Taxon- and vector-specific variation in species richness and abundance during the

transport stage of biological invasions

Elizabeta Briski,1,* Sarah A. Bailey,1 Oscar Casas-Monroy,2 Claudio DiBacco,3 Irena Kaczmarska,4

Janice E. Lawrence,5 Jonas Leichsenring,5 Colin Levings,6,7 Michael L. MacGillivary,4

Christopher W. McKindsey,8 Leslie E. Nasmith,3 Marie Parenteau,2 Grace E. Piercey,6

Richard B. Rivkin,9 Andre Rochon,2 Suzanne Roy,2 Nathalie Simard,8 Bei Sun,9

Candice Way,9 Andrea M. Weise,8 and Hugh J. MacIsaac 10

1 Great Lakes Laboratory for Fisheries and Aquatic Sciences, Fisheries and Oceans Canada, Burlington, Ontario, Canada2 Institut des Sciences de la Mer de Rimouski, Universite du Quebec a Rimouski, Rimouski, Quebec, Canada3 Bedford Institute of Oceanography, Fisheries and Oceans Canada, Dartmouth, Nova Scotia, Canada4 Department of Biology, Mount Allison University, Sackville, New Brunswick, Canada5 Biology Department, University of New Brunswick, Fredericton, New Brunswick, Canada6 Centre for Aquaculture and Environmental Research, Fisheries and Oceans Canada, West Vancouver, British Columbia, Canada7 Institute for Resources, Environment and Sustainability, University of British Columbia, Vancouver, British Columbia, Canada8 Institut Maurice-Lamontagne, Fisheries and Oceans Canada, Mont-Joli, Quebec, Canada9 Ocean Sciences Centre, Memorial University of Newfoundland, St. John’s, Newfoundland, Canada10 Great Lakes Institute for Environmental Research, University of Windsor, Windsor, Ontario, Canada

Abstract

Recent empirical and statistical evidence suggest that propagule pressure (i.e., number of individualsintroduced per event, and the number and frequency of events) and colonization pressure (i.e., number of speciesreleased per event, and the number and frequency of events) are of vital importance to invasion success. Toexplore possible changes in propagule and colonization pressure during the transport stage of the invasionprocess, we examine abundance and species richness of virus-like particles, bacteria, diatoms, dinoflagellates, andinvertebrates transported in commercial ships—a leading vector for global spread of aquatic nonindigenousspecies. We collected 154 ballast water samples from ships that had performed or were exempt from ballast waterexchange (BWE) prior to arrival at Pacific and Atlantic ports in Canada and Laurentian Great Lakes ports. Wefound that abundance and species richness varied across taxa and regions, with ships arriving to the Atlanticregion carrying the highest abundance of taxa. The highest species richness of invertebrates and diatoms wasrecorded from ships arriving to the Pacific, whereas the richest communities of dinoflagellates occurred in theAtlantic region. We also found that BWE had no effect on abundance or species richness of most taxa(dinoflagellates, diatoms, bacteria, and virus-like particles), whereas the effect on abundance of invertebrates wasnot clear. Finally, longer voyages resulted in lower abundance of all taxa except dinoflagellates, and lower speciesrichness of diatoms. Paradoxically, the elevated abundance and species richness of dinoflagellates following BWEsuggest that this group could have enhanced invasion potential when ships manage ballast water by exchange.

The biological invasion process can be divided into aseries of stages: entrainment, transport, introduction,survival, establishment, and spread, where successfultransition between stages is determined by a number offactors, including propagule pressure (i.e., number ofindividuals; Lockwood et al. 2009; Simberloff 2009),physiological tolerance during transport and in therecipient ecosystem, and community interactions (Colauttiet al. 2006b; Simberloff 2009). Recent empirical andstatistical evidence suggest that propagule pressure is ofvital importance for invasion success (Colautti et al. 2006b;Lockwood et al. 2009; Simberloff 2009). The propagulepressure model suggests that the probability of successfulinvasion is improved by increasing the number ofintroduction events (i.e., propagule number) or the numberof propagules released per event (i.e., propagule size),owing to decreased environmental and demographic

stochasticity, respectively (Colautti et al. 2006b; Lockwoodet al. 2009; Simberloff 2009). Similarly, the number ofnonindigenous species (NIS) in a system may be a reflectionof colonization pressure (i.e., number of species released;Lockwood et al. 2009).

Any mechanism for rapid translocation of large volumesof water over great distances has the potential to facilitatemass invasions of plankton and other species, rearrangingglobal biodiversity (Olden et al. 2004). The shippingindustry transports . 90% of the world’s trade (Hulme2009), with ships’ ballast water being a leading mechanismfor global spread of aquatic NIS (NRC 2008; Hulme 2009).To reduce invasion risk, regulations requiring ballast waterexchange (BWE) have been implemented in Canada, theUnited States of America, and numerous additionalcountries (regulation D-1; USCG 1993; IMO 2004;Government of Canada 2006). BWE requires nearlycomplete ($ 95% volumetric exchange) replacement ofcoastal water in filled ballast tanks with oceanic water,* Corresponding author: elizabeta.briski@dfo-mpo.gc.ca

Limnol. Oceanogr., 58(4), 2013, 1361–1372

E 2013, by the Association for the Sciences of Limnology and Oceanography, Inc.doi:10.4319/lo.2013.58.4.1361

1361

which theoretically should reduce species abundance and,to a lesser extent, richness by purging individuals out oftanks and by killing remaining taxa owing to osmotic shock(IMO 2004). In addition, regulation D-2 is a ballast waterperformance standard proposed in an international con-vention that, if ratified, will require ships to manage ballastwater such that they discharge , 10 viable individuals m23

(ind. m23) that are $ 50 mm in minimum dimension, and ,10 viable ind. mL21 for species , 50 mm in maximum and$ 10 mm in minimum dimension, in addition to limits forindicator bacteria (IMO 2004). As species may responddifferently to changes in environmental conditions duringtransport, such as temperature, salinity, phosphate con-centration, and dissolved oxygen concentration (Klein et al.2010; Seiden et al. 2011), propagule pressure and coloni-zation pressure—and invasion risk in consequence—couldbe altered by voyage characteristics including voyage lengthand ship pathway (i.e., trans-pacific, trans-atlantic, coastal-pacific, or coastal-atlantic).

Here we explore changes in propagule and colonizationpressure during the transport stage of the invasion processby examining abundance and species richness of planktonin relation to voyage characteristics, in particular BWE,length of voyage, and ship pathway, using one of theworld’s most comprehensive studies of ships’ ballast watersince regulations were enacted. Our focus was to cover thecomplete array of taxa transported in ballast water of shipsarriving to Pacific and Atlantic ports in Canada and toports on the Laurentian Great Lakes. Additionally, wecollected port samples to compare abundances of taxa inships to that in nature. We test the hypotheses that voyagecharacteristics alter both abundance and species richness ofvirus-like particles (VLPs), bacteria, diatoms, dinoflagel-lates, and invertebrates, and that BWE reduces introduc-tion risk of these taxa to Canada.

Methods

Sampling—Water samples were collected opportunisti-cally between May 2007 and August 2009 from ballasttanks of ships arriving to ports in Pacific and AtlanticCanada or on the Laurentian Great Lakes (Table 1). Shipssampled serviced both domestic and international routesand originated from four continents (20 countries;Table 1), with some having performed BWE. Ships arrivingto the Atlantic and Great Lakes regions typically operatedbetween Atlantic ports, whereas those arriving to Pacificports traveled mainly among Pacific ports (Table 1).Depending on their last port of call and BWE status, shipswere grouped into three categories: transoceanic ex-changed, coastal exchanged, and coastal not-exchangedships. We aimed to sample 20 ships per category arriving oneach coast and to the Great Lakes. Transoceanic exchangedships arrived from any continent except North America.Coastal exchanged ships arrived from American ports,south of Cape Blanco, Oregon, on the Pacific coast andsouth of Cape Cod, Massachusetts, on the Atlantic coast;coastal not-exchanged ships arrived from American andCanadian ports north of these locations and were exemptfrom BWE (USGC 1993; Government of Canada 2006).

Official ballast water reporting forms submitted toTransport Canada, which included dates and locations ofballast uptake and discharge, and last port of call, wereused to discern pathways and ship categories.

We collected 24 transoceanic exchanged, 25 coastalexchanged, and 20 coastal not-exchanged ballast watersamples in the Pacific region, 22 transoceanic exchanged,22 coastal exchanged, and 21 coastal not-exchangedsamples in the Atlantic region, and 17 transoceanicexchanged and 3 coastal exchanged samples in the GreatLakes. Samples of all taxa were collected from the ships’deck through an opened ballast tank manhole hatch.Invertebrate samples were collected with a 30 cm diameter,125 mm Nitex plankton net with flowmeter that waslowered to the maximum accessible depth inside the ballasttank and retrieved at a speed of , 1 m s21. Vertical haulswere repeated until , 1000 liters of ballast water had beenfiltered. Samples were concentrated and preserved in 5%buffered formalin and stored at room temperature untilanalysis. Further, using a 5 liter Niskin bottle, water wascollected at four depths (spread evenly between the surfaceand 1–2 m above bottom) and mixed together in a 20 litercarboy. Three 1 liter subsamples were preserved withacidified Lugol’s solution for analyses of diatoms. At least13.5 liters of water was further filtered through 73 and20 mm sieves, with the contents of the latter rinsed into250 mL glass bottles and preserved with acidified Lugol’ssolution for dinoflagellates. Five hundred milliliters waspreserved with formaldehyde (final concentration 3.7%)and shipped on ice within 5 d for bacterial analyses.Finally, one 250 mL whole water sample was collected andkept on ice for analysis of VLPs. Additionally, we collected12 port samples in Canada, 4 in Europe, and 1 in Asia. Portsamples were collected in the same ports where sampledships had loaded or discharged ballast. All samples wereshipped by commercial couriers to our laboratories foranalysis.

Density counts and identification—As a part of theCanadian Aquatic Invasive Species Network, samplescollected simultaneously in this study were processed bytaxon-specific, expert academic and governmental labora-tories. Invertebrate samples were counted and morpholog-ically identified to the lowest feasible taxonomic level andontogenetic stage (e.g., nauplii) with stereomicroscopy.Only physically intact individuals were counted; no otherdistinction between live or dead individuals was made.

Dinoflagellate samples were kept at 4uC in the dark untilexamination in the laboratory. Each water sample wasmixed by overturning by hand . 20 times, and a volume of50 mL was put in a settling column for 24 h. Dinoflagellatecells were identified and counted in the settled volume usinga Zeiss Axiovert 100 inverted microscope. When necessary,calcofluor was used to help identification by coloring thethecal plates (Fritz and Triemer 1985). The parent solutionwas prepared with 0.1 g of calcofluor from Sigma (F-3543)in 100 mL of distilled water. The working solution wasprepared from 0.1 mL of the parent solution in 50 mL ofdistilled water. One to two drops of this solution weresufficient to color dinoflagellates, which were observed

1362 Briski et al.

Table

1.

Su

mm

ary

da

tafo

rsh

ips

sam

ple

din

thre

ere

gio

ns

(Pa

cifi

c,G

rea

tL

ak

es,

an

dA

tla

nti

c)o

fN

ort

hA

mer

ica

,b

yb

roa

dg

eog

rap

hic

rou

te.

Sh

ipca

teg

ori

esa

retr

an

soce

an

icex

cha

ng

ed(T

OE

),co

ast

al

exch

an

ged

(CE

),a

nd

coa

sta

ln

ot-

exch

an

ged

(CN

E).

Ind

ivid

ua

lsh

iptr

an

sits

ha

ve

bee

ng

rou

ped

acc

ord

ing

toco

un

try

or

sta

teo

fo

rig

ino

fb

all

ast

wa

ter

an

dre

gio

no

fa

rriv

al.

Th

em

ean

vo

ya

ge

len

gth

,m

ean

vo

ya

ge

len

gth

aft

erb

all

ast

wa

ter

exch

an

ge

(BW

E),

nu

mb

ero

fsa

mp

les

coll

ecte

d,

mo

nth

wh

ensa

mp

lew

as

coll

ecte

d,

an

dn

um

ber

of

sam

ple

sw

ith

ou

td

etec

tab

led

ino

fla

gel

late

sa

rere

po

rted

,w

hen

ap

pli

cab

le.

na

den

ote

sn

ot

ap

pli

cab

le.

Reg

ion

Sh

ipca

teg

ory

Ori

gin

of

ba

lla

stw

ate

rM

ean

tota

lv

oy

ag

ele

ng

th(d

)M

ean

vo

ya

ge

len

gth

aft

erB

WE

(d)

No

.o

fsa

mp

les

Mo

nth

wh

ensa

mp

lew

as

coll

ecte

d

No

.o

fsa

mp

les

wit

ho

ut

det

ecta

ble

din

ofl

ag

ella

tes

Pa

cifi

cT

OE

Ch

ina

19

.89

.66

Ma

y,

Jul

3Ja

pa

n2

3.4

11

.71

2M

ay

,Ju

n,

Au

g,

Oct

8S

ou

thK

ore

a2

0.2

58

.75

Jun

–A

ug

,O

ct1

Ha

wa

ii,

U.S

.A.

12

61

Jul

0C

EC

ali

forn

ia,

U.S

.A.

8.5

4.8

17

All

mo

nth

sex

cep

tO

cta

nd

Dec

6E

cua

do

r1

39

1Ju

l1

Gu

ate

ma

la1

26

1Ju

l1

Lo

uis

ian

a,

U.S

.A.

82

1Ju

n0

Mex

ico

11

.87

5F

eb,

Jun

,Ju

l3

CN

EB

riti

shC

olu

mb

ia,

Ca

na

da

6n

a2

Ap

r,N

ov

1O

reg

on

,U

.S.A

.5

na

5A

pr–

Jun

,A

ug

,O

ct,

No

v1

Wash

ingto

n,

U.S

.A.

4.1

na

13

Ap

r–A

ug,

Sep

6G

rea

tL

ak

esT

OE

Bel

giu

m2

4.3

10

3Ju

l,A

ug

,O

ct0

Bra

zil

35

14

1Ju

l1

Co

lom

bia

35

16

1Ju

l0

Ger

ma

ny

17

.31

0.6

3Ju

l,S

ep,

No

v0

Gre

ece

36

21

1Ju

l1

Ita

ly5

02

01

Au

g0

Net

her

lan

ds

21

14

1A

ug

0N

orw

ay

32

17

1N

ov

0S

pa

in1

8.5

11

.52

Au

g,

Sep

1U

nit

edK

ing

do

m2

21

33

Au

g,

Oct

2C

EM

ass

ach

use

tts,

U.S

.A.

86

1S

ep0

New

Jers

ey,

U.S

.A.

87

1A

ug

0P

enn

sylv

an

ia,

U.S

.A.

12

61

Oct

0A

tla

nti

cT

OE

Bel

giu

m1

61

31

Ap

r1

Gre

ece

19

81

Jun

0It

aly

21

.37

.33

Ap

r,M

ay

,A

ug

2Is

rael

16

51

Jun

0N

eth

erla

nd

s1

3.6

7.2

6M

ay

,Ju

n,

Au

g0

Po

lan

d1

71

11

Jun

1S

pa

in1

21

01

Ma

y0

Un

ited

Kin

gd

om

11

.15

.68

Ap

r–Ju

n,

Au

g0

CE

Lo

uis

ian

a,

U.S

.A.

83

1Ju

n0

Ma

ryla

nd

,U

.S.A

.5

.61

.33

Ma

y,

Au

g0

Ma

ssa

chu

sett

s,U

.S.A

.4

2.3

3Ju

ne

Au

g2

New

Ha

mp

shir

e,U

.S.A

.3

1.5

2Ju

n0

New

Jers

ey,

U.S

.A.

31

3Ju

n,

Jul

1N

ewY

ork

,U

.S.A

.5

1.8

5Ju

n–

Au

g1

No

rth

Ca

roli

na

,U

.S.A

.3

21

Jul

0P

enn

sylv

an

ia,

U.S

.A.

63

.25

4M

ay

,Ju

n,

Au

g3

CN

EM

ain

e,U

.S.A

.2

na

4Ju

n,

Jul

0M

ass

ach

use

tts,

U.S

.A.

1.3

na

8Ju

n–

Au

g0

New

Bru

nsw

ick

,C

an

ad

a5

na

5Ju

l3

New

Ha

mp

shir

e,U

.S.A

.1

na

1Ju

n0

No

va

Sco

tia

,C

an

ad

a2

.5n

a3

Jul

1

Transport stage of biological invasions 1363

15 min later using fluorescence microscopy with anexcitation wavelength of 350–380 nm. Cell viability wasnot determined in this study, but only cells with clearlyvisible cell content were counted and assumed viable.

Diatom samples were concentrated to 20 mL and storedat 4uC until analysis. Enumeration and identification wascompleted on a 10 mL subsample of the concentrate usingan Utermohl settling chamber (Hydrobios). To restorechloroplast autofluorescence (quenched by Lugol’s fixative)saturated sodium thiosulfate solution in distilled water wasadded to the sample prior to settling in the chamber.Counts were carried out using a Zeiss Axiovert 200 lightmicroscope with epifluorescence illumination (HBO 50/AC,Mercury Shortarc) and 203 and 403 objectives. Onlyfluorescing cells were counted. Cells and spores $ 10 mm inshortest linear dimension were identified to the lowestpractical taxonomic level, most often to species. Scanningelectron microscopy was used to verify identification whenneeded for the most abundant species. For scanningelectron microscopy examination, after light microscopycounts were completed, the contents of the settling chamberwere transferred to a filtering unit and acid cleaned usingpotassium permanganate and oxalic acid. The cleanedsample was examined using a JEOL JSM-5600 scanningelectron microscope (JEOL USA), operating at 10 kVaccelerating voltage and 8 mm working distance. Furtherdetails of this method may be found in Kaczmarska et al.(2005) and Klein et al. (2010).

Bacteria samples were filtered onto 25 mm diameter,0.2 mm black polycarbonate filters (GE Osmonics Lab-store) and stained with acridine orange (final concentration1.872 3 1025 g L21; Kirchman et al. 1982). Slides werecorrected for non-cell staining by counting 5 mL of 0.2 mmfiltered distilled water before each sample filtration. Toeliminate possible bias introduced by prolonged slidestorage, all slides were counted within 1 h after beingmade. The slides were observed using an Olympus BH2-RFC epifluorescence microscope, equipped with a 10031.30 oil objective (12503 total magnification), a 100 Wmercury lamp, and appropriate filter sets (502 nm forexcitation, 526 nm for emission). The optimal cell directcounting scheme was followed to optimize the accuracy ofthe direct count (Kirchman et al. 1982).

Slides for VLP counts were made and enumeratedaccording to Patel et al. (2007). Using a 12-sample filtrationmanifold (Millipore), VLPs were collected from single800 mL subsamples by filtration onto Anodisc filters(0.02 mm pore size, 25 mm, Anopore aluminum oxidefilter, Whatman) atop backing filters (0.8 mm, MilliporeMF–MilliporeTM AAWG mixed cellulose ester filter,25 mm, MilliporeH) to prevent desiccation. Additionalsubsamples were also prepared by diluting 10-fold in0.02 mm filtered 32% f/2 seawater media (Guillard 1975).Samples were filtered onto the Anodiscs using a vacuum of, 250 mm Hg. After filtration, deoxyribonucleic acidpresent on the Anodiscs was stained by placing the filter onan 80 mL drop of 253 SYBRH Gold (Molecular Probes,Invitrogen) for 15 min in the dark. To prevent fluorescencefading, the filter was then mounted on a glass microscopeslide using 34 ml of 1% v : v of 10% w : v phenylenediamine

(Sigma) and 99% v : v of 50 : 50 phosphate-buffered saline(137 mol L21 NaCl, 2.7 mol L21 KCl, 10 mol L21

Na2HPO4, and KH2PO4 at pH 7.4) in glycerol. Each filterwas topped with a glass cover slip, sealed using anonfluorescing nail polish, and stored at 220uC untilcounting. Estimates of VLP abundance were made bycounting VLPs under a Leica epifluorescence microscope(DM2500 with an I3 filter for epifluorescence: excitation at450–490 nm, emission at 515 nm). Between 30 and 40 VLPswere counted in n squares of an ocular microscope grid,and VLPs in n squares were then counted in at least 10randomly selected fields of view at 10003 magnification,totaling . 200 VLPs slide21 (Patel et al. 2007) to achieve95% confidence in VLP abundance. Control slides weremade from 0.02 mm filtered distilled, deionized H2O and32% f/2 seawater media. Control slides typically did notcontain VLPs. If this was not the case, then the dilutedsamples were not used for VLP enumeration.

As a part of this data has already been published, furtherdetails of methods may be found in DiBacco et al. (2012)for invertebrates, in Roy et al. (2012) and Casas-Monroy(2012) for dinoflagellates, in Klein et al. (2010) for diatoms,and in Sun et al. (2010) for bacteria. Port samples wereanalyzed using the same methods as for ballast watersamples. Data for all port samples were grouped together.

Statistical analysis—Data for each taxonomic groupwere log-transformed to meet assumptions of parametrictests. Variation in total abundance and species richnessamong sampling regions and ship categories was comparedusing two-way multivariate analysis of variance (MAN-OVA), where taxa (invertebrates, dinoflagellates, diatoms,bacteria, and VLPs) were the dependent variables andsampling regions (Pacific, Great Lakes, and Atlantic) andship categories (transoceanic exchanged, coastal ex-changed, and coastal not-exchanged) were the independentvariables. Four separate MANOVAs were performed usingtotal abundance, total species richness, NIS abundance,and NIS species richness. Significance levels for statisticalcomparisons were adjusted for multiple pair-wise compar-isons by Bonferroni-type correction with a family-wiseerror rate of 0.05. Furthermore, to determine if totalabundance or species richness of taxa was related to voyagelength, a series of regression analyses were conducted withtotal abundance and species richness as the dependentvariables and voyage length as the independent variable.The length of voyage was calculated from date of ballastuptake until date of discharge in recipient port for coastalnot-exchanged tanks, and from date of BWE until date ofdischarge in recipient port for transoceanic exchanged andcoastal exchanged tanks. Samples with no detectable taxa(i.e., 51 dinoflagellate and two diatom samples; Table 1)were excluded from regression analyses. Finally, todistinguish between the effect of BWE and other voyagecharacteristics on bacteria and VLPs, data collected in theAtlantic region from transoceanic exchanged ships wascompared to that of transoceanic not-exchanged shipscollected 10 yr ago (Ruiz et al. 2000) using MANOVAwhere taxa (bacteria and VLPs) were the dependentvariables and ship category (transoceanic exchanged and

1364 Briski et al.

transoceanic not-exchanged ships) was the independentvariable; sampling and analysis methods between the twostudies were sufficiently similar to allow the comparison.

Results

Changes in abundance among regions—Our samplesrevealed significant regional variation in total abundanceof invertebrates, dinoflagellates, and bacteria, but not ofdiatoms and VLPs (Table 2; Fig. 1). Mean total abundancesampled from Pacific ships was 1.0 3 104, 2.5 3 104, 4.7 3107, 8.0 3 1011, and 7.3 3 1012 ind. m23 for invertebrates,dinoflagellates, diatoms, bacteria, and VLPs, respectively,whereas that for Great Lakes ships was 6.4 3 103, 3.2 3104, 1.5 3 106, 7.8 3 1011, and 9.0 3 1013 ind. m23. Finally,mean total abundance of invertebrates, dinoflagellates,diatoms, bacteria, and VLPs sampled from Atlantic shipswas 2.4 3 104, 2.7 3 105, 2.3 3 106, 9.8 3 1011, and 7.5 31012 ind. m23, respectively. Even though total abundanceof diatoms was an order of magnitude higher in Pacificships than in those from the Atlantic and Great Lakes,there was no significant difference among regions (Table 2;Fig. 1). Furthermore, mean abundance per species sampled

from Pacific ships was 9.7 3 102, 5.9 3 103, and 4.8 3 106

ind. species21 m23 for invertebrates, dinoflagellates, anddiatoms, respectively, whereas that for Great Lakes shipswas 74, 4.7 3 103, and 1.4 3 105 ind. species21 m23.Finally, mean abundance per species arriving to theAtlantic coast was 1.8 3 103, 9.4 3 103, and 1.4 3 105

ind. species21 m23 for invertebrates, dinoflagellates, anddiatoms, respectively. Invertebrates and bacteria werepresent in all ships; however, dinoflagellates were notdetected in 53%, 25%, and 23% of ships in the Pacific,Great Lakes, and Atlantic regions, respectively, diatoms in1% and 2% of ships in the Pacific and Atlantic regions,respectively, and VLPs in 4% of ships arriving to theAtlantic region. Assuming that higher abundance in ballasttanks represents a higher probability of introduction, ourresults revealed that introduction potential is highest forinvertebrates, dinoflagellates, and bacteria to the Atlanticcoast of Canada, for diatoms to the Pacific coast, and forVLPs to the Great Lakes.

Changes in abundance among ship categories and effectof BWE—Our results showed that there was significantvariation in total abundance of invertebrates, bacteria, and

Table 2. Results of two-way multivariate analysis of variance addressing the effect of region (Pacific, Great Lakes, and Atlantic) andship category (transoceanic exchanged, coastal exchanged, and coastal not-exchanged) on total abundance, total species richness,nonindigenous species (NIS) abundance, and NIS species richness for diversified taxa. Significant p-values after Bonferroni-typecorrection are presented in bold. na denotes not applicable.

Variable

Total abundance Total species richness NIS abundance NIS species richness

Value p Value p Value p Value p

Region

Univariate F-tests

Invertebrates ,0.001 0.779 ,0.001 ,0.001Dinoflagellates 0.003 0.017 0.023 0.082Diatoms 0.050 ,0.001 ,0.001 ,0.001Bacteria 0.001 na na naVLPs 0.190 na na na

Multivariate test

Wilks’ lambda 0.429 ,0.001 0.674 ,0.001 0.605 ,0.001 0.408 ,0.001

Ship category

Univariate F-tests

Invertebrates 0.010 0.875 0.103 0.436Dinoflagellates 0.709 0.111 0.098 0.019Diatoms 0.744 0.310 0.239 0.023Bacteria ,0.001 na na naVLPs 0.001 na na na

Multivariate test

Wilks’ lambda 0.691 ,0.001 0.929 0.409 0.905 0.082 0.811 0.009

Interaction

Univariate F-tests

Invertebrates 0.296 0.933 ,0.001 0.124Dinoflagellates 0.114 0.479 0.033 0.082Diatoms 0.006 0.009 0.172 0.522Bacteria 0.418 na na naVLPs 0.015 na na na

Multivariate test

Wilks’ lambda 0.708 0.005 0.862 0.191 0.605 ,0.001 0.817 0.062

Transport stage of biological invasions 1365

VLPs among ship categories, but not for dinoflagellatesand diatoms (Table 2; Fig. 1). Mean total abundancetransported by transoceanic exchanged ships was 8.1 3103, 8.7 3 104, 1.2 3 106, 7.0 3 1011, and 1.1 3 1012 ind.m23 for invertebrates, dinoflagellates, diatoms, bacteria,and VLPs, respectively, whereas that for coastal exchangedships was 1.4 3 104, 1.8 3 105, 2.3 3 106, 9.1 3 1011, and8.8 3 1012 ind. m23. Finally, mean total abundancetransported by coastal not-exchanged ships was 8.7 3104, 3.1 3 105, 1.3 3 108, 1.1 3 1012, and 9.6 3 1012 ind.m23 for invertebrates, dinoflagellates, diatoms, bacteria,and VLPs, respectively. Total abundance of taxa in

different ship categories ranged from , 1 to several ordersof magnitude per cubic meter (Table 3). Furthermore,mean abundance per species transported by transoceanicexchanged ships was 7.7 3 102, 8.3 3 103, and 3.1 3 105

ind. species21 m23 for invertebrates, dinoflagellates, anddiatoms, respectively, whereas that of coastal exchangedships was 1.1 3 103, 6.7 3 103, and 1.5 3 105 ind.species21 m23. Finally, mean abundance per speciestransported by coastal not-exchanged ships was 2.4 3103, 4.0 3 103, and 1.3 3 107 ind. species21 m23 forinvertebrates, dinoflagellates, and diatoms, respectively.Dinoflagellates were not detected in 35%, 37%, and 39% of

Fig. 1. Mean total and nonindigenous species (NIS) abundance by taxon and region in transoceanic exchanged (TOE), coastalexchanged (CE), and coastal not-exchanged ships (CNE). Standard errors are included. An asterisk denotes a significant difference (p ,0.05) from other ship categories. Note the difference in scale among plots and that the y-axis is log scaled.

1366 Briski et al.

transoceanic exchanged, coastal exchanged, and coastalnot-exchanged ships, respectively, diatoms in 1% oftransoceanic exchanged and 2% of coastal exchangedships, and VLPs in 5% of coastal not-exchanged ships.Although total abundance was not consistently highest forall taxa and in all three regions in coastal not-exchangedships, our multivariate comparison revealed significantlyhigher total abundance of invertebrates in both Pacific andAtlantic regions, and of diatoms and bacteria in the Pacificregion in coastal not-exchanged ships (Fig. 1). Thesefindings suggest that BWE reduced total abundance ofthe three taxa; however, this conclusion may be confound-ed by voyage length. Finally, to distinguish between effectof BWE and that of other voyage characteristics onbacteria and VLPs, comparison of our Atlantic transoce-anic exchanged samples with similar samples collected 10 yrago revealed no significant difference, indicating that BWEdid not reduce total abundance of bacteria nor VLPs intransoceanic exchanged ships (Table 4; Ruiz et al. 2000).Our Atlantic transoceanic exchanged ships contained onaverage 6.7 3 1011 and 3.1 3 1012 bacteria and VLPs m23,respectively (Fig. 1), whereas transatlantic, unmanagedships sampled 10 yr ago contained 8.3 3 1011 and 7.4 31012, respectively (Ruiz et al. 2000). Taking into accountthat there was no difference among ship categories fordinoflagellates and diatoms, and our comparison revealedno effect of BWE on bacteria and VLPs, our study showsthat BWE may only be effective in reducing totalabundance of invertebrates.

Changes in species richness among regions—Our samplesrevealed significant regional variation in species richness ofdinoflagellates and diatoms, but not invertebrates (Table 2;Fig. 2). Mean species richness transported by ships arriving

to the Pacific coast of Canada was 13.7, 8.6, and 17.4species m23 of invertebrates, dinoflagellates, and diatoms,respectively, whereas that of ships arriving to the GreatLakes was 11.5, 6.7, and 4.3 species m23, respectively.Mean species richness transported by ships arriving to theAtlantic coast was 12.0, 17.4, and 14.5 species m23 ofinvertebrates, dinoflagellates, and diatoms, respectively.Assuming that higher species richness corresponds withhigher likelihood that some species will survive transportand adapt to environmental and biological conditions ofthe new area, our study indicates that the Pacific region ofCanada is under highest invasion risk by invertebrates anddiatoms, whereas the Atlantic region appears vulnerable todinoflagellate invasions.

Changes in species richness among ship categories andeffect of BWE—Mean species richness of all three taxaexamined was similar among the three ship categories(Table 2; Fig. 2). Mean species richness transported bytransoceanic exchanged ships was 12.5, 11.1, and 12.3species m23 of invertebrates, dinoflagellates, and diatoms,respectively, whereas that of coastal exchanged ships was13.1, 14.6, and 9.0 species m23, respectively. Mean speciesrichness transported by coastal not-exchanged ships was11.8, 9.0, and 12.4 species m23, respectively. Even thoughtwo-way MANOVA showed that there was no significantdifference in species richness between the three shipcategories for any of the taxa examined (Table 2; Fig. 2),our multiple comparisons revealed significantly higherspecies richness of both invertebrates and dinoflagellatesin Atlantic transoceanic exchanged and coastal exchangedships when compared to coastal not-exchanged ships(Fig. 2), demonstrating that new species were likely addedto tanks during BWE.

Effect of voyage length on abundance and speciesrichness—Mean voyage length varied among regions andship categories (Table 1). Mean voyage length for trans-oceanic exchanged, coastal exchanged, and coastal not-exchanged ships arriving to the Pacific coast of Canada was21.7, 9.5, and 4.3 d, respectively, whereas that of shipsarriving to the Atlantic coast was 14.3, 4.6, and 2.3 d,respectively. Mean voyage length for ships arriving to theGreat Lakes was 26.3 and 9.3 d for transoceanic exchangedand coastal exchanged ships, respectively. Likewise, voyagelength after BWE also varied among regions and shipcategories (Table 1). Mean voyage length after BWE of

Table 3. The ranges of total abundance of diversified taxa in mid-ocean, coastal areas, ports, and in ballast water of different shipcategories (transoceanic exchanged [TOE], coastal exchanged [CE], and coastal not-exchanged ships [CNE]). References for mid-oceanand coastal areas are included. Port samples were collected in this study. na denotes not applicable.

Taxon Mid-ocean (ind. m23) Coastal areas (ind. m23) Ports (ind. m23) TOE (ind. m23) CE (ind. m23) CNE (ind. m23)

Invertebrates 101–103a 102–106b,c 104–105 10–104 10–105 102–105

Dinoflagellates 101–103d 105–106e na 0–106 0–106 0–106

Diatoms 102–103d 104–108e 103–108 0–107 0–107 103–109

Bacteria 1010–1012f 1010–1012f 1011–1012 1010–1012 1011–1012 1011–1012

VLPs 1010–1012f 1011–1013f 1012–1013 1011–1013 1011–1013 0–1013

a Koppelmann and Wiekert 1992; b Fernandez de Puelles et al. 2003; c Hwang et al. 2010; d Head and Pepin 2010; e Cloern 1979; f Culley and Welschmeyer2002.

Table 4. Results of multivariate analysis of varianceaddressing the effect of BWE on total abundance of bacteriaand virus-like particles (VLPs).

Variable

Total abundance

Value p

Univariate F-tests

Bacteria 0.054VLPs 0.098

Multivariate test

Wilks’ lambda 0.818 0.073

Transport stage of biological invasions 1367

ships arriving to the Pacific coast was 10.5 and 5.6 d fortransoceanic exchanged and coastal exchanged ships,respectively, and of those arriving to the Atlantic coastwas 7.1 and 2.0 d, respectively. Mean voyage length afterBWE of ships arriving to the Great Lakes was 13.3 and6.3 d for transoceanic exchanged and coastal exchangedships, respectively. Regression analyses revealed that lengthof voyage influenced invertebrate, diatom, bacteria, andVLP abundances, but not that of dinoflagellates (Fig. 3).With an increase in voyage length, invertebrate totalabundance decreased the fastest, followed by that ofbacteria, diatoms, and VLPs (Fig. 3). Increased voyagelength decreased species richness of diatoms, but not ofinvertebrates or dinoflagellates (Fig. 4).

Abundance of taxa in nature and in ballast tanks—Ourcomparison of abundance of plankton in mid-ocean,coastal, and port areas revealed that mid-ocean areascontain lower total abundance of invertebrates, dinoflagel-lates, diatoms, and VLPs (but not bacteria) when comparedto both coastal and port areas (Table 3). Even though totalabundance of all taxa in coastal and port areas was similar,the range of total abundance of invertebrates in coastalareas was greater than that in ports, and port samples nevercontained very low total species’ abundances, as has beenreported in previous studies from coastal areas (Fernandezde Puelles et al. 2003; Hwang et al. 2010). The lowest totalabundance of invertebrates in coastal and port areas was102 and 104 ind. m23, respectively (Table 3). Furthermore,port samples contained slightly higher total abundances ofbacteria and VLPs than coastal areas (the lowest totalabundance of bacteria and VLPs in ports was 1011 and 1012

ind. m23, respectively; Table 3).Our port samples contained, on average, similar total

species’ abundances as recorded from our ship samples,though ranges were broader with the latter group andoccasionally ships contained no detectable dinoflagellates,diatoms, or VLPs at all (Table 3). Furthermore, totalabundance of all taxa in coastal exchanged and coastal not-exchanged ships was similar to that in coastal and portareas; however, our transoceanic exchanged samplesoccasionally carried three and four orders of magnitudehigher total abundances of dinoflagellates and diatoms,

Fig. 2. Mean total and nonindigenous species (NIS) richnessby taxon and region in transoceanic exchanged (TOE), coastalexchanged (CE), and coastal not-exchanged ships (CNE).Standard errors are included. An asterisk denotes a significantdifference (p , 0.05) from other ship categories. Note thedifference in scale among plots.

Fig. 3. Scatterplots and fitted regression lines with total abundance as the dependent variable and voyage length as the independentvariable. Voyage length of transoceanic and coastal exchanged ships was calculated from ballast water exchange until discharge, whereasthat of coastal not-exchanged ships was calculated from loading of ballast until discharge. An asterisk denotes significant difference(p , 0.05).

1368 Briski et al.

respectively, than was reported for mid-ocean areas(Table 3). As BWE should replace coastal with open-oceantaxa, our results indicate that dinoflagellates and diatomsmay accumulate in ballast tanks.

Discussion

The risk of introduction of new species by ships’ ballastwater was not uniform across regions or taxa despiteimplementation of common, national management prac-tices (USCG 1993; Government of Canada 2006) and mayrequire specific management strategies for biota ofparticular risk in specific regions. Although abundanceswere highest for most taxa in Atlantic ships, and speciesrichness was highest in Pacific ships, both regions wereexposed to propagules of all five taxonomic groups studied.Regional differences in abundance may be partiallyexplained by mortality of species due to voyage length(Klein et al. 2010; Seiden et al. 2011; Simard et al. 2011),with voyages to the Pacific region being nearly twice aslong as those to the Atlantic region. In addition, Seidenet al. (2011) reported regrowth of bacteria during the firstseveral days of voyages, and Klein et al. (2010) observedsome increases in diatom abundances, suggesting possiblereproduction in ballast tanks. However, both authorsreported a decline in abundance of taxa at the end ofvoyages, possibly related to fluctuating and inhospitableconditions in ballast tanks. Environmental conditions canchange dramatically in ballast tanks, including in temper-ature of up to 20uC, in dissolved oxygen of up to 6 mg L21,and salinity of up to 15% (Klein et al. 2010; Seiden et al.2011; Simard et al. 2011). These findings are in agreementwith our results. Regional differences in species richnessobserved in ballast tanks may be influenced by differencesin biodiversity among source regions; however, owing to alack of historical work on biodiversity and biogeography(Carlton 2009), it is difficult to examine this factor further.

Our study revealed that BWE does not effectively reduceabundances of dinoflagellates, diatoms, bacteria, and VLPsin either Pacific or Atlantic ships. As longer voyages also

increase exposure of transported individuals to ambientenvironmental conditions of ballast tanks (Klein et al.2010; Seiden et al. 2011), we could not distinguish the effectof BWE from other factors on invertebrate abundance.Simard et al. (2011) reported that environmental conditionsin ballast tanks explained , 40% of observed reduction inzooplankton abundance, out of a total 72–90% reductionin abundance during two ship voyages. Nonetheless, oncethe international convention is ratified, regulation D-2 willrequire ships to manage ballast water to meet specificmaximum permissible levels for viable organisms containedin discharge (IMO 2004). Taking into account that mostinvertebrates are greater than 50 mm minimum dimension,and most diatoms and dinoflagellates smaller than 50 mm,our study indicates that BWE will not meet the D-2standards. Even though the total number of individualsafter BWE may exceed the D-2 standard, BWE appears toprovide robust protection for freshwater systems at leastfor invertebrates, by eliminating 99% of freshwaterzooplankton through exposure to highly saline ocean water(Gray et al. 2007; Bailey et al. 2011).

Many studies have reported lower abundances ofzooplankton (Koppelmann and Wiekert 1992; Fernandezde Puelles et al. 2003), phytoplankton (Cloern 1979; Headand Pepin 2010), and microbes (Culley and Welschmeyer2002) in mid-ocean than in coastal water; however,abundances in both mid-ocean and coastal areas may varyby several orders of magnitude depending on location,season, and year of collection. Interestingly, while abun-dance of all taxonomic groups in both coastal exchangedand coastal not-exchanged ships were in the same range asin natural coastal areas, abundances of dinoflagellates anddiatoms in transoceanic exchanged ships were occasionallythree and four orders of magnitude higher than thosereported in mid-ocean, respectively. We confirmed that ourtransoceanic exchanged ships did conduct BWE in mid-ocean and our samples were collected across all seasons;therefore, the high abundance of taxa in these vesselscannot be explained by BWE conducted in coastal areas orby sample collections only during peak abundance of

Fig. 4. Scatterplots and fitted regression lines with species richness as the dependent variableand voyage length as the independent variable. Voyage length of transoceanic and coastalexchanged ships was calculated from ballast water exchange until discharge, whereas that ofcoastal not-exchanged ships was calculated from loading of ballast until discharge. An asteriskdenotes significant difference (p , 0.05).

Transport stage of biological invasions 1369

dinoflagellates and diatoms in oceans. Rather, our resultsappear to be related to incomplete BWE and accumulationof these taxa in ships’ ballast tanks. Klein et al. (2010)reported four times higher densities of diatoms inexchanged than not-exchanged ballast tanks on a shipconducting a trans-pacific voyage. Moreover, many dino-flagellate species are mixotrophic and may feed phagotro-phically on ciliates and cryptophytes (Hansen 1991;Carvalho et al. 2008), cyst-forming, or primarily photo-synthetic species that supplement their nutrition by preyingon other cells (Stoecker 1999), which may explain the lackof effect of voyage history on abundance or species richnessof dinoflagellates but not for diatoms. However, repeatedballast uptake events coupled with incomplete BWE canrefresh water and entrain new species in tanks, which mayallow continual accretion of species in tanks. As transoce-anic exchanged ships often load ballast in ports or coastalareas, and then perform BWE in mid-ocean, accumulationof taxa in tanks may keep species’ abundances intransoceanic exchanged ships similar to that of coastalenvironments.

Invertebrate and dinoflagellate species richness weresignificantly higher in ships operating in the Atlantic regionand subjected to BWE than in coastal, not-exchangedships, indicating that new species were likely added to tankswhen BWE took place. Hence, BWE of vessels mayactually increase the risk of invasion by invertebrates anddinoflagellates to coastal areas, although this may beconfounded by ship travel patterns and voyage durationamong other factors. Of primary concern is the spread ofspecies that cause ecological, economical, or health effects(Ruiz et al. 2000; Olden et al. 2004; Colautti et al. 2006a), inparticular harmful and toxic dinoflagellate species thataffect human health, fisheries, and aquaculture (Hallegraeff1998; McCollin et al. 2007; Anderson et al. 2012). Oceanictaxa may not bloom in neritic waters, but as the number ofcoastal dinoflagellate species transported in coastal shipsaccumulates and increases, so too does the likelihood thatsome of those species will be harmful or toxic. The globalfrequency, intensity, and distribution of paralytic shellfishpoisoning has increased in recent years (Hallegraeff 1998;Anderson et al. 2012). The biology of many dinoflagellatesincludes asexual propagation and capacity to formdormant resting stages highly resistant to harsh environ-ments (Lee 2008), factors that when combined with highabundance when transported in ships’ ballast may enhancetheir invasion success relative to other taxonomic groups.Such a flexible life-history attribute can amplify theopportunity for successful establishment in new areas,allowing founding populations to persist under suboptimalconditions and flourish once environmental conditionsimprove. These taxa are likely excellent candidates forsuccessful invasion.

Global shipping is a potent agent of species introduction(NRC 2008; Hulme 2009). Our study indicates thatparticular taxonomic groups with life-history traits suchas dormant resting stages and high physiological plasticitymay be favored during the transport stage. While we arenot aware of any study exploring changes of abundance orspecies richness in vectors other than shipping, a previous

study did report a nonrandom distribution of introducedspecies among insect families due to the type of transportvector (Vazquez and Simberloff 2001). Nevertheless, weexpect that abundance and species richness in otherunintentional vectors, such as species contained in soil ofhorticultural plants (Mack 2004) and on domestic animalsand livestock (Vazquez and Simberloff 2001), might alsochange owing to length of transport, or with theapplication of quarantine or interception measures.

In conclusion, our comprehensive study found that BWEwas ineffective for reducing abundance or species richness ofnearly all taxa surveyed. However, long voyages resulted inlower abundance of all taxa except dinoflagellates, and lowerspecies richness of diatoms. Introduction risk varied by taxaand across regions despite national ballast water manage-ment regulations. The abundance of discharged organisms isan important element of invasion success (Simberloff 2009);thus, substantial reductions in population abundance of taxashould correspond with reduced invasion risk. However, theelevated species richness of dinoflagellates in ships thatexchanged ballast water suggests an enhanced invasionpotential relative to coastal vessels that did not conductballast exchange. Current ballast water regulations (i.e.,BWE to the D-1 standard) appear ineffective for reducingabundance of many taxa (dinoflagellates, diatoms, bacteria,and VLPs), with no clear effect on invertebrates. Thus,transition to the D-2 ballast water performance standardsmay more uniformly reduce species abundances andassociated invasion risks. As species responses to ballasttank environments are variable and affected by voyagecharacteristics, it is important to consider multiple taxo-nomic groups when evaluating the efficacy of ballast waterexchange or treatment.

AcknowledgmentsWe thank participating shipping companies, crews, port

authorities, and the Shipping Federation of Canada for facilitat-ing access to ships, and G. Klein for diatom identification andenumeration. We acknowledge the dedication of our samplingteams. Great thanks to reviewers for helpful comments. Thisresearch was supported by the Natural Sciences and EngineeringResearch Council’s (NSERC) Canadian Aquatic Invasive SpeciesNetwork (CAISN), Transport Canada, Fisheries and OceansCanada, and NSERC Discovery grants to S.A.B., I.K., R.B.F.,and H.J.M., and a NSERC Discovery Accelerator Supplement toH.J.M.

References

ANDERSON, D. M., A. D. CEMBELLA, AND G. M. HALLEGRAEFF.2012. Progress in understanding harmful algal blooms:Paradigm shifts and new technologies for research, monitor-ing, and management. Annu. Rev. Mar. Sci. 4: 143–176,doi:10.1146/annurev-marine-120308-081121

BAILEY, S. A., M. DENEAU, G. MATTHEW, J. LAURENT, C. J. WILEY,B. LEUNG, AND H. J. MACISAAC. 2011. Evaluating efficacy ofan environmental policy to prevent biological invasions.Environ. Sci. Technol. 45: 2554–2561, doi:10.1021/es102655j

CARLTON, J. T. 2009. Deep invasion ecology and the assembly ofcommunities in historical time, p. 13–56. In G. Rilov and J. A.Crooks [eds.], Biological invasions in marine ecosystems.Springer-Verlag.

1370 Briski et al.

CARVALHO, W. F., S. MINNHAGEN, AND E. GRANELI. 2008.Dinophysis norvegica (Dinophyceae), more a predator than aproducer? Harmful Algae 7: 174–183, doi:10.1016/j.hal.2007.07.002

CASAS-MONROY, O. 2012. Introduction des dinoflagelles non-indigenes dans les ecosystemes aquatiques canadiens via lesreservoirs de ballast de navires. Ph.D. thesis. Univ. of Quebecat Rimouski. [Ballast water and sediment mediated introduc-tions of non-indigenous species of dinoflagellates for Cana-dian aquatic ecosystems.]

CLOERN, J. E. 1979. Phytoplankton ecology of the San FranciscoBay system: The status of our current understanding, p.247–264. In T. J. Conomos [ed.], San Francisco Bay: Theurbanized estuary. Pacific Division, American Association forthe Advancement of Science.

COLAUTTI, R. I., S. A. BAILEY, C. D. A. VAN OVERDIJK, K.AMUNDSEN, AND H. J. MACISAAC. 2006a. Characterised andprojected costs of nonindigenous species in Canada. Biol.Invas. 8: 45–59, doi:10.1007/s10530-005-0236-y

———, I. A. GRIGOROVICH, AND H. J. MACISAAC. 2006b.Propagule pressure: A null model for biological invasions.Biol. Invas. 8: 1023–1037, doi:10.1007/s10530-005-3735-y

CULLEY, A. I., AND N. A. WELSCHMEYER. 2002. The abundance,distribution, and correlation of viruses, phytoplankton, andprokaryotes along the Pacific Ocean transect. Limnol.Oceanogr. 47: 1508–1513, doi:10.4319/lo.2002.47.5.1508

DIBACCO, C., D. B. HUMPHREY, L. E. NASMITH, AND C. D.LEVINGS. 2012. Ballast water transport of non-indigenouszooplankton to Canadian ports. ICES J. Mar. Sci. 69:483–491, doi:10.1093/icesjms/fsr133

FERNANDEZ DE PUELLES, M. L., D. GRAS, AND S. HERNANDEZ-LEON. 2003. Annual cycle of zooplankton biomass, abun-dance and species composition in the neritic area of theBalearic Sea, Western Mediterranean. Mar. Ecol. 24:123–139, doi:10.1046/j.1439-0485.2003.03816.x

FRITZ, L., AND R. E. TRIEMER. 1985. A rapid, simple techniqueutilizing calcofluor white M2R for the visualization ofdinoflagellate thecal plates. J. Phycol. 21: 662–664,doi:10.1111/j.0022-3646.1985.00662.x

GOVERNMENT OF CANADA. 2006. Ballast water control andmanagement regulations [Internet]. Canada Gazette 140(l3).Ottawa, Canada: Government of Canada [accessed 2011December 16]. Available from http://gazette.gc.ca/archives/p2/2006/2006-06-28/html/sor-dors129-eng.html

GRAY, G. K., T. H. JOHENGEN, D. F. REID, AND H. J. MACISAAC.2007. Efficacy of open-ocean ballast water exchange as ameans of preventing invertebrate invasions between freshwaterports. Limnol. Oceanogr. 52: 2386–2397, doi:10.4319/lo.2007.52.6.2386

GUILLARD, R. R. L. 1975. Culture of phytoplankton for feedingmarine invertebrates, p. 26–60. In W. L. Smith and M. H.Chanley [eds.], Culture of marine invertebrate animals.Plenum Press.

HALLEGRAEFF, G. M. 1998. Transport of toxic dinoflagellates viaship’s ballast water: Bioeconomic risk assessment and efficacyof possible ballast water management strategies. Mar. Ecol.Prog. Ser. 168: 297–309, doi:10.3354/meps168297

HANSEN, P. J. 1991. Quantitative importance and trophic role ofheterotrophic dinoflagellates in a coastal pelagial food web.Mar. Ecol. Prog. Ser. 73: 253–261, doi:10.3354/meps073253

HEAD, E. J. H., AND P. PEPIN. 2010. Monitoring changes inphytoplankton abundance and composition in the NorthwestAtlantic: A comparison of results obtained by continuousplankton recorder sampling and colour satellite imagery. J.Plankton Res. 32: 1649–1660, doi:10.1093/plankt/fbq120

HULME, P. E. 2009. Trade, transport and trouble: managinginvasive species pathways in an era of globalization. J. Appl.Ecol. 46: 10–18, doi:10.1111/j.1365-2664.2008.01600.x

HWANG, J. S., AND OTHERS. 2010. Patterns of zooplanktondistribution along the marine, estuarine, and riverine portionsof the Danshuei ecosystem in Northern Taiwan. Zool. Stud.49: 335–352.

INTERNATIONAL MARITIME ORGANIZATION (IMO). 2004. Interna-tional convention for the control and management of ships’ballast water and sediments [Internet]. London, UnitedKingdom: International Maritime Organization [accessed2011 December 16]. Available from http://www.imo.org/about/conventions/listofconventions/pages/international-con-vention-for-the-control-and-management-of-ships’-ballast-water-and-sediments-(bwm).aspx

KACZMARSKA, I., M. M. LEGRESLEY, J. L. MARTIN, AND J. M.EHRMAN. 2005. Diversity of the diatom Pseudo-nitzschiaPeragallo in the Quoddy Region of the Bay of Fundy, Canada.Harmful Algae 4: 1–19, doi:10.1016/j.hal.2003.07.001

KIRCHMAN, D., J. SIGDA, R. KAPUSCINSKI, AND R. MITCHELL. 1982.Statistical analysis of the direct count method for enumeratingbacteria. Appl. Environ. Microb. 44: 376–382.

KLEIN, G., K. MACINTOSH, I. KACZMARSKA, AND J. M. EHRMAN.2010. Diatom survivorship in ballast water during trans-Pacific crossings. Biol. Invas. 12: 1031–1044, doi:10.1007/s10530-009-9520-6

KOPPELMANN, R., AND H. WIEKERT. 1992. Full-depth zooplanktonprofile over the deep bathyal of the NE Atlantic. Mar. Ecol.Prog. Ser. 86: 263–272, doi:10.3354/meps086263

LEE, R. E. 2008. Phycology, 4th ed. Cambridge Univ. Press.LOCKWOOD, J. L., P. CASSEY, AND T. BLACKBURN. 2009. The more

you introduce the more you get: The role of colonizationpressure and propagule pressure in invasion ecology. Divers.Distrib. 15: 904–910, doi:10.1111/j.1472-4642.2009.00594.x

MACK, R. N. 2004. Global plant dispersal, naturalization, andinvasions: Pathways, models and circumstances, p. 3–30. InG. M. Ruiz and J. T. Carlton [eds.], Invasive species: Vectorsand management strategies. Island Press.

MCCOLLIN, T., A. M. SHANKS, AND J. DUNN. 2007. The efficiencyof regional ballast water exchange: Changes in phytoplanktonabundance and diversity. Harmful Algae 6: 531–546,doi:10.1016/j.hal.2006.04.015

NATIONAL RESEARCH COUNCIL (NRC). 2008. Great Lakes ship-ping, trade, and aquatic invasive species [Internet]. SpecialReport 291. Washington, D.C.: National Research Council[accessed 2011 December 16]. Available from http://online-pubs.trb.org/Onlinepubs/sr/sr291

OLDEN, J. D., N. L. POFF, M. R. DOUGLAS, M. E. DOUGLAS, AND

K. D. FAUSCH. 2004. Ecological and evolutionary conse-quences of biotic homogenization. Trends Ecol. Evol. 19:18–24, doi:10.1016/j.tree.2003.09.010

PATEL, A., R. T. NOBLE, J. A. STEELE, M. S. SCHWALBACH, I.HEWSON, AND J. A. FUHRMAN. 2007. Virus and prokaryoteenumeration from planktonic aquatic environments byepifluorescence microscopy with SYBR Green I. Nat. Protoc.2: 269–276, doi:10.1038/nprot.2007.6

ROY, S., M. PARENTEAU, O. CASAS-MONROY, AND A. ROCHON.2012. Coastal ship traffic: A significant introductionvector for potentially harmful dinoflagellates in EasternCanada. Can. J. Fish. Aquat. Sci. 69: 627–644, doi:10.1139/f2012-008

RUIZ, G. M., T. K. RAWLINGS, F. C. DOBBS, L. A. DRAKE, T.MULLADY, A. HUQ, AND R. R. COLWELL. 2000. Global spreadof microorganisms by ships. Nature 408: 49–50, doi:10.1038/35040695

Transport stage of biological invasions 1371

SEIDEN, J. M., C. WAY, AND R. B. RIVKIN. 2011. Bacterialdynamics in ballast water during trans-oceanic voyages ofbulk carriers: Environmental controls. Mar. Ecol. Prog. Ser.436: 145–159, doi:10.3354/meps09231

SIMARD, N., S. PLOURDE, M. GILBERT, AND S. GOLLASCH. 2011. Netefficacy of open ocean ballast water exchange on planktoncommunities. J. Plankton Res. 33: 1378–1395, doi:10.1093/plankt/fbr038

SIMBERLOFF, D. 2009. The role of propagule pressure in biologicalinvasions. Annu. Rev. Ecol. Evol. Syst. 40: 81–102,doi:10.1146/annurev.ecolsys.110308.120304

STOECKER, D. K. 1999. Mixotrophy among dinoflagellates. J.Eukaryot. Microbiol. 46: 397–401, doi:10.1111/j.1550-7408.1999.tb04619.x

SUN, B., R. MOULAND, W. CANDICE, AND R. B. RIVKIN. 2010.Redistribution of heterotrophic prokaryotes through ballastwater: A case study from the west coast of Canada. Aquat.Invas. 5: 5–11, doi:10.3391/ai.2010.5.1.2

UNITED STATES COAST GUARD (USCG). 1993. Ballast watermanagement for vessels entering the Great Lakes. Code ofFederal Regulations 33-CFR Part 151.1510. United StatesCoast Guard.

VAZQUEZ, D. P., AND D. SIMBERLOFF. 2001. Taxonomicselectivity in surviving introduced insects in the UnitedStates, p. 103–124. In J. L. Lockwood and M. L. McKinney[eds.], Biotic homogenization. Kluwer Academic PlenumPress.

Associate editor: Alexander D. Huryn

Received: 24 August 2012Accepted: 26 March 2013

Amended: 02 April 2013

1372 Briski et al.