazard. In addition, the medium treatment
concentrations significantly increased the hazard by 1.3 fold relative
to the control concentration, whereas the high concentration
increased it by 1.6 fold (Table 4).
Table 2
K–M means for time-to-event estimates for three coral species: C. delta,Paramur-
icea type B3 and L. glaberrima, in bulk-oil (a) and WAF (b) exposures using a
Mantel-Cox Log-rank analysis. The event was a decline in health rating to: a) 3 or
below, b) 1 or below.
Species Treatment Survival
estimate
Std.
error
95% confidence interval
Lower
bound
Upper
bound
(a) Bulk exposure
C. delta Bulk-oil 90.3 1.88 86.6 94.0
Dispersant 93.5 1.31 90.9 96.0
Oil/disp. 90.5 1.77 87.1 94.0
Overall 91.4 0.95 89.6 93.3
Paramuricea B3 Bulk-oil 91.9 2.25 87.5 96.3
Dispersant 91.0 2.04 87.0 95.0
Oil/disp. 87.6 2.50 82.7 92.5
Overall 90.2 1.27 87.7 92.7
L. glaberrima Bulk-oil 95.4 0.57 94.3 96.6
Dispersant 96.0 0.00 96.0 96.0
Oil/disp. 91.0 1.71 87.7 94.4
Overall 94.1 0.62 92.9 95.4
Overall Overall 92.1 0.54 91.0 93.1
(b) WAF exposure
C. delta Oil WAF 93.5 1.52 90.5 96.5
Dispersant 89.1 2.05 85.1 93.1
Oil/disp. 92.3 1.60 89.1 95.4
Overall 91.6 0.99 89.7 93.6
Paramuricea B3 Oil WAF 96.0 0.00 96.0 96.0
Dispersant 82.5 2.45 77.7 87.3
Oil/disp. 94.5 0.88 92.8 96.2
Overall 90.8 0.98 88.9 92.7
L. glaberrima Oil WAF 96.0 0.00 96.0 96.0
Dispersant 77.8 2.63 72.7 83.0
Oil/Disp. 86.4 2.01 82.4 90.3
Overall 86.4 1.23 84.0 88.8
Overall Overall 89.3 0.64 88.1 90.6
D.M. DeLeo et al. / Deep-Sea Research II 129 (2016) 137–147142
3.5. Comparisons between species
Significant differences between K–M survival estimates were
detected within the dispersant treatment as well as the control
treatment during species comparisons for the bulk-oil series
(po0.05; Table 5). The lowest time-to-event estimate of 90.3 h
was for Paramuricea type B3 compared to an overall time-to-event
estimate of 94.4 h. However, adding species into the Cox regression
model did not improve fit, as species’rate of decline comparisons
were not significantly different.
In the WAF exposures, these differences (po0.001) were also
detected among species in the oil and oil/dispersant treatments.
The lowest time-to-event estimate in oil was for Paramuricea type
B3 at 68.8 h relative to 75.9 h for L. glaberrima, 85.2 h for C. delta
and an overall time-to-event estimate of 76.3 h. Paramuricea type
B3 (64.3 h) and L. glaberrima (65.3 h) had similarly low time-to-
event estimates in the oil/dispersant treatment relative to 93.5 h
for C. delta and an overall time-to-event estimate of 73.1 h. The
Cox regression model also indicated that overall, L. glaberrima
did significantly worse than C. delta by 1.3 fold but decline rates for
Paramuricea type B3 were not significantly different from those of
C. delta (Table 4).
4. Discussion
All three deep-sea coral species examined showed more severe
declines in health in response to dispersant alone and the oil–
dispersant mixtures than the oil-only treatments. The experiments
reported here are the first ever to investigate the effects of oil and
dispersant exposure on live, cold-water corals collected from the
deep sea. Impacted corals have been observed at multiple sites in
the deep GoM (Fisher et al., 2014), some covered with floc linked
to oil from the Macondo well explosion (White et al., 2012).
However, the unprecedented application of chemical dispersants
in the deep-sea may have contributed to the observed pattern of
impact. This exposure series provides crucial insight into the
toxicological impacts of oil and dispersant release on three species
of long-lived, habitat forming corals.
Fig. 3. Box-plots showing time-to-event estimates from the Kaplan–Meier survival analysis for coral fragments in three different treatments: oil, dispersant and oil/
dispersant. (Top row represents bulk-oil exposures and bottom row represents oil WAF exposures.) The event was a decline in health rating to 3 or below (bulk) and 1 or
below (WAF). Box ends represent standard error, line inside the box represents the mean and whiskers represent 95% confidence intervals.
Table 3
Log-rank tests of equality on survival distributions for the different levels of
concentration in the bulk-oil and WAF exposure series. Significant p-values
(po0.05) in bold.
Overall log-rank comparisons among concentrations
Treatment Chi-square df Sig.
Bulk-oil series
Control Log rank (Mantel–Cox) –0–
Bulk-oil Log rank (Mantel–Cox) 0.548 2 0.760
Dispersant Log rank (Mantel–Cox) 15.635 2 0.000
Oil/dispersant Log rank (Mantel–Cox) 21.793 2 0.000
WAF–oil series
Control Log rank (Mantel–Cox) –0–
WAF–oil Log rank (Mantel–Cox) 1.190 2 0.552
Dispersant Log rank (Mantel–Cox) 33.246 2 0.000
Oil/dispersant Log rank (Mantel–Cox) 20.061 2 0.000
D.M. DeLeo et al. / Deep-Sea Research II 129 (2016) 137–147 14 3
Regarding the components of the bulk-oil and WAF mixtures,
hydrocarbon concentrations are likely an overestimate, given crude
oil's variable and complex composition, containing thousands of
compounds differing in hydrophobic and hydrophyllic tendencies
(ClarkJr.andBrown,1977;Singeretal.,2000;DiToroetal.,2007).
Dispersants also contain a variety of polar and non-polar surfactants
and solvents (Singer et al., 1996). It is highly probable that there was
adhesion of oil and dispersant constituents to the mixing flasks used
during serial dilutions, as well as to experimental vials. Moreover,
loss of water-accomodated oil fractions may have occurred through
coalescence and surfacing throughout the exposure period (particu-
larly in the bulk-oil exposure), volatilization during aeration, and/or
biodegradation from the microbial communities associated with
coral tissues (Couillard et al., 2005). Thus, it is difficult to determine
the precise concentrations of oil and dispersant that each coral
fragment may encounter at any given time during the course of
the experiment but clearly actual exposures were lower than target
values, making our results conservative estimates of the effects of oil,
dispersant and oil/dispersant mixtures on deep-sea corals. Indeed,
similar trends in health decline were observed within each treatment
for all three species during four separate experimental trials.
The goal of this experiment was not to reproduce the exact
conditions encountered by deep-water corals during the DWH spill,
but rather to provide experimental evidence of their sensitivity to
various concentrations of oil and dispersant. Reproducing exact
conditions encountered by deep-water corals during the DWH spill
is challenging because oil, dispersant and seawater mixtures form
complex multiphase systems; an organism may then be exposed to
many components of the oil and dispersant in various forms
(National Research Council, 1989; Langevin et al., 2004). It is also
important to note that corals within the vicinity of the DWH may
have been exposed to these pollutants for longer than 96 h. Long-
term exposures may see additional effects but were not feasible due
to the time limitations of experimenting at sea. There is also a low
survival rate when transporting deep-sea corals back to laboratory
aquaria; C. delta and L. glaberrima only survive for approximately 1–3
months, whereas we have had no success keeping Paramuricea type
B3 or P. bi s cay a alive over the long-term.
All three species of corals did surprisingly well in the oil
treatments compared to the dispersant and oil/dispersant treat-
ments (Figs. 1 and 2). In some cases, the corals appeared healthier in
both the bulk-oil and oil–WAF treatments relative to the controls
(e.g. C. delta and L. glaberrima,Figs. 1 and 2). Although corals can be
negatively impacted when covered by oil particulates or floc (White
et al., 2012), it is also possible that corals are deriving some form of
nutrition from hydrocarbon components, a process that is likely to
be mediated by their associated microbial communities. Anecdotal
evidence for this linkage comes from the finding of at least one
species of octocoral (C. delta) with increased abundances around
natural hydrocarbon seeps (Quattrini et al., 2013). Previous studies
of shallow-water octocorals also revealed non-selective hydrocar-
bon uptake of dispersed oil droplets into the gastrovascular cavity of
the coral during water uptake (Cohen et al., 1977). Since additional
food sources were not supplied during the exposure experiments,
and most coral fragments within the oil treatments were frequently
observed with a higher degree of polyp extension, similar uptake of
dispersed oil components might have occurred.
Although our present study suggests MASS crude oil was not
toxic over the range of concentrations tested in these experiments
(Figs. 1 and 2), the effect of oil exposure on corals may be dependent
on life-history stage. Crude oil (from the Macondo well) exposures
of scleratinian coral larvae induced mortality within 24 h, while
reducing settlement capabilities and post-settlement survival
(Goodbody-Gringley et al., 2013). This suggests an increased vulner-
ability for coral planulae larvae and juvenile stages, although there
was an influence of larval size on exposure tolerance. Other studies
have shown premature ejection of planula larvae after exposure to
water-soluble fractions of Iranian crude oil (Loya and Rinkevich,
1979) and sub-lethal oil damage to the female reproductive systems
of scleratinian corals (Rinkevich and Loya, 1979). Similar sub-lethal
impacts may have been imposed on cold-water corals exposed to
oil released from the DWH disaster, although these effects may not
be manifested for a number of years.
Treatments containing dispersants in both exposure experi-
ments were the most toxic to the corals and induced the highest
degree of overall fragment mortality (Figs. 1 and 2). As dispersants
tend to increase the surface area of oil–water interactions, they
may cause increased toxicological effects to marine organisms
(Chandrasekar et al., 2006; Goodbody-Gringley et al., 2013). How-
ever, in the WAF exposure series, dispersant-only solutions were
Table 4
Predictor variables in Cox regression analysis and calculated hazard ratios of the
odds of reaching health-ratings of interest (r3) for the Bulk-oil and WAF–oil
exposure series. Significant differences based on Wald test statistics; Low concen-
tration is not present because values are constant or linearly dependent. Hazard
ratios were calculated relative to the control water treatment, 0 mg/L concentration
and the species C. delta, respectively. Significant p-values (po0.05) in bold.
Variable Cox regression variables
Level Wald df pHazard
ratio
Standard
error
Bulk-oil
Treatment Treatment 14.908 3 0.002 ––
Bulk-oil 0.011 1 0.918 0.963 0.355
Dispersant 7.067 1 0.008 2.322 0.317
Oil/disp. 0.123 1 0.725 1.133 0.367
Concentration Concentration 19.279 2 o0.001 ––
Medium conc. 0.272 1 0.602 0.833 0.350
High conc. 10.795 1 0.0 01 2.500 0.279
Species Species 0.745 2 0.689 ––
Paramuricea B3 0.013 1 0.908 1.027 0.231
L. glaberrima 0.645 1 0.422 1.212 0.239
WAF–oil
Treatment Treatment 95.263 3 o0.0 01 ––
WAF–oil 0.026 1 0.872 0.957 0.276
Dispersant 27.407 1 o0.001 3.404 0.234
Oil/disp. 41.745 1 o0.001 4.417 0.230
Concentration Concentration 10.977 2 0.004 ––
Medium conc. 3.912 1 0.048 1.342 0.149
High conc. 10.977 1 0.0 01 1.608 0.143
Species Species 15.701 2 o0.001 ––
Paramuricea B3 1.807 1 0.179 0.817 0.137
L. glaberrima 4.617 1 0.032 1.342 0.151
Table 5
Log-rank tests on equality of survival distributions for all species in the bulk-oil and
WAF exposure series. Significant p-values (po0.05) in bold.
Overall log-rank comparisons among species
Treatment Chi-square df p-Value
Bulk-oil series
Control Log rank (Mantel–Cox) 11.222 2 0.004
Bulk-oil Log rank (Mantel–Cox) 2.496 2 0.287
Dispersant Log rank (Mantel–Cox) 6.622 2 0.036
Oil/dispersant Log rank (Mantel–Cox) 5.709 2 0.058
WAF–oil series
Control Log rank (Mantel–Cox) 60.353 2 o0.001
WAF–oil Log rank (Mantel–Cox) 7.556 2 0.023
Dispersant Log rank (Mantel–Cox) 2.584 2 0.275
Oil/dispersant Log rank (Mantel–Cox) 22.22 2 o0.001
D.M. DeLeo et al. / Deep-Sea Research II 129 (2016) 137–147144
more lethal than the oil/dispersant mixture treatments (as com-
pared to the bulk-oil exposure series), though both treatments
resulted in some mortality (Figs. 1 and 2). Toxicity of dispersants is
typically attributed to membrane disruption and impairment via
surface-active compounds (Abel, 1974; National Research Council,
198 9). Exposure results in increased permeability of biological
membranes, loss of total membrane function and/or osmoregula-
tion (Benoit et al., 1987;Partearroyo et al., 1990). Although Corexit
9500A was created in an attempt to reduce the toxicity of its
predecessors while increasing effectiveness for dispersing more
vicous oils, studies have shown that exposure effects are similar
to older formulations, Corexit 9527 and 9554 (Singer et al., 1991,
1995, 1996), which are now considered toxic to a variety of marine
organisms.
The results from this toxicological assay suggest that dispersant
addition during the ensuing cleanup efforts following the DWH spill
mayhavecausedmoredamagetocold-watercoralsthantheinitial
release of crude oil into the deep sea. Dispersants were toxic at the
higher concentrations tested here, and dispersed oil solutions proved
to be more toxic than untreated oil solutions (Figs. 1 and 2), as has
been found in previous studies (Epstein et al., 2000; Mitchell and
Holdway, 2000;Shafiretal.,2007;Bhattacharyya et al., 2003;
Milinkovitch et al., 2011; Rico-Martinez et al., 2013). The ability of
different types of dispersants to emulsify petroleum hydrocarbon
components into the water column as well as the relative toxicity of
the dispersants and crude oil, contribute to the overall toxicity of
each solution (Epstein et al., 2000). The dispersant and oil/dispersant
treatments were lethal to all three species in this study, particularly
in the WAF exposure series where dispersant concentrations were
higher.
It has been observed in several toxicology studies that dispersant
additon increases the total concentration of polycyclic aromatic
hydrocarbon (PAH) components in surrounding water (Couillard
et al., 2005; Hodson et al., 2007). Specifically, it increases the
concentration of less water-soluble high-molecular-weight PAHs,
some of which induce enzymatic activity (i.e. cytochrome P4501A)
that can metabolize PAHs into toxic forms causing a variety of
detrimental effects (Henry et al., 1997; Billiard et al., 1999; Couillard
et al., 2005). This could explain the more rapid decline in health for
coral fragments exposed to the bulk-oil/dispersant and oil–WAF/
dispersant mixtures, where it was likely that a larger proportion of
crude oil compounds were made biologically available (Couillard
et al., 2005; Schein et al., 2009). Larval exposure experiments on
two species of shallow-water scleractinian corals, using BP Horizon
source oil and Corexit 9500A, showed a significant decrease in
survival and settlement in dispersant solutions and oil–dispersant
mixtures, with complete mortality after exposure to 50–100 ppm
solutions of dispersant (Goodbody-Gringley et al., 2013). In larvae of
hard and soft coral species exposed to dispersants and Egyptian
crude oil, all dispersant treatments were more toxic than the oil-
only treatments with the highest toxicity observed in oil-dispersed
solutions, which also resulted in abnormal development and tissue
degeneration (Epstein et al., 2000).
Despite these results, it is unclear whether short-term exposures
to oil and dispersant have long-term effects. Following brief (24 h)
exposures to Arabian crude oil or dispersed-oil (with Corexit 9527),
there were no significant long-term effects on the yearly in situ
skeletal growth of shallow water, hermatypic corals in the genus
Diploria and Acropora (Dodge et al., 1984; LeGore et al., 1989). Though
variability in growth rates during that year were not measured,
similar experiments using a different scleractinian coral, Porites
furcata, did reveal reduced growth in exposed fragments relative to
controls (Birkeland et al., 1976). This indicates that although short
exposure to oil and dispersant may not be lethal to these corals,
additional sub-lethal impacts are possible, the extent of which need
to be investigated further.
Oil transport to benthic sediments likely occurred through a
variety of pathways after the DWH spill, including direct particu-
late sinking and absorption into marine snow (Passow et al., 2012).
Exposure to oil-filled particulates may be more damaging to corals
then the dissolved hydrocarbon components when additional
stressors are present. As viscous particulates, such as flocs, settle
onto benthic communities, the unavoidable exposure imposes
many risks (Montagna et al., 2013) including the suffocation of
sessile organisms. Floc was likely trapped in the mucous of corals
(White et al., 2012) and may have also triggered the excretion of
excess mucus in an attempt to remove the debris. This is an
energetically costly mechanism, which may lead to reduced health
when coupled to additional environmental stressors (Crossland et
al., 1980; Riegl and Branch, 1995).
In conclusion, exposure to relatively high concentrations of
crude oil does not appear to be as lethal to these species of deep-
sea corals as dispersant and mixtures of hydrocarbons and
dispersant. However, it is possible that a longer exposure to sub-
lethal oil concentrations may cause adverse effects that could not
be observed in this short-term toxicological assay. Further exam-
ination into the relative effectiveness of different types of dis-
persants, coupled to examinations of their relative toxicity, is
required. To improve future response efforts, alternative methods
of oil cleanup are needed, and caution should be used when
applying oil dispersants at depth, as it may induce further stress
and damage to deep-sea ecosystems.
Author contributions
Conceived and designed the experiments: DMD, DVR-R, IBB,
EEC. Performed the experiments: DMD, DVR-R, IBB. Analyzed the
data: DMD, IBB. Wrote the paper: DMD, IBB, EEC.
Acknowledgments
Thanks to the crews aboard the R/V Falkor and E/V Nautilus and
both the ROV Global Explorer and ROV Hercules for their assistance
with sample collections. Particular thanks to C. Fisher for support.
Additional thanks to R. Dannenberg, D. McKean, A. Anderson, S.
Georgian and A. Durkin for assistance at sea and in the lab as well
as the crews of the Cordes, Baums and Fisher labs for offshore
support. This work was supported by the Gulf of Mexico Research
Initiative's “Ecosystem Impacts of Oil and Gas Inputs to the Gulf”
(ECOGIG) program. This is ECOGIG contribution #279 and the data
fall under GRIIDC accession numbers R1 132.136.0004 and
R1 132.136.0005.
References
Abel, P.D., 1974. Toxicity of synthetic detergents to fish and aquatic invertebrates. J.
Fish Biol. 6, 279–298.
Adkins, J.F., Henderson, G.M., Wang, S.-L., O'Shea, S., Mokadem, F., 2004. Growth
rates of the deep-sea scleractinia Desmophyllum cristagalli and Enallopsammia
rostrata. Ear Planet. Sci. Lett. 227, 481–490.
Almeda, R., Wambaugh, Z., Chai, C., Wang, Z., Liu, Z., et al., 2013. effects of crude oil
exposure on bioaccumulation of polycyclic aromatic hydrocarbons and survival
of adult and larval stages of gelatinous zooplankton. PLoS One 8 (10), e74476.
Anderson, J.W., Neff, J.M., Cox, B.A., Tatem, H.E., Hightower, G.M., 1974. Character-
istics of dispersions and water-soluble extracts of crude and refined oils and
their toxicity to estuarine crustaceans and fish. Mar. Biol. 27 (1), 75–88.
Andrews, A.H., Cordes, E.E., et al., 2002. Age, growth and radiometric age validation
of a deep-sea, habitat-forming gorgonian (Primnoa resedaeformis) from the Gulf
of Alaska. Hydrobiologia 471, 101–110 .
Barron, M., 2012. Ecological impacts of the Deepwater Horizon oil spill: implica-
tions for immunotoxicity. Toxicol. Pathol. 40, 315–320.
Bayer, F.M., Cairns, S.D., Cordeiro, R.T.S., Perez, C.D., 2014. New records of the genus
Callogorgia (Anthozoa: Octocorallia) in the western Atlantic, including the
D.M. DeLeo et al. / Deep-Sea Research II 129 (2016) 137–147 14 5
description of a new species. J. Mar. Biol. Assoc. UK. doi: 10.1017/
S0025315414001957 (Available online).
Bhattacharyya, S., Klerks, P.L., Nyman, J.A., 2003. Toxicity to freshwater organisms
from oils and oil spill chemical treatments in laboratory microcosms. Environ.
Pollut. 122, 205–215.
Benoit, J., Cormier, M., Wepierre, J., 1987. Effect of proteins on the assessment of
surfactant cytotoxicity by an in vitro test. Possible correlation with in vitro data.
Toxicol. in Vitro 1, 91–96.
Billiard, S.M., Querbach, K., Hodson, P.V., 1999. Toxicity of retene to early life stages
of two freshwater fish species. Environ. Toxicol. Chem. 18, 2070–2077.
Birkeland, C., Reimer, A.A., Young, J.R., 1976. Survey of Marine Communities in
Panama and Experiments with Oil. US Environmental Protection Agency,
Ecological Research Series, pp. 60013–76028.
Bonsdorff, E., Bakke, T., Pedersen, A., 1990. Colonization of amphipods and
polychaetes to sediments experimentally exposed to oil hydrocarbons. Mar.
Pollut. Bull. 21 (7), 355–358.
Camilli, R., Reddy, C.M., et al., 2010. Tracking hydrocarbon plume transport and
biodegradation at Deepwater Horizon. Science 330, 201–204.
Camilli, R., Di Iorio., D., et al., 2011. Acoustic measurement of the Deepwater
Horizon Macondo well flow rate. Proc. Natl. Acad. Sci. USA 109 (50),
20235–20239.
Chandrasekar, S., Sorial, G., Weaver, J., 2006. Dispersant effectiveness on oil spills –
impact of salinity. ICES J. Mar. Sci. 63, 1418–1430.
Clark Jr., R., Brown, D., 1977. Petroleum: properties and analyses in biotic and
abiotic systems. In: Malins, D. (Ed.), Effects of Petroleum on Arctic and Subarctic
Marine Environments and Organisms, Vol 1 —Nature and Fate of Petroleum.
Academic Press, New York, NY, USA, pp. 1–89.
Cohen, Y., Nissenbaum, A., Eisler, R., 1977. Effects of Iranian crude iil on the Red Sea
octocoral Heteroxenia Fuscescens. Environ. Pollut. 12, 173–186.
Cordes, E.E., McGinley, M.P., et al., 2008. Coral communities of the deep Gulf of
Mexico. Deep-Sea Res. I—Oceanogr. Res. Pap. 55 (6), 777–787.
Cordes, E.E., Cunha, M.R., et al., 2010. The influence of geological, geochemical, and
biogenic habitat heterogeneity on seep biodiversity. Mar. Ecol. 31, 51–65.
Couillard, C.M., Lee, K., Légaré, B., King, T.L., 2005. Effect of dispersant on the
composition of the water-accommodated fraction of crude oil and its toxicity to
larval marine fish. Environ. Toxicol. Chem. 24, 1496–1504.
Coull, B.C., Chandler, G.T., 1992. Pollution and meiofauna: field, laboratory and
mesocosm studies. Oceanogr. Mar. Biol.: Annu. Rev. 30, 191–271.
Crone, T.J., Tolstoy, M., 2010. Magnitude of the 2010 Gulf of Mexico oil leak. Science
330, 634.
Crossland, A.J.,Barnes, D.J., Borowitzka, M.A., 1980. Diurnal lipid and mucus production
in the staghorn coral Acropora acuminata.Mar.Biol.60(2–3), 81–90.
Davies, A.J., Guinotte, J.M., 2011. Global habitat suitability for framework-forming
cold-water corals. PLoS One 6 (4), e18483.
Di Toro, D., McGrath, J.A., Stubblefield, W.A., 2007. Predicting the toxicity of neat
and weathered crude oil: toxic potential and the toxicity of saturated mixtures.
Environ. Toxicol. Chem. 26, 24–36.
Dodge, R., Wyers, S.C., et al., 1984. The effects of oil and oil dispersants on the
skeletal growth of the hermatypic coral Diploria stringosa. Coral Reefs 3,
191–19 8.
Druffel, E.R.M., Griffin, S., et al., 1995. Gerardia: Bristlecone pine of the deep-sea?
Geochim. Cosmochim. Acta 59, 5031–5036.
Doughty, C.L., Quattrini, A.M., Cordes, E.E., 2014. Insights into the population
dynamics of the deep-sea coral genus Paramuricea in the Gulf of Mexico.
Deep-sea research. Part 2. Topical studies in oceanography 99, 71–82.
Druffel, E.R.M., King, L.L., Belastock, R.A., Buesseler, K.O., 1990. Growth rate of a
deep-sea coral using 210Pb and other isotopes. Geochim. Cosmochim. Acta 54,
149 3–15 0 0.
Dubansky, B., Whitehead, A., Miller, J.T., Rice, C.D., Galvez, F., 2013. Multitissue
molecular, genomic, and developmental effects of the Deepwater Horizon oil
spill on resident gulf killifish (Fundulus grandis). Environ. Sci. Technol. 47,
5074–5082.
Emiliani, C., Hudson, J.H., Shinn, E.A., George, R.Y., 1978. Oxygen and carbon isotopic
growth record in a reef coral from the Florida Keys and a deep-sea coral from
Blake Plateau. Science 202, 627–629.
Epstein, N., Bak, R.P.M., Rinkevich, B., 2000. Toxicity of third generation dispersants
and dispersed Egyptian crude oil on Red Sea coral larvae. Mar. Pollut. Bull. 40
(6), 497–503.
Fisher, C.R., Hsing, P.-Y., Kaiser, C., Yoerger, D., Roberts, H.H., Shedd, W., Cordes, E.E.,
Shank, T.S., Berlet, S.P., Saunders, M., Larcom, E.A., Brooks, J., 2014. Footprint of
Deepwater Horizon blowout impact to deep-water coral communities. Proc.
Natl. Acad. Sci. 111, 11744–11749 .
Fodrie, F.J., Heck Jr., K.L., 2011. Response of coastal fishes to the Gulf of Mexico oil
disaster. PLoS One 6 (7), 1–7.
Freiwald, A., Fossa, J.H., Grehan, A., Koslow, T., Roberts, J.M., 2004. Cold-Water Coral
Reefs. UNEP-WCMC, Cambridge, UK p. 84.
Goodbody-Gringley, G., Wetzel, D.L., et al., 2013. Toxicity of Deepwater Horizon
source oil and the chemical dispersant, Corexit9500, to coral larvae. PLoS One 8
(1), e45574.
Grigg, R.W., 1974. Growth rings: annual periodicity in two gorgonian corals. Ecology
55, 876–881.
Hazen, T., Dubinsky, E.A., et al., 2010. Deep-sea oil plume enriches indigneous oil-
degrading bacteria. Science 330, 204–206.
Henry, T.R., Spitsbergen, J.M., Hornung, M.W., Abnet, C.C., Peterson, R.E., 1997. Early
life stage toxicity of 2,3,7,8-tetrachlorobenzo-p-dioxin in zebrafish (Danio rerio).
Toxicol. Appl. Pharmacol., 142; , pp. 56–68.
Hodson, P., Khan, C., et al., Alkyl PAH in crude oil cause chronic toxicity to early life
stages of larval fish. In: Proceedings of the 28th Arctic and Marine Oilspill
Program (AMOP) Technical Seminar, Environment Canada, Edmonton, AL, 2007,
pp. 291–300.
Husebo, A., Nottestad, L., Fossa, J.H., Furevik, D.M., Jorgensen, S.B., 2002. Distribu-
tion and abundance of fish in deep-sea coral habitats. Hydrobiologia 471,
101–11 0 .
Joye, S.B., MacDonald, I.R., Leifer, I., Asper, V., 2011. Magnitude and oxidation
potential of hydrocarbon gases released from the BP oil well blowout. Nat.
Geosci. 4, 160–164.
Kaplan, E.L., Meier, P., 1958. Nonparametric estimation from incomplete observa-
tions. J. Am. Stat. Assoc. 53, 457–485.
Kessler, J.D., Valentine, D.L., et al., 2011. A persistent oxygen anomaly reveals the
fate of spilled methance in the deep Gulf of Mexico. Science 331, 312–315.
Langevin, D., Poteau, S., Hénaut, I., Argillier, J.F., 2004. Crude oil emulsion properties
and their application to heavy oil transportation. Oil Gas Sci. Technol.—Rev. IFP
59 (5), 511–521.
LeGore, S., Marszalek, D.S., Danek, L.J., Tomlinson, M.S., Hofmann, J.E., 1989. Effect of
chemically dispersed oil on Arabian Gulf corals: a field experiment. In:
Proceedings of 1989 Oil Spill Conference, Hunter Environmental Services, Inc.,
Gainsville, pp. 375–380.
Loya, Y., Rinkevich, B., 1979. Abortion effect of corals induced by oil pollution. Mar.
Ecol.—Prog. Ser. 1, 77–80.
Lunden, J.J., Turner, J.M., McNicholl, C.G., Glynn, C.K., Cordes, E.E., 2014. Design,
development, and implementation of recirculating aquaria for maintenance and
experimentation of deep-sea corals and associated fauna. Limnol. Oceanogr.
Methods 12, 363–372.
Milinkovitch, T., Kanan, R., Thomas-Guyon, H., Le Floch, S., 2011. Effects of
dispersed oil exposure on the bioaccumulation of polycyclic aromatic hydro-
carbons and the mortality of juvenile Liza ramada. Sci. Total Environ. 409,
1643–1650.
Minerals Management Service, 2009. Gulf of Mexico Oil and Gas Production
Forecasts: 2009–2018. OCS Report, MMS.
Mitchell, F.M., Holdway, D.A., 2000. The acute and chronic toxicity of the
dispersants Corexit 9527 and 9500, Water Accommodated Fraction (WAF) of
Crude Oil and Dispersant Enhanced WAF (DEWAF) to Hydra viridissima (green
hydra). Water Res. 34, 343–348.
Montagna, P.A., Baguley, J.G., et al., 2013. Deep-sea benthic footprint of the
Deepwater Horizon blowout. PLoS One 8 (8), e70540.
National Research Council, 1989. In: Using Oil Dispersants on the Sea. National
Academy Press, Washington, DC (Chapter 3).
National Research Council, 2005. Oil Spill Dispersants: Efficacy and Effects. The
National Academies Press, Washington, DC.
Opresko, D.M., Baron-Szabo, R.C., 2001. Re-descriptions of the antipatharian corals
described by E. J. C. Esper with selected English translations of the original
German text. Senckenberg. Biol. 81, 1–21.
Pace, C.B., Clark, J.R., Bragin, G.E., 1995. Comparing crude oil toxicity under standard
and environmentally realistic exposures. In: Proceedings of the 1995 Interna-
tional Oil Spill Conference, American Petroleum Institute, Washington, DC,
pp. 1003–1004.
Partearroyo, M.A., Ostolaza, H., Goni, F.M., Barberá-Guillem, E., 1990. Surfactant-
induced cell toxicity and cell lysis. Biochem. Pharmacol. 40, 1323–1328.
Passow, U., Ziervogel, K., Asper, V., Diercks, A., 2012. Marine snow formation in the
aftermath of the Deepwater Horizon oil spill in the Gulf of Mexico. Environ. Res.
Lett. 7, 035301.
Quattrini, A.M., Georgian, S.E., et al., 2013. Niche divergence by deep-sea octocorals
in the genus Callogorgia across the continental slope of the Gulf of Mexico. Mol.
Ecol. 22,4123–414 0.
Reddy, C.M., Arey, J.S., et al., 2012. Composition and fate of gas and oil released to
the water column during the Deepwater Horizon oil spill. Proc. Natl. Acad. Sci.
USA 109 (50), 20229–20234.
Rico-Martinez, R., Snell, T.W., Shearer, T.L., 2013. Synergistic toxicity of Macondo
crude oil and dispersant Corexit 9500A
s
to the Brachionus plicatilis species
complex (Rotifera). Environ. Pollut. 173, 5–10.
Riegl, B., Branch, G.M., 1995. Effects of sediment on the energy budgets of four
scleractinian (Bourne 1900) and five alcyonacean (Lamouroux 1816) corals.
J. Exp. Mar. Biol. Ecol. 186 (2), 259–275.
Rinkevich, B., Loya, Y., 1979. Laboratory experiments on the effects of crude oil on
the Red Sea coral Stylophora pistillata. Mar. Pollut. Bull. 10, 328–330.
Risk, M.J., Heikoop, J.M., Snow, M.G., Beukens, R., 2002. Lifespans and growth
patterns of two deep-sea corals: Primnoa resedaeformis and Desmophyllum
cristagalli. Hydrobiologia 471, 125–131.
Roark, E.B., Guilderson, T.P., Dunbar, R.B., Fallon, S.J., Mucciarone, D.A., 2009.
Extreme longevity in proteinaceous deep-sea corals. Proc. Natl. Acad. Sci. USA
106, 5204–5208.
Roberts, J.M., Wheeler, A.J., Freiwald, A., 2006. Reefs of the deep: the biology and
geology of cold-water coral ecosystems. Science 312 (5773), 543–547.
Schein, A., Scott, J.A., Mos, L., Hodson, P.V., 2009. Chemical dispersant increases the
apparent bioavailability and toxicity of diesel to rainbow trout (Oncorhynchus
mykiss). Environ. Toxicol. Chem. 28, 595–602.
Silliman, B.R., van de Koppel, J., et al., 2012. Degradation and resilience in Louisiana
salt marshes after the BP–Deepwater Horizon oil spill. Proc. Natl. Acad. Sci. USA
109, 11234–11239.
Singer, M.M., George, S., et al., 1996. Comparison of acute aquatic effects of the oil
dispersant Corexit 9500 with those of other Corexit series dispersants.
Ecotoxicol. Environ. Saf. 35 (2), 183–189.
D.M. DeLeo et al. / Deep-Sea Research II 129 (2016) 137–147146
Singer, M.M., Aurand, D., et al., 2000. Standardization of the preparation and
quantitation of water-accommodated fractions of petroleum fortoxicity testing.
Mar. Pollut. Bull. 40 (11), 1007–1016.
Singer, M.M., George, S., et al., 1995. Acute toxicity of the oil dispersant Corexit
9554 to marine organisms. Ecotoxicol. Environ. Saf. 32, 81–86.
Singer, M.M., Smalheer, D.L., Tjeerdema, R.S., Martin, M., 1991. Effects of spiked
exposure to an oil dispersant on the early life stages of four marine species.
Environ. Toxicol. Chem. 10, 1367–1374.
Shafir, S., VanRijn, J., Rinkevich, B., 2007. Short and long term toxicity ofcrude oil
and oil dispersants to two representative coral species. Environ. Sci. Technol.
41, 5571–5574.
Stark, J.S., Snape, I., Riddle, M.J., 2003. The effects of petroleum hydrocarbon and
heavy metal contamination of marine sediments on recruitment of Antarctic
soft-sediment assemblages: a field experimental investigation. J. Exp. Mar. Biol.
Ecol. 283 (1–2), 21–50.
White, H., Hsing, P., et al., 2012. Impact of the Deepwater Horizon oil spill on a deep-
water coral community in the Gulf of Mexico. Proc. Natl. Acad. Sci. USA 109 (50),
20303–20308.
Whitehead, A., Dubansky, B., et al., 2012. Genomic and physiological footprint of the
Deepwater Horizon oil spill on resident marsh fishes. Proc. Natl. Acad. Sci. USA 109
(50), 20298–20302.
D.M. DeLeo et al. / Deep-Sea Research II 129 (2016) 137–147 147
