Changes in Reef Fish Community Structure Following the Deepwater Horizon Oil Spill


Large-scale anthropogenic disturbances can have direct and indirect effects on marine communities, with direct effects often taking the form of widespread injury or mortality and indirect effects manifesting as changes in food web structure. Here, we report a time series that captures both direct and indirect effects of the Deepwater Horizon Oil Spill (DWH) on northern Gulf of Mexico (nGoM) reef fish communities. We observed significant changes in community structure immediately following the DWH, with a 38% decline in species richness and 26% decline in Shannon-Weiner diversity. Initial shifts were driven by widespread declines across a range of trophic guilds, with subsequent recovery unevenly distributed among guilds and taxa. For example, densities of small demersal invertivores, small demersal browsers, generalist carnivores, and piscivores remained persistently low with little indication of recovery seven years after the DWH. Initial declines among these guilds occurred prior to the arrival of the now-widespread, invasive lionfish (Pterois spp.), but their lack of recovery suggests lionfish predation may be affecting recovery. Factors affecting persistently low densities of generalist carnivores and piscivores are not well understood but warrant further study given the myriad ecosystem services provided by nGoM reef fishes.


The nature, frequency, and intensity of disturbance are important drivers of community structure1,2,3, and it is well established that evolutionary history4, historical disturbance regimes2,4, and the prior state of a community2,5,6 affect its response7. Although natural disturbances can be important for maintaining diverse, resilient species assemblages8, research focused on the impacts of chronic anthropogenic stressors on biodiversity has revealed that even specious communities, presumed to be resilient, can respond unpredictably to natural and anthropogenic disturbances5,9,10. This is particularly true in marine systems which have experienced impacts from human activities for centuries and are severely degraded as a result11,12. Numerous examples exist of long-term community shifts from estuarine13, coral reef14,15, and continental shelf16,17 systems, and it is not uncommon for communities to remain unaffected by localized or moderate disturbances5, only to exhibit a non-linear response following a series of disturbances18 or a single event of sufficient scale or intensity19.

The 2010 Deepwater Horizon Oil Spill (DWH) was the epitome of a large-scale, anthropogenic disturbance capable of producing substantial community-level impacts. Over an 87-day period, approximately 4.9 million gallons of oil20 was released into northern Gulf of Mexico (nGoM) at a depth of ~1,500 m producing a surface slick of ~40,000 km2 at its maximum extent21. Between 4 and 14% of the total discharge was transported to the benthos by contaminated marine snow21,22,23,24, thus exposing numerous pelagic and benthic communities to toxic polycyclic aromatic hydrocarbons (PAHs)25 as well as emulsifying dispersants26. Both lethal and sublethal effects (e.g., compromised immune27 and endocrine function28, developmental abnormalities29, reduced growth30, and impaired olfaction31 and predator avoidance32,33) of oil exposure have been well documented for numerous taxa34,35,36,37, and negative effects at the organismal level had the clear potential to elicit effects on community structure through bottom-up38 or top-down mechanisms39,40.

Much of the effort to document the community-level responses to the DWH was focused on monitoring coastal habitats that provide critical nurseries for several marine taxa41, are widely studied by community ecologists42, and whose proximity favored the rapid collection of critical baseline data43. Despite extensive shoreline oiling44, impacts were mostly relegated to heavily oiled, coastal sites in Louisiana where significant vegetation loss occurred along the marsh edge45. Inshore communities, particularly nekton and fish assemblages, in areas of limited exposure showed little indication of negative effects higher than the organismal level46,47,48,49. Multiple assessments of community-level effects on the nGoM continental shelf also indicated no impacts or a resilient recovery. For example, changes in plankton communities were relatively brief and isolated to the active spill period50,51,52; post-DWH meiofauna diversity and abundance were comparable to pre-spill estimates53; and macroinfauna abundance and diversity showed no signs of impacts54.

The lack of discernable DWH effects on community structure have been documented repeatedly but mostly for communities dominated by short-lived taxa45,55,56 more likely to be resilient to such a distrubance57. Considerably less is known about the impact to and response of neritic fish communities, including reef fish assemblages in the nGoM. These fish communities are diverse58,59, including numerous short-lived, small demersal species, as well as long-lived, fisheries species likely to be resistant to additional sources of mortality60,61. Key to the persistence of many fisheries species is the maintenance of multiple year classes of relatively old, mature individuals (i.e., the storage effect60) that experience low natural mortality, which minimizes biomass loss between periodically strong year classes61. Thus, compensatory population increases following DWH-induced mortality would likely be slow.

Resistance is also conferred through their high vagility and generalist diets which would allow individuals to disperse to avoid acute mortality62 while exploiting locally abundant prey63,64. However, evidence of sub-lethal PAH exposure suggests negative effects reached well beyond the footprint of the surface slick37,65, and recent ecosystem simulations suggest mortality resulting from acute resource limitation was perhaps more severe and widespread than exposure mortality66. Indeed, changes in feeding ecology, trophic position, and condition of red snapper (Lutjanus campechanus) were documented following the DWH63,67 and provide empirical support for a shift in resource availability with possible negative effects on population productivity67. Similar shifts in trophic position and pathways were observed in other reef fish species concomitant with petrocarbon cycling through the nGoM food web68,69. However, to date few data have been presented linking declines in reef fish abundance or community shifts that followed either direct effects of the DWH or indirect effects via food web impacts70.

The lack of information regarding community-level responses following the DWH may stem, in part, from the lack of pre-spill baseline data for many taxa and communities. Here, we analyze a time-series of nGoM reef fish community data that span an eight-year period starting before the DWH. We test whether significant changes occurred in reef fish community structure following the DWH, and if so whether changes occurred evenly across trophic guilds or were more concentrated within specific guilds. Results are presented in the context of acute, direct and chronic, indirect effects of the DWH, as well as factors that may affect the resiliency of reef fish communities.


Site description and data collection

Fish communities were surveyed at 16 natural reefs (Fig. 1) in the nGoM with a VideoRay Pro4 mini remotely operated vehicle (ROV) during 2009–2017. Reefs were randomly selected from a series of sites surveyed in 2008–200971, encompassed a depth range of 17–72 m, and were distributed over an 8,000 km2 area of the continental shelf. Sites were representative of the morphologically variable hard bottom habitat in the region and included low relief ledges, rocky ridges, rock rubble mounds, and flat limestone block reefs72. The epibenthic communities are dominated by coralline algae, soft corals (e.g., black corals, gorgonians, and octocorals), and sponges with limited coverage by azoozanthellate, ahermatypic corals72. Although impacts to epibenthic species were observed west of our sampling region73, sedimentation of contaminated marine snow was patchy24, and we observed no signs of habitat degradation (e.g., oiling or injured/deceased coral colonies). Thus, monitoring of the epibenthic community was not undertaken.

Figure 1

Map of the northern Gulf of Mexico (A) and sampling region (B). Panel B shows the location of the Deepwater Horizon blowout at the Mississippi Canyon-252 wellhead (triangle), and the natural reefs surveyed from 2009 to 2010 (circles). The shaded area represents the cumulative surface oil coverage in days. Maps were produced in R version 3.5.177.

In total, 250 ROV surveys were completed between 2009 and 2017. The pre-DWH portion of our time series consisted of 26 surveys across 11 sites and occurred from the summer of 2009 to the spring of 2010. All 2010 surveys classified as pre-DWH occurred prior to surface oil entering our sampling region; surveys grouped in the 2010 time bin were sampled in November. During 2011–2013, sites were typically surveyed in the spring, summer, and fall, but later in the time series funding constraints limited sampling to summer months. Each ROV survey consisted of 3 to 4 orthogonal, 25-m transects either 1 or 2 m off the bottom, depending on visibility70,71. The ROV was equipped with a 570-line color camera with wide-angle (116°) lens for real-time viewing, twin forward projecting LEDs (3,600 lumens), and rated to a depth of 170 m. An additional forward-facing, high definition (1080p at 60–120 fps) camera (GoPro Hero 2, 3, or 4) was mounted at a 45° angle to the ROV’s float block above the internal camera to record high definition video of the reef fish community.

Videoo samples were analyzed on a high-resolution monitor; observed fishes were identified to the lowest taxonomic level possible and enumerated. Counts were summed across transects to estimate the total abundance for each site. For active, schooling species prone to double counting (e.g., scads, herrings, etc.), the total abundance across all transects was superseded by an estimate of the minimum number of individuals within a school obtained while maneuvering between transect locations. Species abundances were converted to densities by dividing by the total area sampled following the methods of Patterson et al.71. Results presented below only include those taxa identified to the level of species with the exception of purple reeffish (Chromis scotti) and dusky damselfish (Stegastes fuscus), which are difficult to distinguish on video footage and therefore combined into a single group, damselfish.

Community analysis

Permutational multivariate analysis of variance (PERMANOVA) models were computed in PRIMER (v7) to test for temporal changes in reef fish community structure following the DWH. To reduce the influence of abundant species, taxa-specific densities were log(x + 1) transformed. A resemblance matrix was then computed based on the Bray-Curtis dissimilarity, with the inclusion of a dummy species at a density of 1. The dummy species was included because Bray-Curtis can behave erratically if few species are shared between sites74, which is important to consider when evaluating environmental impacts on community structure. The PERMANOVA model had a three-factor hierarchical design with site nested within month nested within year. The nested factors, site and month, were treated as random factors while year was treated as a fixed factor. Our model also included two covariates, depth and longitude, which were z-score transformed. The reasons for this approach were: (1) each covariate represents a gradient along which reef fish community structure naturally varies58,75; (2) cluster analysis of pre-DWH community structure did not identify groups that might justify the use of discrete categorical factors to evaluate depth or longitude effects; (3) the use of covariates as opposed to fixed factors permitted the inclusion of the entire data set; and, (4) at this scale there is not a clear relationship between straight line distance from the well head and impacts. Changes in community structure were also evaluated using common community indices of species richness (S), diversity (Shannon-Weiner H′), and evenness (Pielou’s J′). All indices were calculated using the vegan package76 in R77. Temporal changes were evaluated with linear mixed effects models (LMMs) using the lme4 package78 followed by Dunnett’s multiple comparisons using the multcomp package79.

Trophic guild and species-specific trends

We evaluated temporal changes in density for nine trophic guilds: herbivores, small demersal browsers, large demersal browsers, small demersal invertivores, large demersal invertivores, generalist carnivores, piscivores, reef planktivores, and pelagic planktivores, with small versus large indicating species generally smaller versus larger than 200 mm total length. Species were assigned to guilds based on dietary data, both from the literature (see Appendix A) and recent analyses, and densities were summed by guild for each ROV sample. In the case of small demersal invertivores, we excluded tomtate (Haemulon flavolineatum) from guild-level estimates because densities of this schooling grunt (Haemulidae) were highly variable and often an order of magnitude larger than other guild members. Their inclusion obscured the more general, guild-level pattern (Fig. S1). The species-specific analyses included taxa (n = 52) whose relative frequency of occurrence was >5% before or after DWH and for which sufficient data were available for model convergence.

Temporal trends in trophic guilds and species were assessed by computing standardized density indices with generalized linear mixed effects models (GLMM) following the delta approach80,81. This approach consists of two models, one to model the probability of observing zero individuals (hereafter, presence/absence) and a second to model the density given a guild or species was observed81. The product of the two sub-models was then used as the standardized density index for each guild or taxon. GLMMs included year as a factor and the repeated measures design was specified by including a random intercept parameter for each site. Longitude and depth were also included as covariates following z-score transformation. Least-squares means were calculated for each sub-model as the annual average from a reference grid of predictions across factor levels (i.e., years). Monte Carlo simulations were computed to estimate an annual density index and confidence intervals following the methods in Chagaris et al.82. Briefly, the product of the least-squares mean standard error and 10,000 random normal deviates X N(μ = 0, σ = 1) was added to the least-squares mean estimate of annual density. Error deviates of the log-normal model were adjusted when the log-normal and binomial least-squares mean were correlated (Pearson’s correlation p-value ≤ 0.05). Values of each Monte Carlo simulation were then back-transformed into their original measurement units to obtain a distribution of density values.

The results from binomial and log-normal models were also evaluated separately to infer whether temporal differences resulted from a significant change in presence/absence or non-zero abundance. Multiple comparisons performed using Dunnett’s method, as described above. For guilds and species observed prior to the DWH, comparisons were made between the pre- and post-DWH time periods with non-zero density estimates. For species only observed during the post-DWH time period, comparisons were made between the first and subsequent years with non-zero density estimates.


Species composition

Our ROV dataset included 138 species from 43 families. The highest densities were observed for grunts and snappers (Lutjanidae) reflecting the fact either tomtate or vermilion snapper (Rhomboplites aurorubens) was the most abundant species in a given year (standardized density range 27–136 and 26–195 individuals 1000 km2, respectively) (Supplemental Table S1). Approximately 43% percent of reef fish species were distributed among five other families: Serranidae (15.2%), Carangidae (7.2%), Sciaenidae (5.8%), Sparidae (5.8%), and Pomacentridae (5.1%).

Community analysis

PERMANOVA results indicated community structure significantly differed among years (Table 1). Both covariates and the random effect of month within year were also statistically significant, while the random effect of site within month within year was not significant. Of the eight pairwise comparisons between pre- and post-DWH periods, significant differences in community structure were observed in 2010–2011 and 2013–2017 (Table 2). Pairwise comparisons from the post-DWH portion of the time series (i.e., 2010–2017) also indicated significant interannual differences in community structure. However, significant differences during the post-DWH period were more common for comparisons separated by one or more years.

Table 1 Permutational multivariate analysis of variance results based on Bray-Curtis dissimilarity.
Table 2 Post-hoc pairwise comparisons of community structure among years based on Bray-Curtis dissimilarly.

All three community indices (S, H′, and J′), as well as total fish density, declined following the DWH (Fig. 2) with a significant decline observed for S (Table S2). Species richness showed an upward trend in 2011–2015, going from 14.1 to 21.3 species per site and remained comparable to our pre-DWH baseline of 16.4 in 2016 and 2017. However, H′ and J′ continued to decline and were significantly lower than pre-DWH estimates in 2012, 2013, and 2016 (Table S2). The 2010 decline in total fish density, though not statistically significant, was substantial and represented a decline of ~62%. From 2011 onward, fish density showed a slight positive trend.

Figure 2

Standardized community indices or total fish density (±95% CIs) estimated from generalized linear models with corresponding 95% CIs generated through Monte Carlo simulations. An asterisk (*) denotes a significant pairwise difference between pre- and post-DWH time points (Supplemental Table S2).

Trophic guild and species-specific trends

Following the DWH, densities of all eight trophic guilds observed prior to the oil spill declined (Fig. 3). The magnitude of these declines ranged from 35 to 96% and four of the eight trophic guilds reached their lowest densities in 2010. Although we did not observed significant changes in guild presence/absence (Supplemental Table S3), the initial declines in herbivore, small and large demersal browser, small and large demersal invertivore, generalist carnivore, and piscivore densities were associated with significantly lower abundances when present (Supplemental Table S4). The 52 species for which species-specific trends were evaluated reflect this general pattern (Supplemental Table S1). Forty-six species were observed prior to the DWH and 43 declined between our pre-DWH baseline and 2010. For 29 species, these initial declines reflected either a complete absence, significant change in presence/absence, (Supplemental Table S5), or significantly lower densities when present (Supplemental Table S6).

Figure 3

Standardized trophic guild densities (±95% CIs) estimated from generalized linear mixed effects models with corresponding 95% CIs generated through Monte Carlo simulations. A unique color was assigned to each trophic guild and this color scheme is used in subsequent figures. A stroked t (ŧ) indicates a significant pairwise difference in presence/absence between pre-DWH and post-DWH time bins (Supplemental Table S3) or complete absence. An asterisk (*) denotes a significant difference in density when present (Supplemental Table S4). Trophic guild names are provided in the top right corner of each panel.

Guilds comprised of small-bodied species that forage on benthic prey showed the largest declines immediately following the DWH. Densities of herbivores, small demersal browsers, and small demersal invertivores declined by 96%, 87%, and 82%, and remained persistently low through much of the time series (Fig. 3). The decline in herbivore density almost entirely reflected doctorfish (Acanthurus chirurgus) abundance, while the decline in small demersal browsers resulted from lower densities of the cocoa damselfish (Stegastes variabilis) and seaweed blenny (Parablennius marmoratus) (Fig. 4). The trend displayed by small demersal invertivores was driven by species like the slippery dick (Halichoeres bivittatus) and cubbyu (Paraques umbrosus) and differed from that of the tomtate, which has a looser association with the reef structure (Fig. 4).

Figure 4

Standardized densities of representative species from the herbivore, small demersal browser, large demersal browser, small demersal invertivore, and large demersal invertivore, trophic guilds. Estimates were derived from generalized linear mixed effects models using the delta approach. Corresponding 95% CIs were generated through Monte Carlo simulations. A stroked t (ŧ) indicates a significant pairwise difference in presence/absence between pre-DWH and post-DWH time bins (Supplemental Table S5) or complete absence. An asterisk (*) denotes a significant difference in density when present (Supplemental Table S6). Species names are provided in upper right corner of each panel.

Declines among large-bodied species reliant on benthic production were also evident several years following the DWH, although these declines were less severe (Fig. 3). The blue angelfish (Holacanthus bermudensis) was the most abundant large demersal browser and drove guild-level trends (Fig. 4). The density of large demersal invertivores reflected red porgy (Pagrus pagrus) and gray triggerfish (Balistes capriscus) abundances. However, each species clearly displayed disparate patterns. The initial decline in red porgy was followed by an increase in 2011 and subsequent decline (Fig. 4). Conversely, gray triggerfish density was more variable and temporal changes were not associated with a significant difference in presence/absence or density when present.

Guilds representing higher trophic level consumers showed trends more similar to the small-bodied demersal guilds. Densities of generalist carnivores and piscivores declined by 64% and 73% in 2010 and densities remained persistently low thereafter (Fig. 3). For both guilds, these trends were driven by large-bodied, fisheries species (Fig. 5). Red snapper, gray snapper (L. griseus), and red grouper (Epinephelus morio) declined by 69%, 85%, and 70% following the DWH, and low densities persisted through 2017. The three most abundant piscivores, scamp (Mycteroperca phenax), gag (M. microlepis), and sandbar shark (Carcharhinus plumbeus), all declined after the spill and both scamp and gag remained at densities below pre-DWH baseline values. Densities of smaller bodied generalist carnivores, [e.g., bank seabass (Centropristis ocyurus) and belted sandfish (Serranus subligarius) (Fig. 5)] either displayed no change or failed to recover following the DWH. The one exception was the invasive lionfish (Pterois spp.) which was first observed in 2011 and rapidly increased through 2017 (Fig. 5).

Figure 5

Standardized densities of representative species from the generalist carnivore and piscivore trophic guilds. Density estimates were derived from generalized linear mixed effects models using the delta approach. Corresponding 95% CIs were generated through Monte Carlo simulations. A stroked t (ŧ) indicates a significant pairwise difference in presence/absence (Supplemental Table S5) or complete absence. An asterisk (*) denotes a significant difference in density when present (Supplemental Table S6). Species names are provided in upper right corner of each panel.

Unlike guilds that rely on benthic forage, reef planktivore densities remained unchanged following the DWH (Fig. 3). Similarly, the downward trend through 2012 and subsequent increase were not associated with a significant difference in presence/absence or density when present (Supplemental Tables S3 and S4). The vermilion snapper was the most abundant reef planktivore and displayed a similar temporal pattern (Fig. 6). Of the nine other reef planktivores, four displayed a significant decline in presence/absence or density when present (Supplemental Table S3). However, even species that experienced declines of >90% in 2010 generally increased to pre-DWH densities [e.g., yellowtail reeffish (Chromis enchrysura) and damselfish (Fig. 6)]. Pelagic planktivores were not observed during our pre-DWH surveys and infrequently observed thereafter. No pelagic planktivores met our selection criteria for species-specific analysis (see Methods).

Figure 6

Standardized densities of reef planktivores estimated from generalized linear mixed effects models using the delta approach. Corresponding 95% CIs were generated through Monte Carlo simulations. A stroked t (ŧ) indicates a significant pairwise difference in presence/absence between pre-DWH and post-DWH time bins (Table S5) or complete absence. An asterisk (*) denotes a significant difference density when present (Table S6). Species names are provided in upper right corner of each panel.


Our results indicate reef fish communities exhibited clear signs of negative impacts following the DWH with significant shifts in community structure and declines in species richness, diversity, evenness, and total fish density. This change in community structure was unique in that it that concomitant declines were observed for all eight trophic guilds. At no other point were similar, synchronous declines present nor were significant pairwise differences in community structure evident between successive years. The species composition of the more abundant fishes was similar before and after the spill suggesting declines in species richness resulted from an absence of rare species and changes in community structure, species diversity, evenness, and total fish density resulted from shifts in relative abundances. Declines in species richness did not persist and has remained similar to pre-DWH richness. However, lower estimates of diversity and evenness were evident several years post-spill reflecting lower densities of small demersal browsers, small demersal invertivores, generalist carnivores, and piscivores. These effects are similar to those observed among deep-sea benthic communities where declines in megafauna83, macrofauna84, meiofauna85, and foraminifera86 abundance and diversity followed the DWH. Although the most severe impacts to deep-sea benthic communities were in immediate vicinity of the wellhead84, the negative responses observed among decapod crustacean communities on deep bank formations84 and fish communities on artificial reef70 100 s of km away suggests acute exposure at distant sites was of sufficient intensity to produce observable shifts.

We report declines across a range of reef fish taxa regardless of vagility, trophic position, or diet, which seem improbable without DWH-induced mortality. However, a primary challenge is identifying the extent to which declines resulted from mortality or emigration. Small fishes that live in close proximity to the reef matrix (e.g., small demersal browsers and invertivores) experienced the largest initial declines, which likely reflects a high incident of acute mortality. These species can have limited (<10 m2) home ranges87, are heavily reliant on local resource pools, and would incur the highest cost associated with emigration. These traits not only increase the probability of exposure-related mortality but also mortality associated with resource limitation66. Impacts on pelagic production88,89, increased trophic position68,69, and greater reliance on benthic resources63 suggest small demersal reef fishes experienced increased resource competition and higher predation immediately following the DWH. The effects of resource limitation were possibly exacerbated by sub-lethal exposure37 which can result in physiological stress leading to impaired predator avoidance and foraging ability32,33,90. This inference is further supported by the fact that initial declines among large-bodied reef fishes (e.g., large demersal invertivores, some generalist carnivores, and piscivores) were less severe. These species are quite mobile, exhibit varying degrees of site fidelity on natural reefs91, and their movements can be affected by large-scale disturbances62,92. Additionally, a large area (maximum = 290,000 km2) of the continental shelf was closed to harvest for a few months during the spill, perhaps reducing fishing-related mortality, albeit temporarily. Emigration following the DWH also may have occurred if the cost of exposure or resource limitation outweighed the benefits of staying93,94. For some individuals, exposure may have been sufficient to elicit movement from affected areas. For others, the response was perhaps a shift in foraging behavior63,68, followed by emigration as resources became more scarce. Individuals present after the DWH probably consisted of few residents that survived acute exposure and resource limitation along with new immigrants seeking more favorable conditions93. The fact that guilds comprised of mobile reef species (i.e., generalist carnivores and piscivores) showed little indication of recovery suggests a large number of individuals were either permanently displaced, perished (either from starvation or exposure66), or basal resource pools remained insufficient to support pre-DWH densities.

The effects of natural disturbances within the region (e.g., hurricanes and hypoxia) are often insufficient to produce an observable change in large, mobile reef fish abundance95,96. Thus, it appears the acute impact of the DWH was more severe compared to large-scale natural disturbances typical of our study area. Initial community-wide declines are, however, notably similar to the effects of harmful algal blooms that seasonally occur along the West Florida Shelf (WFS)97,98,99. The 1971 red tide event is the most well-documented case of the impacts on reef fish communities and subsequent recovery. Exposure resulted in the near extirpation of reef fishes across a >1,500 km2 stretch of the WFS97. Conspicuous100 and indiscriminate reef fish mortality produced clear declines in species richness, but the most pronounced effects were changes in species relative abundance97. Recovery followed a predictable pattern of succession initiated by the arrival of small demersal species that recruit directly to reefs, followed by a peak in abundance among early pioneers, and subsequent decline in abundance superior competitors arrived; increases among large mobile fishes that do not recruit directly to reefs occurred later97. Full recovery took several years, but the community that developed was nearly identical to that observed prior to the 1971 red tide. The same general pattern of community level impacts, succession, and recovery were also observed on WFS artificial reefs following a 2005 red tide event98, corroborating observations by Smith97 and others that documented successional patterns among GoM and Caribbean reef fish communities101,102. Although the rate of succession can vary among sites103,104, after seven years of post-DWH monitoring no such pattern emerged despite similarities between community members in the aforementioned studies.

A clear difference between the DWH and natural disturbances was the potential for chronic PAH exposure, which can have long-term, higher-order impacts even at sub-lethal levels105. While sedimentation of contaminated marine snow was the primary vector transporting oil and dispersants to the benthos, this phenomenon was patchy and mostly concentrated off the continental shelf and west of our sampling area106,107. As a result, sediment concentrations of total petroleum and polycyclic aromatic hydrocarbons (PAHs) were typical of background levels within our study area in August 201054. Nonetheless, examination of reef fish tissue samples indicated elevated levels of PAH concentrations in liver and PAH metabolites in bile persisted for years following the DWH37,65,108, thus providing evidence of continual reef fish exposure to toxic petroleum compounds for some time after the spill despite uncertainties about the mechanism of exposure.

The fact that reef fish density remained low for a number of years following the DWH suggests indirect food web effects have played an important role as well66. Consumption of swarming zooplankton by red snapper declined markedly following the DWH, a possible reflection of an initial reduction in pelagic production66. The reduction in zooplankton was balanced by increased foraging on demersal fishes and invertebrates63. This increased reliance on benthic resources continued over weeks and months following DWH, as indicated by declines in δ34S63,68. The associated enrichment of δ15N persisted for several years following the DWH63,68, and similar long-term shifts in δ15N were observed for gray triggerfish, tomtate, red porgy, and vermilion snapper concurrent with declines in δ13C, indicative of petrocarbon cycling through the food web68,109. The 2010 shifts in trophic position and community-wide declines that extended into 2011 also pre-date the arrival and rapid expansion of the invasive lionfish and are in general agreement with previous reports that community-wide impacts following the DWH occurred prior to the nGoM lionfish invasion70.

Although the role of food web effects resulting from a shift in resource availability is apparent, the success of the invasive lionfish may also be an important factor suppressing community recovery. Numerous studies have documented community-level effects of lionfish110,111 via predation112,113 and competition114,115. Impacts are typically most evident among small demersal reef fishes that recruit directly to reef habitats70 and vulnerable to predation both as adults112,116 and newly-settled recruits113. The stunted recovery of small demersal invertivores and browsers, despite their capacity for rapid recolonization following mass mortality events97 and the continued declines among native predators, provides a clear indication that the success of the invasive lionfish is affecting the response among native, small demersal fishes. How lionfish may be affecting the recovery of fisheries species has not been evaluated, but competition between lionfish and native predators (e.g., groupers, snappers, and jacks) may affect population productivity117. Taxonomic resolution of dietary data for fisheries species is often poor and presents challenges when attempting to capture competitive interactions in systems with diverse species assemblages117. Diet information for the majority of small demersal reef fish is also lacking, and the potential food web effects emanating from low densities of small demersal species remains unknown.

The changes in community structure, particularly the persistently low densities among certain groups, provides a clear indication of lasting, community-wide impacts. The available evidence suggests initial declines in 2010 likely reflected both mortality and emigration resulting from exposure and resource limitation. Mortality due to direct or indirect effects of the spill likely drove initial declines of small demersal species, while large-bodied consumers were more likely to be permanently displaced or to suffer delayed mortality. Community-wide declines into 2011 were indicative of protracted resource limitation. However, the lack of recovery in small demersal reef fishes from 2012 onward may, to some extent, reflect top-down pressure from lionfish. How community shifts have altered the flow of energy to higher trophic level fisheries species or impacted system resilience remains uncertain.

A clear challenge moving forward is identifying the underlying mechanisms driving these patterns, estimating the relative impacts of individual stressors (e.g., exposure, resource limitation, lionfish invasion, fisheries harvest, and food web effects), and developing management strategies to facilitate recovery. Efforts along these lines appear particularly relevant considering the dramatic declines observed for higher trophic position consumers that not only serve important ecological roles118 but also provide numerous economic119 and cultural120 benefits. Many of the ecosystems services we ascribe to reef fish assemblages in the nGoM are inextricably linked to fisheries harvest119,120, and fishery-dependent data and assessment of fishery stocks should enable tracking resilience among those species. However, continued funding for fishery-independent surveys, such as the ROV work that forms that basis of the analyses presented herein, is critical to assess the long-term effects of the DWH, lionfish invasion, and the potential for resiliency in the nGoM ecosystem.

Data availability

Data are available through the Gulf of Mexico Research Initiative Information & Data Cooperative (GRIIDC) at under DOI 10.7266/N72J68SF and 10.7266/n7-n4j3-0a26. The remaining portion of the time series can be obtained from the corresponding author upon request.

Change history

  • 07 July 2020

    An amendment to this paper has been published and can be accessed via a link at the top of the paper.


  1. 1.

    Hall, A. R. et al. Diversity–disturbance relationships: frequency and intensity interact. Biol. Lett. 8, 768–771 (2012).

    PubMed  PubMed Central  Google Scholar 

  2. 2.

    Davies, K. W., Svejcar, T. J. & Bates, J. D. Interaction of historical and nonhistorical disturbances maintains native plant communities. Ecol. Appl. 19, 1536–1545 (2009).

    CAS  PubMed  Google Scholar 

  3. 3.

    Svensson, J. R., Lindegarth, M. & Pavia, H. Equal rates of disturbance cause different patterns of diversity. Ecology 90, 496–505 (2009).

    PubMed  Google Scholar 

  4. 4.

    Waples, R. S., Beechie, T. & Pess, G. R. Evolutionary history, habitat disturbance regimes, and anthropogenic changes: what do these mean for resilience of Pacific salmon populations? Ecol. Soc. 14, art3 (2009).

    Google Scholar 

  5. 5.

    Ebeling, A. W., Laur, D. R. & Rowley, R. J. Severe storm disturbances and reversal of community structure in a southern California kelp forest. Mar. Biol. 84, 287–294 (1985).

    Google Scholar 

  6. 6.

    Suding, K. N., Gross, K. L. & Houseman, G. R. Alternative states and positive feedbacks in restoration ecology. Trends Ecol. Evol. 19, 46–53 (2004).

    PubMed  Google Scholar 

  7. 7.

    Reice, S. R. Nonequilibrium dynamics of biological community structure. Am. Sci. 82, 424–434 (1994).

    ADS  Google Scholar 

  8. 8.

    Ebeling, A. W. & Hixon, M. A. Tropical and temperate reef fishes: comparison of community structures. In The Ecology of Fishes on Coral Reefs (ed. Sale, P. F.) 509–563 (Elsevier), (1991).

  9. 9.

    Mumby, P. J. & Steneck, R. S. Coral reef management and conservation in light of rapidly evolving ecological paradigms. Trends Ecol. Evol. 23, 555–563 (2008).

    PubMed  Google Scholar 

  10. 10.

    Scheffer, M. & Carpenter, S. R. Catastrophic regime shifts in ecosystems: linking theory to observation. Trends Ecol. Evol. 18, 648–656 (2003).

    Google Scholar 

  11. 11.

    Lotze, H. K. et al. Depletion, degradation, and recovery potential of estuaries and coastal seas. Science. 312, 1806–1809 (2006).

    ADS  CAS  PubMed  Google Scholar 

  12. 12.

    Jackson, J. B. C. et al. Historical overfishing and the recent collapse of coastal ecosystems. Science. 293, 629–637 (2001).

    CAS  PubMed  Google Scholar 

  13. 13.

    Kemp, W. et al. Eutrophication of Chesapeake Bay: historical trends and ecological interactions. Mar. Ecol. Prog. Ser. 303, 1–29 (2005).

    ADS  Google Scholar 

  14. 14.

    Elmhirst, T., Connolly, S. R. & Hughes, T. P. Connectivity, regime shifts and the resilience of coral reefs. Coral Reefs 28, 949–957 (2009).

    ADS  Google Scholar 

  15. 15.

    Hughes, T. P., Graham, N. A. J., Jackson, J. B. C., Mumby, P. J. & Steneck, R. S. Rising to the challenge of sustaining coral reef resilience. Trends Ecol. Evol. 25, 633–642 (2010).

    PubMed  Google Scholar 

  16. 16.

    Fogarty, M. J. & Murawski, S. A. Large-scale disturbance and the structure of marine systems: Fishery impacts on Georges Bank. Ecol. Appl. 8, 6–22 (1998).

    Google Scholar 

  17. 17.

    Steneck, R. S., Vavrinec, J. & Leland, A. V. Accelerating trophic-level dysfunction in kelp forest ecosystems of the western North Atlantic. Ecosystems 7, 323–332 (2004).

    Google Scholar 

  18. 18.

    Hughes, T. P. Catastrophes, phase shifts, and large-scale degradation of a Caribbean coral reef. Science. 265, 1547–1551 (1994).

    ADS  CAS  PubMed  Google Scholar 

  19. 19.

    Wernberg, T. et al. Climate-driven regime shift of a temperate marine ecosystem. Science. 353, 169–172 (2016).

    ADS  CAS  PubMed  Google Scholar 

  20. 20.

    McNutt, M. K. et al. Review of flow rate estimates of the Deepwater Horizon oil spill. Proc. Natl. Acad. Sci. 109, 20260–20267 (2012).

    ADS  PubMed  Google Scholar 

  21. 21.

    Deepwater Horizon Natural Resource Damage Assessment Trustees. Deepwater Horizon Oil Spill: Final Programmatic Damage Assessment and Restoration Plan and Final Programmatic Environmental Impact Statement. (2016).

  22. 22.

    Valentine, D. L. et al. Fallout plume of submerged oil from Deepwater. Horizon. Proc. Natl. Acad. Sci. 111, 15906–15911 (2014).

    ADS  CAS  PubMed  Google Scholar 

  23. 23.

    Chanton, J. et al. Using natural abundance radiocarbon to trace the flux of petrocarbon to the seafloor following the deepwater horizon oil spill. Environ. Sci. Technol. 49, 847–854 (2015).

    ADS  CAS  PubMed  Google Scholar 

  24. 24.

    Passow, U. & Ziervogel, K. Marine snow sedimented oil released during the Deepwater Horizon spill. Oceanography 29, 118–125 (2016).

    Google Scholar 

  25. 25.

    Hylland, K. Polycyclic aromatic hydrocarbon (PAH) ecotoxicology in marine ecosystems. J. Toxicol. Environ. Heal. Part A 69, 109–123 (2006).

    CAS  Google Scholar 

  26. 26.

    Hemmer, M. J., Barron, M. G. & Greene, R. M. Comparative toxicity of eight oil dispersants, Louisiana sweet crude oil (LSC), and chemically dispersed LSC to two aquatic test species. Environ. Toxicol. Chem. 30, 2244–2252 (2011).

    CAS  PubMed  Google Scholar 

  27. 27.

    Holladay, S. D. et al. Benzo[a]pyrene-induced hypocellularity of the pronephros in tilapia (Oreochromis niloticus) is accompanied by alterations in stromal and parenchymal cells and by enhanced immune cell apoptosis. Vet. Immunol. Immunopathol. 64, 69–82 (1998).

    CAS  PubMed  Google Scholar 

  28. 28.

    Navas, J. M. & Segner, H. Antiestrogenicity of β-naphthoflavone and PAHs in cultured rainbow trout hepatocytes: evidence for a role of the arylhydrocarbon receptor. Aquat. Toxicol. 51, 79–92 (2000).

    CAS  PubMed  Google Scholar 

  29. 29.

    Rhodes, S., Farwell, A., Mark Hewitt, L., MacKinnon, M. & George Dixon, D. The effects of dimethylated and alkylated polycyclic aromatic hydrocarbons on the embryonic development of the Japanese medaka. Ecotoxicol. Environ. Saf. 60, 247–258 (2005).

    CAS  PubMed  Google Scholar 

  30. 30.

    Gilliers, C., Claireaux, G., Galois, R., Loizeau, V. & Pape, O. Le. Influence of hydrocarbons exposure on survival, growth and condition of juvenile flatfish: a mesocosm experiment. J. Life Sci. 4, 113–122 (2012).

    Google Scholar 

  31. 31.

    Schlenker, L. S. et al. Damsels in distress: oil exposure modifies behavior and olfaction in bicolor damselfish (Stegastes partitus). Environ. Sci. Technol. 53, 10993–11001 (2019).

    ADS  CAS  PubMed  Google Scholar 

  32. 32.

    Stieglitz, J. D., Mager, E. M., Hoenig, R. H., Benetti, D. D. & Grosell, M. Impacts of Deepwater Horizon crude oil exposure on adult mahi-mahi (Coryphaena hippurus) swim performance. Environ. Toxicol. Chem. 35, 2613–2622 (2016).

    CAS  PubMed  Google Scholar 

  33. 33.

    Farr, J. A. Impairment of antipredator behavior in Palaemonetes pugio by exposure to sublethal doses of parathion. Trans. Am. Fish. Soc. 106, 37–41 (1977).

    Google Scholar 

  34. 34.

    Whitehead, A. et al. Genomic and physiological footprint of the Deepwater Horizon oil spill on resident marsh fishes. Proc. Natl. Acad. Sci. 109, 20298–20302 (2012).

    ADS  CAS  PubMed  Google Scholar 

  35. 35.

    Barron, M. G. Ecological impacts of the Deepwater Horizon oil spill: implications for immunotoxicity. Toxicol. Pathol. 40, 315–320 (2012).

    CAS  PubMed  Google Scholar 

  36. 36.

    Schwacke, L. H. et al. Health of common bottlenose dolphins (Tursiops truncatus) in Barataria Bay, Louisiana, following the Deepwater Horizon oil spill. Environ. Sci. Technol. 48, 93–103 (2014).

    ADS  CAS  PubMed  Google Scholar 

  37. 37.

    Murawski, S. A., Hogarth, W. T., Peebles, E. B. & Barbeiri, L. Prevalence of external skin lesions and polycyclic aromatic hydrocarbon concentrations in Gulf of Mexico fishes, post-Deepwater Horizon. Trans. Am. Fish. Soc. 143, 1084–1097 (2014).

    CAS  Google Scholar 

  38. 38.

    Moreno, R., Jover, L., Diez, C., Sardà, F. & Sanpera, C. Ten years after the Prestige oil spill: seabird trophic ecology as indicator of long-term effects on the coastal marine ecosystem. PLoS One 8, 1–10 (2013).

    Google Scholar 

  39. 39.

    Bascompte, J., Melián, C. J. & Sala, E. Interaction strength combinations and the overfishing. Proc. Natl. Acad. Sci. USA 102, 5443–5447 (2005).

    ADS  CAS  PubMed  Google Scholar 

  40. 40.

    Fleeger, J. W., Carman, K. R. & Nisbet, R. M. Indirect effects of contaminants in aquatic ecosystems. Sci. Total Environ. 317, 207–233 (2003).

    ADS  CAS  PubMed  Google Scholar 

  41. 41.

    Beck, M. W. et al. The role of nearshore ecosystems as fish and shellfish nurseries. Issues Ecol. 2003, 1–12 (2003).

    Google Scholar 

  42. 42.

    Larkum, A. W. D., Orth, R. J. & Duarte, C. M. Seagrasses: Biology, Ecology, and Conservation. (Springer Netherlands (2006).

  43. 43.

    Able, K. W. et al. Fish assemblages in Louisiana salt marshes: effects of the Macondo oil spill. Estuaries and Coasts 38, 1385–1398 (2015).

    CAS  Google Scholar 

  44. 44.

    Michel, J. et al. Extent and degree of shoreline oiling: Deepwater Horizon oil spill, Gulf of Mexico, USA. Plos One 8, e65087 (2013).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  45. 45.

    Silliman, B. R. et al. Degradation and resilience in Louisiana salt marshes after the BP-Deepwater Horizon oil spill. Proc. Natl. Acad. Sci. 109, 11234–11239 (2012).

    ADS  CAS  PubMed  Google Scholar 

  46. 46.

    Fodrie, F. J. & Heck, K. L. Response of coastal fishes to the Gulf of Mexico oil disaster. Plos One 6 (2011).

  47. 47.

    Moody, R. M., Cebrian, J. & Heck, K. L. Interannual recruitment dynamics for resident and transient marsh species: evidence for a lack of impact by the Macondo oil spill. Plos One 8, 1–11 (2013).

    Google Scholar 

  48. 48.

    Schaefer, J., Frazier, N. & Barr, J. Dynamics of near-coastal fish assemblages following the Deepwater Horizon oil spill in the northern Gulf of Mexico. Trans. Am. Fish. Soc. 145, 108–119 (2016).

    CAS  Google Scholar 

  49. 49.

    Fodrie, F. J. et al. Integrating organismal and population responses of estuarine fishes in macondo spill research. Bioscience 64, 778–788 (2014).

    Google Scholar 

  50. 50.

    Parsons, M. L., Morrison, W., Rabalais, N. N., Turner, R. E. & Tyre, K. N. Phytoplankton and the Macondo oil spill: A comparison of the 2010 phytoplankton assemblage to baseline conditions on the Louisiana shelf. Environ. Pollut. 207, 152–160 (2015).

    CAS  PubMed  Google Scholar 

  51. 51.

    Carassou, L., Hernandez, F. J. & Graham, W. M. Change and recovery of coastal mesozooplankton community structure during the Deepwater Horizon oil spill. Environ. Res. Lett. 9, 124003 (2014).

    ADS  Google Scholar 

  52. 52.

    Abbriano, R. et al. Deepwater Horizon oil spill: a review of the planktonic response. Oceanography 24, 294–301 (2011).

    Google Scholar 

  53. 53.

    Landers, S. C. et al. Meiofauna and trace metals from sediment collections in Florida after the deepwater horizon oil spill. Gulf Mex. Sci. 32, 1–10 (2014).

    Google Scholar 

  54. 54.

    Cooksey, C. et al. Ecological Condition of Coastal Ocean Waters along the U.S. Continental Shelf of Northeastern Gulf of Mexico: 2010 (2014).

  55. 55.

    Roth, A. M. F. & Baltz, D. M. Short-term effects of an Oil spill on marsh-edge fishes and decapod crustaceans. Estuaries and Coasts 32, 565–572 (2009).

    Google Scholar 

  56. 56.

    Paperno, R. et al. The disruption and recovery of fish communities in the Indian River Lagoon, Florida, following two hurricanes in 2004. Estuaries and Coasts 29, 1004–1010 (2006).

    Google Scholar 

  57. 57.

    Soto, L. A., Botello, A. V., Licea-Durán, S., Lizárraga-Partida, M. L. & Yáñez-Arancibia, A. The environmental legacy of the Ixtoc-I oil spill in Campeche Sound, southwestern Gulf of Mexico. Front. Mar. Sci. 1, 1–9 (2014).

    Google Scholar 

  58. 58.

    Murawski, S. A., Peebles, E. B., Gracia, A., Tunnell, J. W. & Armenteros, M. Comparative abundance, species composition, and demographics of continental shelf fish assemblages throughout the Gulf of Mexico. Mar. Coast. Fish. 10, 325–346 (2018).

    Google Scholar 

  59. 59.

    Garner, S. B., Boswell, K. M., Lewis, J. P., Tarnecki, J. H. & Patterson, W. F. III. Effect of reef morphology and depth on fish community and trophic structure in the northcentral Gulf of Mexico. Estuar. Coast. Shelf Sci. 230, 106423 (2019).

    Google Scholar 

  60. 60.

    Warner, R. R. & Chesson, P. L. Coexistence mediated by recruitment fluctuations: a field guide to the storage effect. Am. Nat. 125, 769–787 (1985).

    Google Scholar 

  61. 61.

    Winemiller, K. O. & Rose, K. A. Patterns of life history diversification in North American fishes: implications for population regulation. Can. J. Fish. Aquat. Sci. 49, 2196–2218 (1992).

    Google Scholar 

  62. 62.

    Watterson, J. C., Patterson, W. F. III., Shipp, R. L. & Cowan, J. H. Movement of red snapper, Lutjanus campechanus, in the north central Gulf of Mexico: potential effects of hurricanes. Gulf Mex. Sci. 16, 92–104 (1998).

    Google Scholar 

  63. 63.

    Tarnecki, J. H. & Patterson, W. F. III. Changes in red snapper diet and trophic ecology following the Deepwater Horizon oil spill. Mar. Coast. Fish. 7, 135–147 (2015).

    Google Scholar 

  64. 64.

    McCrawley, J. R. & Cowan, J. H. J. Seasonal and size specific diet and prey demand of Red Snapper on Alabama artificial reefs: implications for management. In Red Snapper (Lutjanus campechanus) Ecology and Fisheries in the Gulf of Mexico (eds. Patterson III., W. F., Cowan, J. H. J., Fitzhugh, G. R. & Nieland, D. L.) 71–96 (American Fisheries Society (2007).

  65. 65.

    Snyder, S. M., Pulster, E. L., Wetzel, D. L. & Murawski, S. A. PAH exposure in Gulf of Mexico demersal fishes, post-Deepwater Horizon. Environ. Sci. Technol. 49, 8786–8795 (2015).

    ADS  CAS  PubMed  Google Scholar 

  66. 66.

    Ainsworth, C. H. et al. Impacts of the Deepwater Horizon oil spill evaluated using an end-to-end ecosystem model. Plos One 13, 1–21 (2018).

    Google Scholar 

  67. 67.

    Herdter, E. S., Chambers, D. P., Stallings, C. D. & Murawski, S. A. Did the Deepwater Horizon oil spill affect growth of red snapper in the Gulf of Mexico? Fish. Res. 191, 60–68 (2017).

    Google Scholar 

  68. 68.

    Patterson, W. F. III. et al. The utility of stable and radioisotopes in fish tissues as biogeochemical tracers of marine oil spill food web effects. In Scenarios and Responses to Future Deep Oil Spills (eds. Murawski, S. A. et al.) 219–238 (Springer International Publishing). (2020).

  69. 69.

    Norberg, M. J. The ecology of tomtate, Haemulon aurolineatum, in the northern Gulf of Mexico and effects of the Deepwater Horizon oil spill. (University of South Alabama (2015).

  70. 70.

    Dahl, K. A., Patterson, W. F. III. & Snyder, R. Experimental assessment of lionfish removals to mitigate reef fish community shifts on northern Gulf of Mexico artificial reefs. Mar. Ecol. Prog. Ser. 558, 207–221 (2016).

    ADS  Google Scholar 

  71. 71.

    Patterson, W. F. III., Tarnecki, J. H., Addis, D. T. & Barbieri, L. R. Reef fish community structure at natural versus artificial reefs in the northern Gulf of Mexico. Proc. 66th Gulf Caribb. Fish. InstituteGulf Caribb. Fish. Inst. (2014).

  72. 72.

    Thompson, M. J., Schroeder, W. W., Phillips, N. W. & Graham, B. D. Ecology of Live Bottom Habitats of the Northeastern Gulf of Mexico: A Community Profile Ecology of Live Bottom Habitats of the Northeastern Gulf of Mexico: A Community Profile. (1999).

  73. 73.

    Silva, M., Etnoyer, P. J. & MacDonald, I. R. Coral injuries observed at mesophotic reefs after the Deepwater Horizon oil discharge. Deep. Res. Part II Top. Stud. Oceanogr. 129, 96–107 (2016).

    ADS  CAS  Google Scholar 

  74. 74.

    Clarke, K. R., Somerfield, P. J. & Chapman, M. G. On resemblance measures for ecological studies, including taxonomic dissimilarities and a zero-adjusted Bray-Curtis coefficient for denuded assemblages. J. Exp. Mar. Bio. Ecol. 330, 55–80 (2006).

    Google Scholar 

  75. 75.

    Semmler, R. F., Hoot, W. C. & Reaka, M. L. Are mesophotic coral ecosystems distinct communities and can they serve as refugia for shallow reefs? Coral Reefs 36, 433–444 (2017).

    ADS  Google Scholar 

  76. 76.

    Oksanen, J. F. et al. vegan: Community Ecology Package. (2018).

  77. 77.

    Team, R. core. R: A language and environment for statistical computing. (2018).

  78. 78.

    Bates, D., Mächler, M., Bolker, B. & Walker, S. Fitting Linear Mixed-Effects Models Using lme4. J. Stat. Softw. 67 (2015).

  79. 79.

    Hothorn, T., Bretz, F., Westfall, P. & Heiberger, R. M. Multcomp: simultaneous inference in general parametric models. Biometrical J. 50, 346–363 (2008).

    MathSciNet  Google Scholar 

  80. 80.

    Lo, N. C., Jacobson, L. D. & Squire, J. L. Indices of relative abundance from fish spotter data based on delta-lognornial models. Can. J. Fish. Aquat. Sci. 49, 2515–2526 (1992).

    Google Scholar 

  81. 81.

    Maunder, M. N. & Punt, A. E. Standardizing catch and effort data: a review of recent approaches. Fish. Res. 70, 141–159 (2004).

    Google Scholar 

  82. 82.

    Chagaris, D., Allen, M. & Camp, E. Modeling temporal closures in a multispecies recreational fishery reveals tradeoffs associated with species seasonality and angler effort dynamics. Fish. Res. 210, 106–120 (2019).

    Google Scholar 

  83. 83.

    Valentine, M. M. & Benfield, M. C. Characterization of epibenthic and demersal megafauna at Mississippi Canyon 252 shortly after the Deepwater Horizon oil spill. Mar. Pollut. Bull. 77, 196–209 (2013).

    CAS  PubMed  Google Scholar 

  84. 84.

    Montagna, P. A. et al. Deep-sea benthic footprint of the Deepwater Horizon blowout. Plos One 8 (2013).

  85. 85.

    Hastings, D. W. et al. Changes in sediment redox conditions following the BP DWH blowout event. Deep. Res. Part II Top. Stud. Oceanogr. 129, 167–178 (2016).

    ADS  CAS  Google Scholar 

  86. 86.

    Felder, D. L. et al. Seaweeds and decapod crustaceans on gulf deep banks after the macondo oil spill. Bioscience 64, 808–819 (2014).

    Google Scholar 

  87. 87.

    Jones, K. M. M. Home range areas and activity centres in six species of Caribbean wrasses (Labridae). J. Fish Biol. 66, 150–166 (2005).

    Google Scholar 

  88. 88.

    Ortmann, A. C. et al. Dispersed oil disrupts microbial pathways in pelagic food webs. Plos One 7, 1–9 (2012).

    Google Scholar 

  89. 89.

    Paul, J. H. et al. Toxicity and mutagenicity of Gulf of Mexico waters during and after the Deepwater Horizon oil spill. Environ. Sci. Technol. 47, 9651–9659 (2013).

    ADS  CAS  PubMed  Google Scholar 

  90. 90.

    Milinkovitch, T. et al. The effect of hypoxia and hydrocarbons on the anti-predator performance of European sea bass (Dicentrarchus labrax). Environ. Pollut. 251, 581–590 (2019).

    CAS  PubMed  Google Scholar 

  91. 91.

    Coleman, F. C., Scanlon, K. M. & Koenig, C. C. Groupers on the edge: shelf edge spawning habitat in and around marine reserves of the northeastern Gulf of Mexico. Prof. Geogr. 63, 456–474 (2011).

    Google Scholar 

  92. 92.

    Addis, D. T., Patterson, W. F. III., Dance, M. A. & Ingram, G. W. Implications of reef fish movement from unreported artificial reef sites in the northern Gulf of Mexico. Fish. Res. 147, 349–358 (2013).

    Google Scholar 

  93. 93.

    Charnov, E. L., Orians, G. H. & Hyatt, K. Ecological implications of resource depression. Am. Nat. 110, 247–259 (1976).

    Google Scholar 

  94. 94.

    Werner, E. E. & Gilliam, J. F. The ontogenetic niche and species interactions in size-structured populations. Annu. Rev. Ecol. Syst. 15, 393–425 (1984).

    Google Scholar 

  95. 95.

    Kirby-Smith, W. W. & Ustach, J. Resistance to hurricane disturbance of an epifaunal community on the continental shelf off North Carolina. Estuar. Coast. Shelf Sci. 23, 433–442 (1986).

    ADS  Google Scholar 

  96. 96.

    Bell, M. & Hall, J. W. Effects of hurricane Hugo on South Carolina’s marine artificial reefs. Bull. Mar. Sci. 55, 836–847 (1994).

    Google Scholar 

  97. 97.

    Smith, G. B. Relationship of eastern Gulf of Mexico reef-fish communities to the species equilibrium theory of insular biogeography. J. Biogeogr. 6, 49 (1979).

    CAS  Google Scholar 

  98. 98.

    Dupont, J. M., Hallock, P. & Jaap, W. C. Ecological impacts of the 2005 red tide on artificial reef epibenthic macroinvertebrate and fish communities in the eastern Gulf of Mexico. Mar. Ecol. Prog. Ser. 415, 189–200 (2010).

    ADS  Google Scholar 

  99. 99.

    Driggers, W. B. et al. Environmental conditions and catch rates of predatory fishes associated with a mass mortality on the West Florida Shelf. Estuar. Coast. Shelf Sci. 168, 40–49 (2016).

    ADS  CAS  Google Scholar 

  100. 100.

    Steidinger, K. A. & Ingle, R. M. Observations on the 1971 summer red tide in Tampa Bay, Florida. Environ. Lett. 3, 271–278 (1972).

    CAS  PubMed  Google Scholar 

  101. 101.

    Ogden, J. & Ebersole, J. Scale and Community Structure of Coral Reef Fishes: A Long-Term Study of a Large Artificial Reef. Mar. Ecol. Prog. Ser. 4, 97–103 (1981).

    ADS  Google Scholar 

  102. 102.

    Smith, L. Small rotenone stations: a tool for studying coral reef fish communities. Am. Museum Nat. Hist. 1 (1973).

  103. 103.

    Dayton, P. K. Experimental evaluation of ecological dominance in a rocky intertidal algal community. Ecol. Monogr. 45, 137–159 (1975).

    Google Scholar 

  104. 104.

    Farrell, T. M. Models and mechanisms of succession: an example from a rocky intertidal community. Ecol. Monogr. 61, 95–113 (1991).

    Google Scholar 

  105. 105.

    Peterson, C. H. et al. Long-term ecosystem response to the Exxon Valdez oil spill. Science. 302, 2082–2086 (2003).

    ADS  CAS  PubMed  Google Scholar 

  106. 106.

    Schwing, P. T. et al. Constraining the spatial extent of marine oil snow sedimentation and flocculent accumulation following the Deepwater Horizon event using an excess 210Pb flux approach. Environ. Sci. Technol. 51, 5962–5968 (2017).

    ADS  CAS  PubMed  Google Scholar 

  107. 107.

    Romero, I. C. et al. Large-scale deposition of weathered oil in the Gulf of Mexico following a deep-water oil spill. Environ. Pollut. 228, 179–189 (2017).

    CAS  PubMed  Google Scholar 

  108. 108.

    Smeltz, M. et al. A multi-year study of hepatic biomarkers in coastal fishes from the Gulf of Mexico after the Deepwater Horizon oil spill. Mar. Environ. Res. 129, 57–67 (2017).

    CAS  PubMed  Google Scholar 

  109. 109.

    Graham, W. M. et al. Oil carbon entered the coastal planktonic food web during the Deepwater Horizon oil spill. Environ. Res. Lett. 5 (2010).

  110. 110.

    Green, S. J., Akins, J. L., Maljković, A. & Côté, I. M. Invasive lionfish drive Atlantic coral reef fish declines. Plos One 7, e32596 (2012).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  111. 111.

    Albins, M. A. Invasive Pacific lionfish Pterois volitans reduce abundance and species richness of native Bahamian coral-reef fishes. Mar. Ecol. Prog. Ser. 522, 231–243 (2015).

    ADS  Google Scholar 

  112. 112.

    Dahl, K. A. & Patterson, W. F. III. Habitat-specific density and diet of rapidly expanding invasive red lionfish, Pterois volitans, populations in the northern Gulf of Mexico. Plos One 9, e105852 (2014).

    ADS  PubMed  PubMed Central  Google Scholar 

  113. 113.

    Albins, M. A. & Hixon, M. A. Invasive Indo-Pacific lionfish Pterois volitans reduce recruitment of Atlantic coral-reef fishes. Mar. Ecol. Prog. Ser. 367, 233–238 (2008).

    ADS  Google Scholar 

  114. 114.

    Raymond, W. W., Albins, M. A. & Pusack, T. J. Competitive interactions for shelter between invasive Pacific red lionfish and native Nassau grouper. Environ. Biol. Fishes 98, 57–65 (2015).

    Google Scholar 

  115. 115.

    Albins, M. A. Effects of invasive Pacific red lionfish Pterois volitans versus a native predator on Bahamian coral-reef fish communities. Biol. Invasions 15, 29–43 (2013).

    Google Scholar 

  116. 116.

    Dahl, K. A., Patterson, W. F. III., Robertson, A. & Ortmann, A. C. DNA barcoding significantly improves resolution of invasive lionfish diet in the northern Gulf of Mexico. Biol. Invasions 19, 1917–1933 (2017).

    Google Scholar 

  117. 117.

    Chagaris, D. et al. An ecosystem-based approach to evaluating impacts and management of invasive lionfish. Fisheries 42, 421–431 (2017).

    Google Scholar 

  118. 118.

    Rooney, N., McCann, K., Gellner, G. & Moore, J. C. Structural asymmetry and the stability of diverse food webs. Nature 442, 265–269 (2006).

    ADS  CAS  PubMed  Google Scholar 

  119. 119.

    National Marine Fisheries Service. Fisheries Economics of the United States, 2010. (2011).

  120. 120.

    Holland, S. M. & Ditton, R. B. Fishing trip satisfaction: a typology of anglers. North Am. J. Fish. Manag. 12, 28–33 (1992).

    Google Scholar 

Download references


We thank Captains Johnny Greene, Gary Jarvis, Sean Kelley, and Seth Wilson and their crews for help in sampling reefs, as well as numerous volunteers for help in the field. We also thank Miaya Glabach, Sarah Friedl, Josh Neese, Mike Norberg, and Brian Klimek for their assistance with ROV operations. This research was made possible in part by a grant from BP/The Gulf of Mexico Research Initiative/C-IMAGE I and II, and in part by funding from the Florida Fish and Wildlife Conservation Commission (Grants FWC-08304 and FWC-16188 to WFP).

Author information




W.F.P. designed the ROV survey. J.H.T., S.B.G. and J.P.L. conducted field work. J.H.T and S.B.G. read the ROV footage. J.P.L. and D.D.C. performed the data analysis. J.P.L. and W.F.P. wrote the main manuscript text. J.P.L. and S.B.G. prepared figures and tables. All authors reviewed the manuscript.

Corresponding author

Correspondence to Justin P. Lewis.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Lewis, J.P., Tarnecki, J.H., Garner, S.B. et al. Changes in Reef Fish Community Structure Following the Deepwater Horizon Oil Spill. Sci Rep 10, 5621 (2020).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.