 |
||||
|
||||
|
|
|
 |
||
|
||||
      |
Assessing Ecological Risks Posed by a Ballast Water Disinfectant
This project is no longer current. Please see the Research Programs page for a list of current research projects.
Collaborators:
S. Bartell, Cadmus Group, Marysville, TN
Dr. Larissa Sano, University of Michigan, MI
Executive Summary
The unintentional release and establishment of nonindigenous aquatic species in the Great Lakes has profoundly impacted this ecosystem through changes in populations of native organisms and alterations of critical ecological processes. Approximately 25% of the nonindigenous species that currently inhabit the Great Lakes were introduced from the ballast water of foreign vessels. To reduce the risk of future invasions, the application of chemical biocides (or 'biocide') to ballast water is currently being evaluated. This work extends the work on glutaraldehyde as a potential disinfectant for ballast water for vessels designated as no-ballast-on-board (NOBOB) by specifically evaluating the risk of discharge of ballast water residuals in Duluth Harbor one of the busiest ports in the upper Great Lakes. This past year a risk assessment for the release of glutaraldehyde into Duluth Harbor using CASM was completed and the manuscript was drafted. In addition, a risk tradeoff analysis for the use of glutaraldehyde as a potential biocide versus invasions was performed to evaluate the practicality of biocides as an approach to prevention for invasions from NOBOB vessels.
Project Rationale
The
unintentional release and establishment of nonindigenous aquatic species
in the Great Lakes has profoundly impacted this ecosystem through changes
in populations of native organisms and alterations of critical ecological
processes. Approximately 25% of the nonindigenous species that currently
inhabit the Great Lakes were introduced from the ballast water of foreign
vessels.
To reduce the risk of future invasions, the application of chemical biocides (or ‘biocide’) to ballast water is currently being evaluated. Biocides may be most effective in treating vessels classified as NOBOB (no ballast on board). These vessels carry between 50 and 210 metric tons of unpumpable ballast water and sediment, which can be important sources of exotic species. A large percentage of the foreign vessels that enter the Great Lakes are classified as NOBOB. Treatment of NOBOBs has great potential because these vessels contain a relatively small amount of ballast material and undergo cross-transfer of lake water into ballast tanks, which dilutes the amount of biocide potentially discharged into the Great Lakes.
One of the major drawbacks of biocide treatment is the potential for ecological impacts following biocide release into receiving waters. For many of the biocides considered for ballast water treatment, acute toxicity may result from the release of high concentrations of the biocide. In addition, many of the impacts will likely be in port and harbor areas, many of which are already subject to other chemical stressors.
Project Objectives
This project will evaluate the potential for environmental effects associated with the use of glutaraldehyde, which has been considered for treatment of NOBOB vessels entering the Great Lakes. The objective of this project is to quantify the release of glutaraldehyde at a Great Lakes’ port and estimate the ecological risks associated with such releases.
Several important issues must be considered in developing protocols for using specific disinfectants or biocides to treat ballast on foreign ships entering the Great Lakes:
- The operational safety of the chemical.
- The chemical potency in controlling a wide range of organisms.
- The effectiveness of the chemical in the ballast tank environment.
- The ability to maintain effective exposures in ballast.
- The risks to non-target species following discharge of the chemical into the aquatic environment.
Previous work completed in FY02 addressed issues (2) – (4) above and demonstrated the potential effectiveness of glutaraldehyde and other disinfectants in imposing significant mortalities to several species of aquatic organisms representative of potentially invasive species that inhabit ballast water and sediments. A summary of the results from the previous work was published in Aquatic Invasions as a means of communicating our research to a broader audience. Effective exposure concentrations of glutaraldehyde have been determined in laboratory studies, including the simulation of ballast tank conditions. Additional studies have characterized rates of glutaraldehyde degradation under ballast tank conditions and resulted in a published manuscript. Based on these studies, risks posed by the discharge of residual glutaraldehyde were estimated for non-target species in Great Lakes receiving waters using the CASM model. The risks were small but there were likely some impacts particularly on phytoplankton because of their sensitivity. The potential hazard was also seasonal with more hazard in the spring and fall when there would be minimal degradation of the glutaraldehyde. The model did not take into account realistic dilutions so the risks are considered the maximum that may occur. The risks are also based on simulated release concentrations which remain to be confirmed through testing in vessels. Should the releases be of the magnitude simulated the use of glutaraldehyde should provide effective and environmentally sound. The risk tradeoff analysis complemented the risk assessment based on the CASM model showing that the risks from the biocide residuals would be far smaller spatially and temporally than the risks from a new invasion.
Project Accomplishments
2006 | 2005 | 2004 | 2003
2006 Accomplishments
During FY 2006, efforts on this project focused on documenting the work that had been completed last year. The project focused on three manuscripts: a manuscript describing decay dynamics of glutaraldehyde, a risk-risk trade off analysis, and an ecological effects model.
Final revisions were made to the manuscript describing the application of a model for simulating the potential decay of the biocide, glutaraldehyde, in the tank of unballasted overseas vessels in the Great Lakes. This model utilized linear degradation equations to simulate varying conditions in tanks and travel time to estimate final concentrations in tanks prior to release of material into the Duluth-Superior Harbor.
The second manuscript, applying a risk-risk tradeoff analysis to biocide treatment of vessels was submitted and reviewed and is currently under revision to address feedback from reviewers. During this fiscal year, efforts were directed at finalizing the manuscript by refining the risk-risk framework applied to ecological effects of two stressors (aquatic invasive species and the biocide, glutaraldehyde). This paper represents a novel application of risk-risk tradeoff approaches, which are increasingly being advocated as a desired approach to compare the range of outcomes associated with different regulatory approaches. In general the manuscript was well received, and efforts are currently underway to address the concerns and suggestions of the reviewers.
Finally, final revisions are underway on a manuscript that describes simulation results from the application of the Comprehensive Aquatic Simulation Model (CASM) to the Duluth-Superior Harbor, Minnesota. This model was calibrated with available environmental and ecological data from this site. The results from the ballast tank decay model were then used to simulate the potential range of release concentrations in the Duluth-Superior Harbor and associated responses of different trophic levels at different times of the year. During this fiscal year, the text for the manuscript was refined. In addition, the modeling program CORMIX (a USEPA-supported mixing zone model) was used to compare concentrations used in the ecological effects model for glutaraldehyde with possible actual release concentrations. This model is a continuous point source discharge model, emphasizing boundary interactions in predicting steady-state mixing behavior. The use of this model represented a necessary refinement to this modeling effort, as very simple assumptions were originally made regarding the dilution of glutaraldehyde once released into the harbor region. The manuscript will be submitted for review in the first quarter for FY07.
2005 Accomplishments
Biocide decay model
A manuscript describing the biocide decay model was written which describes the potential degradation of glutaraldehyde should it be used to treat ballast tanks in unballasted overseas vessels trading on the Laurentian Great Lakes. The manuscript is currently in press in Marine Pollution Bulletin (2005).
Comprehensive Aquatic Simulation Model
A version of the Comprehensive Aquatic Systems Model (CASM) reconfigured for conditions at the Duluth-Superior Harbor, MN, was used to simulate trophic conditions in Duluth Harbor, CASM was structured to include five phytoplankton, five zooplankton, five benthic invertebrate, three planktivorous fish, five omnivorous fish, and three piscivorous fish populations representative of this complex system.
The risks of glutaraldehyde to different trophic levels, as estimated by CASM_Duluth, depended largely on the exposure scenario: Although data from laboratory experiments indicate a high sensitivity of algae to glutaraldehyde, results from this model indicate that annual phytoplankton biomass is largely unaffected under almost all simulation conditions (Table 1). This discrepancy primarily reflects the dominance of cyanobacteria populations in the models. Because the toxicity values for cyanobacteria were relatively high and the populations tended to peak in August and September (prior to significant increases in glutaraldehyde concentrations), these populations were largely unaffected by glutaraldehyde exposure. However, of the individual phytoplankton populations, Melosira spp. was the most sensitive to a range of exposure concentrations. This sensitivity was a function both of the relatively low EC50 data for the species and of the increasing glutaraldehyde concentrations in the spring, which overlapped with peak diatom abundance. If this effect is real, then the potential sensitivity of diatoms may have important implications for zooplankton and therefore also for planktivorous fish populations.
In contrast to phytoplankton, total annual biomass of both zooplankton and planktivorous fish were predicted to decrease only under the most conservative exposure scenario (Table 1). For zooplankton, the risk under this specific exposure scenario was distributed across populations, while for planktivorous fish, only the emerald and spotted shiner populations were predicted to incur risks for decreases in annual biomass. Risks to all other trophic levels were essentially zero, except under the most conservative exposure condition. This suggests that given the current model configuration, ecological risks of glutaraldehyde should be small both across trophic levels and to most populations.
Risk Tradeoff Analysis
The results from both the CASM simulations and from the chronic toxicity bioassays using glutaraldehyde were analyzed in the context of a risk tradeoff framework. This component of the project was added to better understand the nature of the risks posed by environmental release of glutaraldehyde. In this framework, the ecological risks posed by the continued release of non-indigenous species into the Great Lakes (i.e., the "target risk;) were compared to the risks associated with the release of biocide residual (i.e., the "countervailing risk"). Several different elements of the target and countervailing risk were assessed including the potential magnitude and spatial extent of effects, the types of organisms and populations affected, and the level of certainty of impacts. For the target risk, the impacts of current non-indigenous species in the Great Lakes were considered in addition to the likelihood of future invaders. This information was compared with data from laboratory bioassays and an ecosystem-level effects model for the candidate biocide, glutaraldehyde.
Results from this analysis generally indicate that the primary risks from glutaraldehyde come from mortality effects to different taxa, primarily phytoplankton. In contrast, the impacts of invasive species are more diverse and include large-scale restructuring of benthic communities, alterations in foraging patterns, and increases in harmful algal species (Table 2). The impacts of invasive species are also characterized by larger spatial and temporal scales compared to the relatively isolated spatial and temporal distribution of any biocide effects. Estimates of the probability of glutaraldehyde risks are hampered by higher uncertainty of potential environmental concentrations and indirect ecological impacts. For aquatic nonindigenous species, estimates of risk probabilities are particularly difficult given the high uncertainty surrounding propagule pressure, establishment potential, invasiveness, and impact potential. Overall, the ecological risks of glutaraldehyde are low compared to those posed by ANIS, due primarily to the small spatial distribution and limited temporal persistence of any potential effects. These results provide a framework both for evaluating chemical treatment technologies in the Great Lakes and maritime ports, and for characterizing the application of risk tradeoff analysis for comparing ecological impacts.
Table 1. Estimates of risk of change in biomass for different taxa based on several exposure scenarios. Probability values of 1.00 indicate a 100% chance of decreased biomass based on Monte Carlo simulations in CASM_Duluth. Scenarios described as "at 3 sites" refer to conditions in which vessel traffic is assumed to be evenly divided between three different grain elevator sites within the harbor. Blank cells indicate zero probability of biomass change.
Table 2. Risk tradeoff summary for ecological risks of glutaraldehyde versus those posed by ANIS. The risk elements used to compare ecological risks were generally derived from the risk tradeoff analysis proposed by Graham and Wiener 1995 (Confronting risk tradeoffs in Risk versus Risk: Tradeoffs in Protecting Health and the Environment, Editors, J.B. Graham, J.D. Wiener. Harvard University Press, Cambridge MA, 1-41).
2004 Accomplishments
Biocide decay model
Figure 1. Output from Crystal Ball® Monte Carlo simulations of predicted concentrations at port Oswego, NY. The decay dynamics in the initial treatment concentration were similar regardless of whether the worst-case or best-case re-ballasting scenario was modeled.
Figure 2. Output from Crystal Ball® Monte Carlo simulations for predicted concentrations at Duluth Harbor, MN, over the model simulation period. Data are presented for the best-case re-ballasting scenario and the worst-case re-ballasting scenario.
Table 1. Sensitivity parameters for predicted release concentrations at Duluth Harbor, MN. Sensitivity values are given for parameters under both the "best-case" scenario and the "worst-case" scenario.
Comprehensive Aquatic Simulation Model
A version of the Comprehensive Aquatic Systems Model (CASM) was reconfigured for conditions at the Duluth-Superior Harbor, MN, which receives a large volume of ballast water discharge and is thus potentially at higher risk for ecological impacts on resident biota. To simulate trophic conditions in Duluth Harbor, CASM was structured to include five phytoplankton, five zooplankton, five benthic invertebrate, three planktivorous fish, five omnivorous fish, and three piscivorous fish populations representative of this complex system. Reference simulations have been completed and the available toxicity data from glutaraldehyde has been entered into the "effects factor" component of the model. Current simulations are underway to assess the potential impacts of glutaraldehyde release and to characterize the potential for direct toxic effects and indirect ecological effects.
Risk Tradeoff Analysis
The results from both the CASM simulations and from the chronic toxicity bioassays using glutaraldehyde are being analyzed in the context of a risk tradeoff framework. This component of the project was added in response to guidelines set forth by the International Maritime Organization in 2004, in which one of the criteria for treatment standards is that "treatment technology should not cause more, or greater, environmental impacts than it solves." In order to understand better the nature of the risks posed by environmental release of glutaraldehyde, a risk tradeoff typology initially proposed by Gray and Graham (1995) will be applied to biocide treatment. In this framework, the ecological risks posed by the continued release of non-indigenous species into the Great Lakes (i.e., the "target risk;" sensu Gray and Graham 1995) will be compared to the risks associated with the release of biocide residual (i.e., the "countervailing risk"). Several different elements of the target and countervailing risk will be assessed including the potential magnitude and spatial extent of effects, the types of organisms and populations affected, and the level of certainty of impacts. For the target risk, the impacts of current non-indigenous species in the Great Lakes will be considered in addition to the likelihood of future invaders. This information will be compared with data from laboratory bioassays and an ecosystem-level effects model for the candidate biocide, glutaraldehyde. The results from this analysis should provide a framework for evaluating chemical treatment technologies both in the Great Lakes and in other maritime ports.
2003 Accomplishments
The portion of the ballast water work that focused on investigating potential disinfectants for the treatment of ballast water was completed. The work on this portion of the study was generated publications for the research. The first publication on the use of glutaraldehyde as a potential disinfectant was completed and is in press in the Journal of Great Lakes Research. The second publication comparing glutaraldehyde with a surfactant adjuvant, hypochlorite and some work with Sea KleenÔ was completed and submitted to the Journal of Great Lakes Research. In addition, a NOAA Technical Memorandum was produced to describe the degradation of glutaraldehyde under a variety of conditions for use as base information for the risk assessment research.
During the first six months of this project, a glutaraldehyde decay model for unballasted transoceanic vessels (i.e., NOBOBs, no-ballast-on-board) was developed. The model has two components: decay during a hypothetical trans-Atlantic transit and decay through the Great Lakes, after a vessel has entered the Gulf of St. Lawrence. For the first component (the trans-Atlantic segment), the port of origin was assumed to be Antwerp, Belgium. The output from this model segment indicates that glutaraldehyde concentrations do not vary much based on month or length of transit. This lack of decay occurs because of initial concentration of glutaraldehyde is high (500 mg L-1) and the sea surface temperatures of the Atlantic are low, at the relevant latitudes. The minimal decay results in predicted concentrations at the Gulf of St. Lawrence, which vary only by approximately 6% (with the highest average predicted concentration occurring in April (462 mg glutaraldehyde L-1) and the lowest average predicted concentration occurring in August (436 mg glutaraldehyde L-1). For the portion of transit within the Great Lakes, the model was run for a worst-case scenario and a best-case scenario: The former scenario assumes that a vessel reballasts at the port of Windsor, Ontario. Because most of the glutaraldehyde decay occurs after reballasting at lower concentrations, this represents a conservative estimate of biocide concentration at Duluth, since less time elapses between reballasting and release at Duluth Harbor. The best-case scenario involves reballasting at Oswego, Ontario. This port is located on Lake Ontario, and permits more time between reballasting and release at Duluth allowing for longer degradation at the lower concentrations. From these scenarios, the concentration of glutaraldehyde released at Duluth will depend primarily on the dilution factor associated with reballasting: Vessels that take on a large amount of water during reballasting operations will experience more dilution and a lower glutaraldehyde concentration, which facilitates degradation (Figure 2). Other factors such as duration of transit, time of year, and port of reballasting have a more minor effect.
Figure 2. Predicted concentration of glutaraldehyde released at the Duluth-Superior Harbor. Data are presented for a worst case scenario (reballasting at Windsor) and a best case scenario (reballasting at Oswego).
Progress has also been made on identifying a hydrodynamic model for the Duluth Harbor, which will be employed to predict dispersion and dilution effects associated with the release of the biocide. This output will be used to predict glutaraldehyde concentrations for different quadrants of the harbor and these data will form the basis for exposure conditions for the ecological effects model (using CASM). The next six months of work will focus on developing the trophic web model for the outer channel of the harbor and on connecting the output of the hydrodynamic model to the ecological effects model.
References
Gray, G.M. and Graham, J.D. 1995. Regulating Pesticides in Risk vs. Risk (Graham, J.D., and Wiener, J.B., eds). Harvard University Press.
Publications
Sano, LL, Bartell, SM, Landrum, PF. 2005. Decay model for biocide treatment of unballasted vessels: Application for the Laurentian Great Lakes. Marine Pollution Bulletin 50: 1050-1060.
LANDRUM, P. F., L. L. SANO, M. A. MAPILI, E. GARCIA, A. M. KRUEGER and R. A. Moll. Degradation of chemical biocides with application to ballast water treatment. NOAA Technical Memorandum GLERL-123. NOAA, Great Lakes Environmental Research Laboratory, Ann Arbor, MI, 37 pp. (2003). ftp://ftp.glerl.noaa.gov/publications/tech_reports/glerl-123/
Sano, L.L., R.A. Moll, A.M. Krueger, and P.F. Landrum. 2003. Assessing the potential efficacy of glutaraldehyde for biocide treatment of unballasted transoceanic vessels. J. Great Lakes Res. 29:545-557.
Sano, L.L., Mapili, M.A., Krueger, A., Garcia, E., Gossiaux, D., Phillips, K., Landrum, P.F. 2004. Comparative efficacy of potential chemical disinfectants for treating unballasted vessels. J. Great Lakes Res. 30:201-216.
Sano, L.L., Krueger, A.M., and Landrum, P.F. 2005. Chronic Toxicity of glutaraldehyde: Differential sensitivity of three freshwater organisms. Aquat. Toxicol. 71:283-296.
Sano, L.L. and Landrum, P.F. 2005. Evaluation of different biocides for use in treating overseas unballasted vessels entering the Great Lakes. Aquatic Invasions. 16(3): 1-11.
