Predicting invasiveness of freshwater fishes imported into North America: regional differences in models and outcomes

Abstract

Biological invasions driven by international trade heighten the urgency for development of invasion risk models, as the traits and parameters that consistently predict successful invasion remain unresolved. For four regions of North America that include parts of the United States and Canada (Sacramento-San Joaquin River Basins, Lower Colorado River Basin, Great Lakes Region, Mid-Atlantic Region), we construct and compare classification tree models to reveal robust predictors for the establishment and ecological impact stages of freshwater fish invasion. We subsequently apply the models to identify invasive fish species in trade and conduct pathway analyses to determine which trades (aquarium, biological supply, live bait, live food, water garden) and source continents pose the greatest risk to each region. Model results differed by invasion stage and region. Across regions, establishment models shared climate-related predictors including climate match and temperature tolerance. Three of the four impact models contained prior establishment success. The greatest number of species (548) were predicted to establish in the Sacramento-San Joaquin while the fewest (5) were predicted to establish in the Mid-Atlantic. Forty species were predicted to establish in multiple regions, five of which were also predicted to have high impact. The aquarium trade and Asia supplied the most species predicted to establish. Taken together, the results highlight region-specific models, indicating no universal model predicts invasion. Climate-related and prior establishment variables were most useful to risk assessments. The regional models, and identified high-risk pathways and potential invaders, could be applied to prevent future fish invasions in North America.

Full text

Vol.: (0123456789)
Biol Invasions (2025) 27:107
https://doi.org/10.1007/s10530-025-03560-1
ORIGINAL PAPER
Predicting invasiveness offreshwater fishes imported
intoNorth America: regional differences inmodels
andoutcomes
JenniferG.Howeth · SarahA.Amjad· CrystaA.Gantz · NicholasE.Mandrak ·
PaulL.Angermeier · MichaelP.Marchetti · JulianD.Olden · DavidM.Lodge
Received: 20 October 2024 / Accepted: 21 February 2025 / Published online: 22 March 2025
This is a U.S. Government work and not under copyright protection in the US; foreign copyright protection may apply 2025
identify invasive fish species in trade and conduct
pathway analyses to determine which trades (aquar-
ium, biological supply, live bait, live food, water gar-
den) and source continents pose the greatest risk to
each region. Model results differed by invasion stage
and region. Across regions, establishment models
shared climate-related predictors including climate
match and temperature tolerance. Three of the four
impact models contained prior establishment suc-
cess. The greatest number of species (548) were pre-
dicted to establish in the Sacramento-San Joaquin
while the fewest (5) were predicted to establish in the
Mid-Atlantic. Forty species were predicted to estab-
lish in multiple regions, five of which were also pre-
dicted to have high impact. The aquarium trade and
Abstract Biological invasions driven by interna-
tional trade heighten the urgency for development
of invasion risk models, as the traits and parameters
that consistently predict successful invasion remain
unresolved. For four regions of North America that
include parts of the United States and Canada (Sac-
ramento-San Joaquin River Basins, Lower Colo-
rado River Basin, Great Lakes Region, Mid-Atlantic
Region), we construct and compare classification
tree models to reveal robust predictors for the estab-
lishment and ecological impact stages of freshwater
fish invasion. We subsequently apply the models to
Supplementary Information The online version
contains supplementary material available at https:// doi.
org/ 10. 1007/ s10530- 025- 03560-1.
J.G.Howeth(*)· S.A.Amjad
Department ofBiological Sciences, University
ofAlabama, Tuscaloosa, AL35487, USA
e-mail: jennifer_howeth@fws.gov
Present Address:
J.G.Howeth
United States Fish andWildlife Service, San Marcos
Aquatic Resources Center, 500 East McCarty Lane,
SanMarcos, TX78666, USA
C.A.Gantz
Department ofBiological Sciences, University ofNotre
Dame, NotreDame, IN46556, USA
N.E.Mandrak
Department ofBiological Sciences, University ofToronto
Scarborough, Toronto, ONM1C1A4, Canada
P.L.Angermeier
U. S. Geological Survey, Virginia Cooperative Fish
andWildlife Research Unit, Virginia Polytechnic Institute
andState University, Blacksburg, VA24061, USA
M.P.Marchetti
Department ofBiology, St. Mary’s College ofCalifornia,
Moraga, CA94556, USA
J.D.Olden
School ofAquatic andFishery Sciences, University
ofWashington, Seattle, WA98195, USA
D.M.Lodge
Cornell Atkinson Center forSustainability andDepartment
ofEcology andEvolutionary Biology, Cornell University,
Ithaca, NY14853, USA
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J.G.Howeth et al.
107 Page 2 of 26
Vol:. (1234567890)
Asia supplied the most species predicted to estab-
lish. Taken together, the results highlight region-spe-
cific models, indicating no universal model predicts
invasion. Climate-related and prior establishment
variables were most useful to risk assessments. The
regional models, and identified high-risk pathways
and potential invaders, could be applied to prevent
future fish invasions in North America.
Keywords Aquarium· Classification tree· Climate
match· Ecological impact· Establishment success·
Live food· Organisms in trade· Risk assessment
Introduction
Invasive species remain a central driver of global
biodiversity loss (Mollot et al. 2017; Duenas et al.
2021; IPBES 2023). Biological invasions contribute
to biotic homogenization and can compromise eco-
system function and human economies (Baiser etal.
2012; Diagne et al. 2021). Mirroring the growth
in human population size over the last century, an
increasing number of species are transported outside
of their native geographic range through accidental
and intentional anthropogenically-mediated pathways
(Hulme 2009; Gerland et al. 2014; Seebens et al.
2017). A subset of the introduced species become
invasive by successfully establishing reproducing
populations and having a measurable ecological and/
or economic impact in the non-native range (Black-
burn etal. 2011). The ability to predict which species
may become invasive once introduced outside of their
native range serves as the foundation of invasive spe-
cies risk assessment (Lodge etal. 2016; Kumschick
etal. 2024).
Application of invasive species risk assessments
has increased in frequency globally paralleling the
progress in development of risk analyses and the rec-
ognition of heightened invasion risks associated with
international trade (Kumschick et al. 2020; Meyers
etal. 2020). Risk assessment methodologies include
both qualitative and quantitative approaches to pre-
diction of invasive potential, with each framework
carrying its own strengths and limitations (Leung
etal. 2012; Lodge etal. 2016). Categorical scoring
of risk based on prior invasiveness and invasion stage
characteristics of species are included in qualitative
questionnaire-based and horizon scan assessments
(Roy etal. 2014; Vilizzi etal. 2021). In quantitative
approaches, invasion risk parameters and ecological
and life-history traits are often incorporated when
modeling stages of invasion (Kolar and Lodge 2002;
Keller etal. 2011; Strubbe etal. 2023). These models,
typically trained on independent data derived from a
target geographic region, require more resources to
develop but may be more accurate in prediction for a
region than the more generalized qualitative scoring
methods (Ibáñez etal. 2014).
Invasion stage-specific models can be powerful at
predicting invasion success, as key factors often dif-
fer by invasion stage (Capellini etal. 2015; Howeth
et al. 2016; Catford et al. 2019). Important predic-
tive factors are also likely to differ among geographic
regions, but this has rarely been tested using the same
modeling methods across regions (Essl et al. 2015;
Pyšek et al. 2020a). Lack of such studies primarily
stems from limited availability of standardized data
among regions and the large scope of the associated
data and analyses. Yet, findings from such a com-
parative approach may help uncover unified predic-
tors of invasion risk or regional context-dependency,
as uncertainty remains about traits and parameters
linked to successful invasion (Pyšek et al. 2020a).
Application of the developed models could also iden-
tify shared or unique high-risk species and trade path-
ways among regions, which could make prevention
efforts more effective and efficient.
International trade provides source pools of spe-
cies that can be introduced to non-native environ-
ments via intentional and unintentional means from
commodity-driven trade pathways originating from
multiple geographic regions and continents (Har-
rower et al. 2018; Lockwood et al. 2019; Hulme
2021). Species from different source regions likely
vary with respect to their environmental tolerance and
traits, thereby posing differential risks to the receiv-
ing regions (Hubbard etal. 2023). As different trade
pathways select taxa for specific uses and environ-
ments, they may also select for unique trait combina-
tions and correspondingly pose different overall risks
for the recipient regions. Coupling pathway identifi-
cation with invasive species risk assessment can iden-
tify high-risk pathways for target geographic regions
by aggregating species assessment results by spe-
cific pathways (Howeth etal. 2016; Lieurance etal.
2023). This approach can aid decision-makers in dis-
tinguishing which international trade pathways pose
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Predicting invasiveness offreshwater fishes imported intoNorth America: regional differences…
Page 3 of 26 107
Vol.: (0123456789)
the greatest threat of introducing new invaders to a
region, and thus deserve greater management or pol-
icy attention (Pyšek etal. 2011; Lodge etal. 2016).
Determining the relative risk of trade pathways
is crucial for freshwater fishes, which comprise the
majority of live animals imported into the United
States by international trade (Smith etal. 2009; Liv-
engood et al. 2014). Human-aided introduction of
fishes to non-native environments has contributed
to numerous ecological impacts (Cucherousset and
Olden 2011), including decreased species diversity
at regional scales (Campbell and Mandrak 2019; Su
et al. 2021). Multiple international live trade path-
ways supply fishes that have the potential to become
invasive if released into the wild and include the
aquarium, bait, biological supply, live food, and
water garden trades (Keller and Lodge 2007; Bernery
et al. 2022). These pathways may pose differential
threats based upon species characteristics selected
by the pathway and as predicted by quantitative risk
assessment (Bernery et al. 2024). For example, in
the Laurentian Great Lakes region, establishment of
an introduced fish was best predicted by the climate
match between the species’ native range and that of
the Great Lakes region (Howeth etal. 2016). After
evaluating the climatic match of species in trade, the
water garden trade supplied the greatest proportion
of fishes predicted to successfully establish in the
region. Applying this pathway-based risk assessment
approach to multiple regions could identify common-
alities or differences in invasion risk models and the
relative threat of pathways (Lodge etal. 2006; Pyšek
etal. 2020b).
Freshwater fishes are especially suitable for mode-
ling invasion stages due to the availability of data and
knowledge of predictive traits, invasion history, and
geographic range boundaries (Garcia-Berthou 2007;
Frimpong and Angermeier 2009). Variables that fore-
cast non-native fish establishment differ across stud-
ies, but some are consistently correlated with estab-
lishment success (Garcia-Berthou 2007; Bernery
et al. 2022). For example, the climatic similarity
between the species’ native range and recipient region
often predicts establishment (Bomford et al. 2010;
Howeth etal. 2016), as can a history of invasion suc-
cess outside the native range (Kolar and Lodge 2002;
Marchetti etal. 2004a,b; Ribeiro etal. 2008). Traits
predicting establishment include body size (Mar-
chetti etal. 2004a; Ribeiro etal. 2008; Lawson and
Hill 2022), physiological tolerance (Kolar and Lodge
2002; Marchetti et al. 2004a,b), spawning require-
ments (Mandrak 1989; Olden etal. 2006), and trophic
status (Marchetti etal. 2004b; Liu etal. 2017). In con-
trast, traits associated with the ecological impacts of
invasion exhibit greater variation across studies (Gar-
cia-Berthou 2007; Bernery etal. 2022; Britton 2023).
This is due, in part, to challenges in quantifying the
direct and indirect impact of a species in a food web
and whole ecosystem (Ricciardi etal. 2013). There
remains a significant need to better standardize eco-
logical impact measures of freshwater fishes in risk
assessment models to identify robust predictors of
impact (Bernery etal. 2024).
To test for consistent predictors of invasion suc-
cess and contrast stage-specific invasion risk models
among geographic regions, we construct establish-
ment and ecological impact models for freshwater
fishes in four regions of North America using stand-
ardized methods with species traits and invasion risk
parameters. Subsequently, the models are used to
screen non-native fishes in international trade and
identify species that pose a threat of becoming inva-
sive within each region. Species assessment results
are then aggregated to conduct regional pathway risk
analyses to evaluate the relative threats of the aquar-
ium, biological supply, live bait, live food, and water
garden trades and the source continent(s) represent-
ing species native ranges. The results aim to identify
commonalities or differences in traits and param-
eters of invasion success and thus the transferabil-
ity of invasion risk models, the high-risk trades and
source continents, and the high-risk species to North
America. Together, the findings can inform contem-
porary approaches to invasive species risk assessment
in response to the movement of species from interna-
tional trade.
Methods
Establishment and ecological impact models were
developed and utilized to screen potential invad-
ers for four geographic regions located in the United
States (US) and Canada: the Sacramento-San Joaquin
River Basins, Lower Colorado River Basin, Lau-
rentian Great Lakes Region, and the Mid-Atlantic
Region (Fig.1). The regions differ in spatial extent,
native and non-native species, environment, and
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J.G.Howeth et al.
107 Page 4 of 26
Vol:. (1234567890)
anthropogenic land and water use (Maxwell et al.
1995; Page and Burr 2011; Omernik and Griffith
2014; Jones etal. 2022; USGS 2024; TableS1).
A standardized workflow for developing stage-
based invasion models and conducting pathway risk
analyses was applied for all regions. The process for
developing the regional invasion models involved: (1)
compiling data sets on successfully established non-
native fishes and those that were introduced but failed
to establish reproducing populations to train the clas-
sification tree models of establishment; (2) assessing
ecological impact of established fishes using expert
opinion to train classification tree models of impact;
(3) acquiring species trait and invasion risk parameter
data on established and failed fishes as predictor vari-
ables for the establishment and impact models; and
(4) constructing establishment and impact classifica-
tion tree models using collected data. Subsequently,
the approach to predicting high-risk non-native fishes,
trades, and native source continents included: (1) col-
lecting species trait and invasion risk parameter data
required by the developed classification tree mod-
els for freshwater fishes in international trade; (2)
screening species through the models to determine
establishment risk, and if predicted to establish,
impact risk; and (3) aggregating invasion outcome
results by trade and native range source continent
pathways as regional risk analyses.
Model development
Compiling data onestablishment andimpact
byregion
Datasets of established non-native fishes, and fish
species that were documented to be introduced but
failed to establish reproducing populations, in each
region were compiled to build and train models of
invasion. In the Sacramento-San Joaquin River Basins
(hereafter, “Sacramento-San Joaquin”), we identified
65 introduced non-native fish species, of which 42
species have established self-sustaining (reproduc-
ing) populations while the remaining species failed
to establish (Dill and Cordone 1997; Moyle 2002;
Marchetti etal. 2006; TableS2). In the Lower Colo-
rado River Basin (hereafter, “Lower Colorado”), 82
non-native fish species were introduced, of which 68
have established (USGS 2024; TableS3). In the Great
Fig. 1 Geographic location and boundaries of the four study
regions in North America: the Sacramento-San Joaquin River
Basins, Lower Colorado River Basin, Laurentian Great Lakes
Region, and the Mid-Atlantic Region. The regions repre-
sent aggregated United States Geological Survey Hydrologic
Unit Code 6-digit watersheds (Jones etal. 2022) reported in
TableS1. Major river systems in the United States are illus-
trated by thin gray lines
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Predicting invasiveness offreshwater fishes imported intoNorth America: regional differences…
Page 5 of 26 107
Vol.: (0123456789)
Lakes Region (hereafter, “Great Lakes”), 65 non-
native fish species were introduced among the five
lakes, of which 37 species have established (Howeth
et al. 2016; Table S4). In the Mid-Atlantic Region
(hereafter, “Mid-Atlantic”), 108 non-native fish spe-
cies were introduced, of which 77 have established
(Jenkins and Burkhead 1994; Page and Burr 2011;
Angermeier and Pinder 2015; USGS 2024; TableS5).
To evaluate the ecological impact of established
species within each region for the impact model train-
ing data sets, we relied on expert elicitation and a
standardized questionnaire detailing the region-spe-
cific established species list (Appendix S1). Quan-
tifying the magnitude of ecological impacts of non-
native fishes proves challenging as literature-based
assessments of impact only exist for some species
and often overlook regional variation (Blackburn
etal. 2014; Bernery etal. 2024). Expert knowledge
can help overcome these limitations (e.g., Roy etal.
2014). Regional questionnaires were distributed via
an emailed executable PDF to experts at universities,
government agencies, and consulting firms with expe-
rience in the region demonstrated by contributions to
the peer-reviewed scientific literature, engagement in
management, and long-term contributions to conser-
vation practices in addition to substantial knowledge
of aquatic invasive species, fisheries, and/or ecosys-
tems. Questionnaires asked experts to assign each
species to one of four categories ranging from no per-
ceived ecological impact to very high perceived eco-
logical impact and to assign either low or high con-
fidence to their answer as a measure of uncertainty
(Howeth etal. 2016). Respondents could decline to
rank a species if they were not familiar with its impact.
They could additionally contribute names of species
not already on the list of established species for the
region. For the Sacramento-San Joaquin, a total of 31
experts were queried, yielding 21 (68%) completed
questionnaires (Table S6). Respondents reported a
cumulative 474years of experience (mean = 23years)
in the Sacramento-San Joaquin. For the Lower Colo-
rado, a total of 33 experts were queried, yielding 17
(52%) completed questionnaires (TableS7). Respond-
ents reported a cumulative 484 years of experience
(mean = 28 years) in the Lower Colorado. For the
Great Lakes, a total of 33 experts in Canada and the
US were queried, yielding 27 (82%) completed ques-
tionnaires (TableS8). Respondents reported a cumu-
lative 557 years of experience (mean = 21 years) in
the Great Lakes. For the Mid-Atlantic, a total of 51
experts were queried, yielding 28 (55%) completed
questionnaires (Table S9). Respondents reported a
cumulative 571years of experience (mean = 20years)
in the Mid-Atlantic.
Impact categories for each non-native species on
the questionnaire ranged from no to low impact with
a corresponding score of 1, to very high impact with a
corresponding score of 4, and reflected the magnitude
of ecological impacts ranging from small effects on
existing biota to large impacts on whole ecosystems
(Appendix S1). We calculated the mean and 95% con-
fidence interval (CI) of the recorded impact scores for
each species and then ranked the species within each
region from lowest to highest impact based upon the
upper CI value to incorporate variation in respondent
answers (Howeth etal. 2016). Because there were no
obvious breaks in the impact rankings across the spe-
cies within each region (i.e., impact ranks increased
approximately linearly from lowest to highest), for
the modeling analysis we compared the upper third
of species with the highest impact to the lower third
of species with the lowest impact within each region
(TableS10-S13). This approach maximized our abil-
ity to detect differences in species traits and invasion
risk between low- and high-impact species.
Species traits andinvasion risk parameters
Species traits and invasion risk parameters were
scored for established and failed fishes in the regional
species lists to use in classification tree analyses to
predict establishment and impact outcome. We iden-
tified and used 17 species traits and invasion risk
parameters (Table S14) that have been previously
associated with the establishment or impact stages
of invasion (reviewed in Garcia-Berthou 2007) and
for which sufficient data typically exist in publicly
available databases. These 17 variables were catego-
rized a priori as follows. Ecological traits included:
(1) diet breadth (sum of diet items, including algae/
phytoplankton, vascular plants, detritus, aquatic/
terrestrial invertebrates and larval fishes, fishes/
crayfishes/crabs/frogs, blood, eggs); (2) macro-
habitat association (lentic, lotic); (3) salinity toler-
ance (narrow, wide); (4) temperature tolerance [cold
(10–17°C), cold/cool (10–26°C), cool (18–26°C),
cool/warm (18- > 26°C), warm (> 26°C)]; and, (5)
trophic guild (herbivore-detritivore, invertivore,
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J.G.Howeth et al.
107 Page 6 of 26
Vol:. (1234567890)
invertivore-piscivore, omnivore, piscivore). Life-his-
tory traits included (6) maximum total body length
(TL); (7) egg diameter; (8) fecundity (maximum
per spawning season, per female); (9) larval length
(length at hatching); (10) longevity (maximum lifes-
pan); (11) maturation size (proportion, length of
female at maturation as a function of maximum total
length); (12) reproductive guild (nonguarders and
open substratum spawners, nonguarders and brood
hiders, guarders and substratum choosers, guard-
ers and nest spawners, substrate indifferent); and,
(13) spawning frequency (batch, serial). Invasion
risk parameters included: (14) climate match [meas-
ured with the on-line program Climatch (Australian
Bureau of Agricultural and Resource Economics and
Sciences 2020) following a United States Fish and
Wildlife Service (USFWS) protocol (Appendix S2)
based upon Bomford etal. (2010)]; (15) phylogenetic
relatedness according to family membership [i.e.,
the degree of their derived characters, ordered from
most ancestral to most derived from Nelson (2006)]
(Howeth etal. 2016); (16) prior establishment success
(number of countries where the species has estab-
lished); and, (17) size of native range [area; based
upon distribution data reported in FishBase (Froese
and Pauly 2021) and measured with the on-line pro-
gram GeoCAT (Bachman etal. 2011)]. Data for vari-
ables were gathered from on-line databases, including
FishBase and FishTraits (Frimpong and Angermeier
2009), published literature (e.g., Kolar and Lodge
2002; Mims etal. 2010) and, when primary sources
were exhausted, hobbyist gray literature (Tropical
Fish Hobbyist 1979; 1980) and vendor websites (via
Internet searches using vendors listed in up to the top
15 hits for each species queried).
Classification tree models ofestablishment
andimpact
We tested which subsets of the 17 variables were most
strongly associated with the establishment and impact
stages of invasion in each region by using a classifi-
cation tree analysis in the program Salford Predictive
Modeler (v. 8, Minitab, State College, Pennsylva-
nia). Classification tree analysis is a machine learn-
ing method that employs binary recursive partition-
ing to model categorical response variables (Breiman
et al. 1984). The tree is constructed by repeatedly
splitting the data into two groups (nodes) defined by
a threshold value (continuous data) or category (cat-
egorical data) of a single independent variable that
maximizes homogeneity of outcome (e.g., established
versus not established) within the two groups created
by the split (De’ath and Fabricius 2000). Classifica-
tion trees are particularly well suited for the analy-
sis of complex ecological datasets that may support
nonlinear relationships that interact hierarchically,
and where data may be missing for some independent
variables (Olden etal. 2008).
We developed classification tree models for each
stage of invasion with binary dependent outcomes:
establish or fail for the establishment stage, and low
or high for the ecological impact stage (Howeth etal.
2016). Node-splitting criteria were based on the Gini
homogeneity index, with no constraints on the mini-
mum node sample size (De’ath and Fabricius 2000).
We chose the optimal classification tree by perform-
ing 10-fold cross-validation and selecting the smallest
tree within one standard error of relative cost (mis-
classification rate) that maximized predictive ability
based upon the correct classification rate.
Predicting invasive species by region
Datasets from the US and Canada that document live
fishes imported through legal, permit-based processes
were compiled to identify high-risk fish invaders in
international trade (Howeth etal. 2016) and to rep-
resent commonly traded species in North America
(Livengood etal. 2014; Lieurance etal. 2023). From
the USFWS Law Enforcement Management Informa-
tion System, a database that tracks import and export
data for all US ports authorized to process live animal
shipments, we used a list of live fish imports from
1 October 2004 to 30 November 2005 (Romagosa
etal. 2009). We also used a list of Canadian live fish
imports from 1 October 2004 to 30 September 2005
from the Canadian Border Service Agency’s Facility
for Information Retrieval Management, which serves
all ports in Canada (Mandrak etal. 2014). The online
databases FishBase and WoRMS (WoRMS Editorial
Board 2025) were used to confirm or update the cur-
rent scientific name. In cases where only the common
name was reported in the trade lists, it was used to
identify the scientific name.
For the importation datasets, each species was
assigned to one or more of five trade pathways, aquar-
ium, biological supply, live bait, live food (including
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Predicting invasiveness offreshwater fishes imported intoNorth America: regional differences…
Page 7 of 26 107
Vol.: (0123456789)
aquaculture), and water garden, based on informa-
tion in the documents from the importer’s website.
Our analysis excluded fishes imported for zoos, pub-
lic aquaria, or academic research, fish species listed
on the Convention on International Trade of Endan-
gered Species and therefore not regularly traded, and
marine fishes. Standardized Internet searches for spe-
cies in the biological supply, live bait, and water gar-
den trades were performed because they were poorly
represented in the federal importation species data-
sets (Howeth etal. 2016). The search strings “trade
pathway name supply” and “trade pathway name sup-
ply Canada” were used for each of the three trades.
Vendors listed in up to the top 15 hits were used to
obtain identities of species available from major sup-
pliers. Four non-native species were documented for
the live bait trade from Internet searches. Additional
bait species were obtained from published surveys
of bait shops (Litvak and Mandrak 1999; Drake and
Mandrak 2014).
The final list of species in all five trade pathways
was customized for each region by excluding fishes
already in the established/failed data set for the region
(Tables S2–S5) or that were native to the region. For
these species in trade, invasion risk parameters and
trait data required by the developed classification tree
models were collected using the methods described
above. For larval length, surrogate species (usually
congeners) were used to estimate body length for 16
species for which data were not available.
We used the establishment classification tree
model, and collected species-specific traits and
parameters, to predict which species in trade would
successfully establish in a region. For those species
predicted to establish from each regional model, we
estimated ecological impact using the region-specific
impact classification tree model and relevant traits
and parameters of screened species.
Pathway risk analysis
Trades posing the greatest threat to each region
were identified by: (1) the proportion of species pre-
dicted to establish within each trade pathway from
the classification tree model results as a function of
the total number of species per pathway; and, (2)
the absolute number of species predicted to estab-
lish and have high impact within each trade pathway
from model screening. To determine the geographic
source of species predicted to establish from the
regional classification tree models, the continent(s)
representing the native range was assigned for each
species in trade from FishBase. The number of
established species contributed by each continent to
each region was determined to identify geographic
sources that may be supplying successful invaders.
To generate a list of species at high risk of estab-
lishing in much of North America, we compiled a
list of species predicted to successfully establish in
two or more regions.
Results
Establishment models by region
Classification tree models predicting establishment
differ in complexity and topology by geographic
region but share climate-related invasion risk
parameters including climate match and tempera-
ture tolerance, albeit in different orders of influence
(Fig. 2, Table 1). Cross-validation analyses indi-
cate that models tend to classify established species
more accurately than failed species (60–82% versus
43–75% correct, respectively; Table2). The Sacra-
mento-San Joaquin establishment classification tree
is the most complex of the four regional models. It
identifies a root node of maturation size and four
other variables including temperature tolerance,
maximum total body length, trophic guild, and phy-
logeny that predict establishment success (Fig.2a,
Table2). The Lower Colorado establishment model
identifies temperature tolerance as the root node
splitter and prior invasion success and size of the
native geographic range as subsequent distinguish-
ing variables (Fig. 2b, Table2). The Great Lakes
classification tree for establishment, the simplest
of the four regional models, indicates that a cli-
matic match > 71.7% between the native range of
the species and the Great Lakes region determines
establishment success (Fig.2c, Table 2). The Mid-
Atlantic establishment model specifies that a cli-
matic match > 79% between the species native range
and the Mid-Atlantic region yields establishment
success, whereas species with a match ≤ 79% are
further differentiated by maturation size (Fig. 2d,
Table2).
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J.G.Howeth et al.
107 Page 8 of 26
Vol:. (1234567890)
Impact models by region
Classification trees produced for ecological impact
also differ in composition and topology among
regions, but three of the four regions contain prior
establishment success as the root node (Table 1,
Fig.3). Cross-validation analyses indicate that mod-
els classify high- and low-impact species with similar
accuracy (50–83% versus 48–88% correct, respec-
tively; Table2). The Sacramento-San Joaquin classifi-
cation tree for the ecological impact stage designates
prior establishment success as the root node and addi-
tionally differentiates impact based upon diet breadth,
larval length, and climate match (Fig. 3a, Table2).
The Lower Colorado impact tree also identifies prior
establishment success as the root node and incorpo-
rates trophic guild, larval length, and temperature
tolerance into the model (Fig.3b, Table2). The Great
Lakes impact classification tree specifies trophic
guild as the root node and additionally incorporates
fecundity (Fig.3c, Table2). The Mid-Atlantic impact
classification tree, the simplest and most robust of
the four regional models, highlights prior establish-
ment success as the sole predictor of impact (Fig.3d,
Table2).
Predicting invasive species by region
Of the species supplied by international trade with the
US and Canada, 797 species pose a potential invasion
risk to Sacramento-San Joaquin (TableS15), 789 spe-
cies to the Lower Colorado (TableS16), 787 species
to the Great Lakes (TableS17), and 786 species to the
Mid-Atlantic (TableS18). The number of freshwater
Fig. 2 Classification trees predicting establishment of non-
native fish species in the (a) Sacramento-San Joaquin, (b)
Lower Colorado, (c) Great Lakes, and (d) Mid-Atlantic
regions. The number of fish species from the model training
data set located in each response category is indicated at every
terminal node
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Predicting invasiveness offreshwater fishes imported intoNorth America: regional differences…
Page 9 of 26 107
Vol.: (0123456789)
fishes predicted by the classification tree models to
establish in each region differed greatly (TableS19).
Sacramento-San Joaquin had the largest number of
species predicted to successfully establish, with 514
species predicted to have a low ecological impact and
34 species predicted to have a high impact. In the
Lower Colorado, 73 species were predicted to estab-
lish, with 40 species having a low impact and 33 spe-
cies having a high impact. The Great Lakes and the
Mid-Atlantic regions had the fewest species predicted
to successfully establish with seven and five species,
respectively. In the Great Lakes, three species were
predicted to have a low ecological impact while four
species were predicted to have a high impact. In the
Mid-Atlantic, three species were predicted to have a
low impact and two species were predicted to have a
high impact.
Pathway risk analysis
Overall, contributions to establishment success from
the different trade pathways vary by geographic
region (Fig.4a,c,e,g), but the aquarium trade poses
the greatest risk to all regions in terms of the abso-
lute number of fish species predicted to establish
and have a high ecological impact (Fig.4b,d,f,h). In
proportion to the total number of species per trade
pathway, the aquarium trade supports the highest
proportion of species likely to establish in Sacra-
mento-San Joaquin followed by equal contributions
from the live bait and water garden trades (Fig.4a).
In the Lower Colorado, the biological supply and
live food trades have the greatest proportion of
species predicted to establish from trade pathways
(Fig. 4c). In the Great Lakes, the highest propor-
tion of species is predicted to establish from the
water garden trade followed by the live food trade
(Fig.4e). The live food trade supplies the greatest
proportion of species likely to establish in the Mid-
Atlantic, distantly followed by the aquarium trade
(Fig.4g).
The source continents supplying native species
predicted to establish also vary by region (Fig. 5).
In Sacramento-San Joaquin, species native to South
America, Africa, and Asia have the greatest establish-
ment potential (Fig.5a). In contrast, Asia and South
America source the greatest number of species pre-
dicted to establish in the Lower Colorado (Fig.5b).
In the Great Lakes, Europe and Asia supply the most
species predicted to establish with no species pre-
dicted to establish from Africa, Oceania, or South
America (Fig.5c). Most species likely to establish in
the Mid-Atlantic are native to Asia, with none pre-
dicted to establish from Africa (Fig.5d).
Forty species were predicted to establish in more
than one geographic region while no species was
predicted to establish in all four regions (Table3).
Of the 40 species, most are found in the aquarium
trade (36 species), followed by the live food (5 spe-
cies), water garden (3 species), and biological sup-
ply (2 species) trades, with five of the species rep-
resented in more than one trade. Five species, all of
which occur in the aquarium trade, were predicted
to establish and have high impact in more than one
region. Wels Catfish (Silurus glanis) is predicted to
have a high impact in all three regions in which it is
predicted to establish, the Sacramento-San Joaquin,
Great Lakes, and Mid-Atlantic. Bronze Corydoras
(Corydoras aeneus), Pearl Danio (Danio albolinea-
tus), and Siamese Algae-eater (Gyrinocheilus aymo-
nieri) are predicted to have a high impact in the
Sacramento-San Joaquin and Lower Colorado. Ide
Table 1 Summary of traits and invasion risk parameters iden-
tified in the developed models predicting establishment success
(Fig.2) and ecological impact (Fig. 3) of non-native freshwa-
ter fishes in four regions. The regions include Sacramento-San
Joaquin (SS), Lower Colorado (LC), Great Lakes (GL), and
Mid-Atlantic (MA). Traits and parameters are ordered by fre-
quency of use among regional models for each stage of inva-
sion
Invasion stage Trait/Parameter Region
Establishment Climate match GL, MA
Maturation size MA, SS
Temperature tolerance LC, SS
Maximum total length SS
Phylogeny SS
Prior establishment success LC
Size of native range LC
Trophic guild SS
Impact Prior establishment success LC, MA, SS
Larval length LC, SS
Trophic guild GL, LC
Climate match SS
Diet breadth SS
Fecundity GL
Temperature tolerance LC
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J.G.Howeth et al.
107 Page 10 of 26
Vol:. (1234567890)
Table 2 Classification tree model summaries for prediction of establishment (ESTABL) and ecological impact (IMPACT) of non-native fishes in four regions
Model Region Established
or
high impact
N
Failed
or
low impact
N
ROC Relative cost # Nodes Root Node Rules of establishment or high
impact
% Correct
classification
ESTABL SS 42 23 0.53 0.97 6 Maturation size (≤ 35%
or > 35%…)
if ≤ 35% and temperature tolerance
is cold, cold-cool, cool, or warm
the species is likely to establish;
if > 35% and maximum total
length is ≤ 88 cm and trophic
guild is herbivore-detritivore,
invertivore, or piscivore the spe-
cies is likely to establish OR if an
invertivore-piscivore or omnivore
with phylogeny ≤ 147 the species
is likely to establish
60% of established,
43% of failed
ESTABL LC 68 14 0.65 0.69 4 Temperature tolerance (if cold,
cool-warm, cool, warm and…) prior success in ≥ 1 country and size
of native range is > 412,419 km2
the species is likely to establish
81% of established,
50% of failed
ESTABL GL 37 28 0.77 0.44 2 Climate match (> 72% likely to
establish)
None additional 81% of established,
75% of failed
ESTABL MA 77 31 0.76 0.50 3 Climate match (> 79% likely to
establish…)
if ≤ 79% and maturation size
is ≤ 22% the species is likely to
establish
82% of established,
68% of failed
IMPACT SS 14 14 0.57 0.71 5 Prior success
(≥ 1 country and…) diet breadth is > 2.50 and climate
match is > 6% the species is
likely to have high impact OR diet
breadth ≤ 2.50 and larval length
is > 3.05 mm the species is likely
to have high impact
50% of high impact,
79% of low impact
IMPACT LC 21 21 0.56 0.86 5 Prior success (≤ 13 countries
or > 13 countries…)
if ≤ 13 countries and trophic guild
is omnivore or piscivore the spe-
cies is likely to have high impact;
if > 13 countries and larval length
is > 3.9 mm and temperature
tolerance is cold, cold-cool,
cool-warm, or warm the species is
likely to have high impact
62% of high impact,
48% of low impact
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Predicting invasiveness offreshwater fishes imported intoNorth America: regional differences…
Page 11 of 26 107
Vol.: (0123456789)
(Leuciscus idus) is predicted to have a high impact
in the Sacramento-San Joaquin and Great Lakes.
Discussion
The present study highlights strong context-depend-
ency in regional risk assessment models, whereby
the trait and parameter predictors of non-native
species establishment and ecological impact varied
substantially for freshwater fishes across four geo-
graphic regions of North America. Despite these
differences, there were important overarching pre-
dictors for each stage of invasion shared among
most regions. In the establishment models of all
regions, climate match was either the most impor-
tant predictor or a significant predictor through the
influence of specific traits (e.g., temperature toler-
ance). In the impact models, prior establishment
success was a key predictor in three of the four
regional models. Collectively, the model results
suggest that applying a single quantitative inva-
sive species risk assessment to multiple geographic
regions could yield inaccurate predictions but that
integrating climate-related variables and prior suc-
cess into model construction can be highly informa-
tive for most regions. The number of species pre-
dicted to establish differed among regions, with the
greatest number of species (548) predicted to estab-
lish in the Sacramento-San Joaquin and the fewest
number (5) in the Mid-Atlantic. Forty species were
predicted to establish in more than one region while
five of those species (Wels Catfish, Bronze Cory-
doras, Ide, Pearl Danio, and Siamese Algae-eater)
were predicted to have high impact in those regions
and thus are of considerable concern to national
biosecurity. From the pathway risk analysis, the
models identified the aquarium trade as the origin
of the most fishes predicted to establish and have a
large impact in all regions. The relative contribution
of native source continents to establishment suc-
cess varied by region with Asia supplying the high-
est risk species overall. Taken together, the findings
from this study utilizing a novel comparative multi-
region approach to test for robust predictors of inva-
sion success provide strong guidance to developing
risk assessments to prevent biological invasions.
The regions include Sacramento-San Joaquin (SS), Lower Colorado (LC), Great Lakes (GL), and Mid-Atlantic (MA). The sample sizes used to train the model (N), the Area
Underneath the Receiver Operating Curve (ROC), the relative cost, the number of terminal nodes (# Nodes), the identity of the root node splitter (Root Node), the model rules for
predicting establishment success (ESTABL) and high impact (IMPACT), and the proportion of correct classification from 10-fold cross-validation are reported for each model.
Model-identified traits and parameters are in bold and categories are italicized
Table 2 (continued)
Model Region Established
or
high impact
N
Failed
or
low impact
N
ROC Relative cost # Nodes Root Node Rules of establishment or high
impact
% Correct
classification
IMPACT GL 12 12 0.79 0.42 3 Trophic guild
(invertivore-piscivore or piscivore
likely to have high impact…)
if an herbivore-detritivore, inver-
tivore, or omnivore and fecun-
dity > 1,013,000 eggs the species
is likely to have high impact
83% of high impact,
75% of low impact
IMPACT MA 24 24 0.76 0.29 2 Prior success
(> 1 country, likely to have high
impact)
None additional 83% of high impact,
88% of low impact
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J.G.Howeth et al.
107 Page 12 of 26
Vol:. (1234567890)
Establishment and importance of climate
Climate-related predictors were important in all
regional establishment models. In the coolest
regions, the Great Lakes and the Mid-Atlantic, a
close match (72%, 79%) between the climate in
the native range and recipient region was the most
important predictor of establishment and with high
accuracy. This result is consistent with the generally
recognized importance of physiological tolerance
in shaping species distributions (Liu et al. 2020;
Strubbe etal. 2023), often making climate match a
robust predictor of non-native species establishment
(Hayes and Barry 2008). In these two regions, high
accuracy in prediction is likely driven by failure of
introduced warm-water species to establish repro-
ducing populations in the colder climates (Kolar and
Lodge 2002; Rixon etal. 2005; Chan etal. 2019). In
contrast, in the warmer and more arid climates of
the Sacramento-San Joaquin and Lower Colorado,
establishment models included more predictors
and supported lower accuracy in prediction. These
environments may have precluded such distinct pat-
terns of species sorting from invasion (Marchetti
etal. 2004a, 2006; Olden etal. 2006; Chang etal.
2009), but models still invoked temperature toler-
ance as an important determinant of establishment.
Smaller maturation size was identified as a predic-
tor of establishment in the Sacramento-San Joaquin
and Mid-Atlantic, highlighting a role for opportun-
istic life history strategies in invasion success (Mar-
chetti etal. 2004a; Olden etal. 2006). Other vari-
ables contributing to prediction of establishment
success were unique to a single region and included
smaller maximum total length, greater prior suc-
cess, larger size of native range, trophic guild (i.e.,
Fig. 3 Classification trees predicting ecological impact of
non-native fish species in the (a) Sacramento-San Joaquin,
(b) Lower Colorado, (c) Great Lakes, and (d) Mid-Atlantic
regions. The number of fish species from the model training
data set located in each response category is indicated at every
terminal node
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Predicting invasiveness offreshwater fishes imported intoNorth America: regional differences…
Page 13 of 26 107
Vol.: (0123456789)
Fig. 4 Proportion of non-
native fish species predicted
to establish in the (a)
Sacramento-San Joaquin,
(c) Lower Colorado, (e)
Great Lakes, and (g) Mid-
Atlantic from the aquarium,
biological (Biol.) supply,
live bait,live food, and
water garden trades. The
total number of species in
each trade pathway for a
region is reported above
the associated pathway bar.
The number of fish species
predicted to establish in
the (b) Sacramento-San
Joaquin, (d) Lower Colo-
rado, (f) Great Lakes, and
(h) Mid-Atlantic from each
trade pathway and their
corresponding predicted
ecological impact. For
(b), (d), (f), and (h), note
differences in y-axis scales.
Within all four geographic
regions, some species are
represented in more than
one trade pathway. Identi-
ties of species in each trade
pathway are in TableS15-
S18
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J.G.Howeth et al.
107 Page 14 of 26
Vol:. (1234567890)
herbivore-detritivore, invertivore, piscivore), and
phylogeny thus indicating context-dependent factors
influencing establishment (Marchetti et al. 2004a;
Peoples and Midway 2018). The models classified
established species more accurately than failed spe-
cies indicating the potential contribution of prop-
agule pressure to establishment failure in addition
to the incorporated traits and parameters (Lockwood
etal. 2005; Peoples and Goforth 2017). The direct
and indirect role of climate in establishment success
invoked by all four models strongly suggests that
integrating climate into future risk assessment tools
is essential for accurate prediction of establishment,
particularly under projections of regional and global
climate change (Sorte et al. 2013; Hubbard et al.
2024).
Impact and importance of prior establishment
The impact models reflect the establishment models
in geographic patterns of complexity but were more
uniform through identifying prior establishment suc-
cess as a key predictor of impact in three of the four
regions. Prior success in more than one country was
sufficient to predict high impact in the Mid-Atlantic
(83–88% correct classification), the highest predict-
ability of the four models. In the Sacramento-San
Joaquin, no prior success predicted low impact align-
ing with an earlier model for the region (Marchetti
et al. 2004b), while diet breadth, larval length, and
climate match contributed to predicting high impact.
In the Lower Colorado, the root node of prior suc-
cess was differentiated by trophic guild, larval length,
and temperature tolerance to predict impact. Longer
Fig. 5 Proportion of fish species predicted to establish in the
(a) Sacramento-San Joaquin, (b) Lower Colorado, (c) Great
Lakes, and (d) Mid-Atlantic, by contributing source continent
based upon the native range of the species. Within all four geo-
graphic regions, some species predicted to establish are native
to more than one continent and therefore represented more
than once
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Predicting invasiveness offreshwater fishes imported intoNorth America: regional differences…
Page 15 of 26 107
Vol.: (0123456789)
Table 3 Non-native fish species predicted to establish in multiple regions (with corresponding impact; L = low, H = high), and with
the contributing native source continent(s) and trade pathway(s)
Species Common name Source Trade pathway Regions (Impact)
Andinoacara pulcher
Gill 1858
Blue Acara North America, South
America
Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(H)
Aphyocharax nattereri
Steindachner 1882
Dawn Tetra South America Aquarium Mid-Atlantic (L), Sacra-
mento-San Joaquin (L)
Aplocheilus lineatus
Valenciennes 1846
Striped Panchax Asia Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(H)
Aplocheilus panchax
Hamilton 1822
Blue Panchax Asia Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(H)
Barbonymus schwanen-
feldii Blecker 1854
Tinfoil Barb Asia Aquarium Lower Colorado (H),
Sacramento-San Joaquin
(L)
Betta imbellis Ladiges
1975
Crescent Betta Asia Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(L)
Betta pugnax Cantor
1849
Penang Betta Asia Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(H)
Betta splendens Regan
1910
Betta Asia Aquarium, water garden Lower Colorado (L),
Sacramento-San Joaquin
(L)
Carassius carassius Lin-
naeus 1758*
Crucian Carp Asia, Europe Live food Great Lakes (L), Sacra-
mento-San Joaquin (H)
Cirrhinus molitorella
Valenciennes 1844
Mud Carp Asia Live food Lower Colorado (H),
Sacramento-San Joaquin
(L)
Cobitis taenia Linnaeus
1758
Spined Loach Europe Aquarium Great Lakes (L), Sacra-
mento-San Joaquin (H)
Copella arnoldi Regan
1912
Splash Tetra South America Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(L)
Corydoras aeneus Gill
1858
Bronze Corydoras North America, South
America
Aquarium Lower Colorado (H),
Sacramento-San Joaquin
(H)
Cyclocheilichthys apo-
gon Valenciennes 1842
Beardless Barb Asia Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(H)
Danio albolineatus
Blyth 1860
Pearl Danio Asia Aquarium Lower Colorado (H),
Sacramento-San Joaquin
(H)
Danio rerio Hamilton
1822
Zebra Danio Asia Aquarium, biological
supply, water garden
Lower Colorado (L),
Sacramento-San Joaquin
(H)
Dawkinsia filamentosa
Valenciennes 1844
Blackspot Barb Asia Aquarium Lower Colorado (H),
Sacramento-San Joaquin
(L)
Devario malabaricus
Jerdon 1849
Malabar Danio Asia Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(H)
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J.G.Howeth et al.
107 Page 16 of 26
Vol:. (1234567890)
Table 3 (continued)
Species Common name Source Trade pathway Regions (Impact)
Gymnocorymbus ternetzi
Boulenger 1895
Black Tetra South America Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(L)
Gyrinocheilus aymonieri
Tirant 1883
Siamese Algae-eater Asia Aquarium, biological
supply
Lower Colorado (H),
Sacramento-San Joaquin
(H)
Hemigrammus ocellifer
Steindachner 1882
Head-and-Taillight Tetra South America Aquarium Lower Colorado (H),
Sacramento-San Joaquin
(L)
Hoplosternum littorale
Hancock 1828
Atipa North America, South
America
Aquarium Lower Colorado (H),
Sacramento-San Joaquin
(L)
Hyphessobrycon eques
Steindachner 1882
Jewel Tetra South America Aquarium Lower Colorado (H),
Sacramento-San Joaquin
(L)
Lates calcarifer Bloch
1790
Barramundi Asia, Oceania Live food Lower Colorado (L), Mid-
Atlantic (L), Sacra-
mento-San Joaquin (L)
Leporinus fasciatus
Bloch 1794
Banded Leporinus South America Aquarium Lower Colorado (H),
Sacramento-San Joaquin
(L)
Leuciscus idus Linnaeus
1758
Ide Asia, Europe Aquarium, water garden Great Lakes (H), Sacra-
mento-San Joaquin (H)
Misgurnus fossilis Lin-
naeus 1758
European Weatherfish Asia, Europe Aquarium Great Lakes (L), Sacra-
mento-San Joaquin (H)
Nannostomus unifascia-
tus Steindachner 1876
Oneline Pencilfish South America Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(L)
Oxyeleotris marmorata
Bleeker 1852
Marble Goby Asia Live food Lower Colorado (L),
Sacramento-San Joaquin
(H)
Parambassis ranga
Hamilton 1822
Indian Glassy Fish Asia Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(L)
Parambassis siamensis
Fowler 1937
Asiatic Glassfish Asia Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(H)
Pethia conchonius Ham-
ilton 1822
Rosy Barb Asia Aquarium Lower Colorado (H),
Sacramento-San Joaquin
(L)
Potamotrygon motoro
Müller & Henle 1841
South American Fresh-
water Stingray
South America Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(H)
Pterophyllum scalare
Schultze 1823
Freshwater Angelfish South America Aquarium, live food Lower Colorado (L),
Sacramento-San Joaquin
(H)
Puntigrus partipenta-
zona Fowler 1934
Tiger Barb Asia Aquarium Lower Colorado (H),
Sacramento-San Joaquin
(L)
Rasbora borapetensis
Smith 1934
Blackline Rasbora Asia Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(L)
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Predicting invasiveness offreshwater fishes imported intoNorth America: regional differences…
Page 17 of 26 107
Vol.: (0123456789)
larval length at hatching corresponded to a higher
impact in the two western regions and may be related
to greater resource consumption and larval growth in
the novel environment (Pepin etal. 2014). The Great
Lakes impact model served as the regional outlier,
with trophic guild as the root node. Consistent with
observations of top-down community and ecosystem
effects in food webs, high-impact non-native fishes
are likely to be top predators (piscivores, invertivore-
piscivores; Great Lakes) and those with a wide diet
breadth (omnivores; Lower Colorado) (Cucherous-
set and Olden 2011). In the Great Lakes, the impor-
tance of top predators may be amplified by the long
food chains supported by the unique ecosystems
of the large deep lakes (Vander Zanden et al. 1999;
Havel etal. 2015). This outlier model could also be
informed by the vast knowledge of Great Lakes inva-
sion history, food webs and ecosystems from over a
century of research (Mandrak and Cudmore 2010;
Sturtevant etal. 2019). The robust prediction of prior
success in three regional models, alone or in tandem
with other variables, highlights consistency in impact
of invasive species globally (Kulhanek etal. 2011).
Establishment and impact model conclusions
Models produced for the establishment and impact
stages of invasion differed in traits and parameters
among regions emphasizing context-dependency
in predictors of invasion success and supporting the
development and application of region-specific mod-
els. Unknown and unincorporated variables such as
propagule pressure influencing establishment success,
and residence time influencing ecological impact
estimates, likely contribute to some of the regional
variation and context-dependency in predictive traits
and parameters (Pyšek etal. 2020a). Collectively, the
composition of the four regional models challenges
the notion of a unified predictor of invasion success,
the holy grail of invasion biology (Elton 1956), and
suggests the transferability of quantitative risk assess-
ment models is limited in part by regional contexts
and their interactions with species invasion history
(Kolar and Lodge 2002; Pyšek etal. 2020a). Despite
the potential for improved accuracy in invasion pre-
diction using quantitative regional species risk assess-
ments, a lack of region-specific data on prior success-
ful and failed invasions for a given taxonomic group
often precludes development of regional invasive risk
models thereby limiting widespread implementation
of this approach.
The results identifying climate-related variables
as predictors of establishment success in multiple
regions highlights a robust role for climate in inva-
sion outcome (Liu et al. 2020). These findings are
consistent with previous analyses of fishes (Bomford
etal. 2010) and other taxa including amphibians and
reptiles (Bomford etal. 2009; Van Wilgen and Rich-
ardson 2012), birds (Duncan et al. 2001; Strubbe
etal. 2023), mammals (Forsyth etal. 2004; Broenni-
mann etal. 2021), and plants (Petitpierre etal. 2012;
Liu etal. 2020). Adopting a simple risk assessment
framework that integrates climate match for evaluat-
ing establishment success could predict invasiveness
with sufficient accuracy for cooler temperate regions
but may be less effective for warmer subtropical and
Table 3 (continued)
Species Common name Source Trade pathway Regions (Impact)
Rasbora trilineata Stein-
dachner 1876
Three-lined Rasbora Asia Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(L)
Silurus glanis Linnaeus
1758*
Wels Catfish Asia, Europe Aquarium Great Lakes (H), Mid-
Atlantic (H), Sacra-
mento-San Joaquin (H)
Symphysodon aequifas-
ciatus Pellegrin 1904
Blue Discus South America Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(H)
Trichopsis vittata Cuvier
1831
Croaking Gourami Asia Aquarium Lower Colorado (L),
Sacramento-San Joaquin
(H)
*Denotes species prohibited from importation into the United States by the authority of the Lacey Act (18 U.S.C 42)
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J.G.Howeth et al.
107 Page 18 of 26
Vol:. (1234567890)
tropical regions (Hill et al. 2024). Regularly updat-
ing risk assessments to address changes in climatic
conditions and the climatic match between the native
and introduced range to re-evaluate the prediction for
establishment of previously thermally-limited species
could improve predictive accuracy (Wittmann et al.
2014; Hubbard etal. 2024). Future assessments could
also integrate the potential for adaptation and plas-
ticity in thermally responsive species traits (Strubbe
etal. 2023).
Additionally, based upon the model results, evalu-
ating prior establishment success provides a straight-
forward and accessible method for measuring eco-
logical impact in a region (Hayes and Barry 2008;
Kulhanek etal. 2011). Documented geographic dis-
tributions of species in their native and non-native
ranges are publicly available in biodiversity and inva-
sive species databases (e.g., Bargeron and Moorhead
2007; GBIF 2024; USGS 2024). If a species has no
prior history of invasion from the absence (or record)
of any past introduction events, however, it could
negate the utility of this metric and yield a false nega-
tive allowing a species that may become invasive to
screen as likely to fail. Thus, there may be species in
this study that are predicted to fail that have the poten-
tial to successfully establish if introduced. Interest-
ingly, the current USFWS risk assessment tool inte-
grates both a climate comparison and the history of
species invasiveness to determine invasive potential
in the US (i.e., Ecological Risk Screening Summary
(ERSS); Meyers etal. 2020). The establishment and
impact models developed here support this approach
to risk screening for fishes but additionally offer
quantitative model alternatives to identify high risk
invaders for a subset of regions in North America.
Finally, the classification tree approach to devel-
oping invasion stage models presents both strengths
and weaknesses. Strengths include the application
of nonparametric methods, ability to handle miss-
ing data and outliers, and simplicity of model inter-
pretation (Olden et al. 2008). Weaknesses include
the inability to repeat predictor variables or invoke
combined effects, the lack of confidence intervals or
probability values associated with node splits, and the
potential for overfit models. The impact models are
trained on impact values derived from expert opinion;
thus, a different suite of regional experts could lead
to changes in species-specific impact estimates and
thus overall impact model construction. Therefore,
the establishment and impact models in this study
should be interpreted with caution and considered
a statistical representation of the included traits and
parameters with respect to invasion stage rather than
a definitive causal relationship.
Pathways across regions
The pathway analysis identified the aquarium trade
as posing the greatest risk to all regions in terms of
the number of species predicted to establish and have
high impact. The number of species predicted to
establish varied by region with 548 (68.7%) species
predicted to establish in Sacramento-San Joaquin, 73
(9.3%) species in the Lower Colorado, seven (0.9%)
species in the Great Lakes, and five (0.6%) spe-
cies in the Mid-Atlantic. This result likely reflects
temperature suitability for species in the aquarium
trade (Gertzen et al. 2008; Chang etal. 2009; Ven-
ezia etal. 2017), which supplies the majority of fish
species imported into North America due to wide-
spread consumer demand (Bradie etal. 2013; Chan
etal. 2019). The commercial and domestic aquarium
fish industry has grown in popularity thus increas-
ing propagule pressure from aquarium trade imports
(Smith et al. 2017; Olden et al. 2021). With global
warming, improved climate matches to non-native
environments for warm-water aquarium species are
predicted to expand the overall establishment risk
from this pathway in North America (Muñoz‐Mas
etal. 2023; Hubbard etal. 2024). Once established,
non-native aquarium fishes can have large impacts on
community structure and ecosystem function (Britton
2023). They may also possess unique traits or charac-
teristics relative to native species thereby occupying a
novel niche in the receiving environment (Capps and
Flecker 2013; Xu etal. 2022). Much of the economic
cost of fish invasion impacts in North America, esti-
mated at $32 billion per year, can be attributed to spe-
cies originating from the aquarium trade (Haubrock
etal. 2022; Turbelin etal. 2022). These considerable
costs emphasize the importance of addressing the
consequences of aquarium species release, and acci-
dental escape, to curb biological invasions (Lock-
wood etal. 2019).
Region-specific differences in invasion success
related to species and traits selected by specific path-
ways materialized when examining the proportion of
species within each pathway predicted to establish.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Predicting invasiveness offreshwater fishes imported intoNorth America: regional differences…
Page 19 of 26 107
Vol.: (0123456789)
The aquarium trade poses the greatest risk to the Sac-
ramento-San Joaquin when evaluating proportional
establishment. The threat of the aquarium trade to this
region may be further enhanced given the numerous
pet stores supplying aquarium taxa in the San Fran-
cisco metropolitan area (Chang etal. 2009). Fish spe-
cies native to South America, a major origin of warm-
water aquarium fishes (Moreau and Coomes 2007),
comprise the largest proportion of species predicted
to establish in Sacramento-San Joaquin. In contrast,
the greatest threats elsewhere are the biological sup-
ply and live food trades to the Lower Colorado, the
water garden trade to the Great Lakes, and the live
food trade to the Mid-Atlantic. These regional trade
pathway patterns can be better understood by recog-
nizing species’ continental source(s), with Asia sup-
plying the most species predicted to establish among
all regions. The continent conveying the greatest risk
to the Lower Colorado and the Mid-Atlantic is Asia,
where high human consumption of fishes and associ-
ated aquaculture provides an ongoing supply chain of
Asian fishes for the live food pathway (Naylor etal.
2021). In the Great Lakes, the water garden trade has
the highest proportion of species likely to establish;
all of which are predicted to have high impact. The
water garden industry intentionally selects species
that survive well outdoors and supplies many retail
markets in the Great Lakes region (Keller and Lodge
2007). The origins of 90% of the species predicted
to establish in the Great Lakes are Europe and Asia,
reflective of regions with similar cool temperate cli-
mates (Hubbard etal. 2024). Collectively, the path-
way results suggest that the aquarium trade poses the
largest overall risk, but region-specific differences in
predicted establishment by pathway highlight unique
biological requirements and management challenges
posed by different regional environments.
Risk at the continental scale
From the perspective of the North American con-
tinent, the 40 fish species predicted to establish in
more than one region could be considered higher-
risk invaders. The aquarium trade supplies 90% of
these species, while 83% are predicted to estab-
lish only in the warmer climates of the two west-
ern regions of the US. The five species predicted
to both establish and have a high ecological impact
in multiple regions pose the greatest concern to
the integrity of North American ecosystems. Wels
Catfish was the only species predicted to estab-
lish and have a high impact in three regions, the
Sacramento-San Joaquin, Great Lakes, and Mid-
Atlantic. As a large-bodied piscivore that has been
introduced outside of its native range in Europe
(Copp et al. 2009), Wels Catfish can have strong
top-down effects from a broad diet breadth that
includes fishes, crayfishes, aquatic insects, and
semi-aquatic prey such as amphibians, birds, and
mammals (Vejřík etal. 2017). Consequently, Wels
Catfish was prohibited from importation into the US
in 2016 under the authority of the Lacey Act (Dean
etal. 2024). Another species, Ide, was predicted to
successfully establish and have high impact in the
Sacramento-San Joaquin and Great Lakes. This
omnivore and habitat generalist has invasive popu-
lations in Europe but multiple failed introductions
are recorded from intentional stocking in the US
in the 1800s (Rohtla etal. 2021; Nico etal. 2024).
While there are no known established populations
of Ide in North America, the risk of invasion suc-
cess increases with hobbyist demand and propagule
pressure from the aquarium trade. The remain-
ing three species, Bronze Corydoras, Pearl Danio,
and Siamese Algae-eater are predicted to estab-
lish and have high impact in the Sacramento-San
Joaquin and Lower Colorado. Bronze Corydoras,
a small-bodied tropical catfish that is a facultative
air breather, has established invasive populations
in Hawaii (Nelson 2014; Nico and Schofield 2024).
Pearl Danio is a small-bodied tropical minnow that
has established non-native populations in Asia adja-
cent to its native range (Hui etal. 2020). The species
was first imported into Europe in the early 1900s for
cultivation and circulation in the aquarium trade
(Novák etal. 2020). The Siamese Algae-eater ranks
as one of the most popular traded aquarium spe-
cies in North America due to its ability to graze and
control algae in tanks (Maceda‐Veiga etal. 2016).
The species has established non-native populations
in Puerto Rico and, based on an invasion risk model
integrating propagule pressure, is also predicted
to establish in Hawaii and Florida (Venezia et al.
2017; Rodríguez-Barreras et al. 2020). Given the
biology and history of these five fish species pre-
dicted to become invasive in multiple regions, addi-
tional regulations and surveillance for these species
in North America may be warranted.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J.G.Howeth et al.
107 Page 20 of 26
Vol:. (1234567890)
Management implications
The fish species predicted to become invasive from
this study would benefit from a detailed risk assess-
ment (e.g., USFWS ERSS) and, in the interim, a
rapid assessment through horizon scans to further
evaluate threats to North American ecosystems
(e.g., Daniel etal. 2025). None of the five identi-
fied highest risk species are regulated at the federal
level by Canada while only Wels Catfish is prohib-
ited from importation in the US. Monitoring fresh-
water ecosystems for these species using eDNA
(Roy etal. 2018), combined with traditional sam-
pling methods, could provide early warning of any
incipient invasion. Additionally, the regional risk
assessment models produced here may also be uti-
lized to screen any new species entering trade path-
ways in North America. Scaling up our modeling
approach in the future, through acquisition of data
on successful and failed invasions among different
regions across continents, could yield global cover-
age of analogous invasion stage models and have
the potential to provide additional insight into driv-
ers and correlates of biological invasion along with
high-risk species.
Global transport of species from international
trade continues to operate at levels that negatively
impact native biodiversity and drive biotic homog-
enization (Young et al. 2016; Hulme 2021). Con-
tributing to the demand for species imports, the
purchase of ornamental fishes by hobbyists in the
aquarium industry has surged over the last decade
largely stimulated by new supply from Internet
providers (Olden etal. 2021; Sinclair etal. 2021).
Additionally, freshwater fish imports from aquacul-
ture sources have continued to increase globally for
human consumption and are predicted to double by
2050 (Naylor etal. 2021). The two associated trade
pathways supplying these fishes, the aquarium trade
and the live food trade, will likely provide continu-
ous delivery of potential new invaders to North
America. Results from this study suggest that these
pathways supply some of the most high-risk species
to North American ecosystems. Based upon his-
torical trends of continental accumulations of inva-
sive species, new non-native fishes are predicted to
establish and become invasive in North America
through 2050 from these pathways and other trades
and transportation pipelines (Seebens etal. 2021).
Conclusions
This is one of the first studies to use a compara-
tive multi-region approach to test for robust traits
and parameters predicting invasion success by con-
structing models for the establishment and impact
stages of fish invasion in four regions of North
America. The models are subsequently applied to
identify potentially invasive species in trade and
conduct pathway analyses to determine which trades
and source continents pose the greatest risk. Models
produced for the establishment and impact stages
of invasion differed in composition among regions
emphasizing context-dependency in quantitative
risk assessments and supporting the need to develop
and apply region-specific models. The consistent
predictors of climate-related variables and prior
establishment success among regions, however,
suggests their inclusion into future risk assessments
may improve efficacy. From the pathway analysis,
fish species in the aquarium trade and Asia have
a greater risk of becoming invasive if introduced
with a subset of the highest-risk species posing a
multi-region threat. Ultimately, invasion prevention
informed by risk assessments such as those devel-
oped here yields the least costly approach to stop
establishment of these higher-risk species (Lodge
etal. 2016; Cuthbert etal. 2021). Prevention meas-
ures can include prohibitions or regulations on the
importation, possession, or release of potentially
invasive species and public education. Success-
ful conservation of North American ecosystems
depends upon implementing proactive invasion pre-
vention measures informed by current invasion sci-
ence and global change.
Acknowledgements U.S. Environmental Protection Agency
Great Lakes Restoration Initiative funded this work via United
States Fish and Wildlife Service award 30181AJ261 to the
University of Notre Dame. Additional support was provided
by National Science Foundation Division of Environmen-
tal Biology award 1645137 to JGH, and Cornell Atkinson
Center for Sustainability to DML. Thanks to Matt Bury and
Mariana Castaneda-Guzman for developing the map. Com-
ments from Hanno Seebens, Wes Daniel, and an anonymous
reviewer greatly improved the manuscript. The Virginia Coop-
erative Fish and Wildlife Research Unit is jointly sponsored
by the U.S. Geological Survey, Virginia Polytechnic Institute
and State University, and the Virginia Department of Wild-
life Resources. Any use of trade, firm, or product names is for
descriptive purposes only and does not imply endorsement by
the U.S. Government. The findings and conclusions in this
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Predicting invasiveness offreshwater fishes imported intoNorth America: regional differences…
Page 21 of 26 107
Vol.: (0123456789)
article are those of the authors and do not necessarily represent
the views of the U.S. Fish and Wildlife Service.
Author contributions All authors contributed to study
design. J.G.H, S.A.A, C.A.G. and N.E.M. collected, collated,
and analyzed species trait and invasion risk parameter data,
and conducted the screening analyses. P.L.A, N.E.M, M.P.M.,
and J.D.O. contributed regional data sets. D.M.L. and J.G.H.
funded the study. J.G.H. led writing of the manuscript, with
contributions to the conceptual framework and revisions from
all authors. All authors read and approved the final manuscript.
Funding U.S.Environmental Protection Agency Great Lakes
Restoration Initiative funded this work via United States Fish
and Wildlife Service award 30181AJ261 to the University of
Notre Dame. Additional support was provided by National Sci-
ence Foundation Division of Environmental Biology award
1645137 to JGH, and Cornell Atkinson Center for Sustainabil-
ity to DML.
Data availability The data that supports this study will be
shared upon requestto the corresponding author.
Declarations
Conflict of interest The authors have no relevant financial or
non-financial interests to disclose.
Open Access This article is licensed under a Creative Com-
mons Attribution 4.0 International License, which permits
use, sharing, adaptation, distribution and reproduction in any
medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Crea-
tive Commons licence, and indicate if changes were made. The
images or other third party material in this article are included
in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not
included in the article’s Creative Commons licence and your
intended use is not permitted by statutory regulation or exceeds
the permitted use, you will need to obtain permission directly
from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/.
References
Angermeier PL, Pinder MJ (2015) Viewing the status of Vir-
ginia’s environment through the lens of freshwater
fishes. Va J Sci 66:147–169. https:// doi. org/ 10. 25778/
yyy2- g953
Bachman S, Moat J, Hill AW, De La Torre J, Scott B (2011)
Supporting Red List threat assessments with GeoCAT:
geospatial conservation assessment tool. ZooKeys
150:117. https:// doi. org/ 10. 3897/ zooke ys. 150. 2109
Baiser B, Olden JD, Record S, Lockwood JL, McKinney ML
(2012) Pattern and process of biotic homogenization in
the New Pangaea. Proc R Soc B: Biol Sci 279:4772–
4777. https:// doi. org/ 10. 1098/ rspb. 2012. 1651
Bargeron CT, Moorhead DJ (2007) EDDMapS—early detec-
tion and distribution mapping system for the southeast
exotic pest plant council. Wildland Weeds 10:4–8
Bernery C, Bellard C, Courchamp F, Brosse S, Gozlan RE,
Jarić I etal (2022) Freshwater fish invasions: a compre-
hensive review. Annu Rev Ecol Evol Syst 53:427–456.
https:// doi. org/ 10. 1146/ annur ev- ecols ys- 032522- 015551
Bernery C, Bellard C, Courchamp F, Brosse S, Leroy B (2024)
A global analysis of the introduction pathways and char-
acteristics associated with non-native fish species intro-
duction, establishment, and impacts. Ecol Process 13:22.
https:// doi. org/ 10. 1186/ s13717- 024- 00495-8
Blackburn TM, Pyšek P, Bacher S, Carlton JT, Duncan RP,
Jarošík V et al (2011) A proposed unified framework
for biological invasions. Trends Ecol Evol 26:333–339.
https:// doi. org/ 10. 1016/j. tree. 2011. 03. 023
Blackburn TM, Essl F, Evans T, Hulme PE, Jeschke JM, Kühn
I et al (2014) A unified classification of alien species
based on the magnitude of their environmental impacts.
PLoS Biol 12:e1001850. https:// doi. org/ 10. 1371/ journ al.
pbio. 10018 50
Bomford M, Kraus F, Barry SC, Lawrence E (2009) Predicting
establishment success for alien reptiles and amphibians:
a role for climate matching. Biol Invasions 11:713–724.
https:// doi. org/ 10. 1007/ s10530- 008- 9285-3
Bomford M, Barry SC, Lawrence E (2010) Predicting estab-
lishment success for introduced freshwater fishes: a role
for climate matching. Biol Invasions 12:2559–2571.
https:// doi. org/ 10. 1007/ s10530- 009- 9665-3
Bradie J, Chivers C, Leung B (2013) Importing risk: quantify-
ing the propagule pressure–establishment relationship at
the pathway level. Divers Distrib 19:1020–1030. https://
doi. org/ 10. 1111/ ddi. 12081
Breiman L, Friedman JH, Olshen A, Stone CG (1984) Clas-
sification and regression trees. Wadsworth International
Group, Belmont, California.
Britton JR (2023) Contemporary perspectives on the ecologi-
cal impacts of invasive freshwater fishes. J Fish Biol
103:752–764. https:// doi. org/ 10. 1111/ jfb. 15240
Broennimann O, Petitpierre B, Chevalier M, González-Suárez
M, Jeschke JM, Rolland J etal (2021) Distance to native
climatic niche margins explains establishment success of
alien mammals. Nat Commun 12:2353. https:// doi. org/
10. 1038/ s41467- 021- 22693-0
Australian Bureau of Agricultural and Resource Economics
and Sciences. (2020). Climatch v2.0. https:// clima tch.
cp1. agric ulture. gov. au/
Campbell SE, Mandrak NE (2019) Temporal dynamics of tax-
onomic homogenization in the fish communities of the
Laurentian Great Lakes. Divers Distrib 25:1870–1878.
https:// doi. org/ 10. 1111/ ddi. 12986
Capellini I, Baker J, Allen WL, Street SE, Venditti C (2015)
The role of life history traits in mammalian invasion suc-
cess. Ecol Lett 18:1099–1107. https:// doi. org/ 10. 1111/
ele. 12493
Capps K, Flecker KA (2013) Invasive aquarium fish transform
ecosystem nutrient dynamics. Proc R Soc B Biol Sci.
https:// doi. org/ 10. 1098/ rspb. 2013. 1520
Catford JA, Smith AL, Wragg PD, Clark AT, Kosmala M,
Cavender-Bares J etal (2019) Traits linked with species
invasiveness and community invasibility vary with time,
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J.G.Howeth et al.
107 Page 22 of 26
Vol:. (1234567890)
stage and indicator of invasion in a long-term grassland
experiment. Ecol Lett 22:593–604. https:// doi. org/ 10.
1111/ ele. 13220
Chan FT, Beatty SJ, Gilles AS Jr, Hill JE, Kozic S, Luo D etal
(2019) Leaving the fish bowl: the ornamental trade as a
global vector for freshwater fish invasions. Aquat Eco-
syst Health Manag 22:417–439. https:// doi. org/ 10. 1080/
14634 988. 2019. 16858 49
Chang AL, Grossman JD, Spezio TS, Weiskel HW, Blum
JC, Burt JW et al (2009) Tackling aquatic invasions:
risks and opportunities for the aquarium fish industry.
Biol Invasions 11:773–785. https:// doi. org/ 10. 1007/
s10530- 008- 9292-4
Copp GH, Britton JR, Cucherousset J, Garcia-Berthou E, Kirk
R, Peeler E, Stakenas S (2009) Voracious invader or
benign feline? A review of the environmental biology of
European Catfish Silurus glanis in its native and intro-
duced ranges. Fish Fish 10:252–282. https:// doi. org/ 10.
1111/j. 1467- 2979. 2008. 00321.x
Cucherousset J, Olden JD (2011) Ecological impacts of nonna-
tive freshwater fishes. Fisheries 36:215–230. https:// doi.
org/ 10. 1080/ 03632 415. 2011. 574578
Cuthbert RN, Pattison Z, Taylor NG, Verbrugge L, Diagne C,
Ahmed DA etal (2021) Global economic costs of aquatic
invasive alien species. Sci Total Environ 775:145238.
https:// doi. org/ 10. 1016/j. scito tenv. 2021. 145238
Daniel WM, Sofaer HR, Jarnevich CS, Erickson RA, DeGre-
gorio BA, Engelstad PS etal (2025) Vertebrates in trade
that pose high invasion risk to the United States. Biol
Conserv 302:110887. https:// doi. org/ 10. 1016/j. biocon.
2024. 110887
Dean EM, Jordon A, Agnew AC, Hernandez ND, Morning-
star CR, Neilson M etal (2024) America’s Most Wanted
Fishes: cataloging risk assessments to prioritize invasive
species for management action. Manag Biol Invasion
15:1–20. https:// doi. org/ 10. 3391/ mbi. 2024. 15.1. 01
De’ath G, Fabricius KE (2000) Classification and regression
trees: a powerful yet simple technique for ecological
data analysis. Ecology 81:3178–3192. https:// doi. org/ 10.
1890/ 0012- 9658(2000) 081[3178: CARTAP] 2.0. CO;2
Diagne C, Leroy B, Vaissière A, Gozlan RE, Roiz D, Jarić I
etal (2021) High and rising economic costs of biological
invasions worldwide. Nature 592:571–576. https:// doi.
org/ 10. 1038/ s41586- 021- 03405-6
Dill WA, Cordone AJ (1997) History and status of introduced
fishes in California, 1871–1996. Fish Bulletin 178. Cali-
fornia Department of Fish and Game, Sacramento, Cali-
fornia, USA
Drake DAR, Mandrak NE (2014) Ecological risk associated
with live bait fisheries: a new angle on selective fishing.
Fisheries 39:201–211. https:// doi. org/ 10. 1080/ 03632 415.
2014. 903835
Duenas MA, Hemming DJ, Roberts A, Diaz-Soltero H (2021)
The threat of invasive species to IUCN-listed critically
endangered species: A systematic review. Glob Ecol
Conserv 26:e01476. https:// doi. org/ 10. 1016/j. gecco.
2021. e01476
Duncan RP, Bomford M, Forsyth DM, Conibear L (2001) High
predictability in introduction outcomes and the geo-
graphical range size of introduced Australian birds: a role
for climate. J Anim Ecol 70:621–632. https:// doi. org/ 10.
1046/j. 1365- 2656. 2001. 00517.x
Elton CS (1956) The ecology of invasions by animals and
plants. Methuen, London
Essl F, Bacher S, Blackburn TM, Booy O, Brundu G, Brunel
S etal (2015) Crossing frontiers in tackling pathways of
biological invasions. Bioscience 65:769–782. https:// doi.
org/ 10. 1093/ biosci/ biv082
Forsyth DM, Duncan RP, Bomford M, Moore G (2004) Cli-
matic suitability, life-history traits, introduction effort,
and the establishment and spread of introduced mammals
in Australia. Conserv Biol 18:557–569. https:// doi. org/
10. 1111/j. 1523- 1739. 2004. 00423.x
Frimpong EA, Angermeier PL (2009) Fish traits: a database of
ecological and life-history traits of freshwater fishes of
the United States. Fisheries 34:487–495. https:// doi. org/
10. 1577/ 1548- 8446- 34. 10. 487
Froese R, Pauly D (2021) FishBase. www. fishb ase. org version
(02/2021)
García-Berthou E (2007) The characteristics of invasive fishes:
what has been learned so far? J Fish Biol 71:33–55.
https:// doi. org/ 10. 1111/j. 1095- 8649. 2007. 01668.x
GBIF: The Global Biodiversity Information Facility (2024)
What is GBIF? Available from https:// www. gbif. org/
what- is- gbif
Gerland P, Raftery AE, Ševčíková H, Li N, Gu D, Spoorenberg
T et al (2014) World population stabilization unlikely
this century. Science 346:234–237. https:// doi. org/ 10.
1126/ scien ce. 12574 69
Gertzen E, Familiar O, Leung B (2008) Quantifying invasion
pathways: fish introductions from the aquarium trade.
Can J Fish Aquatic Sci 65:1265–1273. https:// doi. org/ 10.
1139/ F08- 056
Harrower CA, Scalera R, Pagad S, Schonrogge K, Roy HE
(2018) Guidance for interpretation of CBD categories on
introduction pathways. Report to the European Commis-
sion https:// www. cbd. int/ doc/c/ 9d85/ 3bc5/ d640f 059d0
3acd7 17602 cd76/ sbstta- 22- inf- 09- en. pdf
Haubrock PJ, Bernery C, Cuthbert RN, Liu C, Kourantidou
M, Leroy B etal (2022) Knowledge gaps in economic
costs of invasive alien fish worldwide. Sci Total Environ
803:149875. https:// doi. org/ 10. 1016/j. scito tenv. 2021.
149875
Havel JE, Kovalenko KE, Thomaz SM, Amalfitano S, Kats
LB (2015) Aquatic invasive species: challenges for the
future. Hydrobiologia 750:147–170. https:// doi. org/ 10.
1007/ s10750- 014- 2166-0
Hayes KR, Barry SC (2008) Are there any consistent predic-
tors of invasion success? Biol Invasions 10:483–506.
https:// doi. org/ 10. 1007/ s10530- 007- 9146-5
Hill JE, Tuckett QM, Lawson KM (2024) Climate match fails
to explain variation in establishment success of non-
native freshwater fishes in a warm climate region. Aquat
Invasions 19:73–83. https:// doi. org/ 10. 3391/ ai. 2024.
19.1. 117603
Howeth JG, Gantz CA, Angermeier PL, Frimpong EA, Hoff
MH, Keller RP etal (2016) Predicting invasiveness of
species in trade: climate match, trophic guild and fecun-
dity influence establishment and impact of non-native
freshwater fishes. Divers Distrib 22:148–160. https:// doi.
org/ 10. 1111/ ddi. 12391
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Predicting invasiveness offreshwater fishes imported intoNorth America: regional differences…
Page 23 of 26 107
Vol.: (0123456789)
Hubbard JA, Drake DAR, Mandrak NE (2023) Estimating
potential global sources and secondary spread of fresh-
water invasions under historical and future climates.
Divers Distrib 29:986–996. https:// doi. org/ 10. 1111/ ddi.
13695
Hubbard JA, Drake DAR, Mandrak NE (2024) Climate change
alters global invasion vulnerability among ecoregions.
Divers Distrib 30:26–40. https:// doi. org/ 10. 1111/ ddi.
13778
Hui TH, Peng KLK, Huan LJ, Wei LB, Hing RLB, Beng JKT,
Yeo DC (2020) The non-native freshwater fishes of Sin-
gapore: an annotated compilation. Raffles Bull Zool.
https:// doi. org/ 10. 26107/ RBZ- 2020- 0016
Hulme PE (2009) Trade, transport and trouble: managing inva-
sive species pathways in an era of globalization. J Appl
Ecol 46:10–18. https:// doi. org/ 10. 1111/j. 1365- 2664.
2008. 01600.x
Hulme PE (2021) Unwelcome exchange: International trade
as a direct and indirect driver of biological invasions
worldwide. One Earth 4:666–679. https:// doi. org/ 10.
1016/j. oneear. 2021. 04. 015
Ibáñez I, Diez JM, Miller LP, Olden JD, Sorte CJ, Blumen-
thal DM etal (2014) Integrated assessment of biologi-
cal invasions. Ecol App 24:25–37. https:// doi. org/ 10.
1890/ 13- 0776.1
IPBES (2023) Summary for Policymakers of the Thematic
Assessment Report on Invasive Alien Species and their
Control of the Intergovernmental Science-Policy Plat-
form on Biodiversity and Ecosystem Services. Roy HE,
Pauchard A, Stoett P, Renard Truong T, Bacher S, Galil
BS, et al (eds.). IPBES secretariat, Bonn, Germany.
https:// doi. org/ 10. 5281/ zenodo. 74306 92
Jenkins RE, Burkhead NM (1994) Freshwater fishes of Vir-
ginia. American Fisheries Society, Bethesda, MD, p
1079
Jones KA, Niknami LS, Buto SG, Decker, D (2022) Federal
standards and procedures for the national Watershed
Boundary Dataset (WBD) (5 ed.): U.S. Geological Sur-
vey Techniques and Methods 11-A3, 54 p.,https:// pubs.
usgs. gov/ tm/ 11/ a3/
Keller RP, Lodge DM (2007) Species invasions from com-
merce in live aquatic organisms: Problems and possible
solutions. Bioscience 57:428–436. https:// doi. org/ 10.
1641/ B5705 09
Keller RP, Kocev D, Džeroski S (2011) Trait-based risk assess-
ment for invasive species: high performance across
diverse taxonomic groups, geographic ranges and
machine learning/statistical tools. Divers Distrib 17:451–
461. https:// doi. org/ 10. 1111/j. 1472- 4642. 2011. 00748.x
Kolar CS, Lodge DM (2002) Ecological predictions and risk
assessment for alien fishes in North America. Science
298:1233–1236. https:// doi. org/ 10. 1126/ scien ce. 10757 53
Kulhanek SA, Ricciardi A, Leung B (2011) Is invasion history
a useful tool for predicting the impacts of the world’s
worst aquatic invasive species? Ecol App 21:189–202.
https:// doi. org/ 10. 1890/ 09- 1452.1
Kumschick S, Wilson JRU, Foxcroft LC (2020) A framework
to support alien species regulation: the risk analysis for
alien Taxa (RAAT). NeoBiota 62:213–239. https:// doi.
org/ 10. 3897/ neobi ota. 62. 51031
Kumschick S, Bertolino S, Blackburn TM, Brundu G, Costello
KE, De Groot M etal (2024) Using the IUCN Environ-
mental Impact Classification for Alien Taxa to inform
decision-making. Conserv Biol 38:e14214. https:// doi.
org/ 10. 1111/ cobi. 14214
Lawson KM, Hill JE (2022) Life history strategies differenti-
ate established from failed non-native freshwater fish in
peninsular Florida. Divers Distrib 28:160–172. https://
doi. org/ 10. 1111/ ddi. 13448
Leung B, Roura-Pascual N, Bacher S, Heikkilä J, Brotons L,
Burgman MA etal (2012) TEASIng apart alien species
risk assessments: a framework for best practices. Ecol
Lett 15:1475–1493. https:// doi. org/ 10. 1111/ ele. 12003
Lieurance D, Canavan S, Behringer DC, Kendig AE, Minteer
CR, Reisinger LS etal (2023) Identifying invasive spe-
cies threats, pathways, and impacts to improve biosecu-
rity. Ecosphere 14:e4711. https:// doi. org/ 10. 1002/ ecs2.
4711
Litvak MK, Mandrak NE (1999) Baitfish trade as a vector of
aquatic introductions. In: Claudi R, Leach JH (eds) Non-
indigenous freshwater organisms: vectors, biology, and
impacts. Lewis Publishers, New York, pp 163–180
Liu C, Comte L, Olden JD (2017) Heads you win, tails you
lose: Life-history traits predict invasion and extinction
risk of the world’s freshwater fishes. Aquat Conserv: Mar
Freshw Ecosyst 27:773–779. https:// doi. org/ 10. 1002/ aqc.
2740
Liu C, Wolter C, Xian W, Jeschke JM (2020) Most invasive
species largely conserve their climatic niche. Proc Natl
Acad Sci 117:23643–23651. https:// doi. org/ 10. 1073/
pnas. 20042 89117
Livengood EJ, Funicelli N, Chapman FA (2014) The applica-
bility of the US Law Enforcement Management System
(LEMIS) database for the protection and management
of ornamental fish. Aquacult Aquarium Conserv Legis
7:268–275
Lockwood JL, Cassey P, Blackburn T (2005) The role of prop-
agule pressure in explaining species invasions. Trends
Ecol Evol 20:223–228. https:// doi. org/ 10. 1016/j. tree.
2005. 02. 004
Lockwood JL, Welbourne DJ, Romagosa CM, Cassey P, Man-
drak NE, Strecker A et al (2019) When pets become
pests: the role of the exotic pet trade in producing inva-
sive vertebrate animals. Front Ecol Environ 17:323–330.
https:// doi. org/ 10. 1002/ fee. 2059
Lodge DM, Williams S, MacIsaac HJ, Hayes KR, Leung B,
Reichard S et al (2006) Biological invasions: recom-
mendations for US policy and management. Ecol App
16:2035–2054. https:// doi. org/ 10. 1890/ 1051- 0761(2006)
016[2035: BIRFUP] 2.0. CO;2
Lodge DM, Simonin PW, Burgiel SW, Keller RP, Bossenbroek
JM, Jerde CL etal (2016) Risk analysis and bioeconom-
ics of invasive species to inform policy and management.
Annu Rev Environ Resour 41:453–488. https:// doi. org/
10. 1146/ annur ev- envir on- 110615- 085532
Maceda-Veiga A, Domínguez-Domínguez O, Escribano-Alacid
J, Lyons J (2016) The aquarium hobby: can sinners
become saints in freshwater fish conservation? Fish Fish
17:860–874. https:// doi. org/ 10. 1111/ faf. 12097
Mandrak NE (1989) Potential invasion of the Great Lakes by
fish species associated with climatic warming. J Great
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J.G.Howeth et al.
107 Page 24 of 26
Vol:. (1234567890)
Lakes Res 15:306–316. https:// doi. org/ 10. 1016/ S0380-
1330(89) 71484-2
Mandrak NE, Cudmore B (2010) The fall of native fishes and
the rise of non-native fishes in the Great Lakes Basin.
Aquat Ecosyst Health Manag 13:255–268. https:// doi.
org/ 10. 1080/ 14634 988. 2010. 507150
Mandrak NE, Gantz C, Jones LA, Marson D, Cudmore B
(2014) Evaluation of five freshwater fish risk assess-
ment protocols and application to non-indigenous organ-
isms in trade in Canada. DFO Can Sci Adv Sec Res Doc.
2013/122. v + 125 p
Marchetti MP, Moyle PB, Levine R (2004a) Alien fishes in
California watersheds: characteristics of successful and
failed invaders. Ecol App 14:587–596. https:// doi. org/ 10.
1890/ 02- 5301
Marchetti MP, Moyle PB, Levine R (2004b) Invasive species
profiling? Exploring the characteristics of non-native
fishes across invasion stages in California. Freshw Biol
49:646–661. https:// doi. org/ 10. 1111/j. 1365- 2427. 2004.
01202.x
Marchetti MP, Lockwood JL, Light TS (2006) Effects of urban-
ization on California’s fish diversity: differentiation,
homogenization and the influence of spatial scale. Biol
Conserv 127:310–318. https:// doi. org/ 10. 1016/j. biocon.
2005. 04. 025
Maxwell JR, Edwards CJ, Jensen ME, Paustian SJ, Parrot H,
Donley M (1995) A hierarchical framework of aquatic
ecological units in North America. U.S. Department of
Agriculture, Forest Service, North Central Forest Experi-
ment Station, Report NC-176, St. Paul, Minnesota
Meyers NM, Reaser JK, Hoff MH (2020) Instituting a national
early detection and rapid response program: needs for
building federal risk screening capacity. Biol Invasions
22:53–65. https:// doi. org/ 10. 1007/ s10530- 019- 02144-0
Mims MC, Olden JD, Shattuck ZR, Poff NL (2010) Life his-
tory trait diversity of native freshwater fishes in North
America. Ecol Freshw Fish 19:390–400. https:// doi. org/
10. 1111/j. 1600- 0633. 2010. 00422.x
Mollot G, Pantel JH, Romanuk TN (2017) The effects of inva-
sive species on the decline in species richness: a global
meta-analysis. Adv Ecol Res 56:61–83. https:// doi. org/
10. 1016/ bs. aecr. 2016. 10. 002
Moreau MA, Coomes OT (2007) Aquarium fish exploitation in
western Amazonia: conservation issues in Peru. Environ
Conserv 34:12–22. https:// doi. org/ 10. 1017/ S0376 89290
70035 66
Moyle PB (2002) Inland Fishes of California (Second ed.).
University of California Press, Berkeley 502 pp
Muñoz-Mas R, Essl F, van Kleunen M, Seebens H, Dawson W,
Casal CMV, García-Berthou E (2023) Two centuries of
spatial and temporal dynamics of freshwater fish intro-
ductions. Glob Ecol Biogeogr 32:1632–1644. https:// doi.
org/ 10. 1111/ geb. 13714
Naylor RL, Kishore A, Sumaila UR, Issifu I, Hunter BP, Belton
B etal (2021) Blue food demand across geographic and
temporal scales. Nat Commun 12:5413. https:// doi. org/
10. 1038/ s41467- 021- 25516-4
Nelson JS (2006) Fishes of the world. John Wiley and Sons,
Incorporated, Hoboken, New Jersey
Nelson JA (2014) Breaking wind to survive: fishes that breathe
air with their gut. J Fish Biol 84:554–576. https:// doi. org/
10. 1111/ jfb. 12323
Nico L, Schofield P (2024) Corydoras aeneus (Gill, 1858):
U.S. Geological Survey, Nonindigenous Aquatic Spe-
cies Database, Gainesville, FL, https:// nas. er. usgs. gov/
queri es/ facts heet. aspx? Speci esID= 336 , Revision Date:
2/2/2012, Peer Review Date: 2/2/2012, Accessed on 17
May 2024
Nico L, Fuller P, Neilson M, Fusaro A, Davidson A, Alame K
etal (2024) Leuciscus idus(Linnaeus, 1758): U.S. Geo-
logical Survey, Nonindigenous Aquatic Species Data-
base, Gainesville, FL, and NOAA Great Lakes Aquatic
Nonindigenous Species Information System, Ann Arbor,
MI, https:// nas. er. usgs. gov/ queri es/ great lakes/ FactS
heet. aspx? Speci es_ ID= 557& Poten tial= Y& Type= 2&
HUCNu mber= , Revision Date: 3/21/2018, Access Date:
5/17/2024
Novák J, Kalous L, Patoka J (2020) Modern ornamental
aquaculture in Europe: early history of freshwater fish
imports. Rev Aquac 12:2042–2060. https:// doi. org/ 10.
1111/ raq. 12421
Olden JD, Poff NL, Bestgen KR (2006) Life-history strate-
gies predict fish invasions and extirpations in the Colo-
rado River Basin. Ecol Monogr 76:25–40. https:// doi.
org/ 10. 1890/ 05- 0330
Olden JD, Lawler JJ, Poff NL (2008) Machine learning meth-
ods without tears: a primer for ecologists. Q Rev Biol
83:171–193. https:// doi. org/ 10. 1086/ 587826
Olden JD, Whattam E, Wood SA (2021) Online auction mar-
ketplaces as a global pathway for aquatic invasive spe-
cies. Hydrobiologia 848:1967–1979. https:// doi. org/ 10.
1007/ s10750- 020- 04407-7
Omernik JM, Griffith GE (2014) Ecoregions of the contermi-
nous United States: evolution of a hierarchical spatial
framework. Environ Manage 54:1249–1266. https://
doi. org/ 10. 1007/ s00267- 014- 0364-1
Page LM, Burr BM (2011) Peterson Field Guide to Freshwa-
ter Fishes of North America North of Mexico, 2nd edn.
Houghton Mifflin Harcourt, Boston
Peoples BK, Goforth R (2017) Commonality in traits and
hierarchical structure of vertebrate establishment suc-
cess. Divers Distrib 23:854–862. https:// doi. org/ 10.
1111/ ddi. 12584
Peoples BK, Midway SR (2018) Fishing pressure and species
traits affect stream fish invasions both directly and indi-
rectly. Divers Distrib 24:1158–1168. https:// doi. org/ 10.
1111/ ddi. 12747
Pepin P, Robert D, Bouchard C, Dower JF, Falardeau M,
Fortier L etal (2014) Once upon a larva: revisiting the
relationship between feeding success and growth in fish
larvae. ICES J Mar Sci 72:359–373. https:// doi. org/ 10.
1093/ icesj ms/ fsu201
Petitpierre B, Kueffer C, Broennimann O, Randin C, Dae-
hler C, Guisan A (2012) Climatic niche shifts are rare
among terrestrial plant invaders. Science 335:1344–
1348. https:// doi. org/ 10. 1126/ scien ce. 12159 33
Pyšek P, Jarošík V, Pergl J (2011) Alien plants introduced by
different pathways differ in invasion success: uninten-
tional introductions as a threat to natural areas. PLoS
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Predicting invasiveness offreshwater fishes imported intoNorth America: regional differences…
Page 25 of 26 107
Vol.: (0123456789)
ONE 6:e24890. https:// doi. org/ 10. 1371/ journ al. pone.
00248 90
Pyšek P, Bacher S, Kühn I, Novoa A, Catford JA, Hulme PE,
Pergl J etal (2020a) MAcroecological framework for
invasive aliens (MAFIA): disentangling large-scale
context dependence in biological invasions. NeoBiota
62:407–461. https:// doi. org/ 10. 3897/ neobi ota. 62. 52787
Pyšek P, Hulme PE, Simberloff D, Bacher S, Blackburn TM,
Carlton JT etal (2020b) Scientists’ warning on invasive
alien species. Biol Rev 95:1511–1534. https:// doi. org/
10. 1111/ brv. 12627
Ribeiro F, Elvira B, Collares-Pereira MJ, Moyle PB (2008)
Life-history traits of nonnative fishes in Iberian water-
sheds across several invasion stages: a first approach.
Biol Invasions 10:89–102. https:// doi. org/ 10. 1007/
s10530- 007- 9112-2
Ricciardi A, Hoopes MF, Marchetti MP, Lockwood JL (2013)
Progress toward understanding the ecological impacts
of nonnative species. Ecol Monogr 83:263–282. https://
doi. org/ 10. 1890/ 13- 0183.1
Rixon CA, Duggan IC, Bergeron NM, Ricciardi A, Macisaac
HJ (2005) Invasion risks posed by the aquarium trade
and live fish markets on the Laurentian Great Lakes. Bio-
divers Conserv 14:1365–1381. https:// doi. org/ 10. 1007/
s10531- 004- 9663-9
Rodríguez-Barreras R, Zapata-Arroyo C, Falcón LW, Olmeda
MDL (2020) An island invaded by exotics: a review of
freshwater fish in Puerto Rico. Neotrop Biodivers 6:42–
59. https:// doi. org/ 10. 1080/ 23766 808. 2020. 17293 03
Rohtla M, Vilizzi L, Kováč V, Almeida D, Brewster B, Britton
JR etal (2021) Review and meta-analysis of the environ-
mental biology and potential invasiveness of a poorly-
studied cyprinid, the ide Leuciscus idus. Rev Fish Sci
Aquac 29:512–548. https:// doi. org/ 10. 1080/ 23308 249.
2020. 18222 80
Romagosa CM, Guyer C, Wooten MC (2009) Contribution of
the live-vertebrate trade toward taxonomic homogeni-
zation. Conserv Biol 23:1001–1007. https:// doi. org/ 10.
1111/j. 1523- 1739. 2009. 01194.x
Roy HE, Peyton J, Aldridge DC, Bantock T, Blackburn TM,
Britton R et al (2014) Horizon scanning for invasive
alien species with the potential to threaten biodiversity
in Great Britain. Glob Chang Biol 20:3859–3871. https://
doi. org/ 10. 1111/ gcb. 12603
Roy M, Belliveau V, Mandrak NE, Gagné N (2018) Develop-
ment of environmental DNA (eDNA) methods for detect-
ing high-risk freshwater fishes in live trade in Canada.
Biol Invasions 20:299–314. https:// doi. org/ 10. 1007/
s10530- 017- 1532-z
Seebens H, Blackburn TM, Dyer EE, Genovesi P, Hulme PE,
Jeschke JM etal (2017) No saturation in the accumula-
tion of alien species worldwide. Nat Commun 8:14435.
https:// doi. org/ 10. 1038/ ncomm s14435
Seebens H, Bacher S, Blackburn TM, Capinha C, Dawson W,
Dullinger S etal (2021) Projecting the continental accu-
mulation of alien species through to 2050. Glob Chang
Biol 27:970–982. https:// doi. org/ 10. 1111/ gcb. 15333
Sinclair JS, Stringham OC, Udell B, Mandrak NE, Leung B,
Romagosa CM etal (2021) The international vertebrate
pet trade network and insights from US imports of exotic
pets. Bioscience 71:977–990. https:// doi. org/ 10. 1093/
biosci/ biab0 56
Smith KF, Behrens M, Schloegel LM, Marano N, Burgiel S,
Daszak P (2009) Reducing the risks of the wildlife trade.
Science 324:594–595. https:// doi. org/ 10. 1126/ scien ce.
11744 60
Smith KM, Zambrana-Torrelio C, White A, Asmussen M,
Machalaba C, Kennedy S etal (2017) Summarizing US
wildlife trade with an eye toward assessing the risk of
infectious disease introduction. EcoHealth 14:29–39.
https:// doi. org/ 10. 1007/ s10393- 017- 1211-7
Sorte CJ, Ibáñez I, Blumenthal DM, Molinari NA, Miller LP,
Grosholz ED etal (2013) Poised to prosper? A cross-sys-
tem comparison of climate change effects on native and
non-native species performance. Ecol Lett 16:261–270.
https:// doi. org/ 10. 1111/ ele. 12017
Strubbe D, Jiménez L, Barbosa AM, Davis AJ, Lens L, Rahbek
C (2023) Mechanistic models project bird invasions with
accuracy. Nat Comm 14:2520. https:// doi. org/ 10. 1038/
s41467- 023- 38329-4
Sturtevant RA, Mason DM, Rutherford ES, Elgin A, Lower E,
Martinez F (2019) Recent history of nonindigenous spe-
cies in the Laurentian Great Lakes; an update to Mills
etal., 1993 (25 years later). J Great Lakes Res 45:1011–
1035. https:// doi. org/ 10. 1016/j. jglr. 2019. 09. 002
Su G, Logez M, Xu J, Tao S, Villéger S, Brosse S (2021)
Human impacts on global freshwater fish biodiversity.
Science 371:835–838. https:// doi. org/ 10. 1126/ scien ce.
abd33 69
Tropical Fish Hobbyist (1979) TFH Publications. Neptune
City, NJ
Tropical Fish Hobbyist (1980) TFH Publications. Neptune
City, NJ
Turbelin AJ, Diagne C, Hudgins EJ, Moodley D, Kourantidou
M, Novoa A etal (2022) Introduction pathways of eco-
nomically costly invasive alien species. Biol Invasions
24:2061–2079
US Geological Survey (2024) Nonindigenous Aquatic Species
Database, Gainesville, FL. http:// nas. er. usgs. gov
Van Wilgen NJ, Richardson DM (2012) The roles of climate,
phylogenetic relatedness, introduction effort, and repro-
ductive traits in the establishment of nonnative reptiles
and amphibians. Conserv Biol 26:267–277. https:// doi.
org/ 10. 1111/j. 1523- 1739. 2011. 01804.x
Vander Zanden MJ, Shuter BJ, Lester N, Rasmussen JB (1999)
Patterns of food chain length in lakes: a stable isotope
study. Am Nat 154:406–416. https:// doi. org/ 10. 1086/
303250
Vejřík L, Vejříková I, Blabolil P, Eloranta AP, Kočvara L,
Peterka J etal (2017) European catfish (Silurus glanis) as
a freshwater apex predator drives ecosystem via its diet
adaptability. Sci Rep 7:15970. https:// doi. org/ 10. 1038/
s41598- 017- 16169-9
Venezia LD, Samson J, Leung B (2017) The rich get richer:
invasion risk across North America from the aquarium
pathway under climate change. Divers Distrib 24:285–
296. https:// doi. org/ 10. 1111/ ddi. 12681
Vilizzi L, Copp GH, Hill JE, Adamovich B, Aislabie L, Akin
D et al (2021) A global-scale screening of non-native
aquatic organisms to identify potentially invasive species
under current and future climate conditions. Sci Total
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J.G.Howeth et al.
107 Page 26 of 26
Vol:. (1234567890)
Environ 788:147868. https:// doi. org/ 10. 1016/j. scito tenv.
2021. 147868
Wittmann ME, Jerde CL, Howeth JG, Maher SP, Deines AM,
Jenkins JA etal (2014) Grass carp in the Great Lakes
region: establishment potential, expert perceptions, and
re-evaluation of experimental evidence of ecological
impact. Can J Fish Aquatic Sci 71:992–999. https:// doi.
org/ 10. 1139/ cjfas- 2013- 0537
WoRMS Editorial Board (2025) World Register of Marine
Species. Available from https:// www. marin espec ies. org
at VLIZ. Accessed on 15 Jan 2025. https:// doi. org/ 10.
14284/ 170
Xu M, Li SP, Dick JT, Gu D, Fang M, Yang Y et al (2022)
Exotic fishes that are phylogenetically close but
functionally distant to native fishes are more likely to
establish. Glob Chang Biol 28:5683–5694. https:// doi.
org/ 10. 1111/ gcb. 16360
Young HS, McCauley DJ, Galetti M, Dirzo R (2016) Patterns,
causes, and consequences of anthropocene defaunation.
Annu Rev Ecol Evol Syst 47:333–358. https:// doi. org/ 10.
1146/ annur ev- ecols ys- 112414- 054142
Publisher’s Note Springer Nature remains neutral with regard
to jurisdictional claims in published maps and institutional
affiliations.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”),
for small-scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are
maintained. By accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use
(“Terms”). For these purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or
a personal subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or
a personal subscription (to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the
Creative Commons license used will apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data
internally within ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking,
analysis and reporting. We will not otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of
companies unless we have your permission as detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that
Users may not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to
circumvent access control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil
liability, or is otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by
Springer Nature in writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer
Nature journal content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates
revenue, royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain.
Springer Nature journal content cannot be used for inter-library loans and librarians may not upload Springer Nature journal
content on a large scale into their, or any other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any
information or content on this website and may remove it or features or functionality at our sole discretion, at any time with or
without notice. Springer Nature may revoke this licence to you at any time and remove access to any copies of the Springer Nature
journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express
or implied with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or
warranties imposed by law, including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be
licensed from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other
manner not expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com