The Center for the Integrated Modeling and Analysis of the Gulf Ecosystem

A tale of two Gulf spills: A research consortium of 19 institutions from 5 countries studying the impacts of oil spills on the Gulf of Mexico.

Task 6 Publications

Task 6 Publications (9)

One of the more important steps in understanding the ecosystem-level effects of anthropogenic disturbances on resident species is developing an accurate representation of the lethal and sub-lethal effects of these stressors. We develop methods for describing the impacts of oil on growth and mortality rates in fishes. We conducted a literature search to determine potential relationships between direct and indirect effects of exposure to oil, based on the frequency of lesions and body growth reduction. Data examining these effects with different exposure mediums were assessed and then input into four potential response models (a linear, step-wise, hockey-stick, and exponential model). We assessed the models using the Akaike Information Criterion. The most parsimonious and best fit model was the hockey-stick. This analysis will aid in identifying where future research on the impact of oil on fish should focus and also aid the development of ecosystem models on impacts of oil spills.

Dornberger, L.; Ainsworth, C.; Gosnell, S.; Coleman, F; Developing a polycyclic aromatic hydrocarbon exposure dose-response model for fish health and growth, Marine Pollution Bulletin, Available online 10 June 2016


Harmful algal blooms of the dinoflagellate Karenia brevis require an upwelling circulation to manifest along the coastline of the West Florida Continental Shelf. Too much upwelling, however, can impede bloom formation by increasing inorganic nutrient levels to the point where faster growing phytoplankton such as diatoms may out-compete the slower growing K. brevis, as occurred in 1998 and 2010. Both 2012 and 2013 experienced persistent upwelling, but only 2012 exhibited a robust harmful algal bloom. Here we examine the subtle differences in the coastal ocean circulation between those two years that led to the disparate bloom evolutions.

Source: Weisberg, R.H., Zheng, L., Liu, Y., Corcoran, A. A., Lembke, C., Hu, C., Lenes, J.M., Walsh, J.J., Karenia brevis blooms on the West Florida Shelf: A comparative study of the robust 2012 bloom and the nearly null 2013 event, Continental Shelf Research, Volume 120, 1 June 2016, Pages 106-121

Fig 5. - Karenia brevis abundance along the southwest Florida Shelf (A) between September and November 2012, (B) during December 2012, (C) between January to March 2013, and (D) between September and November 2013. Dots represent single samples collected via routine monitoring and multiple event response cruises conducted during each period. In cases where multiple concentrations were present at a single location, the samples with the greatest K. brevis abundance are overlaid onto samples with a lower abundance.

Over 50 years of multiple anthropogenic perturbations, Florida zooplankton stocks of the northeastern Gulf of Mexico declined ten-fold, with increments of mainly dominant toxic dinoflagellate harmful algal blooms (HABs), rather than diatoms, and a shift in loci of nutrient remineralization and oxygen depletion by bacterioplankton, from the sea floor to near surface waters. Yet, lytic bacterial biomass and associated ammonification only increased at most five-fold over the same time period, with consequently little indication of new, expanded “dead zones” of diatom-induced hypoxia. After bacterial lysis of intact cells of these increased HABs, the remaining residues of zooplankton biomass decrements evidently instead exited the water column as malign aerosolized HAB asthma triggers, correlated by co-traveling mercury aerosols, within wind-borne sea sprays. To unravel the causal mechanisms of these inferred decadal food web transitions, a 36-state variable plankton model of algal, bacterial, protozoan, and copepod component communities replicated daily time series of each plankton group's representatives on the West Florida shelf (WFS) during 1965–2011. At the lower phytoplankton trophic levels, 52% of the ungrazed HAB increments, between 1965–1967 and 2001–2002 before recent oil spills, remained in the water column to kill fishes and fuel bacterioplankton. But, another 48% of the WFS primary production then left the ocean’s surface as a harbinger of increased public health hazards during continuing sea spray exports of salts, HAB toxins, and Hg poisons. Following the Deepwater Horizon petroleum releases in 2010, little additional change of element partition among the altered importance of WFS food web components of the trophic pyramid then pertained between 2001–2002 and 2010–2011, despite when anomalous upwelled nutrient supplies instead favored retrograde benign, oil-tolerant diatoms over the HABs during 2010. Indeed, by 2011 HABs were back, with biomass accumulations equivalent to those found in 2001.

Ref: Walsh, J, J., Lenes, J. M., Darrow, B., Parks, A., Weisberg, R. H. (2016) Impacts of combined overfishing and oil spills on the plankton trophodynamics of the West Florida shelf over the last half century of 1965–2011: A two-dimensional simulation analysis of biotic state transitions, from a zooplankton- to a bacterioplankton-modulated ecosystem. Continental Shelf Research, 116, 54-73.

The Gulf of Mexico supports a high biological diversity and biomass of fish, seabirds, and mammals; in this region, multiple commercial and recreational fishing fleets operate providing economic resources for local populations. The Gulf is also the site of important oil and gas production and tourism. As a result of intensive human use, the Gulf is subject to various impacts, includig oil spills, habitat degradation, and anoxia. Management of this Large Marine Ecosystem requires an ecosystem-based management approach that provides a holistic approach to resource management. The Gulf of Mexico is managed as part of NOAA's Integrated Ecosystem Assessment Program (IEA). This program considers the developent of ecosystem models as a tool for ecosystem-based fisheries management (EBFM) and to support the different stages in the IEA process, particularly testing the effects of alternative management scenarios. As part of this program, we have parameterzed an Atlantis ecosysem model for the Gulf of Mexico, including major functional groups, physiographic dynamics, and fishing fleets. The Gulf of Mexico (GOM) Antanis model represents a collaboration between the University of South Florida, the University of Miami, the Southeast Fishereis Science Center, the National Coastal Data Developemnt Center, and other contributors.

The full article is available at the link below:

Ainsworth, C.H., Schirripa, M.J. and Morzaria-Luna, H. (eds.), (2015). An Atlantis ecosystem model for the Gulf of Mexico supporting Integrated Ecosystem Assessment. US Dept. Comm. NOAA Technical Memorandum NMFS-SEFSC-676. 149 pp.

A two-dimensional (2-D) ecosystem model, set within the De Soto Canyon ecotone of the Northern Gulf of Mexico (NGOM) and driven by 3-D flow fields from decoupled water and air circulation models, explores the daily food web and sedimentary consequences, as well as potential public health implications, of oil-, nutrient-, and overfishing-induced transitions of dominant particle transports to the sea floor: from changing vectors of copepod fecal pellets to those of marine snow over the annual period of 2010–2011. Recent spilled petrochemicals are found to minimally impact already decimated zooplankton populations on the West Florida shelf (WFS). They facilitate instead formation of marine snow macroaggregates. These recent oil effects just exacerbate other results of prior overfishing in the absence of major eutrophication along the eastern side of this ecotone. East of the De Soto Canyon ecotone, overly optimistic removals of piscivore fish stocks over the last half-century had already caused a trophic cascade, with ~95% of WFS marine pelagic herbivore losses occurring by 2010, before the Deepwater Horizon [DWH] blowout. By contrast, west of De Soto Canyon, zooplankton on the eutrophic Louisiana shelf were less impacted by overfishing, retaining order of magnitude more stocks of the same genera of copepods. Yet, during descent of modeled marine snow to the ~1200-m isobath of the upper slope, ~98% of the particulate import to the benthos is now mainly clay minerals of Mississippi River origin along the ecotone. The lithogenic particles are scavenged by aggregates from the whole water column, not only biotic plankton from just the near-surface euphotic zone. The model results here replicate concurrent time series of: 1) annual sediment accumulations, measured at the sea floor; 2) bimonthly onshore nutrient supplies, due to upwelling, forced remotely by the Loop Current; and 3) near-surface weekly phytoplankton changes, seen by satellite. Because of such minimal grazing stresses, after the most recent DWH oil reductions to ~50% of the remaining few WFS copepods, their decreased herbivory amounts to only an 8% fecal pellet contribution to the model's particle fluxes, with 92% settling as marine snow to the sea bottom of the continental slope in 2010–2011 By contrast, 12% of the biotic particle exports, entrained within marine snow, result from increased loadings of Mississippi River nutrients via opened flood gates, while 50% of the ungrazed sinking phytoplankton are fueled by decadal anomalies of increased nutrient supplies from greater upwelling by the Loop Current. Finally, of the exiting phytodetritus embedded within marine snow from the 2010–2011 water columns, 38% are due to uptake of autocthonous nutrients in the slope waters. But, explicit upward fluxes of nutrient-poor, petrochemical dissolved organic carbon [DOC] substrates for use by aphotic chemolithoautotrophic bacterial bioremediators, responding to DWH oil releases and external dissolved nutrients, are now ignored in the present model. These model results also relate downstream, wind-borne trajectories of evaded potential marine aerosolized toxins to both human asthma episodes and total mercury amounts in coastal soils, without explicit sea-air exchanges of initial aerosols. Thus, future simulation analyses, with instead extant numerical descriptions of breaking wave exports of oil, harmful algal bloom [HAB], and Hg aerosol poisons, must fully couple onshore 3-D aerial imports of this suite of marine toxins to direct causal factors of adjacent human health impacts, constrained by known surrogate asthma hospitalization rates. They must next deconvolve net multiyear oil and top-down NGOM indirect losses of herbivores, within now unbalanced marine food webs of an overfishing-induced trophic cascade, from other concurrent anthropogenic forcings, due to in situ mercury, pesticide, and radionuclide poisonings, within linked habitats of both water and air, subject to continued climate and biotic changes of the southeastern U.S. seaboard.

The full article is available at the link below:

Walsh, J. J., Lenes, J. M., Darrow, B. P., Parks, A. A., Weisberg, R. H., Zheng, L., et al. (2015). A simulation analysis of the plankton fate of the Deepwater Horizon oil spills. Continental Shelf Research, 107, 50–68.

Ammonium is a nutrient frequently preferred by microorganisms that photosynthesize at the base of the marine food web and remove atmospheric carbon dioxide via carbon fixation. Because their photosynthesis is concentrated in the ocean's thin euphotic zone, its nutrient concentrations are critical to oceanic carbon fixation. Identification of replacement processes for euphotic-zone ammonium thus becomes important. These processes were investigated in a two-experiment, Lagrangian field study that produced results consistent with an apparent inverse effect of wind forcing on upper-ocean ammonium concentrations. At low wind speeds (especially ≤ 4 ms− 1), continuous seawater sampling, supported by sulfur-hexafluoride (SF6) water-mass tracing and meteorological measurements, detected 1.1–4.4-km-wide boluses of surface seawater exhibiting ammonium enrichments that were 5 - to-10-fold above background. In the first experiment, ammonium maxima comprising the enrichment event disappeared at higher wind speeds. In the second experiment, which had consistently higher wind speeds, an ammonium event composed of such maxima was never found. The apparent correlation between elevated ammonium concentrations and low wind stress could therefore be viewed as potentially important for understanding ammonium cycling and carbon fixation in the ocean.

The full article is available at the link below:

Simulating cell death in the termination of Karenia brevis blooms: implications for predicting aerosol toxicity vectors to humans

ABSTRACT: To predict both waterborne and aerosolized toxin vectors associated with harmful algal blooms (HABs) of Karenia spp. in European, Asian, and North American waters, loss processes associated with distinct stages of bloom development, maintenance, and termination must be defined in relation to their toxins. In the case of Karenia brevis, exposure to brevetoxins (PbTx during maintenance phase is detrimental to marine life. In addition, release of PbTx-2,3 during cell death leads to respiratory difficulties in mammals. Human asthma attacks and chronic obstructive pulmonary disease occur once HAB toxins are aerosolized and transported to the coast. Here, we tested the hypothesis that heterotrophic bacterioplankton are a major source of mortality for Karenia HABs. A non-linear lysis term for simulation of K. brevis HAB termination on the West Florida Shelf was introduced, with the assumption that particle encounters of planktonic microalgae and bacteria can be described as the square of the phytoplankton biomass. This formulation also accounts for nutrient-limitation of K. brevis as a precondition for susceptibility to bacterial and viral attack, and potentially programmed cell death. Two model simulations were run of linear and non-linear lysis cases. Model output was compared against observed weekly maximum K. brevis concentrations, with statistical metrics calculated over 3 HAB phases during 2001. Introduction of the non-linear lysis term increased the modeling efficiency by 0.68 due to improved reproduction of the bloom termination.

Ref: Lenes, J. M., Walsh, J. J., Darrow, B. P. Simulating cell death in the termination of Karenia brevis blooms: implications for predicting aerosol toxicity vectors to humans, Marine Ecology Progress Series, 493: 71-81, 2013.



A probabalistic representation of fish diet compositions from multiple data sources: A Gulf of Mexico case study

Ecosystem models can be extremely useful tools for ocean resource managers to determine how commercial fisheries can respond to a specific management strategy. C-IMAGE researchers out of the University of South Florida with colleagues from the Florida Fish and Wildlife Conservation Commission looked at fish diet for two non-commercially relevant species in the Gulf of Mexico to produce Maximum Liklihood Estimates (MLEs).  MLEs describe how much a specific prey item constitutes a predator's diet.  These values are used to produce a food web diagram that is ultimately used to fine tune the Atlantis ecosystem model.

Ref: Masi, M., Ainsworth, C., Chagris, D. A Probabilistic Representation of Fish Diet Compositions from Multiple Data Sources: A Gulf of Mexico Case Study, Ecological Modeling, 284, 60-74, 2014.



Generalized Additive Models Used to Predict Species Abundance in the Gulf of Mexico: An Ecosystem Modeling Tool

Spatially explicit ecosystem models of all types require an initial allocation of biomass, often in areas where fisheries independent abundance estimates do not exist. A generalized additive modelling (GAM) approach is used to describe the abundance of 40 species groups (i.e. functional groups) across the Gulf of Mexico (GoM) using a large fisheries independent data set (SEAMAP) and climate scale oceanographic conditions. Predictor variables included in the model are chlorophyll a, sediment type, dissolved oxygen, temperature, and depth. Despite the presence of a large number of zeros in the data, a single GAM using a negative binomial distribution was suitable to make predictions of abundance for multiple functional groups. We present an example case study using pink shrimp (Farfantepenaeus duroarum) and compare the results to known distributions. The model successfully predicts the known areas of high abundance in the GoM, including those areas where no data was inputted into the model fitting. Overall, the model reliably captures areas of high and low abundance for the large majority of functional groups observed in SEAMAP. The result of this method allows for the objective setting of spatial distributions for numerous functional groups across a modeling domain, even where abundance data may not exist.

Ref: Drexler, M., Ainsworth, C. The use of zero-inflated generalized additive models to predict species abundance distributions in the Gulf of Mexico: a tool for ecosystem modeling, PLOS ONE, 2013



Podcasts from The Loop

The Loop is a series of podcasts which take an in depth look at C-IMAGE research. Partnering with Mind Open Media reporters Ari Daniel Shapiro and David Levin, our researchers share the importance of their studies and how they help our understanding of oil spills. David and Ari have produced eight podcasts and have more in the queue. The podcasts are linked below. Plug in and learn about our research!

Listen to More Podcasts!