^^PRESS] Environmental Toxicology and Chemistry, Vol. 28, No. 2, pp. 346-353, 2009 ? 2009 SETAC Printed in the USA 0730-7268/09 $12.00 + .00 CONCENTRATED SODIUM CHLORIDE BRINE SOLUTIONS AS AN ADDITIONAL TREATMENT FOR PREVENTING THE INTRODUCTION OF NONINDIGENOUS SPECIES IN THE BALLAST TANKS OF SHIPS DECLARING NO BALLAST ON BOARD SCOTT SANTAGATA,*! KAROLINA BACELA,? DAVID F. REID,? KEVIN A. MCLEAN,! JILL S. CoHEN,t JEFFERY R. CORDELL,|| CHRISTOPHER W. BROWN,! THOMAS H. JOHENGEN,# and GREGORY M. Ruizf I'Marine Invasions Laboratory, Smithsonian Environmental Research Center, 647 Contees Wharf Road, Edgewater, Maryland 21037, USA ^Department of Invertebrate Zoology and Hydrobiology, University of Lodz, 90 237 Lodz, Poland ?Great Lakes Environmental Research Laboratory, National Oceanic and Atmospheric Administration, 2205 Commonwealth Boulevard, Ann Arbor, Michigan 48105, USA ||School of Aquatic and Fishery Sciences, University of Washington, 1122 NE Boat Street, Seattle, Washington 98195, USA #Cooperative Institute for Limnology and Ecosystem Research, University of Michigan, 440 Church Street, Ann Arbor, Michigan 48109, USA (Received 24 July 2008; Accepted 11 August 2008) Abstract?Currently, seawater flushing is the only management strategy for reducing the number of viable organisms in residual sediments and water of ballast tanks of vessels declaring no ballast on board (NOBOB) that traffic ports of the eastern United States. Previously, we identified several species of freshwater and brackish-water peracarid crustaceans able to survive the osmotic shock that occurs during open-ocean ballast water exchange and, potentially, to disperse over long distances via ballasted ships and NOBOB vessels. We tested the efficacy of concentrated sodium chloride brine solutions as an additional treatment for eradicating the halotolerant taxa often present in the ballast tanks of NOBOB ships. The lowest brine treatments (30 ppt for 1 h) caused 100% mortality in several species of cladocerans and copepods collected from oligohaline habitats. Several brackish-water peracarid crustaceans, however, including some that can survive in freshwater as well, required higher brine concentrations and longer exposure durations (45?60 ppt for 3?24 h). The most resilient animals were widely introduced peracarid crustaceans that generally prefer mesohaline habitats but do not tolerate freshwater (required brine treatments of 60?110 ppt for 3?24 h). Brine treatments (30 ppt) also required less time to cause 100% mortality for eight taxa compared with treatments using 34 ppt seawater. Based on these experiments and published data, we present treatment strategies for the ballast tank biota often associated with NOBOB vessels entering the Great Lakes region. We estimate the lethal dosage of brine for 95% of the species in our experiments to be 110 ppt (95% confidence interval, 85?192 ppt) when the exposure time is 1 h and 60 ppt (95% confidence interval, 48?98 ppt) when the exposure duration is 6 h or longer. Keywords?Introduced species Ballast water Zooplankton Great Lakes Salinity tolerance INTRODUCTION The main vector responsible for many of the nonindige- nous species established in the estuaries and freshwater hab- itats of the eastern United States is ballast water of com- mercial ships originating from ports of the North, Baltic, Mediterranean, and Black-Azov seas [1?3]. Ports of the Bal- tic and North seas dominate the routes of U.S.-bound ship- ping traffic from Europe, and many of these ports are found in freshwater or mesohaline systems that provide an envi- ronmental match sufficient to allow the establishment of non- native organisms [4] (http://www.glerl.noaa.gov/seagrant/ GLERLUpdates/GLERLUpdates2007.html). Overall, ships carrying low-salinity water from source ports in northern Europe pose a significant risk to the freshwater and estuarine habitats of the eastern United States through the introduction of nonnative species. A proportion of ships entering the Great Lakes and estuarine port systems of the eastern United States are not in a ballasted condition, reporting no ballast on board (NOBOB) [5,6]. Sed- * To whom correspondence may be addressed (scott.santagata@gmail.com). The present address of S. Santagata is Biology Department, C.W Post Campus of Long Island University, 720 Northern Boulevard, Brookville, NY 11548-1300, USA. Published on the Web 9/23/2008. iments and residual water in the ballast tanks of these ships often contain adults, larvae, and resting stages of many taxa [7]. As a consequence of off-loading and loading cargo, many commercial ships take on and discharge ballast water while in port, potentially releasing nonnative species [6,8]. To min- imize the risk of species transfers from NOBOB vessels, ships entering the Great Lakes and St. Lawrence Seaway recently have been required to treat residuals with open-ocean seawater (seawater flushing [9]) before entering this region. Previously, we identified several freshwater and brackish-water taxa that can survive exposure to oceanic seawater (34 ppt) and are likely to be dispersed via the operations of commercial ships [10]. Furthermore, some ships entering the Great Lakes through the St. Lawrence Seaway are unable to follow the seawater flushing policy because of load-line restrictions and seaway draft constraints [11] (http://www.glerl.noaa.gov/res/ projects/nobob/products/NOBOBFinalReport.pdf), and these ships may have low-salinity residual water in their tanks. Both of these factors create a potential risk of introducing nonin- digenous species to the Great Lakes, especially if the ballast tank organisms originate from distant ports of similar envi- ronmental conditions [6,12]. Regulators responsible for the Great Lakes presently are faced with limited choices for deal- ing with such ships and may require them to return to an 346 Eradicating ballast tank organisms with sodium chloride Environ. Toxicol. Chem. 28, 2009 347 alternate exchange zone either at or beyond the mouth of the St. Lawrence River (a 2- to 4-d round trip) or to seal their tanks while in the Great Lakes. For these reasons, we con- ducted experiments to evaluate the efficacy of concentrated sodium chloride brine solutions as a treatment method to aug- ment the current ballast water management strategy for NO- BOB vessels based on seawater flushing. This treatment may provide a simple and effective method for eliminating halo- tolerant taxa within the sediments and residual water found in the ballast tanks of NOBOB vessels with minimal impacts on adjacent native flora and fauna of the Great Lakes. MATERIALS AND METHODS The goal of the present study was to build on our previous experiments investigating the salinity tolerance of native and introduced species [10]. In particular, we wanted to evaluate the use of concentrated sodium chloride brine solutions as a means of quickly eliminating freshwater and mesohaline taxa capable of surviving exposure to the average salinity of open- ocean seawater (34 ppt). We used our previously established experimental protocol based on seawater to identify seawater- intolerant and seawater-tolerant taxa and to compare their sur- vivorship among several brine treatments. Animal collection and sites Initial experiments focused on native and introduced taxa of the Great Lakes (Lake Erie, Toledo, OH, USA; Lake Huron, Alpena, MI, USA; and Lake Michigan, Traverse City, MI, USA). The geographic scope was then expanded to source populations of species known to have been introduced to fresh- water and brackish-water habitats of the United States (from the port of Rotterdam, The Netherlands; Curonian Lagoon, Lithuania; and the Vistula and Oder rivers, Poland). We also conducted experiments on native and introduced species of mesohaline habitats from two estuaries on the east and west coasts of the United States (Rhode River, MD, USA, and San Francisco and Suisun bays, CA, USA). Most animals were collected directly from the field using a variety of sampling methods (plankton net, light trap, or seine net), sorted in the laboratory adjacent to the collection site, and then used within 12 h in the experiments outlined below. A small subset of species were reared in the laboratory from broods of adult animals {Rhithropanopeus harrisii and Palaemonetes pugio), which were held for a short time (days) in the laboratory until they developed into planktonic stages for experiments. Seawater-tolerance protocols Experiments were conducted in the laboratory with treat- ments simulating the two commonly used ballast water ex- change (BWE) methods, flow-through (F-T) and empty-refill (E-R). Although for the present study we were not specifically interested in the osmotic effects of BWE, we used this protocol because it discriminated between seawater-tolerant and sea- water-intolerant species. Animals were assigned randomly to one of three treatments: Control animals were kept at the sa- linity and temperature observed in their habitat at the time of collection, E-R animals experienced an abrupt shift to 34 ppt seawater, and F-T animals experienced a stepwise increase in salinity from ambient conditions to 14, 24, and 34 ppt seawater over a total period of 3 h (1 h per step). Each experiment included four replicates (bowls of animals) for each BWE method (F-T and E-R) and control treatment. Each replicate contained 10 individuals for each salinity treatment in a 100- ml glass bowl (120 total animals required for the control and experimental treatments). Each bowl was 4 cm in height and 8 cm in diameter, with a high surface to volume ratio to sustain high dissolved oxygen concentrations. Seawater solutions were made from filtered water (pore size, 0.2 jjim) from the animal collection site and adjusted to the target salinity with artificial sea salt. Each experiment had a time limit of 48 h that included plausible minimum exposure times before discharge for ships conducting BWE in open ocean [2,13]. All treatments were maintained in temperature-controlled incubators set at the am- bient water temperatures measured during field collections. The water in the F-T treatment was changed at 1, 2, and 3 h, adjusting the salinity upward at each time point. Water in all treatments was changed at 24 h to limit any buildup of me- tabolites that might stress the organisms. No food was provided for these short-duration experiments. Survivorship was recorded for all treatments after exposure times of 1, 2, 3, 24, and 48 h. If 100% mortality occurred in all replicates of a particular treatment (F-T or E-R), further observations at any subsequent time points usually were dis- continued for that treatment. In all cases, survivorship was assessed visually using a stereomicroscope (X10-60) to detect any muscular contractions or respiratory currents produced by the animals. In the absence of any noticeable movement, even after contact with a probe, an animal was considered to be dead. At this point, the tissues of most animals were compro- mised and lost their natural pigment. When all individuals in a given treatment appeared to be dead or when 48 h had passed, animals were transferred back to their habitat water to assess whether any recovery occurred during the next 1 to 24 h. This latter step was particularly important for determining the sur- vivorship of copepods and peracarid crustaceans. Brine-tolerance protocol We modified our seawater-tolerance protocol to track the survivorship over time of animals abruptly exposed to sodium chloride solutions of 30, 45, and 60 ppt. A group of seawater- tolerant species collected from the San Francisco and Suisun bays was exposed to sodium chloride solutions of 85 and 110 ppt to evaluate the efficacy of short-term brine exposures (1-3 h). Brine experiments had the same number of replicate bowls (n = 4) per control and treatment concentration and 10 individuals per bowl, as listed previously. Cumulative survivorship was recorded at exposure times of 1, 2, 3, 6, 24, and 48 h. In a few experiments, when op- portunity allowed, survivorship also was scored at 12 h. Brine solutions were composed of filtered water (pore size, 0.2 u,m) from the collection site and adjusted to the target concentra- tions of sodium chloride (S-9888; Sigma-Aldrich, St. Louis, MO, USA) using a YSI 85 conductivity meter (YSI Incor- porated, Yellow Springs, OH, USA) except for concentrations at 85 and 100 ppt, which exceeded the measuring range of the meter and were based on weight per volume. To ensure that any mortality was caused by changes in salt concentration and not by decreases in dissolved oxygen, the medium in all trials was replaced at 24 h. All treatments were conducted at the ambient temperature of the habitat from which the animals were collected. Survivorship was scored as discussed previ- ously. 348 Environ. Toxicol. Chem. 28, 2009 S. Santagata et al. Data analyses Previously, we assessed the proportion of species from low- salinity habitats that could withstand exposure to 34 ppt sea- water [10]. In the present study, we summarized these exper- imental results as a function of mortality rate and treatment. For the purposes of the present study, if any individuals in either the F-T and E-R treatments survived after 48 h, this species was considered to be seawater tolerant. The effective- ness of the brine treatments was evaluated among seawater- intolerant and seawater-tolerant species by the relative expo- sure time necessary to cause 100% mortality in the experi- mental treatments as compared to the survivorship in the con- trol treatments. The E-R and various brine treatments all abruptly exposed animals to elevated salt solutions, so mor- tality rate was only compared among these treatments and not to the F-T treatments. Using Probit (Ver 1.5; U.S. Environ- mental Protection Agency, Washington, DC), we estimated the sodium chloride concentrations capable of causing 100% mor- tality in 95% of the species in our experiments for several treatment durations. This analysis was preferred over other toxicological analyses, such as listing the time required to cause 95% mortality of individuals for each species, because we wished to estimate treatment levels that would eradicate groups of taxa. Based on published data regarding the preferred habitat salinity ranges of species in our experiments [4], we used a logistic regression analysis to explore which factors (overall salinity range, habitat salinity and temperature at the time of collection) best predicted the ability of a species to survive in either the seawater-tolerance experiments (48 h) or interme- diate brine exposures (45 ppt sodium chloride) for 3 h or more. In general, we categorized species into three salinity groups based on their habitat preferences: Freshwater and oligohaline animals were those confined to habitats of 0 to 5 ppt, brackish- water animals were those found in habitats that range from 0 to 20 ppt, and mesohaline species were those that usually occupy a salinity range of 5 to 25 ppt. Finally, the results from all our brine experiments were summarized as a function of the habitat salinity preferences of the species included, the cumulative proportion of the species eliminated, the effective brine concentration, and the exposure time. Additional sup- porting data for these regression models were gathered from the literature [7,14-26]. RESULTS Seawater-tolerance experiments Overall, 44% (F-T method) and 32% (E-R method) of the 54 taxa in our seawater-tolerance experiments had some sur- vivorship at 48 h (Fig. 1). Among the species without 100% mortality, 42% (F-T method) and 53% (E-R method) of these taxa were peracarid crustaceans. These results support the need for additional treatments along with the current mandatory usage of BWE or seawater flushing, especially for eradicating seawater-tolerant taxa such as peracarid crustaceans. The fol- lowing sections summarize our results for the sodium chloride treatments on a variety of seawater-intolerant and seawater- tolerant taxa. Freshwater and oligohaline species Adults of six freshwater cladoceran species, two copepod species (Eucyclops and Eurytemora), and one rotifer species experienced 100% mortality after a 1-h exposure, even at the / Empty-Refill Treatment ? ? ? Flow-Through Treatment 123 24 43 Time (Hours) Fig. 1. The effect of osmotic shock using seawater on ballast water taxa. Both lines (dashed and solid) represent the cumulative number of taxa experiencing 100% mortality at specific dosages (ppt) and exposure durations (h). Dashed and solid lines indicate results from flow-through and empty-refill ballast water exchange protocols, re- spectively (see Materials and Methods). Model for flow-through ex- periments is based on single rectangular, two-parameter, hyperbolic nonlinear regression. Model for empty-refill experiments is a linear regression. Both adjusted r2 values are greater than 0.98 (n = 54 taxa). Additional seawater-tolerance experiments were added to our previ- ously published data [10] for this analysis. lowest brine concentration (30 ppt). No differences were found in the time required to cause 100% mortality between the 30 ppt brine treatment and the 34 ppt seawater treatment for any of these species (Table 1). When placed back into their habitat water after the brine treatments, however, adults of two cla- doceran species, Eurycercus lamellatus and Bosmina longi- rostris, burst and released completely developed, brooded ju- veniles. These juveniles survived in each of the three brine treatment concentrations (30, 45, and 60 ppt) when the ex- posure periods were limited to 1 h. Experiments repeated using reproductive adults of both species with late-stage broods, however, did not result with any survivorship when the ex- posure duration was at least 3 h (Table 1). Mussel veligers (mix of quagga and zebra mussels) retract- ed into their larval shells when exposed to either 34 ppt sea- water or any of the brine solutions. Based on observations of velar lobe ciliary activity, mussel veligers survived for 48 h in 34 ppt seawater but only 3 h in 30 ppt sodium chloride. Mussel veligers were killed by both the 45 and 60 ppt brine treatments within 1 h. Two freshwater gammarid amphipods from Poland, Gammarus fossarum and Gammarus roeseli, were similar to each other in their tolerance to seawater (3 h) and levels of brine that caused 100% mortality (1-3 h). In- dividuals of G. fossarum, however, were eliminated with short- er exposure duration in the 30 ppt sodium chloride treatment as compared to individuals of G. roeseli. Brackish-water and mesohaline species Animals we classified as either brackish-water or meso- haline species included 12 peracarid crustacean species, one copepod species (Acartia tonsa), and the larvae of two deca- pod crustacean species. Newly released, stage-I zoeae from adult broods of the decapods, R. harrisii and P. pugio, gen- erally were tolerant of 34 ppt seawater but differed in their tolerance to sodium chloride. Rhithropanopeus harrisii zoeae survived for 2 h at the lowest brine treatment (30 ppt), but P. pugio zoeae survived this treatment for at least 12 h (Table Eradicating ballast tank organisms with sodium chloride Environ. Toxicol. Chem. 28, 2009 349 Table 1. Species, sites, and results for the brine-tolerance experiments" Exposure time (h) Collection site, Species, seawater tolerance and Habitat by concentration' dateb taxonomic group0 (?C, ppt) C ? iDh 30 ppt 45 ppt 60 ppt OH, 6/05/06 Bosmina longirostrisd (C) 21, 0.1 1 ? 0 1 1 1 OH, 6/05/06 Bosmina longirostrisd brooded juveniles (C) 21, 0.1 1 ? 0 3 3 3 OH, 6/01/06 Daphnia retrocur\>ad (C) 21, 0.1 1 ? 0 1 1 OH, 6/01/06 Quagga and zebra mussel larvaee (M) 21, 0.1 1 ? 0 3 1 TC, 8/04/06 Asplanchna priodontad (R) 23, 0.1 1 ? 0 1 1 TC, 8/05/06 Alona quadrangularisd (C) 23, 0.1 1 ? 0 1 1 TC, 8/08/06 Bythotrephes longimanusd (C) 23, 0.1 0.95 ? 0.1 1 1 TC 8/06/06 Cercopagis pengoid (C) 23, 0.1 1 ? 0 1 1 TC, 8/05/06 Eurycercus lamellatusd (C) 23, 0.1 0.98 ? 0.1 1 1 TC, 8/05/06 Eurycercus lamellatusd brooded juveniles (C) 23, 0.1 0.95 ? 0.1 3 3 LH, 5/03/07 Chydorus sphaericusd (C) 10, 0.1 0.98 ? 0.1 1 1 LH, 5/03/07 Eucyclops speratusd (CP) 10, 0.1 1 ? 0 1 1 LH, 4/30/07 Eurytemora affinisd (CP) 5, 0.1 1 ? 0 1 1 LH, 5/01/07 Chaetogammarus ischnusf (A) 5, 0.1 1 ? 0 24 3 2 VP, 7/18/07 Gammarus fossarumd (A) 15, 0.2 1 ? 0 2 2 2 VP, 7/28/07 Dikerogammarus villosuss (A) 18, 0.3 0.98 ? 0.1 24 6 6 VP, 7/18/07 Gammarus roeselid (A) 17, 0.3 1 ? 0 3 2 1 VP, 7/17/07 Pontogammarus robustoides! (A) 18, 0.4 0.98 ? 0.1 24 6 3 RM, 7/25/06 Neomysis integer* (MY) 24, 1.8 1 ? 0 1 1 1 CL, 8/30/06 Gammarus tigrinus1 (A) 17, 5 1 ? 0 24 24 2 CB, 7/29/06 Acartia tonsad (CP) 20, 7.1 1 ? 0 1 1 1 CB, 7/24/06 Gammarus tigrinus1 (A) 20, 7.1 0.90 ? 0.1 48 24 3 CB, 7/06/06 Rhithropanopeus harrisii' zoeae (D) 20, 9 0.90 ? 0.1 2 1 1 CB, 6/19/06 Palaemonetes pugio1 zoeae (D) 20, 12.3 0.75 ? 0.1 12 2 1 SU, 10/22/07 Grandidierella japonicaf (A) 16, 11 0.88 ? 0.1 X X 48 a Exposure times (h) refer to the amount of time required to kill all individuals of a species within a particular sodium chloride concentration (30, 45, and 60 ppt). The details of the seawater-tolerance experiments have been previously reported [10], and some new data were added here. b CB = Chesapeake Bay, Maryland, USA; CL = Curonian Lagoon, Lithuania; LH = Lake Huron (Alpena), Michigan, USA; OH = Lake Erie, Toledo, Ohio, USA; RM = Port of Rotterdam, The Netherlands; SU = Suisun Bay, California, USA; TC = Lake Michigan (Traverse City), Michigan, USA; VP = Oder and Vistula rivers, Poland. c A = amphipods; C = cladocera; CP = copedpoda; D = decapoda; M = molluska; MY = Mysidae. d All individuals killed within 1 to 3 h with exposure to 34 ppt seawater. e Although zebra and quagga mussel larvae survived in 34 ppt seawater for 48 h, larvae died once they were put back into freshwater from their habitat. f Survivorship greater than 50% after exposure to 34 ppt seawater for 48 h and placed back into habitat water. g All individuals killed within 24 h with exposure to 34 ppt seawater. h C ? SD = proportion of survivorship ? standard deviation in the control treatments for the brine experiments. 1 x = treatment ineffective in killing all individuals of Grandidierella japonica (83 and 65% of individuals survived the 30 and 45 ppt sodium chloride treatments after 48 h, respectively). 1). Exposure durations causing 100% mortality also differed for these species in the 45 ppt brine treatments, but both species were killed within 1 h in the 60 ppt treatments. Brackish-water peracarid crustaceans (except Neomysis in- teger) required much greater exposure times and sodium chlo- ride concentrations to cause 100% mortality compared with other taxa. The introduced gammarid amphipod, Chaetogam- marus (=Echino gammarus) ischnus was somewhat resilient in our brine experiments. Previous experiments showed that 5% of individuals of this species survived in 34 ppt seawater for at least 48 h [10]. In our brine experiments, specimens of C. ischnus were killed after 24-, 3-, and 2-h exposures in the 30, 45, and 60 ppt brine treatments, respectively. Although the amphipods Dikerogammarus villosus and Pontogammarus robustoides were both collected from freshwater rivers in Po- land, they are found in brackish-water habitats as well. These two species survived for relatively long periods in the seawater treatments (24 and >48 h, respectively) but were killed within 24 h when treated with 30 ppt sodium chloride. Both species were eliminated within 3 to 6 h in 45 and 60 ppt sodium chloride treatments (Table 1). Two of the most resilient animals in our experiments were the gammarid amphipods, Gammarus tigrinus and Grandidierella japonica. Specimens of Gam- marus tigrinus used in these experiments were collected from brackish-water habitats of both the Chesapeake Bay and Baltic Sea. Adults of this species tolerated the 30 and 45 ppt sodium chloride treatments for 12 to 24 h but were killed rather quickly in the 60 ppt treatments (2-3 h). Greater salinity tolerances were observed for adults of Grandidierella japonica, which that had high survivorships in the 30 and 45 ppt sodium chlo- ride treatments after 48 h (83 and 65% survivorship, respec- tively) but experienced 100% mortality at 60 ppt after 48 h (Table 1). The results with Gammarus tigrinus and Grandidierella japonica prompted us to focus more closely on short-term, high-salinity brine treatments capable of eliminating resistant brackish-water and mesohaline peracarid crustaceans. Popu- lations of introduced and native peracarid species from the San Francisco and Suisun bays provided a useful source of test animals. Except for the amphipods Hyalella spp., all the spe- cies of peracarid crustaceans used in this portion of the present study were tolerant of 34 ppt seawater for 48 h. The amphipods Hyalella spp., Americorophium spinicorne, and Gammarus daiberi as well as the isopod Synidotea laevidorsalis and the 350 Environ. Toxicol. Chem. 28, 2009 S. Santagata et al. Table 2. Species, sites, and results for the concentrated brine-tolerance experiments3 Collection site, dateb Exposure time (h) by concentration0 Species and taxonomic group0 Habitat (?C, ppt) SDd 85 ppt 110 ppt SF, 7/05/07 SU, 7/04/07 SU, 7/03/07 SU, 7/04/07 SU, 7/15/07 SU, 7/06/07 SU, 7/06/07 SU, 7/07/07 Synidotea laevidorsalis (I) Hyallela spp. (A) Gammarus daiberi (A) Americorophium spinicorne (A) Neomysis kadiakensis (MY) Grandidierella japonica (A) Sinocorophium alienese (A) Nippoleucon hinumensis (CU) 22, 25.6 22, 1.8 22,2.2 22, 1.8 22, 13 22, 11.8 22, 11.8 22, 11.8 0.95 0.98 0 0 0 0 0 0 0.1 0.1 3 3 >3f a All species except Hyalella spp. had survivorships of greater than 50% after exposure to 34 ppt seawater for 48 h and a subsequent transfer to water from their collection habitat (all individuals of Hyalella spp. died at 24 h). b SF = San Francisco Bay, CA, USA; SU = Suisun Bay, CA, USA. c A = amphipoda; CU = cumacea; I = isopoda; MY = Mysidae. d C ? SD = proportion of survivorship ? standard deviation in the control treatments for the brine experiments. e Exposure times (h) refer to the amount of time required to kill all individuals of a species within a particular brine concentration (85 and 110 ppt). f 30% of individuals of Nippoleucon hinumensis survived a 3-h exposure to 85-ppt brine. mysid Neomysis kadiakensis were all killed by 1-h exposures to the 85 and 110 ppt brine treatments (Table 2). Two intro- duced amphipod species, Corophium alienese and Grandi- dierella japonica, and an introduced cumacean, Nippoleucon hinumensis, required exposure durations of at least 3 h or longer in the 85 ppt brine treatments and of 1 to 3 h exposures in the 110 ppt brine treatments to cause 100% mortality. Data analyses Comparing the results from the seawater-tolerance exper- iments to those of the 30 ppt sodium chloride treatments, the lowest sodium chloride concentration required less time to cause 100% mortality in eight freshwater and brackish-water taxa than did treatments with 34 ppt seawater. Sodium chloride concentrations causing 100% mortality in 95% of the species in our experiments are listed for exposure durations of 1, 3, and 6 h in Table 3. The results of the logistic regression analyses of the three main species and habitat characteristics (salinity range, am- bient salinity, and ambient temperature) were different be- tween the treatment types. The best predictor of survivorship in our seawater-tolerance experiments was the overall salinity range documented for a particular species (t ratio = 2.43, p = 0.02). A species' salinity range, however, was not a good predictor of survivorship in the intermediate brine treatments. The remaining two factors, habitat salinity and temperature at the time of animal collection, were not significant predictors of survivorship for either treatment method. Proposed brine treatment levels and their effectiveness are summarized in Figure 2. Cladocerans, copepods, and peracarid crustaceans limited to freshwater or oligohaline habitats (0-5 ppt) were killed within 1 to 3 h, even at the lowest brine concentration (30 ppt). Taxa that may occur in freshwater but that generally prefer brackish-water habitats (0-20 ppt) require higher brine concentrations (45-60 ppt) and longer treatment durations (3-24 h) to cause 100% mortality. The most resilient invertebrates in our experiments were peracarid crustaceans, such as Grandidierella japonica and Nippoleucon hinumen- sis, that prefer mesohaline habitats but do not usually survive in freshwater habitats [27] (http://www.southbayrestoration. org). These species required a minimum sodium chloride con- centration of 60 ppt for at least 24 to 48 h to cause significant mortality. If the sodium chloride treatment was reduced to 3 h, then the brine concentrations needed to cause 100% mor- tality were 85 and 110 ppt for Grandidierella japonica and Nippoleucon hinumensis, respectively. DISCUSSION Despite the numerous proposals for ballast water treatments and systems [28-30], various protocols involving seawater flush- ing of commercial ship ballast tanks likely will remain an active, if not the primary, management strategy for preventing the spread of nonnative species during the next several years [31,32] (http://www.slc.ca.gov/Spec.Pub/MFD/Ballast.Water/ Documents/Final.TechReport.121307_2.pdf and http://www.lr. org/NR/rdonlyres/55A3C43A-0AlB-43E0-9F3 360 52F5BAE29D/80591/LRIMOMEPC56Reportl.pdf). Based on this scenario, we previously demonstrated the dual effects of propagule reduction [33] (http://www.serc.si.edu/labs/marine_ invasions/publications/PortOakfinalrep.pdf) and osmotic shock [10] occurring during BWE for coastal organisms within ballast tanks. From the present research, two overall conclusions can be drawn: First, the combined effects of propagule reduction and osmotic shock make ballast-water exchange and seawater flushing useful management tools for reducing the spread of nonindige- nous species, especially among freshwater and oligohaline hab- Table 3. Recommended minimum treatment concentrations for sodium chloride based on exposure time (n = 26 species)3 Exposure duration 1 h 3 h 6 h LC95 NaCl concentration (95% confidence interval) 110 (85-192) ppt 75 (59-137) ppt 60 (48-98) ppt 1 A chi-square test of heterogeneity was not significant for any of the concentration?mortality distributions and supported the usage of a PROBIT analysis. LC95 = lethal concentration of sodium chloride causing 100% mortality for 95% of the species tested. Eradicating ballast tank organisms with sodium chloride Environ. Toxicol. Chem. 28, 2009 351 Mesohaline Animals 5-25 ppt Brackish-water Animals 0-20 ppt / Freshwater Animals ' / 0-5 ppt 1-h exposure duration 6-h exposure duration Sodium Chloride Concentration (ppt) Fig. 2. Exposure time and dosage models for sodium chloride treat- ments. Each curve represents the total number of species experiencing 100% mortality for various brine concentrations at two exposure times. The 1-h exposure model is based on a four-parameter logistic curve. The 6-h exposure model is based on a single rectangular II, three-parameter curve. Both adjusted r2 values are greater than 0.99. Freshwater and oligohaline organisms (those preferring salinities of 0?5 ppt) generally are intolerant of low-concentration brine exposures even at the shortest treatment duration. Brackish-water animals that also occur in freshwater (those preferring habitats of 0?20 ppt) require longer exposure durations and brine concentrations near 60 ppt. An- imals that prefer mesohaline habitats (5?25 ppt) but generally are incapable of surviving and reproducing in freshwater require the great- est exposure times and brine concentrations (n = 26 species). itats, and second, invertebrates more likely to overcome the os- motic effects of BWE and to disperse via ballast water discharges include peracarid crustaceans (amphipods, isopods, cumaceans, and mysids). In addition to the wide salinity tolerances exhibited by peracarid crustaceans, several other life-history characteristics contribute to their ability to expand their geographic range, in- cluding relatively high reproductive output [34] and, perhaps more importantly, their ability to establish populations in fresh- water habitats. The latter trait is not unique to peracarid crusta- ceans; however, tolerance of both freshwater and hypersaline con- ditions may be more prevalent among introduced peracarid spe- cies than among other introduced crustacean species. Toxicity experiments using solutions of equal concentration based on artificial sea salt or only sodium chloride against other freshwater macroinvertebrates have yielded results sim- ilar to those of the present study. The increased toxicity of sodium chloride solutions to that of seawater is linked to the elevated levels of sodium and chloride ions relative to those of other biologically important ions, such as calcium and po- tassium [20,21,26]. The differences in tolerance among various taxonomic groups in the present study also were observed for other species of freshwater macroinvertebrates in Australia [18,19]. In general, common freshwater invertebrates, such as fly larvae, were intolerant of increased salinity, but decapods, peracarids, and gastropods were the most resilient taxa, tol- erating salinities ranging of 76 to 160 mS/cm (~50-100 ppt). Sodium chloride treatment also was evaluated for killing several species of fish, mollusks, and aquatic plants from the fresh- water fishing nets of commercial vessels in New Zealand [35] (http: // yearofthemountains.org.nz/upload/documents/science- and-technical/drds261.pdf), and the recommended dosage was 70 ppt for 1 to 3 h. Although not included in the present study, we also have preliminary data that show sodium chloride treat- ments of 45 to 60 ppt kill the introduced round goby, Apollonia melanostoma (formerly Neogobius melanostomus, Lake Michigan population) within 1 h. The effectiveness of salt toxicity on microorganisms, such as bacteria and viruses, is more difficult to interpret. Some halotolerant bacteria from freshwater habitats can survive at 1 to 20 ppt?and in some species, up to saturation [24]. The physiological tolerances of viruses tend to coevolve along with their hosts and thus, in some cases, exhibit wide halotolerance [36]. Although many species were quickly killed by lower brine concentrations (45 ppt for 1-6 h) in our experiments, we con- servatively propose, considering the additional salinity toler- ance data cited here, that all sodium chloride treatments aim for a minimum exposure concentration of 80 ppt for at least 3 h. We list both the lower and upper 95% confidence limits for each treatment concentration, but we prefer to remain con- servative regarding the effectiveness of sodium chloride treat- ments. Therefore, we suggest that the lower concentration limit for each of the three treatment durations be ignored. In some cases, it may be possible to treat the sediment and residual water in the ballast tanks of NOB OB vessels for longer periods (e.g., 12-24 h). Despite longer treatment durations, we do not recommend brine treatment concentrations less than 60 ppt for any exposure period. The feasibility of this treatment strategy has been evaluated for use on commercial NOBOB ships in the Great Lakes region [37]. Several facilities can deliver highly concentrated sodium chloride solutions (up to 230 ppt) via trucks to many of the common ports of the region. Brine solutions could then be pumped into the ballast tanks of NOBOB vessels containing residual sediment and water. Residuals on most ships will be only a few metric tons per tank [11] and, thus, would require only a few metric tons of 230 ppt sodium chloride to produce a final salt concentration of at least 80 ppt. If the treatment is conducted at ports where cargo will be off-loaded, it should be possible to treat the tanks for 3 to 6 h before cargo off- loading would necessitate addition of ballast water to the tank. One possible concern is that posttreatment discharge of highly saline water into freshwater habitats may negatively impact native species. After off-loading cargo and completion of the brine treatment, NOBOB vessels would need to fill their tanks with new freshwater ballast. In most cases, this would reduce the salinity in these tanks to much less than that of oceanic seawater and, along with the effects of dispersion, should not cause any greater environmental risk than the dis- charge of 30 ppt seawater already allowed under Great Lakes ballast regulations. Also, based on an analysis of ship traffic, the increased amount of sodium chloride introduced into the Great Lakes would not be significant compared to the levels already introduced via the deicing of roads using rock salt [37]. Furthermore, previous studies have found that the short- term toxic effects of sodium chloride as a result of road salt discharges do not threaten native freshwater taxa [14,15]. Another issue requiring further exploration is the effect of high-concentration sodium chloride on the metal structure of ballast tanks. This aspect has been addressed briefly, and in general, major tank-coating manufacturers have stated that their coatings would not be adversely affected by short ex- posures to sodium chloride solutions [37]. Tank coatings can become cracked over time, however, and such cracks would expose the underlying metal to the corrosion-causing brine solution. We acknowledge that this is a concern. On the other hand, the use of sodium chloride is not proposed as a regular treatment but, rather, as one to be used occasionally and for 352 Environ. Toxicol. Chem. 28, 2009 S. Santagata et al. short durations. This may be a more acceptable alternative for ships that otherwise would have to return to an alternate ex- change site (usually a 2- to 4-d round trip from Montreal, where most tank inspections are performed) or to seal their tanks. CONCLUSION If followed rigorously, current ballast water management strategies are a useful tool for reducing the spread of nonin- digenous species, especially in freshwater systems. We have identified several species of peracarid crustaceans, however, with life-history characteristics that allow them to survive these management strategies and, thus, to exploit ballast tank transport as a means of long-distance dispersal that would otherwise be unavailable to them. In addition, some NOBOB ships are unable to conduct seawater flushing for safety reasons and may be carrying low-salinity, residual ballast water in some tanks. Our proposal to use concentrated sodium chloride solutions as an additional treatment for the residual organisms in the tanks of NOBOB ships provides a simple, economical, and effective method for eliminating numerous organisms, in- cluding halotolerant taxa. By timing the brine treatment to occur before ballasting, the resulting dilution will minimize both the corrosion potential in the treated ballast tanks and the potential for negative impacts on native flora and fauna during subsequent discharge (deballasting). Approval of brine as an alternative, limited-use treatment could be incorporated into existing ballast water management programs with minimal ef- fort and cost to shipping operations, providing an additional tool for regulators to use against the spread of nonnative spe- cies. Acknowledgement?We thank the following institutions and individ- uals: Z.R. Gasiunaite, Coastal Research and Planning Institute, Klai- peda University, Klaipeda, Lithuania; C. Stepien, Lake Erie Center, University of Toledo; Northwestern Michigan College Great Lakes Water Studies Institute; M. Laterveer, The Rotterdam Zoo/Aquarium; National Oceanographic Atmospheric Administration's laboratories at Alepna and Muskegon, Michigan, USA; and the Romberg Tiburon Center for Environmental Studies, San Francisco State University. Funding was provided by the National Oceanic and Atmospheric Ad- ministration, Aquatic Invasive Species Program (National Center for Research on Aquatic Invasive Species), through the Cooperative In- stitute for Limnology and Ecosystems Research administered by the University of Michigan. Contribution 1476 of the Great Lakes En- vironmental Research Laboratory. REFERENCES 1. Leppakoski E, Gollasch S, Gruszka P, Ojaveer H, Olenin S, Panov V. 2002. The Baltic?A sea of invaders. Can J Fish Aquat Sci 59:1175-1188. 2. Maclsaac HI, Robbins TC, Lewis MA. 2002. Modeling ships' ballast water as invasion threats to the Great Lakes. Can J Fish Aquat Sci 59:1245-1256. 3. Reid D, Orlova M. 2002. Geological and evolutionary underpin- nings for the success of Ponto-Caspian species invasions in the Baltic Sea and North American Great Lakes. Can J Fish Aquat Sci 59:1144-1158. 4. Santagata S, Ruiz GM. 2007. Characterization of source inver- tebrate populations and assessing salinity toxicity as a barrier to prevent transfers of "high risk" species to the Great Lakes in ballast tanks. Great Lakes Environmental Research Laboratory, Ann Arbor, MI, USA. 5. Colautti RI, Niimi AJ, van Overdijk CDA, Mills EL, Holeck K, Maclsaac HI. 2003. Spatial and temporal analysis of transoceanic shipping vectors to the Great Lakes. In Ruiz GM, Carlton IT, eds, Invasive Species: Vectors and Management Strategies. Island, Washington, DC, USA, pp 227-246. 6. Duggan IC, Van Overdijk CDA, Bailey SA, lenkins PT, Limen H, Maclsaac HI. 2005. Invertebrates associated with residual bal- last water and sediments of cargo-carrying ships entering the Great Lakes. Can J Fish Aquat Sci 62:2463?2474. 7. Bailey SA, Nandakumar K, Maclsaac HI. 2006. Does saltwater flushing reduce viability of diapausing eggs in ship ballast sed- iment? Divers Distrib 12:328-335. 8. Duggan IC, Bailey SA, van Overdijk CDA, Maclsaac HI. 2006. Invasion risk of active and diapausing invertebrates from residual ballast in ships entering Chesapeake Bay. Mar Ecol Prog Ser 324:57-66. 9. Transport Canada. 2006. Ballast water control and management regulations. Canada Gazette 140:705?723. 10. Santagata S, Gasiunaite ZR, Verling E, Cordell IR, Eason K, Cohen IS, Bacela K, Quilez-Badia G, lohengen TH, Reid OR, Ruiz GM. 2008. Effect of osmotic shock as a management strat- egy to reduce transfers of nonindigenous species among low- salinity ports by ships. Aquatic Invasions 3:61?76. 11. lohengen TH, Reid OF, Fahnenstiel GL, Maclsaac HI, Dobbs FC, Doblin M, Ruiz GM, lenkins PT. 2005. Assessment of transoce- anic NOBOB vessels and low-salinity ballast water as vectors for nonindigenous species introductions to the Great Lakes. Final Report. Great Lakes Environmental Research Laboratory, Ann Arbor, MI, USA. 12. Drake LA, lenkins PT, Dobbs FC. 2005. Domestic and interna- tional arrivals of NOBOB (no ballast on board) vessels to Lower Chesapeake Bay. Mar Pollut Bull 50:560-565. 13. Choi KH, Kimmerer W, Smith G, Ruiz GM, Lion K. 2005. Post- exchange zooplankton in ballast water of ships entering the San Francisco Estuary. J Plankton Res 27:707?714. 14. Benbow ME, Merritt RW. 2004. Road-salt toxicity of select Mich- igan wetland macroinvertebrates under different testing condi- tions. Wetlands 24:68-76. 15. Blasius BJ, Merritt RW. 2002. Field and laboratory investigations on the effects of road salt (NaCl) on stream macroinvertebrate communities. Environ Pollut 120:219?231. 16. Hopkins SH. 1973. Annotated bibliography on effects of salinity and salinity changes on life in coastal waters. NTIS AD-767 562. Technical Report. National Technical Information Service, Springfield, VA, USA. 17. Kefford BJ, Palmer CG, Nugegoda D. 2005. Relative salinity tolerance of freshwater macroinvertebrates from the southeast Eastern Cape, South Africa, compared with the Barwon Catch- ment, Victoria, Australia. Mar Freshw Res 56:163?171. 18. Kefford BJ, Nugegoda D, Metzeling L, Fields EJ. 2006. Vali- dating species sensitivity distributions using salinity tolerance of riverine macroinvertebrates in the southern Murray?Darling Basin (Victoria, Australia). Can J Fish Aquat Sci 63:1865?1877. 19. Kefford BJ, Nugegoda D, Zalizniak L, Fields EJ, Hassell KL. 2007. The salinity tolerance of freshwater macroinvertebrate eggs and hatchlings in comparison to their older life-stages: A diversity of responses?The salinity tolerance of freshwater macroinver- tebrate eggs and hatchlings. Aquat Ecol 41:335?348. 20. Martinez-Ieronimo F, Martinez-Ieronimo L. 2007. Chronic effect of NaCl salinity on a freshwater strain of Daphnia magna Straus (Crustacea: Cladocera): A demographic study. Ecotoxicol Envi- ron Saf 67:411-416. 21. Piscart C, Usseglio-Polatera P, Moreteau IC, Beisel JN. 2006. The role of salinity in the selection of biological traits of freshwater invertebrates. Arch Hydrobiol 166:185?198. 22. Pillard DA, DuFresne DL, Caudle DD, Tietge JE, Evans JM. 2000. Predicting the toxicity of major ions in seawater to mysid shrimp (Mysidopsis bahia), sheepshead minnow (Cyprinodon variegatus), and inland silverside minnow (Menidia beryllina). Environ Toxicol Chem 19:183?191. 23. Sarma SSS, Nandini S, Morales-Ventura J, Delgado-Martinez I, Gonzalez-Valverde L. 2006. Effects of NaCl salinity on the pop- ulation dynamics of freshwater zooplankton (rotifers and cladoc- erans). Aquat Ecol 40:349-360. 24. Tiquia SM, Davis D, Hadid H, Kasparian S, Ismail M, Sahly R, Shim I, Singh S, Murray KS. 2007. Halophilic and halotolerant bacteria from river waters and shallow groundwater along the Rouge River of southeastern Michigan. Environ Technol 28:297? 307. 25. Varsamos S. 2002. Tolerance range and osmoregulation in hy- persaline conditions in the European sea bass (Dicentrarchus labrax). J Mar Biol Assoc U K 82:1047-1048. 26. Zalizniak L, Kefford BJ, Nugegoda D. 2006. Is all salinity the Eradicating ballast tank organisms with sodium chloride Environ. Toxicol. Chem. 28, 2009 353 same? I. The effect of ionic compositions on the salinity tolerance of five species of freshwater invertebrates. Mar Freshw Res 57: 75-82. 27. Hansen SR. 2003. Assessment of impact to aquatic life associated with circulation of saline pond water during the initial stewardship period. Technical Report 42. South Bay Restoration Project, San Francisco, CA, USA. 28. Perrins JC, Cordell JR, Perm NC, Grocock JL, Herwig RR 2006. Mesocosm experiments for evaluating the biological efficacy of ozone treatment of marine ballast water. Mar Pollut Bull 52:1756? 1767. 29. Raikow DF, Reid DF, Maynard EE, Landrum PE. 2006. Sensitivity of aquatic invertebrate resting eggs to SeaKleen? (menadione): A test of potential ballast tank treatment options. Environ Toxicol Chem 25:552-559. 30. Raikow DF, Reid DF, Blatchley ER, Jacobs G, Landrum PF. 2007. Effects of proposed physical ballast tank treatments on aquatic invertebrate resting eggs. Environ Toxicol Chem 26:717?725. 31. Dobroski N, Takata L, Scianni C, Falkner M. 2007. Assessment of the efficacy, availability and environmental impacts of ballast water treatment systems for use in California waters. Technical Report. California State Lands Commission, Marine Facilities Division, Long Beach, CA, USA. 32. Lloyds Register. 2007. Ballast water treatment technology, current status. Technical Report 56. London, UK. 33. Ruiz GM, Smith G. 2005. Biological study of container vessels at the Port of Oakland: (a) biota associated with ballast water of container ships arriving to the Port of Oakland, (b) ballast water exchange efficacy on eight container ships, and (c) analysis of biofouling organisms associated with the hulls of container ships arriving to the Port of Oakland. Final Report. Port of Oakland, CA, USA. 34. Grabowski M, Bacela K, Konopacka A. 2007. How to be an invasive gammarid (Amphipoda: Gammaroidea)?Comparison of life history traits. Hydrobiologia 590:75?84. 35. Matheson FE, Dugdale AE, Wells RDS, Taumoepeau A, Smith JR 2007. Efficacy of saltwater solutions to kill introduced fresh- water species and sterilize freshwater fishing nets. DOC Research and Development Series 261. Technical Report. Science & Tech- nical Publishing, Department of Conservation, Wellington, New Zealand. 36. Weinbauer MG. 2004. Ecology of prokaryotic viruses. FEMS Microbiol Rev 28:127-181. 37. Jenkins PT 2007. Brine as a treatment solution for the control of aquatic nuisance species introductions into the Great Lakes by NOBOB vessels. Transport Canada T8275-05-0249. Technical Report. Transport Canada-Marine Safety, Ottawa, ON.