Environmental Pollution 316 (2023) 120450

Contents lists available at ScienceDirect

Environmental Pollution
journal homepage: www.elsevier.com/locate/envpol

Simulating oil-driven abundance changes in benthic marine invertebrates

using an ecosystem model☆
L.N. Dornberger a, P.A. Montagna b, C.H. Ainsworth a, *
a
College of Marine Science. University of South Florida. 140 7th St. Petersburg, FL. 33701, USA
b
Harte Research Institute. Texas A&M University-Corpus Christi, 6300 Ocean Drive, Corpus Christi, TX, 78412-5869, USA

A R T I C L E I N F O A B S T R A C T

Keywords: Field studies showed that benthic macrofauna and meiofauna abundances increased with sediment oil concen­
DWHOS tration in areas affected by the Deepwater Horizon (DWH) oil spill. Benthic invertebrate biomass shows a dome-
Oil spill shaped relationship with respect to petrogenic hydrocarbon concentrations suggesting a positive effect on
Atlantis
biomass at low-to-medium oil concentrations and a negative effect at high concentrations. If this is due to
Gulf of Mexico
MOSSFA
enrichment of the benthic food web, then this adds to an emerging picture of a food web response over a large
Ecosystem based management spatial area with both abundance increases and decreases as a result of DWH. We would be obliged to consider
long term multispecies effects beyond the initial pulse disturbance in modeling impacts and recovery of
economically valuable species. An Atlantis ecosystem model of the Gulf of Mexico is used to simulate three
mechanisms that could explain observed changes in the invertebrate community. Scenario 1 is that stimulation of
surface primary productivity occurred as a result of nutrient loading caused by diversion of Mississippi River
water into Barataria Bay (a mitigation action taken during the DWH oil spill). Scenario 2 is that enrichment of the
benthos occurred due to detrital loading from marine oil snow sedimentation and flocculent accumulation
(MOSSFA). Scenario 3 is that predator declines and/or avoidance of oiled areas caused a release of predation
mortality on benthic invertebrates. Scenario 2 (MOSSFA) stimulated the detritus-driven food web and was best
able to cause a net increase in invertebrate biomass despite a realistic amount of oil toxicity. Scenario 3 (predator
release) plausibly could have contributed to changes in benthic invertebrates. Scenario 1 (nutrient loading) had
little impact on the benthos suggesting the benthic food web is decoupled from local pelagic production sources.

1. Introduction policies impacted the rate of deposition and affected benthic ecology.
The increased flow contributed to deposition of sediment and hydro­
The Deepwater Horizon (DWH) oil spill in the Gulf of Mexico (GOM) carbons near De Soto Canyon and other areas (White et al., 2012;
released an estimated 3.2–4.6 million barrels of oil (Griffiths, 2012, US Montagna et al., 2013; Valentine et al., 2014; Brooks et al., 2015;
vs BP Exploration and Production Inc., 2015). Mitigation actions taken Chanton et al., 2015; Romero et al., 2015). The mass sediment flux to the
to reduce the impact of the spill included diversions of freshwater from seafloor is from marine oil snow formation during and after the spill
the Mississippi River, surface oil burns, and dispersant applications at (Brooks et al., 2015) through a pathway now known as marine oil snow
the surface and wellhead (Kujawinski et al., 2011; Lubchenco et al., sedimentation and flocculent accumulation (MOSSFA, Daly et al., 2016;
2012; McNutt et al., 2012). The freshwater outflow from the Mississippi Passow, 2016). Indeed, 0.5–47% of the oil accumulated on the seafloor
River, a large source of sediment and particulates in the region (Corbett (Valentine et al., 2014; Chanton et al., 2015, Romero et al., 2017) in a
et al., 2006; Bianchi et al., 2007), was elevated above normal levels spotty distribution across the continental shelf, slope, and onto the deep
between May and October 2010 due to intentional releases of water ocean seafloor. Potential effects of MOSSFA include direct toxicity from
meant to reduce the influx of oil into Barataria Bay (Bianchi et al., 2011; crude oil and its by-products, and smothering and shading of the
Kourafalou and Androulidakis, 2013). It is potentially the case therefore benthos. With the potentially large volume of oil that reached the sea­
that mitigation decisions made during DWHOS and river management floor, chronic effects could manifest in demersal communities due to

☆
This paper has been recommended for acceptance by Maria Cristina Fossi.
* Corresponding author.
E-mail address: [email protected] (C.H. Ainsworth).

https://doi.org/10.1016/j.envpol.2022.120450
Received 27 July 2022; Received in revised form 8 October 2022; Accepted 14 October 2022
Available online 20 October 2022
0269-7491/© 2022 Published by Elsevier Ltd.
L.N. Dornberger et al. Environmental Pollution 316 (2023) 120450

resuspension of oil from disturbance events such as storms (Latimer abundance through predation release or diet shifts (Gray, 1990; Martin,
et al., 1999; Yang et al., 2008; Roberts, 2012). 2017; Ainsworth et al., 2018). Specific to the GOM and DWH, common
Montagna et al. (2013) compared effects of oil release on meiofauna fish invertivores of interest include grunts (Haemulidae), porgies
and macrofauna communities at the DWH wellhead and at natural seeps (Sparidae), snappers (Lutjanidae), and triggerfish (Balistidae) (Dance
in the same vicinity. What they found were diversity reductions in et al., 2011).
soft-bottom benthic meiofauna and macrofauna in the 24 km2 sur­ This study aimed first to explore whether a domed relationship
rounding the DWH wellhead, with more moderate effects observed as (Fig. 1) exists between oil concentrations and benthic invertebrate
distance from the wellhead increased. They found a significant corre­ abundance by examining sediment core data (Montagna et al., 2013).
lation between this effect and oil indicators, as well as distance from the Total petrogenic hydrocarbons (TPH) and Barium concentrations were
wellhead. No comparable correlation was found between diversity and both tested as predictors of abundance. Secondly, this study tested three
oil indicators associated with natural hydrocarbon seeps. A follow-up mechanisms that could potentially produce a domed oil response
study conducted the subsequent year revealed persistent contamina­ (Fig. 1), where invertebrate biomass is elevated at medium oil concen­
tion and continued declines in biodiversity (Montagna et al., 2017). This trations. The three scenarios simulate nutrient loading from the Mis­
result suggests that the system had long-term community structure sissippi River, benthic detritus loading from MOSSFA, and release from
shifts. predation. These scenarios were chosen because they could potentially
The direct and indirect effects of oil exposure on benthic-dwelling explain why there are increases in invertebrate abundance at medium oil
organisms varies with species group and habitat. Fishes are negatively concentrations, and because Scenarios 1 and 2 involve processes that are
affected by oil toxins, such as polycyclic aromatic hydrocarbons (PAHs). influenced by the oil spill mitigation decisions. We used an ‘end-to-end’
PAHs can reduce growth and fecundity in fishes and are correlated with ecosystem model, Atlantis, to examine effects on the food web (Fulton
increased prevalence of lesions and cancers (Collier et al., 2013; Mur­ et al., 2004, 2011; Ainsworth et al., 2015). Other applications of the Gulf
awski et al., 2014; Vignet et al., 2014; Brown-Peterson et al., 2015; of Mexico Atlantis model (GOM Atlantis) in studying oil spills can be
Hedgpeth & Griffitt, 2016; Herdter et al., 2017). found in Ainsworth et al. (2018), Ortega-Ortiz et al., 2020, Dornberger
Some populations of organisms may actually benefit from oil expo­ et al. (2020), and Morzaria-Luna et al. (2018). Food web effects have
sure through organic enrichment of the food web (Jewett et al., 1999) or been shown to be significant following the initial oil spill pulse distur­
other mechanisms. Studies of natural oil seeps reveal toxic effects of bance, and may potentially increase recovery time and translate impacts
exposure on benthic invertebrates within seeps, but microbial and of the oil spill to distant fish populations and fisheries (Ainsworth et al.,
meiofaunal enrichment at seep peripheries (Montagna & Spies, 1985; 2018; Ortega-Ortiz et al., 2020).
Montagna et al., 1986, 1987). These findings suggest a nuanced dynamic In Scenario 1 (nutrient loading), Mississippi nutrient loading stim­
between toxicity and enrichment. The pioneer experiment by Spies et al. ulates primary production in surface waters. High river flow and
(1988) examined dose-responses of meiofauna and macrofauna abun­ intentional discharges of freshwater diversion systems such as Davis
dance in sediments containing various concentrations of oil. They found Pond and Caernarvon into Barataria Bay may have contributed to
a pronounced domed response in macrofaunal abundance, with abun­ elevated nutrient inputs during DWH (Bianchi et al., 2011; Hu et al.,
dance being highest at intermediate concentrations of oil and declining 2011, O’Connor et al., 2016). This was potentially exacerbated by
as oil concentrations increased (Fig. 1). They determined that direct oil storms (O’Connor et al., 2016). In Scenario 2 (MOSSFA), increased
exposure had little effect on meiofaunal abundance, although other detritus inputs from MOSSFA stimulates the detritus-driven benthic food
studies have shown elevated meiofaunal mortality associated with oil web (Brooks et al., 2015; Romero et al., 2015). In Scenario 3 (predation
exposure (Ustach, 1979; McLachlan and Harty, 1982; Frithsen et al., release), benthic predators are either killed by oil or avoid the area,
1985; Danovaro et al., 1995; Fleeger et al., 2018). reducing predation mortality on invertebrates (Gray, 1990; Martin,
Additional studies have found similar domed-shaped relationships in 2017; Ainsworth et al., 2018). Scenarios 1 and 2 explain the invertebrate
relation to natural oil seeps (Montagna et al., 1986; Palmer et al., 1988; biomass increase through bottom-up mechanisms and involve, respec­
Montagna et al., 1995). Benthic macrofauna abundances at oil seep sites tively, enrichment of phytoplankton- and detritus-based food webs.
in the Gulf of Mexico are higher than adjacent non-seep sites, but there Scenario 3 explains the invertebrate biomass increase via a top-down
are no community or bioindicators that are unique to seeps (Washburn mechanism. Each scenario was tested in GOM Atlantis to determine
et al., 2018). A leading explanation is that in the presence of oil, the which mechanisms could cause an increase in invertebrate biomass with
abundance of oil-degrading microbes increases, providing additional and without realistic oil toxicity present. Ecosystem network charac­
food for meiofauna, which in turn are food for macrofauna (Olsgard & teristics help us understand the implications of these mechanisms on
Gray, 1995; Steichen et al., 1996; Jewett et al., 1999). Alternatively, food web dynamics (Ulanowicz, 1986, 2000). If enrichment of the
top-down shifts through fish physiological (e.g., reduced growth and benthic food web occurs then this implies a nuanced ecosystem response
fecundity, increased mortality) and predator declines or behavioral re­ to oil spills. We would be obliged then to look beyond the initial pulse
sponses to oil exposure could indirectly boost benthic invertebrate disturbance in predicting impacts and recovery.

2. Methods

2.1. Analysis of benthic invertebrate relationship

Invertebrate abundance data are based on sediment cores from 68

stations in the northern GOM (Montagna et al., 2013). Sampling stations
were arranged in a radial pattern extending up to 125 miles from the
wellhead (UAC, 2010). Two cruises were conducted in September and
October 2010. Water depths ranged from 76 m to 2767 m. Sediment
samples were collected using an OSIL multicorer (Montagna et al.,
2013). Montagna et al. assess meiofaunal and macrofaunal abundances
and other biotic and chemical data such as TPH. Chemical contaminant
data was based on the top 3 cm of the sediment and it was collected at
Fig. 1. A conceptual model of the potential domed relationship between the same time as the infauna data in 2010. Methods for sampling and
invertebrate abundance and environmental oil concentration. chemical analyses are described in UAC (2010) and at https://www.

2
L.N. Dornberger et al. Environmental Pollution 316 (2023) 120450

nodc.noaa.gov/deepwaterhorizon/ship.html. were used. For scenario 1 (nutrient loading), Mississippi surface nutrient
These data were used to compare invertebrate abundances against loading rates were targeted to increase for silica, nitrate, and ammonium
TPH values. However, relationships of invertebrate abundances with surface loading rates (each tracked by Atlantis) by a factor of six
water depth could be a confounding influence. We examined the rela­ compared to a control simulation. This rate is based approximately on
tionship between water depth and benthic invertebrate abundances. relative flow rates from the Davis Pond Sediment Diversion at points
Abundances were log transformed and regressed against depth using during and after the spill (Bianchi et al., 2011). We apply this rate to six
linear model least squares residuals. Because the negative correlations polygons surrounding the Mississippi River Delta to produce an
with depth were significant (meiofauna: R2 = 0.09, p < 0.05, macro­ area-averaged response. Inshore polygons of Atlantis implicitly include
fauna: R2 = 0.24, p < 0.0001), two other key environmental variables both estuarine and shelf processes. Results from nutrient addition in this
were examined against depth and each other: TPH and barium. TPH was scenario will stimulate production in the surface depth layer. Results are
the measured oil exposure variable, and barium was another potential sensitive to an uptake process specified in Atlantis and other variables so
oil-related variable that could be used to describe the domed relation­ an additional Scenario 1 simulation was run increasing the nutrient
ship, as barium is found in relation to offshore drilling operations loading rate up to thousands of times greater to help map the model
(Kennicutt et al., 1996). TPH and barium were tested against each other, response to this variable. Another example of nutrient forcing in Atlantis
and each were tested against depth to determine tendencies of covari­ is provided by Bossier et al. (2018).
ation using Spearman’s rank correlation. Because TPH and barium were For scenario 2 (MOSSFA), detrital loading rates on the seafloor were
significantly positively correlated (ρ = 0.48, p < 0.001), TPH alone was increased by a factor of 2–5 in the seven polygons closest to the wellhead
used in model building. Depth was not correlated with oil variables, so (Brooks et al., 2015). The elevated nutrient and detrital input rates were
partial residuals were generated from a fitted linear model to remove the maintained constant for 167 days then returned to normal. This is
effect of depth (Rex et al., 2006). To preserve the appropriate scale for consistent with the duration of the DWH oil spill (April 20th and July
the abundances, the y-intercept from the depth regression was added to 15th, 2010) and comparable to previous simulations studies (Ainsworth
the depth-controlled data (Rex et al., 2006). et al., 2018, Morzaria-Luna et al., 2018). There are other potential ef­
To test the dome relationship, the depth controlled benthic inverte­ fects of MOSSFA we are not modeling including smothering and light
brate abundances were fit to four models. We used sediment core data effects on phytobenthos. For scenario 3 (predator release), we repre­
where TPH values were measured and present. The constant, linear, sented predation release by removing benthic predators from oiled
exponential and quadratic models are provided in Eqs. (1)–(4). polygons using a spatial mortality effect applied in proportion to oil
concentration by Atlantis polygon and depth layer. This approach re­
log(abundance) = a (1)
duces predator biomass wherever oil is present and so could serve as a
first order approximation of predator declines due to oil impacts and/or
log(abundance) = log(TPH) + a (2)
predator avoidance of oiled areas. The predator mortality forcing lasted
for the duration of the oil spill. Functional groups were considered a
log(abundance) = b ∗ e(log(TPH∗c)) (3)
benthic predator fish if their diet consisted of 20% or more from benthic
groups. An additional scenario 3 simulation was run removing all fishes
log(abundance) = b ∗ log(TPH)2 + c ∗ log(TPH) + a (4)
from the oiled region (not just benthic predators) to help map the model
Here, a represents log of abundance with no oil effect, and b and c are response to this variable.
modifiers for the rate of change of abundance. All models were built and To help us understand how new production is routed through the
fitted in R using least squared residuals for both macro- and meiofauna ecosystem, we calculate network characteristics throughput, ascen­
data sets. The models were then assessed for parsimony using AIC dency, and overhead (Ulanowicz, 1986, 2000). Ascendency (A) is a
(Akaike, 1974). We calculate Akaike weight (ωAIC), which can be measure of average mutual constraint in the ecosystem. It represents the
interpreted as the probability that the model would be the best fit under size and organization of the ecosystem in information-theoretic terms. It
repeated sampling among all the models tested (Pechenik, 1987). is estimated by Ulanowicz and Norden (1990) (eq. (5)):
∑ ( )
Tij T
2.2. Atlantis simulations A= Ti,j log (5)
i,j
Ti Tj
Atlantis is a spatially explicit, three-dimensional, deterministic, end- Ascendancy increases with total system throughput (T), the sum of
to-end ecosystem model (Fulton et al., 2004, 2011). The GOM imple­ material flows (Tij ) between prey (i) to predator (j) (eq. (6)). Prey
mentation is described in Ainsworth et al. (2015) with recent im­ biomass (Bi ) and predation mortality (Mij ) are taken from Atlantis.
provements in Morzaria-Luna et al. (2018). Hydrodynamics are
provided by the American Seas model based on the National Research Tij = Bi • Mij (6)
Laboratory developed NCOM model (Martin, 2000). These affect
High ascendency ecosystems have dominant pathways of material
nutrient cycling, primary production, and organism physiology and
flow and are associated with mature ecosystems containing many spe­
distribution. Atlantis polygon geometry is determined based on biore­
cialists. These ecosystems make efficient use of available primary pro­
gional features and fisheries management areas. The GOM imple­
duction and host large biomasses but are brittle because they have few
mentation uses up to six water column depth layers per polygon and a
alternative pathways of material exchange around major nodes. Low
sediment layer, and 91 functional groups. Of these, 61 are
ascendency ecosystems have smaller exchanges and more densely con­
age-structured vertebrates and invertebrates, 19 are non-age-structured
nected nodes due to the presence of generalists. They incur thermody­
invertebrates, 6 are primary producers, 2 are bacteria and 3 are detritus.
namic losses due to the presence of inefficient horizontal exchanges. The
Each functional group uses sub-models describing consumption, pro­
quantity overhead (Φ) measures the disordered information that a sys­
duction, respiration, reproduction, and movement. Vertebrates and
tem retains (eq. (7)). A system with high overhead retains many ineffi­
some exploited invertebrates are tracked by numbers and body weight,
cient pathways for material flow and can easily restructure after
while non-exploited invertebrates are only tracked as biomass pools. We
calamity.
therefore will report on invertebrate biomass changes here, not abun­ ( )
dance changes. Ainsworth et al. (2015) describes how the model was ∑ Tij2
validated against observational catch and biomass data. A review of the φ= − Tij log (7)
i,j
Ti Tj
system equations is available in Link et al. (2011).
To drive biomass increases in scenarios 1 and 2, point source inputs Each of the three scenarios were first simulated in Atlantis in the

3
L.N. Dornberger et al. Environmental Pollution 316 (2023) 120450

absence of oil toxicity to determine if the tested mechanisms were indicate an abundance peak at higher TPH concentrations than macro­
capable of inducing benthic invertebrate biomass increases. Biomass faunal abundances.
increase was determined by comparing the scenarios at the polygon
level to a control simulation of the Gulf of Mexico, with no enrichment 3.2. Atlantis simulations
or oil mortality. Each scenario was run with and without oil toxicity.
These were applied in conjunction with oil concentration estimates. We Of the three scenarios, scenario 2 (MOSSFA) showed the largest in­
calculate oil forcing functions on the invertebrates in the same way as crease in invertebrate biomass, scenario 3 (predation release) showed an
Ainsworth et al. (2018) and Dornberger et al. (2020) via a spatial intermediate increase, and scenario 1 (nutrient loading) showed a small
mortality scaler updated at 24-h time steps. Resolved at the level of increase (Fig. 3A). Even though nutrient loading resulted in little stim­
polygons and depth layers, it represents a realistic rendition of the ulation of the benthic food web, it had a large effect on the ecosystem
Deepwater Horizon oil spill. Our interest here is to maximize the heu­ overall as indicated by a large change in total throughput (Fig. 3B).
ristic value of the Atlantis model by simulating a realistic set of oil spill Ascendency for scenario 1 is also high, which suggests the uptake and
conditions. The mortality function is based on a dose-response model incorporation of nutrients into primary producers and the food web
(Dornberger et al., 2016) and oil densities from CIMAGE Lagrangian oil happens in a small number of high-volume exchanges between Atlantis
fate simulations from the Connectivity Modeling System (CMS: Paris functional groups (Fig. 3C). Ascendency responds to eutrophication
et al., 2012, 2013; Aman & Paris, 2013). CMS applies tri-cubic inter­ (Patrício et al., 2006), although the result is sensitive to Atlantis group
polation of circulation velocities from the Hybrid Coordinate Ocean structure. The pelagic primary producers (large phytoplankton (LPP)
Model (Chassignet et al., 2003) accounting for factors like buoyancy, and small phytoplankton (SPP)) are aggregated high biomass groups.
tidal transport and turbulence. Toxicity spatial forcing was generated for These primary producers benefit from the nutrient uptake more directly
the 17 benthic associated invertebrate functional groups using toxico­ than do consumers, which tend to be more resolved in the group
logical functions based on fish lesion frequencies (Dornberger et al., structure. Since there was little effect of nutrient loading on in­
2016). Assumptions are that invertebrates have similar sensitives to oil vertebrates, the benthic food web appears decoupled from local pelagic
as fish, that oil concentrations below a threshold of lethality have no production. Overhead in Fig. 3D supports the idea that there are a set of
impact on the population, and that oil concentrations above that stimulated pelagic pathways of exchange unique to scenario 1 which are
threshold cause a log-linear increase in mortality. Oil effect thresholds not present in the control or in scenarios 2 or 3. Whereas the nutrient
tested were 100, 200, 300, 500, 1000, 2000, and 3000 ppb. The calcu­ stimulation affected a small number of high biomass groups, the detritus
lation is performed for each polygon and depth layer and at 24-h in­ stimulation of scenario 2 affects a larger number of relatively low
crements. We estimate the exposure of benthic fauna to oil in a manner biomass detritivorous fish and invertebrate groups involved in
similar to Ainsworth et al. (2018). Those authors integrated over each low-volume exchanges. This helps explains why scenario 2 (MOSSFA)
depth layer the cumulative oil concentration in the water column from has a lower ascendency than scenario 1 (nutrient loading) (Fig. 3D). In
the oil fate model and applied a scaler of 1000. This scaler represents scenario 3 (predation release) the food web is not being enriched as it is
concentration of oil on the benthic surface and was determined by in scenarios 1 and 2, yet there is an increase in ascendency. This reflects
comparing sedimentary oil deposition data with oil fate data (Romero prey increases ongoing at the end of the 1-year simulation.
et al., 2015). Simulations in Atlantis were one-year duration. When these three scenarios were combined with oil toxicity effects,
there was a tendency for the mortality effect to overwhelm whatever
3. Results increases in biomass were attributable to each mechanism under study.
Of all of the oil effect thresholds used in simulations, only scenario 2
3.1. Benthic invertebrate abundance and TPH relationship (MOSSFA) demonstrated both increased biomass in some polygons and
decreased biomass (due to toxicity) in other polygons (Fig. 4), and then
The results of fitting the established linear model of abundance by only at the lowest level of oil effect, the 3000 ppb threshold. In a more
depth (reviewed by Rowe, 1983; Rex et al., 2006) indicated that for this extreme simulation, when all fish were removed from the oiled region
data set, depth had a significant (R2 = 0.24, p < 0.0001 for macrofauna, (not just benthic predators), no polygons increased in invertebrate
and R2 = 0.09, p < 0.05 for meiofauna) relationship with log abundance biomass except one Atlantis functional group, infaunal meiobenthos
of both groups, consistent with past research (reviewed by Rowe, 1983; INF. Based on these results, only scenario 2 (MOSSFA) produced poly­
Rex et al., 2006). TPH and Barium were strongly correlated (ρ = 0.48, p gons representing both the ascending and descending limbs of the
< 0.001), which is why Barium was excluded from further modeling domed relationship (Fig. 5).
efforts. For TPH and depth, the Spearman’s ρ was − 0.02 and the p value In scenario 2, the benthic food web is stimulated by the addition of
was 0.87, indicating that these variables are not correlated and would detritus. The importance of shrimp is reduced in the diets of benthic
act independently on abundance. The results from the linear and predators like mullet MUL as diets shift toward to other benthic inver­
nonlinear least squares fitted models for TPH against both macrofaunal tebrate prey resources. Stimulation of the benthic detrital food web
and meiofaunal depth-controlled abundances from core samples moves across the model through neighboring polygons, the blue poly­
selected the quadratic equation as the best fit with the lowest sum of gons in the southwest GOM in scenario 2 are mostly a result of these
squared residuals, as well as the most parsimonious based on the AIC shrimp biomass increases (Fig. 4). Oysters were removed from Fig. 4, as
results (Table 1). A domed relationship exists between oil concentration their biomass increase was orders of magnitude greater than other
and benthic invertebrate abundance (Fig. 2). Meiofaunal abundances functional groups, but that result is likely sensitive to unmodeled as­
sumptions on salinity and sedimentation. As the only scenario indicating
both increases and decreases in biomass based on oil concentration,
Table 1 scenario 2 invertebrate functional groups’ biomass was plotted against
Sum of squared residuals (SS) and Akaike Information Criterion (AIC) results for simulated oil exposure to determine if the expected dome shape was
macrofauna and meiofauna invertebrate abundances, ωAIC is Akaike weight. present, and it is for many groups (Fig. 5). Other groups indicated an
Macrofauna Meiofauna inverted dome reflecting indirect trophic dynamic responses to the
Model SS AIC ωAIC SS AIC ωAIC benthic enrichment.
Constant 2.542 − 10.243 2.97E-05 6.310 41.313 1.22E-04
Linear 2.539 − 8.311 1.13E-05 5.195 32.811 0.00859 4. Discussion
Exponential 2.423 − 10.929 4.19E-05 5.898 39.674 2.78E-04
Parabola 1.631 − 31.089 1 4.197 23.296 1
This study aligns with previous field-based results indicating that

4
L.N. Dornberger et al. Environmental Pollution 316 (2023) 120450

Fig. 2. Benthic invertebrate abundances against sediment total petrogenic hydrocarbon sediment concentrations, with best fit quadratic models. The data points
show abundance partial residuals after removing the depth effect, with the intercept added back to values for scale preservation. The fitted line shows the quadratic
model. A) Macrofaunal abundances in numbers per square meter. B) Meiofaunal abundances in numbers per 10 square centimeters. Raw data from Montagna
et al. (2013).

Fig. 3. Atlantis invertebrate biomass and ecosystem

network characteristics for scenarios 1 to 3 (nutrient
loading, MOSSFA, and predation release). Results are
presented after a one-year simulation relative to a no
enrichment control. A) Average invertebrate biomass
change was as calculated using the six common
polygons where direct forcing occurred in each sce­
nario. The oyster functional group was removed from
the MOSSFA scenario. B) total ecosystem throughput,
C) ascendency, D) overhead.

5
L.N. Dornberger et al. Environmental Pollution 316 (2023) 120450

Fig. 4. Benthic invertebrate density differences in oil scenarios 1–3 compared to a control simulation with no scenario forcings or toxicity. A) Mississippi nutrient
loading and toxicity, B) Increased detritus from MOSSFA and toxicity, oyster functional group removed as an extreme outlier, C) Predators altered behavior to avoid
oil exposure and toxicity. Results are summed across functional groups, from October 2010 in the 3000 ppb threshold simulation. Blue polygons indicate a net
increase in biomass, red polygons indicate a net decrease as a result of the scenario drivers. (For interpretation of the references to colour in this figure legend, the
reader is referred to the Web version of this article.)

Fig. 5. Invertebrate functional group biomass differences in scenario 2 (MOSSFA) relative to a no oil control. Threshold of oil effect is 3000 ppb. The dotted lines
were added to illustrate possible domed relationships; BSH – Brown Shrimp, WSH – White Shrimp, PSH – Pink Shrimp, OSH – Other Shrimp, CMB – Carnivorous
Macrobenthos, INF – Infaunal Meiobenthos, OYS – Oysters, SES – Sessile Filter Feeders, BIV – Bivalves.

exposure to oil can cause increases in some benthic invertebrate pop­ to possess a domed relationship with respect to oil concentration. The
ulations (Spies et al., 1988; Steichen et al., 1996; Olsgard and Gray, meiofauna abundance maximum occurred at a higher oil concentration
1995; Jewett et al., 1999; Guarinello et al., 2022) and cause a transition than the macrofauna abundance maximum, possibly as a result of tro­
in benthic ecosystem function towards heterotrophy (Lee and Lin, phic interactions. Of the three mechanisms examined here, 1) nutrient
2013). The Montagna et al. (2013) invertebrate abundances were shown loading, 2) MOSSFA, and 3) predation release, all three were able to

6
L.N. Dornberger et al. Environmental Pollution 316 (2023) 120450

cause an increase in biomass and so represent plausible mechanisms to model salinity or sedimentation effects (Powers et al., 2017). Never­
explain observed invertebrate increases. Nutrient loading seems least theless, the modeled increase in oyster biomass and decline in sessile
likely to have contributed as we see evidence of large amounts of new filter feeders demonstrates the sensitivity of the food web in these
production elsewhere in the ecosystem and a weak linkage between the polygons to environmental factors. Other potential factors not repre­
benthic food web and local pelagic production. Linkage between the sented in these simulations include bioaccumulation, the potential
benthic and pelagic food webs may be sensitive to environmental chemical effects of dispersants, benthic oxygen limitation, and the
changes over the long term, and also short-term variability, especially by-products of the surface oil burns (Kujawinski et al., 2011; Lubchenco
affecting stratification and food concentrating mechanisms. The envi­ et al., 2012; McNutt et al., 2012). These could be important aspects to
ronment can also motivate changes in migration and feeding behavior in consider in further simulations.
fish (Bertrand et al., 2020).
Our findings consistent with the case where there are somewhat 5. Conclusions
independent pelagic and benthic food webs. In such cases, ecosystem
ascendency can be used to test for the presence of distinct pathways Nutrient loading and detritus input rates may be affected by de­
since it increases with the heterogeneity and magnitude of internal flows cisions made in response to an oil spill, as they were with DWH. A
(Ainsworth and Walters, 2015). Atlantis was able to produce a large realization emerged from DWH study that the MOSSFA phenomenon
invertebrate increase from detritus loading in scenario 2. Only scenario may be common in deltaic systems and could be expected to accompany
2 could cause an increase in invertebrate biomass when matched with a deep water oil spills (Murk et al. 2015). Enrichment processes stimulate
realistic amount of oil toxicity, and then only at the lowest oil effect the ecosystem from the bottom-up, while the ecosystem may already be
tested. Nevertheless, these findings validate MOSSFA as a key pathway in a state of flux from having lost predators to oil mortality (Lewis et al.,
of oil transport to benthic ecosystems (Brooks et al. 2015; Daly et al., 2020). Thus, this study highlights the need to consider multispecies ef­
2016) to the extent that our model can represent this complex process. fects in modeling oil spill impacts and recovery. MOSSFA had a large
Predation release (scenario 3) creates a smaller increase in benthic food impact on the benthic food web in our study, reaffirming that it is an
web exchanges when predators are removed from the system. important component in understanding the ecosystem response to the
Using the same dataset examined in the current modeling work, DWH oil spill (Brooks et al., 2015; Romero et al., 2015). It is possible
Balthis et al. (2017) created toxicological response models that that any or all three of these tested mechanisms contributed to the
confirmed that concentrations of TPH and PAH near the DWH wellhead response of benthic communities. These realizations can be added to a
were sufficient to cause a change in taxa richness because of a loss of growing appreciation of the nuanced outcomes of mitigation decisions
species sensitive to oil exposure. Forcing toxicity across invertebrate that has emerged since the oil spill.
functional groups, instead of just those identified from the sampling in
Montagna et al. (2013), allowed the Atlantis model to determine which Author statement
groups demonstrate a domed response.
Addition of nutrients at the Mississippi River Delta had the smallest L. Dornberger and C. Ainsworth both contributed to conceptualiza­
effect on invertebrate populations, and never produced a domed rela­ tion, formal analysis of invertebrate and model data, and writing the
tionship in Atlantis, even at the highest nutrient levels. For this reason, manuscript. P. Montagna was responsible for data curation and formal
scenario 1 must be eliminated as a major driver of short-term (<1 y) analysis of the invertebrate data.
benthic invertebrate biomass increases in the Atlantis GOM oil spill
model. This suggests that surface nutrient source input alone associated
with high flow rates and any potential mitigation actions involving di­ Declaration of competing interest
versions is less likely to explain the observed effects on benthic in­
vertebrates from the DWH oil spill. The authors declare that they have no known competing financial
For scenario 3, only benthic predators were removed from testing to interests or personal relationships that could have appeared to influence
avoid confounding food web effects. Although benthic predator removal the work reported in this paper.
caused slightly greater increases in invertebrate biomass than scenario
1, any increase in invertebrate biomass was overwhelmed by the oil Data availability
toxicity effect so there was not a domed relationship. This suggests that
predator decreases or behavior of benthic predators to avoid oil expo­ Data will be made available on request.
sure is less likely to be a major driver of benthic invertebrate abundance
increases. This effect alone is not sufficient to recreate the observed Acknowledgements
invertebrate responses.
Enrichment of the detrital food web appears to be the best candidate Funding for LD was provided by the Garrels Memorial Fellowship in
as a potential driver of benthic invertebrate population increase. This Marine Science, the Jack and Katharine Ann Lake Fellowship, the Wells
fits with other research indicating that increased oil provides food for Fargo Fellowship in Marine Science, the USF University Graduate
microbes, which in turn provide food for meiofauna, then macrofauna Fellowship, and Center for Integrated Modeling and Analysis of Gulf
(Steichen et al., 1996; Olsgard & Gray, 1995; Jewett et al., 1999). The Ecosystems (C-IMAGE II). C-IMAGE II is funded by a grant from the Gulf
functional group biomass increases supported the pattern seen in the of Mexico Research Initiative (GRI2001-I-072). Funding for CA and for
data from Montagna et al. (2013) for both macrofauna and meiofauna. model development was provided by C-IMAGE II.
However, not all macrofaunal functional groups had biomass increases.
The sessile filter feeders functional group indicated a concave relation­ References
ship with oil concentration, suggesting non-linear effects are caused by
food web interactions. Oysters, which had a uniquely large biomass Ainsworth, C., Walters, C., 2015. Ten common mistakes made in Ecopath with Ecosim
increase and mid-oil concentrations, could be outcompeting the other modelling. Ecol. Model. 308, 14–17.
Ainsworth, C.H., Schirripa, M.J., Morzaria-Luna, H. (Eds.), 2015. An Atlantis Ecosystem
sessile filter feeding functional group, which would otherwise benefit Model for the Gulf of Mexico Supporting Integrated Ecosystem Assessment. US Dept.
from increased detritus in the water. Comm. NOAA Technical Memorandum NMFS-SEFSC-676, p. 149.
It should be noted that our oyster dynamics do not match observed Ainsworth, C.H., Paris, C., Perlin, N., Dornberger, L.N., Patterson, W., Chancellor, E.,
Murawski, S., Hollander, D., Daly, K., Romero, I., Coleman, F., Perryman, H., 2018.
trends (Powers et al., 2017). In fact, under the conditions of the simu­ Impacts of the Deepwater Horizon oil spill evaluated using an end-to-end ecosystem
lation, the increase in oyster biomass is not unexpected since we did not model. PLoS One 13 (1), e0190840, 2018 Jan 25.

7
L.N. Dornberger et al. Environmental Pollution 316 (2023) 120450

Akaike, H., 1974. A new look at the statistical model identification. IEEE Trans. Automat. Herdter, E.S., Chambers, D.P., Stallings, C.D., Murawski, S.A., 2017. Did the deepwater
Control 19 (6), 716–723. Horizon oil spill affect growth of red snapper in the Gulf of Mexico? Fish. Res. 191,
Aman, Z.M., Paris, C.B., 2013. Response to comment on “Evolution of the Macondo well 60–68.
blowout: simulating the effects of the circulation and synthetic dispersants on the Hu, C., Weisberg, R.H., Liu, Y., Zheng, L., Daly, K.L., English, D.C., Zhao, J., Vargo, G.A.,
subsea oil transport”. Environ. Sci. Technol. 47 (20), 11906–11907. 2011. Did the northeastern Gulf of Mexico become greener after the Deepwater
Balthis, W.L., Hyland, J.L., Cooksey, C., Montagna, P.A., Baguley, J.G., Ricker, R.W., Horizon oil spill? Geophys. Res. Lett. 38 (9), L09601.
Lewis, C., 2017. Sediment quality benchmarks for assessing oil-related impacts to the Jewett, S.C., Dean, T.A., Smith, R.O., Blanchard, A., 1999. ’Exxon Valdez’ oil spill:
deep-sea benthos. Integrated Environ. Assess. Manag. 13, 840–851. impacts and recovery in the soft-bottom benthic community in and adjacent to
Bertrand, A., Lengaigne, M., Takahashi, K., Avadí, A., Poulain, F., Harrod, C., 2020. El eelgrass beds. Mar. Ecol. Prog. Ser. 185, 59–83.
Niño Southern Oscillation (ENSO) Effects on Fisheries and Aquaculture. FAO, Rome. Kennicutt II, M.C., Boothe, P.N., Wade, T.L., Sweet, S.T., Rezak, R., Kelly, F.J., Brooks, J.
FAO Fisheries and Aquaculture Technical Paper No. 660. M., Presley, B.J., Wiesenburg, D.A., 1996. Geochemical patterns in sediments near
Bianchi, T.S., Galler, J.J., Allison, M.A., 2007. Hydrodynamic sorting and transport of offshore production platforms. Can. J. Fish. Aquat. Sci. 53 (11), 2554–2566.
terrestrially derived organic carbon in sediments of the Mississippi and Atchafalaya Kourafalou, V.H., Androulidakis, Y.S., 2013. Influence of Mississippi River induced
Rivers. Estuarine. Coastal Shelf Sci. 73 (1), 211–222. circulation on the Deepwater Horizon oil spill transport. J. Geophys. Res.: Oceans
Bianchi, T.S., Cook, R.L., Perdue, E.M., Kolic, P.E., Green, N., Zhang, Y., Smith, R.W., 118 (8), 3823–3842.
Kolker, A.S., Ameen, A., King, G., Ojwang, L.M., 2011. Impacts of diverted Kujawinski, E.B., Kido Soule, M.C., Valentine, D.L., Boysen, A.K., Longnecker, K.,
freshwater on dissolved organic matter and microbial communities in Barataria Bay, Redmond, M.C., 2011. Fate of dispersants associated with the Deepwater Horizon oil
Louisiana, USA. Mar. Environ. Res. 72 (5), 248–257. spill. Environ. Sci. Technol. 45 (4), 1298–1306.
Bossier, S., Palacz, A.P., Nielsen, J.R., Christensen, A., Hoff, A., Maar, M., Gislason, H., Latimer, J.S., Davis, W.R., Keith, D.J., 1999. Mobilization of PAHs and PCBs from in-
Bastardie, F., Gorton, B., Fulton, E.A., 2018. The Baltic Sea Atlantis: an integrated place contaminated marine sediments during simulated resuspension events. Estuar.
end-to-end modelling framework evaluating ecosystem-wide effects of human- Coast Shelf Sci. 49 (4), 577–595.
induced pressures. PLoS One 13 (7), e0199168. Lee, L., Lin, H., 2013. Effects of an oil spill on benthic community production and
Brooks, G.R., Larson, R.A., Schwing, P.T., Romero, I., Moore, C., Reichart, G.J., respiration on subtropical intertidal sandflats. Marine Pollution Bulletin 73 (1),
Jilbert, T., Chanton, J.P., Hastings, D.W., Overholt, W.A., Marks, K.P., 2015. 291–299.
Sedimentation pulse in the NE Gulf of Mexico following the 2010 DWH blowout. Lewis, J.P., Tarnecki, J.H., Garner, S.B., Chagaris, D.D., Patterson, W.F., 2020. Changes
PLoS One 10 (7), e0132341. in reef fish community structure following the Deepwater Horizon oil spill. Science
Brown-Peterson, N.J., Krasnec, M., Takeshita, R., Ryan, C.N., Griffitt, K.J., Lay, C., Reports 10, 5621, 2020.
Mayer, G.D., Bayha, K.M., Hawkins, W.E., Lipton, I., Morris, J., 2015. A multiple Link, J.S., Gamble, R.J., Fulton, E.A., 2011. NEUS—Atlantis: Construction, Calibration,
endpoint analysis of the effects of chronic exposure to sediment contaminated with and Application of an Ecosystem Model with Ecological Interactions, Physiographic
Deepwater Horizon oil on juvenile Southern flounder and their associated Conditions, and Fleet Behavior. NOAA Technical Memorandum, NMFS NE, p. 218.
microbiomes. Aquat. Toxicol. 165, 197–209. Lubchenco, J., McNutt, M.K., Dreyfus, G., Murawski, S.A., Kennedy, D.M., Anastas, P.T.,
Chanton, J.P., Zhao, T., Rosenheim, B., Joye, S.B., Bosman, S., Brunner, C., Yeager, K., Chu, S., Hunter, T., 2012. Science in support of the deepwater Horizon response.
Hollander, D., 2015. Radiocarbon tracing of the flux of petrocarbon to the sea floor Proc. Natl. Acad. Sci. USA 109 (50), 20212–20221.
associated with the Deepwater Horizon event. Environ. Sci. Technol. 49, 847–854. Martin, P.J., 2000. Description of the Navy Coastal Ocean Model Version 1.0, 749 NRL/
Chassignet, E.P., Smith, L.T., Halliwell, G.R., Bleck, R., 2003. North Atlantic simulation FR/7322-00-9962. Naval Research Laboratory, p. 42.
with the HYbrid Coordinate Ocean Model (HYCOM): Impact of the vertical Martin, C.W., 2017. Avoidance of oil contaminated sediments by estuarine fishes. Mar.
coordinate choice, reference density, and thermobaricity. Journal of Physical Ecol. Prog. Ser. 576, 125–134.
Oceanography. McLachlan, A., Harty, B., 1982. Effects of crude oil on the supralittoral meiofauna of a
Collier, T.K., Anulacion, B.F., Arkoosh, M.R., Dietrich, J.P., Incardona, J.P., Johnson, L. sandy beach. Mar. Environ. Res. 7 (1), 71–79.
L., Ylitalo, G.M., Myers, M.S., 2013. Effects on fish of polycyclic aromatic McNutt, M.K., Chu, S., Lubchenco, J., Hunter, T., Dreyfus, G., Murawski, S.A.,
hydrocarbons (PAHs) and naphthenic acid exposures. Fish Physiol.: Organ. Chem. Kennedy, D.M., 2012. Applications of science and engineering to quantify and
Toxicol. Fishes: Fish Physiol. 33, 195–255. control the Deepwater Horizon oil spill. Proc. Natl. Acad. Sci. USA 109 (50),
Corbett, D.R., McKee, B., Allison, M., 2006. Nature of decadal-scale sediment 20222–20228.
accumulation on the western shelf of the Mississippi River delta. Continent. Shelf Montagna, P.A., Spies, R.B., 1985. Meiofauna and chlorophyll associated with Beggiatoa
Res. 26 (17), 2125–2140. mats of a natural submarine petroleum seep. Mar. Environ. Res. 16 (4), 231–242.
Daly, K.L., Passow, U., Chanton, J., Hollander, D., 2016. Assessing the impacts of oil- Montagna, P.A., Bauer, J.E., Prieto, M.C., Hardin, D., Spies, R.B., 1986. Benthic
associated marine snow formation and sedimentation during and after the metabolism in a natural coastal petroleum seep. Mar. Ecol. Prog. Ser. 31–40.
Deepwater Horizon oil spill. Anthropocene 13, 18–33. Montagna, P.A., Bauer, J.E., Toal, J., Hardin, D., Spies, R.B., 1987. Temporal variability
Dance, M.A., Patterson, I.I.I., William, F., Addis, D.T., 2011. Fish community and trophic and the relationship between benthic meiofaunal and microbial populations of a
structure at artificial reef sites in the northeastern Gulf of Mexico. Bull. Mar. Sci. 87 natural coastal petroleum seep. J. Mar. Res. 45 (3), 761–789.
(3), 301–324. Montagna, P.A., Bauer, J.E., Hardin, D., Spies, R.B., 1995. Meiofaunal and microbial
Danovaro, R., Fabiano, M., Vincx, M., 1995. Meiofauna response to the Agip Abruzzo oil trophic interactions in a natural submarine hydrocarbon seep. Oceanogr. Lit. Rev. 10
spill in subtidal sediments of the Ligurian Sea. Mar. Pollut. Bull. 30 (2), 133–145. (42), 882.
Dornberger, L., Ainsworth, C., Gosnell, S., Coleman, F., 2016. Developing a polycyclic Montagna, P.A., Baguley, J.G., Cooksey, C., Hartwell, I., Hyde, L.J., Hyland, J.L.,
aromatic hydrocarbon exposure dose-response model for fish health and growth. Kalke, R.D., Kracker, L.M., Reuscher, M., Rhodes, A.C., 2013. Deep-sea benthic
Mar. Pollut. Bull. 109 (1), 259–266. footprint of the Deepwater Horizon blowout. PLoS One 8 (8), e70540.
Dornberger, L.N., Ainsworth, C.H., Coleman, F., Wetzel, D.L., 2020. A synthesis of top Montagna, P.A., Baguley, J.G., Cooksey, C., Hyland, J.L., 2017. Persistent impacts to the
down and bottom up impacts of the Deepwater Horizon oil spill using ecosystem deep soft-bottom benthos one year after the Deepwater Horizon event. Integrated
modeling. In: Murawski, S.A., Ainsworth, C., Gilbert, S., Hollander, D., Paris, C.B., Environ. Assess. Manag. 13 (2), 342–351.
Schlüter, M., Wetzel, D. (Eds.), Deep Oil Spills – Facts, Fate and Effects. Springer Morzaria-Luna, H., Ainsworth, C.H., Tarnecki, J., Grüss, A., 2018. Diet composition
(Chapter 31). uncertainty determines impacts on fisheries following an oil spill. Ecosyst. Serv. 33,
Fleeger, J.W., Riggio, M.R., Mendelssohn, I.A., Lin, Q., Hou, A., Deis, D.R., 2018. 187–198.
Recovery of saltmarsh meiofauna six years after the Deepwater Horizon oil spill. Murawski, S.A., Hogarth, W.T., Peebles, E.B., Barbeiri, L., 2014. Prevalence of external
J. Exp. Mar. Biol. Ecol. 502, 182–190. skin lesions and polycyclic aromatic hydrocarbon concentrations in Gulf of Mexico
Frithsen, J.B., Elmgren, R., Rudnick, D.T., 1985. Responses of benthic meiofauna to long- fishes, post-Deepwater Horizon. Trans. Am. Fish. Soc. 143 (4), 1084–1097.
term, low-level additions of No. 2 fuel oil. Mar. Ecol. Prog. Ser. 1–14. O’Connor, B.S., Muller-Karger, F.E., Nero, R.W., Hu, C., Peebles, E.B., 2016. The role of
Fulton, E.A., Parslow, J.S., Smith, A.D.M., Johnson, C.R., 2004. Biogeochemical marine Mississippi River discharge in offshore phytoplankton blooming in the northeastern
ecosystem model II: the effect of physiological detail on model performance. Ecol. Gulf of Mexico during August 2010. Rem. Sens. Environ. 173, 133–144.
Model. 173, 371–406. Olsgard, F., Gray, J.S., 1995. A comprehensive analysis of the effects of offshore oil and
Fulton, E.A., Link, J.S., Kaplan, I., Savina-Rolland, M., Johnson, P., Ainsworth, C.H., gas exploration and production on the benthic communities of the Norwegian
Horne, P., Gorton, R., Gamble, R.J., Smith, A.D.M., Smith, D.C., 2011. Lessons in continental shelf. Mar. Ecol. Prog. Ser. 122, 277–306.
modelling and management of marine ecosystems: the Atlantis experience. Fish Fish. Ortega-Ortiz, J.G., Ainsworth, C.H., Gracia, A., 2020. Comparing ecosystem model
12 (2), 171–188. outcomes between Ixtoc 1 and Deepwater Horizon oil spills. In: Murawski, S.A.,
Gray, R.H., 1990. Fish behavior and environmental assessment. Environ. Toxicol. Chem. Ainsworth, C., Gilbert, S., Hollander, D., Paris, C.B., Schlüter, M., Wetzel, D. (Eds.),
9 (1), 53–67. Deep Oil Spills – Facts, Fate and Effects. Springer (Chapter 32).
Griffiths, S.K., 2012. Oil release from Macondo well MC252 following the Deepwater Palmer, M.A., Montagna, P.A., Spies, R.B., Hardin, D., 1988. Meiofauna dispersal near
Horizon accident. Environ. Sci. Technol. 46 (10), 5616–5622. natural petroleum seeps in the Santa Barbara Channel: a recolonization experiment.
Guarinello, M.L., Sturdivant, S.K., Murphy, A.E., Brown, L., Godbold, J.A., Solan, M., Oil Chem. Pollut. 4 (3), 179–189.
Carey, D.A., Gemano, J.D., 2022. Evidence of Rapid Functional Benthic Recovery Paris, C.B., Hénaff, M.L., Aman, Z.M., Subramaniam, A., Helgers, J., Wang, D.P.,
Following the Deepwater Horizon Oil Spill. ACS EST Water. https://doi.org/ Kourafalou, V.H., Srinivasan, A., 2012. Evolution of the Macondo well blowout:
10.1021/acsestwater.2c00272. Aug 31 2022. simulating the effects of the circulation and synthetic dispersants on the subsea oil
Hedgpeth, B.M., Griffitt, R.J., 2016. Simultaneous exposure to chronic hypoxia and transport. Environ. Sci. Technol. 46 (24), 13293–13302.
dissolved polycyclic aromatic hydrocarbons results in reduced egg production and Paris, C.B., Helgers, J., Van Sebille, E., Srinivasan, A., 2013. Connectivity Modeling
larval survival in the sheepshead minnow (Cyprinodon ariegatus). Environ. Toxicol. System: a probabilistic modeling tool for the multi-scale tracking of biotic and
Chem. 35 (3), 645–651. abiotic variability in the ocean. Environ. Model. Software 42, 47–54.

8
L.N. Dornberger et al. Environmental Pollution 316 (2023) 120450

Passow, U., 2016. Formation of rapidly-sinking, oil-associated marine snow. Deep Sea UAC (Unified Area Command), 2010. Deepwater Horizon MC 252 Response Unified Area
Res. Part II Top. Stud. Oceanogr. 129, 232–240. Command – Strategic Plan for Sub-sea and Sub-surface Oil and Dispersant Detection,
Patrício, J., Ulanowicz, R., Pardal, M.A., Marques, J.C., 2006. Ascendency as ecological Sampling, and Monitoring. U.S. Coast Guard and BP Exploration and Production,
indicator for environmental quality assessment at the ecosystem level: a case study. Inc., New Orleans, LA. USA. November 13, 2010 Final.
Hydrobiologia 555, 19–30. Ulanowicz, R.E., 1986. Growth and Development: Ecosystems Phenomenology. Springer-
Pechenik, J.A., 1987. Environmental influences on larval survival and development. Verlag, NY, p. 203.
Reproduction of Marine Invertebrates 9, 551–608. Ulanowicz, R.E., 2000. Ascendancy: a measure of ecosystem performance. In:
Powers, S.P., Grabowski, J.H., Roman, H., Geggel, A., Rouhani, S., Oehrig, J., Baker, M., Jørgensen, S.E., Müller, F. (Eds.), Handbook of Ecosystem Theories and
2017. Consequences of large-scale salinity alteration during the Deepwater Horizon Management. Lewis Publishers, Boca Raton, pp. 305–315.
oil spill on subtidal oyster populations. Mar. Ecol. Prog. Ser. 576, 175–187. Ulanowicz, R.E., Norden, J.S., 1990. Growth and Development: Ecosystems
Rex, M.A., Etter, R.J., Morris, J.S., Crouse, J., McClain, C.R., Johnson, N.A., Stuart, C.T., Phenomenology. Excel Press, Lincoln (NE), p. 203.
Deming, J.W., Thies, R., Avery, R., 2006. Global bathymetric patterns of standing Ustach, J.F., 1979. Effects of sublethal oil concentrations on the copepod, Nitocra affinis.
stock and body size in the deep-sea benthos. Mar. Ecol. Prog. Ser. 317, 1–8. Estuaries 2 (4), 273–276.
Roberts, D.A., 2012. Causes and ecological effects of resuspended contaminated Valentine, D.L., Fisher, G.B., Bagby, S.C., Nelson, R.K., Reddy, C.M., Sylva, S.P., Woo, M.
sediments (RCS) in marine environments. Environ. Int. 40, 230–243. A., 2014. Fallout plume of submerged oil from Deepwater Horizon. Proc. Natl. Acad.
Romero, I.C., Schwing, P.T., Brooks, G.R., Larson, R.A., Hastings, D.W., Ellis, G., Sci. USA 111 (45), 15906–15911.
Goddard, E.A., Hollander, D.J., 2015. Hydrocarbons in deep-sea sediments following Vignet, C., Le Menach, K., Mazurais, D., Lucas, J., Perrichon, P., Le Bihanic, F., Devier, M.
the 2010 Deepwater Horizon blowout in the northeast Gulf of Mexico. PLoS One 10 H., Lyphout, L., Frère, L., Bégout, M.L., Zambonino-Infante, J.L., 2014. Chronic
(5), e0128371. dietary exposure to pyrolytic and petrogenic mixtures of PAHs causes physiological
Romero, I.C., Toro-Farmer, G., Diercks, A.R., Schwing, P., Muller-Karger, F., disruption in zebrafish-part I: survival and growth. Environ. Sci. Pollut. Control Ser.
Murawski, S., Hollander, D.J., 2017. Large-scale deposition of weathered oil in the 21 (24), 13804–13817.
Gulf of Mexico following a deep-water oil spill. Environmental Pollution. Washburn, T.W., Demopoulous, A.W.J., Montagna, P.A., 2018. Macrobenthic infaunal
Environmental Pollution 228, 179–189. communities associated with deep-sea hydrocarbon seeps in the northern Gulf of
Rowe, G.T., 1983. Benthic Production and Processes off Baja California, Northwest Africa Mexico. Mar. Ecol. 39 (3), e12508.
and Peru: a Classification of Benthic Subsystems in Upwelling Ecosystems. White, H.K., Hsing, P.Y., Cho, W., Shank, T.M., Cordes, E.E., Quattrini, A.M., Nelson, R.
Spies, R.B., Hardin, D.D., Toal, J.P., 1988. Organic enrichment or toxicity? A comparison K., Camilli, R., Demopoulos, A.W., German, C.R., Brooks, J.M., 2012. Impact of the
of the effects of kelp and crude oil in sediments on the colonization and growth of Deepwater Horizon oil spill on a deep-water coral community in the Gulf of Mexico.
benthic infauna. J. Exp. Mar. Biol. Ecol. 124 (3), 261–282. Proc. Natl. Acad. Sci. USA 109 (50), 20303–20308.
Steichen Jr., D.J., Holbrook, S.J., Osenberg, C.W., 1996. Distribution and abundance of Yang, Z., Feng, J., Niu, J., Shen, Z., 2008. Release of polycyclic aromatic hydrocarbons
benthic and demersal macrofauna within a natural hydrocarbon seep. Mar. Ecol. from Yangtze River sediment cores during periods of simulated resuspension.
Prog. Ser. 138, 71–82. Environ. Pollut. 155 (2), 366–374.