Ballast Water Exchange

Alternate Areas or

Alternative Methods?

Lessons from the Gulf of St. Lawrence

YVES DE LAFONTAINESt. Lawrence CentreEnvironment Canada105 McGill St.Montreal, QuébecCanada H2Y 2E7yves.delafontaine@ec.gc.ca

NATHALIE SIMARDMaurice Lamontagne InstituteDepartment of Fisheries and OceansCanada850 route de la MerPO Box 1000Mont-Joli, QuébecCanada G5L 3A1SimardN@dfo-mpo.gc.ca

Note: The alternative treatment discussion is not anendorsement for this process (ed.).

INTRODUCTION

The problem of aquatic species introductionsis certainly as old as the transport of goods andmerchandise by ship. The replacement of “solidballast” used in the old days by “liquid ballast”in modern ships has exacerbated the problem,such that ballast water discharge is now recog-nized as the main pathway of species introduc-tion and transfer in aquatic ecosystems (Carlton1985, 1996; Carlton and Geller 1993; Wiley andClaudi 2000). Although the importance of thisvector is now widely accepted, there are stillnumerous questions and unverified statementsrelative to the mechanisms determining the suc-cess of species transport and transfer via ballastwater. Notwithstanding these unknowns, ballastwater management was rapidly foreseen by sci-entists, governments, international maritimeagencies and the shipping industry as a necessarytool to resolve the environmental problem of

aquatic species introductions (IMO 2004).Theoretically, ballast water can be managedeither via ballast water exchange (BWE) at sea orvia ballast water treatment methods. In practice,due to its ease of application and relatively inex-pensive operational cost, BWE was adoptedquickly by the shipping industry and it currentlyrepresents the main method of preventing theintroduction and transfer of aquatic species.

BWE in mid-ocean may pose a risk to shipsafety, so the use of alternate zones in coastalenvironments has been recommended in caseswhere BWE cannot be performed at offshorelocations. Ideally, these alternate zones should bedefined and delineated after rigorous assessmentof their appropriateness to minimize the risk ofintroductions. Since the use of alternate zonesshould be minimal and restricted to particularcases, the option of using alternative methods totreat ballast water is worth considering. The pre-sent paper reviews information on the appropri-ateness, effectiveness and benefits of alternatezones versus alternative treatment technologiesto manage ballast water, using the Gulf of St.Lawrence alternate exchange zone as an exam-ple. First, we present an historical overview ofthe “back-up” zone for BWE in the Gulf of St.Lawrence; secondly, we provide a preliminaryassessment of the impact and effectiveness ofBWE in the Great Lakes - St. Lawrence basin;and, finally, we discuss the use of alternativetreatment methods in the foreseeable future.

THE “BACK-UP” EXCHANGE ZONE IN THEGULF OF ST. LAWRENCE

In response to the environmental crisisresulting from the introduction and invasion ofthe Laurentian Great Lakes by zebra mussels(Dreissena polymorpha) in the mid-1980s and atthe urging of the Great Lakes FisheryCommission, the Canadian Coast Guard, withthe scientific guidance of several departmentsand agencies, established in May 1989 “volun-tary” interim guidelines for BWE for ships enter-ing the Great Lakes. These guidelines were firstset for ships going upbound of Montreal harbouronly. These voluntary guidelines designated asecondary (or “alternate”) BWE zone within the

63

Gulf of St. Lawrence. The area was initially des-ignated in the Laurentian Channel waters east ofthe Gaspé peninsula at 64oW longitude and oversounding depths >300 m (Figure 1). The easternlimit of the alternate exchange zone remainedundefined but apparently extended into theAtlantic Ocean beyond Cabot Strait.

In 1990, a scientific expert group commis-sioned by the Department of Fisheries andOceans recommended to redefine the boundariesof the alternate zone for exchange as theLaurentian Channel waters located between63oW and 61oW longitude and over depths >300 m(Kerr 1990). Reasons for this proposed changewere that the western boundary was consideredtoo “risky”, due to the proximity of the Gaspécurrent which could potentially entrain surfacewater over the Magdalen Shallows and nearbycoastal waters. The proposed eastern boundarywas set at 61oW longitude was set to minimizethe risk of entrainment of ballast water beingdischarged in Cabot Strait over the coastalwaters of the Nova Scotian shelf (Kerr 1990).This proposition would have substantiallyreduced the size of the zone, so that ship transittime through it might be shorter than the amount

of time required to complete BWE. On the otherhand, the smaller size of the area would presum-ably minimize or eliminate its use on a “routine”basis, rather than under exceptional conditionsonly, as originally intended. The recommendationfor an eastern limit to the alternate exchangezone never really came into force.

Ships steaming to ports in the St. LawrenceRiver (Montreal and downstream) as well asports in the estuary and the Gulf of St. Lawrencewere exempt from these guidelines and remaineda risk with regards to non-indigenous speciesintroduction. In 1993, the Canadian “voluntary”guidelines were changed to accommodate theU.S. mandatory regulation for ships entering theGreat Lakes, so that all ships passing 63oW lon-gitude were authorized to use the alternateexchange zone. This would presumably helpprotect the freshwater section of the St.Lawrence River downstream of Montreal. TheCanada Shipping Act is currently under reviewand enforcement of new legislation should makeBWE mandatory for ships entering the GreatLakesSt. Lawrence River system. The westernlimit of the alternate zone has remainedunchanged at 63oW longitude.

64

Figure 1. Map of the St. Lawrence River, the estuary and the Gulf of St. Lawrence. The shaded area correspondsto the alternative zone for ballast exchange and vertical lines indicate the various limits of application of theCanadian voluntary guidelines, (adapted from Bourgeois et al. 2001).

The changes in the definition and applicationcriteria of the voluntary guidelines and the limitsof the alternate zone merely reflect the complex-ity of this issue. The problem arises principallyfrom the desire to apply a single rule for protect-ing three different ecosystems: the Great Lakes,the freshwater St. Lawrence River and themarine sector comprising the estuary and theGulf of St. Lawrence. In order to assess thepotential risk of species introductions into eachof these ecosystems, Bourgeois et al. (2001)compiled the information on foreign shippingtraffic entering the Gulf of St. Lawrence between1978 and 1996. Results on the number of ships,ballast capacity and effective volume of ballastbeing discharged into each ecosystem (Table 1)revealed that the risk of introduction would beobviously much higher within the marine sector.This sector accounted for over 50% of the esti-mated risk due to the presence of much largerships (with much larger ballast capacity) poten-tially discharging at ports of destination. Thefreshwater sector of the St. Lawrence Riverranked second with a risk estimate of 38%,while the risk for the Great Lakes sector was8%. When considering the percentage of shipsreported to carry ballast on board (BOB) duringa survey made between 1994 and 1996 (Harveyet al. 1999), the total volume of ballast waterdischarged into the estuary and the Gulf wasestimated to be approximately 10 million tonsper year, or 78% of the total amount of ballastwater discharged by foreign vessels in the GreatLakes and St. Lawrence basin. Paradoxically, thepotential risk was lowest in the Great Lakes sec-tor where the annual discharge of ballast waterwas 0.25 million tons (~2% of the total). Thedifference between sectors was essentially due tothe percentage of ships without ballast on board.While the vast majority (87%) of ships stoppingin the marine sector declared carrying ballast onboard, the percentage dropped to 20% for shipsat destinations along the St. Lawrence River andto 4% for ships entering the Great Lakes in years1994-1996 (Table 1).

An analysis of the transoceanic shipping intothe Great Lakes since the opening of the seawayin 1959 showed that the percentage of ships withno ballast on board or in partial ballast(NOBOB), has increased slightly since the estab-

lishment of the Canadian voluntary guidelines in1989 and even more since that of the mandatoryU.S. rules in 1993 (Grigorovich et al. 2003).Assuming that misreporting was not a problem,reasons for this increase could presumably berelated either to economic pressure which forcedthe shipping industry to maximize cargo inrecent years or to the fact that ships, in order tocomply to guidelines and legislation, do dis-charge ballast water but not necessarily take newballast on-board before entering the Great Lakes.Data from the U.S. Coast Guard on ship surveil-lance at the entrance of the Great Lakes indicat-ed that full compliance with BWE rules andguidelines has been relatively high and exceeded90% since 1993 (Figure 2). Depending on theyear, up to 7% of ships were considered techni-cally non-compliant. These ships did complywith salinity standards (>26 ppt), but not withthe zone of exchange (>200 nautical miles off-shore and over depth >2000 meters) and maytherefore include ships using the alternate zonefor salinity compliance. The above values areindeed very similar to the percentage of ships(7%) using the Gulf alternate zone for ballastexchange in 1989 immediately following theestablishment of the voluntary guidelines (Kerr1990). Given the number of foreign ships enter-ing the Great Lakes, the number of ships thatused the alternate exchange zone each yearwould probably be very small. The compliancerate varies seasonally and is usually lower inearly spring and late fall, probably a result ofharsh weather and poor sea conditions rendering

65

Table 1. Information on foreign shipping traffic to theGreat Lakes and the St. Lawrence basin between1978 and 1996 (data from Bourgeois et al. 2001 andHarvey et al. 1999).

Ships Ballast Capacity Ballast discharged in 1994-1996

Destination (Nb. yr1) % (10

6 T yr

-1) % % BOB (10

6 T yr

-1)

Great Lakes 249 39 13 2.3 0.5 8 4 0.25 (2%)

St. Lawrence River

1048 104 53 10.8 2.3 38 20 2.70 (20%)

Montréal 735 70 6 60

Québec 179 17 3 27

Other ports 135 13 1.8 13

Estuary and Gulf

674 75 34 15.6 2.8 54 87 10.46 (78%)

*

* Ballast on board

ballast water exchange more difficult and risky(data not shown). Unfortunately, comparabledata on ship compliance are lacking for otherCanadian sectors (the St. Lawrence River, theestuary and Gulf of St. Lawrence). In theabsence of mandatory rules for Canadian waters,monitoring or surveillance programs for shipballast are still nonexistent and there is no for-mal checkpoint for ships entering the St.Lawrence River and the maritime sector in east-ern Canada. Preliminary survey results fromTransport Canada during year 2002 indicatedthat only 3 out of 573 foreign ships (0.7%) atdestination ports in the Quebec region only (notincluding ships entering Gulf of St. Lawrenceports in the remaining provinces) did reportexchanging ballast water in the alternate zone(D. Duranceau, Transport Canada, QuebecRegion, pers. comm.). This would confirm thatthe alternate back-up zone in the Gulf of St.Lawrence is not frequently used for BWE.

Knowing the potential risk of species intro-ductions associated with shipping traffic, wecompiled the available information relative tothe number of exotic species introduced intoeach of the three sectors in the Great Lakes – St.Lawrence basin in order to assess and comparethe relative risk of species establishment amongsectors. A recent analysis by de Lafontaine andCostan (2002) showed that the number of non-indigenous species introduced since 1820 in theupstream Great Lakes (160 species) was almosttwice that reported in the downstream St.Lawrence River (85 species). The relative pro-portion of the various taxonomic groups (algae,

vascular plants, invertebrates, fish) differed sig-nificantly between the two regions, suggestingdifferent dynamics of introduction and establish-ment. Considering that the risk of introductionby shipping (simply based on traffic and ballastvolume estimates) was presumably higher in theriver than in the lakes, this result may be surpris-ing and somewhat counterintuitive. From an eco-logical perspective, riverine systems are far moredynamic than lakes and would not necessarilyfavour the establishment of all forms of organ-isms. The rapid flow rate of the St. LawrenceRiver may indeed accelerate the drift of planktonand other larval forms toward the saline estuary,therefore reducing the risk of species establish-ment and invasion in the river. In fact, mostintroduced species reported in the river wereeither benthic organims (vascular plants andmolluscs) or mobile animals (fish, benthic crus-taceans). Moreover, nearly all species (98%)introduced in the St. Lawrence River since 1960were first reported in the Great Lakes, suggest-ing that these species were introduced and estab-lished in the Great Lakes first before being sub-sequently transferred and spread into the St.Lawrence River by either passive or active trans-fer. Although several ecological reasons could becalled for explaining this regional difference inthe introduction and establishment success ofnon-indigenous species, it does not seem to par-allel the potential risk associated with shipping(estimated on proportion of BOB ships and bal-last capacity). In fact, these results support thehypothesis put forward by Grigorovich et al.(2003) that NOBOB ships after re-ballastingwith water in the Great Lakes may indeed be themajor cause of species introductions. Thisimplies that untreated residual waters and sedi-ments remaining at the bottom of tanks inNOBOB ships would present a higher risk factorand should be managed with adequate treatmentmethods.

On the positive side however, the number ofspecies introductions associated with the shippingvector in freshwater (Great Lakes and St.Lawrence River) seems to have declined slightlyover the last decade (1990-2000) since theimplementation of the exchange guidelines byCanada and the U.S. during the early 1990s (deLafontaine and Costan 2002; Grigorovich et al.

66

Figure 2. Ballast water compliance rates.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1990 1992 1994 1996 1998

Navigation Season

Compliant Non-Compliant Technical Non-Compliant

2003). If this decline is significant and persistsover the next decade, it would strongly suggestthat BWE does effectively help minimizing theinput of new species. However, the objective ofno new introductions is still far from being met.

Information on non-indigenous species intro-duction in Canadian marine environments ismore scarce, in particular for the estuary andGulf of St. Lawrence. Despite the presumablyvery high risk associated with the high propor-tion of BOB ships arriving at destinations in theGulf of St. Lawrence harbours, the total reportednumber of exotic species in both marine andestuarine environments is 26 (Table 2), which ismuch less than that reported for the Great Lakes.Algae accounted for nearly half of that number.Shipping would account for 25% (6 species) ofthe total number of these introductions, which isclose to the value of 33% estimated for the GreatLakes according to Wiley and Claudi (2002).Reports of introduced species are very localized,but the spatial coverage of sampling effort andinventory is much less thorough for the entireGulf of St. Lawrence than that realized in theGreat Lakes. Given the paucity of informationavailable, no conclusions may be reached regard-ing the risk and the impact of BWE within theGulf of St. Lawrence. Evidence indicates, how-ever, that species introductions and establish-ment in the marine sector is not as important noras frequent as in the upstream freshwater environ-ments. Alternatively the theory that cummulativeinvaders facilitate new introductions might sug-gest that the Great Lakes is in the midst of“invasion meltdown,” whereas the Gulf of St.Lawrence system may not have attained thisphase of accelerating rates of invasion(Simberloff nad Von Halle, 1999).

ALTERNATIVE TREATMENT METHODS

The above results highlight that BWE cannotbe fully satisfactory to eliminate the risk ofspecies introduction and protect the variousaquatic environments. BWE at sea was originallyrecommended for protecting freshwater systems,by replacing freshwater in ballast tanks witholigotrophic marine waters to be dischargedeventually in freshwater ports. The effectiveness

of the method for protecting marine coastalwaters is questionable, however. Shipping activi-ties along the North American coastal watersrepresent a good example of ships transiting andtransferring ballast water between differentmarine systems. In this regard, the possibility ofusing the Gulf alternate zone more frequently inthe future would increase the risk of speciesintroductions in the area. In scenarios whereBWE would not be possible, the use of alterna-tive treatment methods is recommended.

Excluding the use of on-shore treatment thatwould require the installation of treatment facili-ties in every port, various on-board ballast watertreatment technologies have been developed butonly a few have been tested adequately. Thesetechnologies can be grouped into either physi-cal/mechanical or chemical processes, and thepros and cons of each method have beenreviewed in different reports (Pollutech 1992;Hay et al. 1997; Mountfort et al. 1999; Calvé2001; Tamburri et al. 2002). Most physical treat-ment methods require major re-fitting of shipsand some may be unsafe to use. The proposedchemical methods are often based on the use ofbiocides and toxic chemicals which make themenvironmentally unacceptable because thesetreated waters have to be discharged in naturalenvironments. A notable exception is the deoxy-genation process by which the concentration ofdissolved oxygen in ballast water is rapidlyreduced to very low levels inducing a massivekill of most aquatic living organisms (Stalderand Marcus 1997; Mountfort et al. 1999;Tamburri et al. 2002). The anoxic environment

67

Table 2. Introduced non-indigenous species reportedin the Gulf of St Lawrence. Data compiled from vari-ous sources by N. Simard (DFO-Mont-Joli), and A.Locke and M. Hanson (DFO-Moncton).

Taxonomicgroup

Total Presumed vector of introduction

Shipping Naturalor cryptic

Deliberate Unknown

Algae 11 4 2 - 5Inverte-brates 9 2 2 2 3

FreshwaterFish 2 - - 2 -

Parasites/Pathogens 4 - 4 - -

in ballast tanks could also help prevent corrosionof a ship’s structure, which could confer someadvantage in using deoxygenation methods.Oxygen removal can be achieved either by satu-rating ballast waters with the addition of nitro-gen gas as shown by Tamburri et al. (2002) or bybio-reactive processes as suggested by Mountfortet al. (1999).

While nitrogen addition has proven to befeasible and effective, its use is relatively costly,however, and might present some safety risk.Preliminary trials by Mountfort et al. (1999) sug-gested the potential of deoxygenating biologicalprocesses, but the development and use of suchtechniques can encounter many challenges. Thetreatment has to be effective in both fresh- andsaltwater environments and at temperatures rang-ing from nearly 0 to ~30oC. The bio-reactiveprocess has to be fully activated and completedover a short time period (<5 days) in order to beused during short voyages. The process shouldnot release any toxic degradation product andtreated ballast waters should be safe for releaseinto the receiving environment. From a cost/ben-efit point of view, the technology should requireno or minimal ship modification and it should besimple to handle and safe to use by the ship’screw.

Recent developments in biotechnological pro-cesses offer a new methodology that appears to bemeeting the challenges of the criteria (see Table 1)and may become an effective and viable treatmentalternative to ballast water exchange. Laboratorytesting in 200 L vessels to assess the effectivenessand environmental impact of biological oxygena-tion, has shown promise in meeting Canadian dis-charge threshholds. Pilot-scale trials of this tech-nique onboard ship are planned for the next year.We offer an example of one treatment to illustratethe type of data we expect from an alternativetreatment prospective (Figure 3).

The development and testing of new tech-niques will eventually offer a variety of effectivetreatment methods to manage ballast waterappropriately. The proposed existing techniques(including ballast water exchange) can be com-pared and evaluated with a list of different per-formance criteria. Based on literature information(Pollutech 1992; Hay et al. 1997; Calvé 2001;Moore and Ryan 2003; A. Valois, PolyGo,

Montreal, pers. comm.), the characteristics ofeach method relative to each criterion were com-piled in an information matrix (Tables 3 and 4).The estimated cost (in Canadian dollars) per tripwas calculated for a 34,000 DWT vessel making40 trips per year and carrying ballast 33% of thevoyages. A 7-year depreciation time wasassumed for capital investment. As expected,ballast water exchange was the least expensivemethod ($ 250 per trip) but its control effective-ness to treat and control species introductions isconsidered generally poor (Locke et al. 1991).On the other hand, the three deoxygenating tech-niques can meet a large number of criteria, andranked very high in terms of control effective-ness. The deoxygenation techniques are presum-ably also capable of treating residual waters,which normally remained untreated after ballastwater discharge (as in the NoBOB condition).The total cost per trip was quite variablebetween the different treatment methods, but thebiological deoxygenating treatment method wasthe second least expensive ($ 1800 per trip)

68

Fresh Water at 15oC

Time

09-08 09-09 09-10 09-11 09-12 09-13 09-14 09-15 09-16 09-17 09-18

Dis

solv

ed o

xyg

en (

mg

/L)

0

2

4

6

8

10

12

ControlTreatment 1.0%Treatment 0.67%Treatment 0.33%

Salt Water at 10oC

Time

07-19 07-20 07-21 07-22 07-23 07-24 07-25 07-26 07-27 07-28 07-29

Dis

solv

ed o

xyg

en (

mg

/L)

0

2

4

6

8

10

12

14

Figure 3. Concentration of dissolved oxygen as afunction of time in both fresh- and saltwater experi-ments using a bio-technological process from 200 Lvessel testing.

after ballast water exchange. Of course, thesecalculations did not account for any socio-economic and environmental cost for controllingand mitigating the impact of species once theyare introduced into an ecosystem.

TOWARD BETTER MANAGEMENT PRACTICES

In summary, 15 years after the establishmentof the Canadian voluntary guidelines for BWE,it is quite obvious that accurate information to

69

Table 3. Comparative assessment of performance characteristics of different alternative technologies for treatingballast water.

Table 4. Comparative assessment of performance characteristics of different alternative technologies for treatingballast water (suite from Table 3).

Criteria BWE Filtration UV Ozone Chlorine VacuumDeoxygenation

NitrogenDeoxygenation

BiologicalDeoxygenation*

Capital investment(K$CAD) - 1000 1000 2000 500 1400 1000 ?

Depreciationexpenses (K$CAD) - 3.6 3.6 7.1 1.7 5.0 3.6 ?

Maintenance(K$CAD) ? 1.0 1.0 2.0 0.5 1.4 1.0 ?

Perishables(K$CAD) - - - - 1.0 - - 1.8 (est.)

Time loss ? ? - - - - - ?

Total cost per trip(K$CAD) 0.30 4.6 4.6 9.1 3.2 6.4 4.6 1.8 (est.)

Control efectiveness Poor Partial Medium High High High High High

Residual waterstreated No Partial No No Partial Yes ? Yes ? Yes ?

Criteria BWE Filtration UV Ozone Chlorine VacuumDeoxygenation

NitrogenDeoxygenation

BiologicalDeoxygenation*

Energy cost Small High High High Medium High Medium Null

Risk to safety Medium Medium High High High High High Null

Training personnel Low Medium High High High High High Low

Effect on ship lifespan

Almostnull Null Null Corrosion Corrosion Positive Positive ? Positive

Possibility forchange Yes No No No No No No Yes

Environmentalimpact ? ? Null Null Unacceptable Acceptable Acceptable Acceptable

Equipmentspace (m2) Null 10 4 15 2 8 4 Null

Perishablespace (m2) Null Some ? 0 0 ? 0 0 0.4

*Based on 200 L vessel trials.

*Based on 200 L vessel trials.

adequately assess the control effectiveness ofballast water exchange is still largely limited.Circumstantial evidence suggests that BWEimplementation has not resulted in an absolutedecline in species introduction in freshwater sys-tems and it cannot be truly effective for coastalshipping. Moreover, the problems of untreatedresidual waters after BWE and the lack of com-pliance monitoring and surveillance make diffi-cult the management of ballast water to eventu-ally stop the introduction and transfer of aquaticspecies via shipping. To meet this objective,BWE (even in alternate zones) may not representthe best method since the risk of introductionwill always remain higher than with any alterna-tive treatment method. Up to now, the switch tousing other technologies for treating ballastwater has been slow, primarily for two reasons.First, the lack of environmental standards fordischarging the content of ballast tanks in aquat-ic systems does not provide a definition of refer-ence criteria for comparing and testing the vari-ous treatment technologies, including BWE. It ishoped that the new convention proposed by theInternational Maritime Organization (IMO 2004)will be rapidly accepted in order to establishsuch standards at an international level. This willdefinitely serve as a first and essential steptoward better management practices.

Second, in the absence of environmentalstandards, the shipping industry had no strongincentive to solve the problem of species intro-ductions. Thus, BWE is still the best currentpractice due to its cheap operational cost. Inorder to meet our environmental objective ofreducing the number of future introductions andto protect the integrity of our ecosystems, wemust work toward finding and implementingeffective treatment technologies that could alsobenefit the industry. With dedicated effort in thisdirection, the dream may come true more quicklythan many think.

LITERATURE CITED

Bourgeois, M., M. Gilbert, and B.Cusson. 2001. Évolution du traf-ic maritime en provenance de l’étranger dans le Saint-Laurentde 1978 à 1996 et implications pour les risques d’introductiond’espèces aquatiques non indigènes. Rapport technique cana-dien des sciences halieutiques et aquatiques No. 2338, 34 p.

Calvé, R. 2001. A review of potential alternative ballast water

treatments. Pollution prevention. Year 1. Report presented toEnvironment Canada, Atlantic Region, 24 p.

Carlton, J.T. 1985. Transoceanic and interoceanic dispersal ofcoastal marine organisms: the biology of ballast water.Oceanogr. Mar. Biol. Ann. Rev. 23: 313-371.

Carlton, J.T. 1996. Biological invasions and cryptogenic species.Ecology 77: 1653-1655.

Carlton, J.T. and J.B. Geller. 1993. Ecological roulette: the globaltransport of nonindigenous marine organisms. Science 261:78-82.

de Lafontaine, Y., and G. Costan. 2002. Introduction and transferof alien aquatic species in the Great Lakes - St. LawrenceRiver drainage basin. In: R. Claudi, P. Nantel, and E. Muckle-Jeffs (eds), Alien Invaders in Canada’s Waters, Wetlands, andForests. Natural Resource Canada, Canadian Forestry Service,Science Branch, Ottawa, Canada. Pp. 73-92.

Grigorovich, I.A., R.I. Colautti, E.L. Mills, K. Holeck, A.G.Ballert and H.J. MacIsaac. 2003. Ballast-mediated animalintroductions in the Laurentian Great lakes: retrospective andprospective analyses. Can. J. Fish. Aquat. Sci. 60: 740-756.

Harvey, M., M. Gilbert, D. Gauthier and D.M. Reid. 1999. A pre-liminary assessment of risks for the ballast water-mediatedintroduction of nonindigeneous marine organisms in theEstuary and Gulf of St. Lawrence. Canadian Technical Reportof Fisheries and Aquatic Sciences. No. 2268, 56p.

Hay, C., S. Handley, T. Dodgshun, M. Taylor, and W. Gibbs. 1997.Cawthron’s ballast water research: final report. CawthronInstitute Report No. 417, 135p.

IMO. 2004. Alien invaders in ballast water – new conventionadopted at the International Maritime Organization.International Conference on Ballast water Management, 9-13February 2004. http://www.imo.org/home.asp

Kerr, S.R. 1990. The risk to Atlantic Canadian waters of unwantedspecies introductions carried in ships’ ballast. CanadianAtlantic Fisheries Scientific Advisory Committee (CAFSAC)Research Document 90/34, Ottawa, Canada, 11p.

Locke, A., D.M. Reid, W.G. Sprules, J.T. Carlton, and H.C. vanLeeuwen. 1991. Effectiveness of mid-ocean exchange in con-trolling freshwater and coastal zooplankton in ballast water.Canadian Technical Report of Fisheries and Aquatic SciencesNo. 1822, 92p.

Moore, K. and E. Ryan. 2003. Shipping traffic analysis and costassessment for ballast water exchange en route to the UnitedStates. Poster presented at the 12th International Conferenceon Aquatic Invasive Species, June 2003, Windsor, ON.

Mountfort, D.O., C. Hay, T. Dodgshun, S. Buchanan and W.Gibbs. 1999. Oxygen deprivation as a treatment for ships’ bal-last water - laboratory studies and evaluation. J. Marine Env.Engg. 5: 175-192.

Pollutech Environmental Ltd. 1992. A review and evaluation ofballast water management and treatment options to reduce thepotential for the introduction of non-native species to theGreat Lakes. Appendix A - Ballast Water Characterization,Report to the Canadian Coast Guard, Ship safety Branch.Ottawa, March 1992.

Stalder, I.C. and N.H. Marcus. 1997. Zooplankton responses tohypoxia: behavioral patterns and survival of three species ofcalanoid copepods. Mar. Biol. 127: 599-607.

Tamburri, M.N., K. Wasson and M. Matsuda. 2002. Ballast deoxy-genation can prevent aquatic introductions while reducingship corrosion. Biological Conservation 103: 331-341.

Wiley, C.J. and R. Claudi. 2000. The role of ships as a vector ofintroduction for no-indigenous freshwater organisms, withfocus on the Great Lakes. In: Claudi, R. and J.H. Leach (eds),Nonindigenous freshwater organisms. Vectors, biology andimpacts. Lewis Publishers, Boca Raton, USA. Pp. 203-214.

Wiley, C.J. and R. Claudi. 2002. Alien species transported inships’ ballast water: from known impact to regulations. In: R.Claudi, P. Nantel, and E. Muckle-Jeffs (eds), Alien Invaders inCanada’s Waters, Wetlands, and Forests. Natural ResourceCanada, Canadian Forestry Service, Science Branch, Ottawa,Canada. Pp. 233-242.

70

Suitability of the

Gulf of St. Lawrence as an Alternate

Zone for Ballast Water Exchange by

Foreign Ships Proceeding up the

St. Lawrence Seaway

MICHEL GILBERTNATHALIE SIMARDFisheries and Oceans CanadaMaurice Lamontagne Institute850 Rte de la MerMont-Joli, QuébecCanada G5L 3A1SimardN@dfo-mpo.gc.ca

FRANÇOIS J. SAUCIERUniversité du Québec à Rimouski300 des UrsulinesRimouski, QuébecCanada

INTRODUCTION

VOLUNTARY GUIDELINES AND LOCATION OF THEALTERNATE BALLAST WATER EXCHANGE ZONE

Offshore ballast water exchanges are currentlyused as a voluntary and, in some cases, mandatorymeasure to reduce risks for the ballast water-mediated introduction of nonindigenous marineorganisms in the coastal and inland waters of agrowing number of countries. Some voluntaryguidelines for offshore ballast water exchangemake provision for an alternate exchange zone toprotect as much as possible those particularlyvulnerable receiving environments while at thesame time allowing these exchanges to be con-ducted safely. The existing “VoluntaryGuidelines for the Control of Ballast WaterDischarges from Ships Proceeding to the St.Lawrence River and Great Lakes” provide forsuch an alternate zone, located in the Gulf of St.Lawrence (Figure 1).

The Voluntary Guidelines recommend that,if not feasible in the Atlantic Ocean, ships arriv-ing from outside the Exclusive Economic Zone(EEZ) can exchange their ballast water in analternate zone located in the Gulf of St. Lawrence,within the Laurentian Channel southeast of 63° W

71

Note: This research was presented by NathalieSimard, who may be contacted by email:SimardN@dfo-mpo.gc.ca

59°62°65°68°

46°

47°

48°

49°

50°

51°

46°

47°

48°

49°

50°

51°

Magdalen

Islands

CapeBretonIsland

Prince EdwardIsland

Chaleur Bay

Northumberland Strait

St. George's Bay

Anticosti Island

Gaspé Peninsula

Laurentian Channel

Esq

uim

anC

han

nel

Newfoundland

Cabot

Strait

Strait of B

elle Is

le

Sept-Îles

Rimouski

Gaspé

Baie-Comeau

Saguenay Fjord St. Lawre

nce E

stuar

y

Charlottetown

Pugwash

Newcastle

Bathurst

Stephenville

Corner Brook

Gulf of

St. Lawrence

0 - 200 m

> 200 m

Backup Exchange Zone

Port-Cartier

Pointe-Noire

Matane

Les Escoumins

Cacouna

Pointe-au-Pic

Dalhousie

Chandler

Belledune

Summerside

Pictou

Georgetown

Souris

Havre-St-Pierre

59°62°65°68°

46°

47°

48°

49°

50°

51°

46°

47°

48°

49°

50°

51°

Magdalen

Islands

CapeBretonIsland

Prince EdwardIsland

Chaleur Bay

Northumberland Strait

St. George's Bay

Anticosti Island

Gaspé Peninsula

Laurentian Channel

Esq

uim

anC

han

nel

Newfoundland

Cabot

Strait

Strait of B

elle Is

le

Sept-Îles

Rimouski

Gaspé

Baie-Comeau

Saguenay Fjord St. Lawre

nce E

stuar

y

Charlottetown

Pugwash

Newcastle

Bathurst

Stephenville

Corner Brook

Gulf of

St. Lawrence

0 - 200 m

> 200 m

Backup Exchange Zone

Port-Cartier

Pointe-Noire

Matane

Les Escoumins

Cacouna

Pointe-au-Pic

Dalhousie

Chandler

Belledune

Summerside

Pictou

Georgetown

Souris

Havre-St-Pierre

68 65 62 59

Longitude (ºW)

Lat

itude

(ºN

)

46

4

7

4

8

4

9

5

0

5

1

Figure 1. Current alternate ballast water exchange zone for ships proceeding up the Seaway.

longitude and at water depths exceeding 300 m.The Guidelines also recommend ballast waterexchange for ships proceeding to Estuary andGulf ports that are located west of 63° W longi-tude, which represents more than 80% of theentire maritime traffic from a foreign origin inthe area on an annual basis.

An important oceanographic feature of theGulf that has relevance for this study is theinward net transport of deep waters upstream tothe head of the Laurentian Channel. The pre-scribed area is also located relatively close to thecoastal waters of Anticosti Island, the GaspéPeninsula and the Magdalen Islands which sup-port important fisheries, including snow crab,lobster, and shrimp, in the Laurentian Channel.Thus, the location of the alternate exchangezone, particularly its upstream part, may repre-sent and additional risk for the introduction ofnonindigenous marine organisms in coastal areasof the Gulf of St. Lawrence.

OBJECTIVES

The objectives of this study were to assessrisks for the introduction of nonindigenousmarine organisms associated with ballast waterexchange by foreign ships in the alternate zoneof the Gulf of St. Lawrence and to minimizethese risks in order to maintain an alternate bal-last water exchange zone (ABWEZ) in the Gulfof St. Lawrence.

METHODS

A three-dimensional model of circulation andthermodynamics in the Gulf of St. Lawrence(Saucier et al. 2003) was used to: (1) simulatethe seasonal dispersion and fate of planktoninoculated with ballast water discharges in theABWEZ and (2) identify areas from where sur-face waters are quickly flushed out of the Gulfso as to identify areas that could be used foralternate ballast water exchange (ABWE). Thedetermination of these areas was put into contextwith other factors such as ballast water salinityand time of the year to derive a set of conditionsunder which backup exchanges could be allowed

with an acceptable level of risk for the introduc-tion of nonindigenous species in the Gulf of St.Lawrence.

DESCRIPTION OF THE MODEL

The model is a three-dimensional prognosticcoastal model (Backhaus 1985) for currents,temperature, salinity, turbulence, and ice thicknessand concentration in the Estuary and Gulf of St.Lawrence. The model has a horizontal resolutionof 5 km and its vertical resolution is composedof 73 layers. Surface meteorological forcing isprescribed by six-hourly air temperature, winds,dew point, relative humidity, pressure, dry/liquidprecipitation, and radiation. These data are pro-vided by 20 virtual stations distributed throughoutthe Estuary and Gulf and derived from actualobservations. Hydrological forcing is derivedfrom monthly runoff at Quebec City and dailyrunoff for all major tributaries, and is adjusted toaccount for net basin drainage. An open boundaryforcing at Belle-Isle Strait and Cabot Straitincludes 15 water level tidal constituents and cli-matological water mass properties for enteringAtlantic Ocean and Labrador Sea waters.

In terms of validation, the model was shownto reproduce well conditions that were observedin 1986 and 1987, including temperature andsalinity profiles, current meter and tide gaugerecords, summer coastal temperature measure-ments, all of which with a relatively good preci-sion. For example, the model was able to recreatefreshwater pulses, storm events, and the forma-tion, stability and dispersion of the GaspéCurrent, which is one of the major features ofcirculation in the Gulf.

SIMULATIONS

All simulations of the dispersion of planktonwere run by simultaneously releasing a predeter-mined number of particles over a given area thatwas limited to the upstream part of the alternatearea for clarity purposes and to address greatestconcerns. All particles were tracked by recordingtheir position at the beginning of each day ofsimulation. Two-dimensional simulations wererun to follow the discharge and surface dispersion

72

of phytoplankton. A total of 8 simulations over90 days were run, each starting at the beginningof the months of April to November. No simula-tions were run for winter months because of iceconditions which affect surface transport, partic-ularily at small scales. For the dispersion of zoo-plankton, three-dimensional simulations integrat-ing vertical migrations were run. However, giventhe CPU requirements for multi-variable simula-tions, these were limited to four, (starting inApril, June, August, and October) and lasted for60 days. For these simulations, the integration ofdaily vertical migrations was achieved by driv-ing vertical movements of particles at dusk anddawn at a speed of 2 cm.s-1. Vertical migrationswere forced down to 150 meters in theLaurentian Channel to allow particles to reachdeep waters as usually observed, and to the bot-tom when water depths decreased to less than150 m. Finally, to determine suitable areas forexchanges, simulations of the flushing of surfacewaters out of the Gulf were run in conjunctionwith phytoplankton dispersion simulations toidentify initial positions of surface water parti-cles that would be evacuated from the Gulf andthese positions were recorded at 15-day intervalsduring these simulations.

RESULTS AND DISCUSSION

PHYTOPLANKTON AND ZOOPLANKTONDISPERSION SIMULATIONS

Phytoplankton Surface Dispersion SimulationsThe results of animations of two simulations

of the surface dispersion of phytoplankton withcontrasting results were presented to illustratethe large seasonal variability of transport anddispersion patterns. The first one started in Apriland shows the transport of particles first towardthe Magdalen Islands, then upstream to coastalareas of Anticosti Island, and further upstreamwhere some particles are caught up by theAnticosti Gyre. The patch is then transporteddownstream again with the arrival of the fresh-water pulse around mid-May. The second startedin September and shows a net downstream trans-port of inoculated particles, first towards theMagdalen Islands then further downstream toCape Breton Island, as a result of typically dom-inant northwesterly winds at this time of theyear. Most particles are ultimately flushed out ofthe Gulf through Cabot Strait and only a fewremained at the end of the simulation.

73

April - June May - July June - August July - September

August - October September - November October - December November - January

Number of Particles (per cell) Reaching Coastal Areas over 90 Days

0 30 60 90 120 150

Figure 2. Estimated number of particles reaching coastal areas for 8 phytoplankton surface dispersion simulations.

To illustrate coastal areas at risks in the Gulfof St. Lawrence, results of the eight simulationswere summarized using three parameters: (1) thetotal number of particles reaching cells of themodel that are adjacent to coastlines, (2) theminimum time required to reach the cells of themodel that are adjacent to coaslines, and (3) thenumber of particles that are retained in the Gulfthroughout the simulation.1) Total number of particles. Results of all 8

phytoplankton surface dispersion simulationsshow that several coastal areas would bereached, some quite intensively, at differenttimes of the year (Figure 2). Dispersiontowards upstream areas such as AnticostiIsland would occur in spring and late sum-mer while downstream areas, namely theMagdalen Islands, southwesternNewfoundland and Cape Breton Island,would be reached in almost all cases. Theentire west coast of Newfoundland would beparticularily at risk in summer.

2) Time to reach coastal areas. In several cases,inoculated particles would reach coastalareas after 60 days, particularly on the westcoast of Newfoundland (Figure 3). However,

this is not the case in the fall and early winterfor the Magdalen Islands and Cape BretonIsland, which would be reached generallywithin 30 days. Anticosti Island would bereached within 45 days in spring/early sum-mer and in early fall.

3) Retention in the Gulf. These figures showthe change in retained particles with increas-ing time of simulation for the 8 seasonalruns (Figure 4). Most of the particles are stillpresent in the Gulf after 45 days in springand late summer, and at least half of theinoculated particles remain after completionof the 90-day simulations. This indicates thatsignificant retention within the Gulf wouldoccur for phytoplankton cells that would beinoculated with ballast water discharges inthe upstream part of the ABWEZ duringthese periods. However, a significant numberof surface particles would be flushed rapidlyout of the Gulf in early summer and in fall;in extreme conditions (late fall), almost allof the inoculated particles would be lost toAtlantic waters. These times of the year arecharacterized by the passage of the springfreshwater runoff pulse in early summer, and

74

April - June

May - July June - August July - September August - October

September - November October - December November - January

M in im u m T im e (d ays) to R each C o asta l A reas

0 15 30 45 60 75 90

Minimum Time (days) to Reach Coastal Areas

0 12 24 36 48 60

Figure 3. Estimated time needed to reach coastal areas for 8 phytoplankton surface dispersion simulations.

by strong northwesterly winds in fall whichenhance the downstream transport of surfacewater towards the Atlantic Ocean.

Zooplankton Dispersion SimulationsThe first four maps of the three-dimensional

zooplankton dispersion simulations present thenumber of particles reaching coastal areas over

75

April - June

Nb

f

0

250

500

750

1000

1250

1500

1750

2000May - July June - August July - September

August - October

15 30 45 60 75 900

250

500

750

1000

1250

1500

1750

2000

September - November

Time (days)

15 30 45 60 75 90

October - December

15 30 45 60 75 90

November - January

15 30 45 60 75 90

April - May

April - May June - July August - September October - November

June - July August - September October - November

Number of Particles (per cell) Reaching Coastal Areas over 90 Days

0 30 60 90 120 150

Minimum Time (days) to Reach Coastal Areas

0 12 24 36 48 60

Figure 5. Estimated number of particles reaching coastal areas for 4 zooplankton dispersion simulations.

Figure 4. Estimated number of surface particles retained in the Gulf of St. Lawrence for 8 phytoplanktonsurface dispersion simulations.

Time (days)

Num

ber o

f sur

face

par

ticle

s

the 4 simulations (Figure 5). These results showthat dispersion patterns for zooplankton areclearly different from those observed for phyto-plankton. In all cases, inoculated particles aredispersed upstream towards coastal areas of theGaspé Peninsula, Anticosti Island and even thenorthwestern coast of the Gulf, whereas veryfew particles are dispersed towards downstreamcoastal areas. In contrast to phytoplankton simu-lations, the Magdalen Islands do not seem to bethreatened by the inoculation of zooplankton inthe upstream part of the ABWEZ. The four bot-tom maps show the minimum time required toreach these coastal areas and indicate that, inmost cases, it takes at least 30 to 40 days for theinoculated particles to reach coastal areas exceptfor the southeast coast of Anticosti Island. Thesedifferent dispersion patterns are clearly related todaily migrations of zooplankton into deep watersof the Laurentian Channel, which show a netinward transport from Cabot Strait up to thehead of the Laurentian Channel. This pattern isconsistent with the observed zooplanktondynamics in the Gulf of St. Lawrence, wherezooplankton is known to passively migrateupstream and accumulate at the head of theLaurentian Channel near Tadoussac. Thus, it canbe assumed that zooplankton inoculated in theLaurentian Channel with ballast water dischargeswould be subjected to this passive upstreamtransport and ultimately could be incorporated inthe observed zooplankton dynamics in this area,if conditions were favorable for their survivaland reproduction.

SummaryIn summary, phytoplankton dispersion simu-

lations showed that several coastal areas are atrisk from phytoplankton inoculation in theLaurentian Channel, including the MagdalenIslands, Southwest Newfoundland, NorthernCape Breton Island, and Anticosti Island. Someof these coastal areas, namely the MagdalenIslands, Anticosti Island, Cape Breton Island,would be reached within 30 days. A significantdownstream transport would occur in late spring/early summer, because of the spring freshet, andin fall as a result of predominant northwesterlywinds, while a significant retention of surface

phytoplankton within the Gulf would occur inspring and late summer.

Three-dimensional dispersion simulations ofzooplankton yielded further retention of particlesin the Gulf as a result of the net inward transportof deep waters in the Laurentian Channel. Thus,inoculated zooplankton would be mainlyretained within the Laurentian Channel andupstream coastal areas would be particularly atrisk (Gaspé Peninsula, Anticosti Island).Ultimately, inoculated zooplankton could beincorporated into the zooplankton dynamics ofthe Estuary and Gulf ecosystems.

ABWE southeast of Anticosti Island repre-sents a risk from a Gulf perspective, because ofthe significant retention of inoculated planktonicorganisms within the Gulf and for their disper-sion towards coastal areas. However, it was rec-ognized that there was a need for ballast waterregulating agencies in Canada and the U.S. andfor the industry to have an alternate zone in theGulf of St. Lawrence, and the authors tried toidentify such a ABWEZ by looking at areas fromwhere surface waters would be quickly flushedout of the Gulf.

FLUSHING OF SURFACE WATERSOUT OF THE GULF (90 DAYS)

Figure 6 shows the initial position of surfaceparticles that would be flushed out of the Gulf attwo week intervals during the simulations for the8 phytoplankton surface dispersion simulations.Deep blue and blue areas represent areas flushedout within 15 and 30 days respectively, lightblue and green areas represent areas flushed after30 and 45 days, and yellow and red areas are for75 and 90-day flushing time. These maps clearlyshow that surface waters in the Cabot Strait areaof the Gulf are rapidly flushed out to theAtlantic, usually within 30 days. This is particu-larly the case in summer, when high surface tem-perature would allow survival of a greater num-ber of inoculated species, and in late fall whenthe need for a ABWEZ is highest because of badweather conditions in the North Atlantic at thistime of the year.

76

CONCLUSION AND RECOMMENDATIONS

CONCLUSION

The Gulf area that is located around CabotStrait east of the Magdalen Islands could beacceptable for ABWE from a Gulf perspective.However, zooplankton inoculations in this areawould probably still be subjected to a netupstream transport and, more importantly, thefate of inoculated organisms that would beflushed out of the Gulf with surface waters isunknown, but may pose a threat to the CanadianAtlantic Coast. This hypothesis has been con-firmed with recent studies. However, theapproach presented in the present study shouldbe applied to the Atlantic Coast in order to iden-tify alternate zones for ballast water exchange.

RISK ANALYSIS AND DECISION SUPPORT SYSTEM

There is no ideal solution to this problemand some level of risk must be accepted.However, such an acceptable level of risk wouldhave to be minimized as much as possible; the

suitability of ballast water exchanges aroundCabot Strait (or other areas) would have to bebased also upon other conditions that influencethe survival of inoculated organisms, whichpoint to a risk analysis approach and a decisionsupport system. For ships proceeding up the St.Lawrence River and to the Great Lakes, the firstfactor to be considered would be the salinity ofballast waters transported by ships. Exchanges inthe Gulf of St. Lawrence could be allowed in theABWEZ when ships are transporting freshwaterin their ballast tanks. On the other hand,exchanges would not be required for ships trans-porting marine waters with salinity exceeding 30PSU, because these exchanges would not furtherreduce risks for ships proceeding up to freshwaterports but could increase those for the Gulf of St.Lawrence. If brackish waters were transported inballast tanks, then other factors would have toconsidered, the second most important being thetime of the year. In this case, the need forABWEZ would be particularly important duringfall and winter because of stormy conditions inthe North Atlantic. This is also a time of the yearwhen surface temperature in the Gulf decreasessignificantly to near freezing temperatures, which

77

< 15 Days

15 - 30 Days

30 - 45 Days

45 - 60 Days

60 - 75 Days

75 - 90 Days

Figure 6. Estimated time required for flushing surface waters out of the Gulf of St. Lawrence for the phytoplankton dispersion simulations.

limit the potential for survival of inoculatedorganisms. As a result, ABWE could be allowedfrom late fall through spring around Cabot Strait(or other areas) while for the rest of the year, athird factor would be incorporated in the riskanalysis, namely the origin of the ship. If shipsoriginate from tropical or subtropical areas, thenABWE would be allowed, whereas for shipsoriginating from temperate and subarctic ports,other means such as storage facilities and treat-ment options or sealed tanks (exchanges notallowed) would have to be considered. Such arisk analysis, if incorporated in voluntary guide-lines in conjunction with ABWEZ around CabotStrait, would maintain risks for ballast water-mediated introductions in the Great Lakes at alow level, but would significantly reduce theserisks for the Estuary and Gulf of St. Lawrence. Asimilar risk analysis could also be developed forships proceeding to marine ports of the Estuaryand Gulf. In conclusion, this approach would beconsistent with a precautionary approach andwould definitely facilitate the development ofregulations for ballast water exchange at thenational level in Canada.

LITERATURE CITED

Backhaus, J. O. 1985. A three-dimensional model for the simula-tion of shelf-sea dynamics. Dtsch. Hydrogr. Z. 38: 165-187.

Saucier, F.J., F. Roy, D. Gilbert, P. Pellerin, H. Ritchie. 2003.Modeling the formation and circulation processes of watermasses and sea ice in the Gulf of St. Lawrence. J. Geophys.Res. 108(C8): 3269-3289.

Risk Assessment for Species

Introduced from Ballast Water

Exchange

BERNARD KELLYGeocentric Mapping ConsultingFalmouth, NS Canadakellybk@atcon.com

ABSTRACT

The geographical information system (GIS)mapping of geo-referenced ballast waterexchange volume endpoints recorded on BallastWater Report Forms submitted to Marine Safety,Transport Canada, provides insight in ballastactivities reported off Eastern Canada during theyear 2002.

Vessel movement tracks and seasonal com-puter-modelled potential (interpolation) surfacethematic maps rendered location information onpreferred vessel exchange corridors and delineat-ed high-value areas for ballast exchange volumes(reported endpoints) for vessel traffic approach-ing Eastern Canada from the United States,Europe and other international departure points.

The same Transport Canada ballast waterdata provided for a comparative, statistical anal-ysis of shipping trends and patterns for EasternCanada, with years 1998 and 2000.

Please see the full 85-page report, GISMapping of Marine Vessel Ballast WaterExchange Enpoint Data in Atlantic Canada, forthe 2002 Shipping Season, which appears inthese proceedings as Appendix VII.

78