PRIMARY RESEARCH PAPER
Cryptic intercontinental dispersal, commercial retailers,
and the genetic diversity of native and non-native cattails
(Typha spp.) in North America
Claudia Ciotir .Joanna Freeland
Received: 2 June 2015 / Revised: 7 October 2015 / Accepted: 8 October 2015 / Published online: 23 October 2015
ÓSpringer International Publishing Switzerland 2015
Abstract Although cattails (Typha spp.) are impor-
tant components of wetlands around the world, the
three most widespread species (T. angustifolia, T.
domingensis, T. latifolia) are becoming increasingly
dominant. We used global phylogenetic and phylo-
geographic assessments to test the hypotheses that
each species has experienced multiple introductions of
divergent lineages into North America and that
commercial retailers are aiding long-distance dispersal.
Our analyses identiﬁed T. angustifolia as a paraphyletic
species with a highly divergent lineage. We found
evidence for at least one introduced T. angustifolia
lineage in wild populations and garden centres of North
America. Although potentially complicated by incom-
plete lineage sorting, our data suggest dispersal of T.
domingensis between Europe and Australia, and further
investigation should assess a possible introduction of a
non-native T. domingensis lineage into North America.
T. latifolia has experienced bidirectional dispersal
between North America and Europe, and a sample of
T. latifolia purchased in a Canadian garden centre was
an Asian lineage. Interspeciﬁc hybridization and novel
intraspeciﬁc admixture have been repeatedly impli-
cated in biological invasions, includinginvasions by the
hybrid cattail Typha 9glauca, and future work should
focus on the potential contributions of non-native
lineages to regional patterns of invasion by Typha spp.
in North America.
Keywords Typha spp. Intercontinental dispersal
Cryptic introductions Biological invasions
Wetlands Garden centres
Propagule pressure is recognized as a potentially
important determinant of the invasive success of
introduced species, partly because strong propagule
pressure should increase the likelihood of a suit-
able match between genotype and environment in the
introduced range (Simberloff, 2009). In addition, when
multiple provenances are involved, strong propagule
pressure can increase the likelihood of admixture
between formerly allopatric conspeciﬁc lineages that
have become sympatric in their introduced range, and
the resulting evolutionary novelty may facilitate bio-
logical invasions (Kolbe et al., 2004; Lavergne &
Molofsky, 2007). Although multiple founders and high
Handling editor: John Havel
Electronic supplementary material The online version of
this article (doi:10.1007/s10750-015-2538-0) contains supple-
mentary material, which is available to authorized users.
Environmental and Life Sciences Graduate Program,
Trent University, Peterborough, ON K9J 7B8, Canada
J. Freeland (&)
Department of Biology, Trent University, 2140 East Bank
Drive, Peterborough, ON K9J 7B8, Canada
Hydrobiologia (2016) 768:137–150
genetic diversity provide only one explanation for
successful biological invasions (Dlugosch & Parker,
2008), genetic diversity can be an important contribut-
ing factor (Roman & Darling, 2007) and may be
particularly likely to play a role in invasions by species
that have been repeatedly—and sometimes deliber-
ately—introduced to sites around the world.
Plant nurseries provide one important route for
introducing non-native species, including aquatic
macrophytes (Martin & Coetzee, 2011; Hussner,
2012). These macrophytes may include cattails
(species in the genus Typha) which are perennial
semi-aquatic plants important in temperate and tropical
wetlands of the northern and southern hemispheres.
There are approximately 10–15 species of Typha
worldwide (Smith, 2000; Kim & Choi, 2011), occur-
ring in all major land masses except Greenland and
Antarctica (Geze, 1912; Smith, 2000). The three most
widely distributed cattail species, which are also the
only three species growing in the wild in North
America, are Typha angustifolia L. (narrow-leaf cat-
tail), Typha domingensis Pers. (southern cattail), and
Typha latifolia L. (broad-leaf cattail); each of these,
plus their interspeciﬁc hybrids, has in recent decades
become increasingly dominant in multiple geographic
regions (Grace & Harrison, 1986; Beare & Zedler,
1987; Galatowitsch et al., 1999; Parsons & Cuthbert-
son, 2001; Miao, 2004; Freeland et al., 2013; Zapfe &
Although the three species are partially sympatric
(see Figs. 1,2,3for geographic ranges), their latitu-
dinal and altitudinal limits vary. Typha angustifolia
reaches its northern limit around 51°N and grows
throughout temperate North America and Eurasia
(Grace & Harrison, 1986; Smith, 2000; Stevens &
Hoag, 2000) at elevations up to 2000 m (Hickman,
1993). This species was introduced into Australia and
North America (AVH, 2013; Ciotir et al., 2013) and is
considered invasive in parts of North America (Gala-
towitsch et al., 1999). Typha domingensis is limited to
the pantropical latitudes of Eurasia, North, Central, and
South America, southern Paciﬁc Islands, New Zealand,
and Australia (Briggs, 1987; Paczkowska, 1994;
Smith, 2000; Gupta, 2013), reaching its limits at
around 40°N (Smith, 2000) and 1500 m in altitude
(Hickman, 1993; Vibrans, 2004; Stevens & Hoag,
2006). This species is considered introduced and
invasive in Hawaii (USDA-NRCS, 2014) and,
although native to central and southern North America,
is considered invasive in the Florida Everglades
(Newman et al., 1998; Miao, 2004), parts of California
(Beare & Zedler, 1987), and Costa Rica (Osland et al.,
2011). Typha latifolia is more widely distributed and
tolerates more northerly latitudes and higher altitudes
(up to 2300 m) than the other two species (Grace &
Harrison, 1986; Smith, 2000). Typha latifolia has a
native range in temperate latitudes of Europe (USDA-
ARS, 2005), North America (Smith, 2000; Sawada
et al., 2003), Africa (Geze, 1912), and Asia (Kun &
Simpson, 2010) and has been introduced to Australia
(Briggs, 1987; Parsons & Cuthbertson, 2001), New
Zealand (ISSG, 2006), Tasmania (Barnard, 1882), and
the Caribbean (USDA-ARS, 2005). Australia (Parsons
& Cuthbertson, 2001; Zedler & Kercher, 2004), New
Zealand (Champion & Clayton, 2001), Tasmania
(Parsons & Cuthbertson, 2001), and Hawaii (HISC,
2013) consider T. latifolia an invasive species.
A quick search on the internet using the search terms
‘‘Typha’’ and ‘‘nurseries’’ identiﬁed nearly 70,000
websites from around the world, many of which
advertised one or more of T. latifolia, T. angustifolia,
and T. domingensis (along with additional congeneric
species) for sale. This high level of Typha trafﬁcking
suggests that cryptic intercontinental dispersal of Typha
lineages may be more common than realized, which in
turn may have implications for the growing invasive-
ness of Typha spp. In addition, if multiple non-native
lineages of Typha spp. have been introduced into North
America, they could help guide potential avenues of
investigation into phenomena such as regional varia-
tions in the success of non-native Typha (Freeland et al.,
2013), or the transition to invasiveness by native Typha
species (Newman et al., 1998; Miao, 2004).
The primary goal of this study was therefore to test
the hypothesis that multiple non-native lineages of T.
angustifolia, T. domingensis, and T. latifolia have been
introduced into North America. We also tested the
hypothesis that plant nurseries are facilitating long-
distance dispersal of Typha lineages.
Materials and methods
Sampling, DNA extraction, and DNA sequencing
This study was primarily based on T. angustifolia, T.
latifolia, and T. domingensis sequences from three
sources. The ﬁrst source was 82 samples collected
138 Hydrobiologia (2016) 768:137–150
from North America, Europe, and Africa between
2007 and 2013, from which we identiﬁed 42 novel
haplotypes (see below). Throughout this study, hap-
lotype refers to a speciﬁc cpDNA sequence. The
second source was eight plant nurseries in Canada
(Ontario and British Columbia) from which we
purchased 25 plants. Nineteen of these purchases
were identiﬁed by the nurseries as T. angustifolia, one
as T. latifolia, four as Typha laxmannii Lepech., and
one as Typha minima Funck; these are the only four
Typha species that we found for sale in Canada, and
border restrictions meant that we were unable to order
plants from nurseries in the USA or other countries.
The third source was 38 sequences previously pub-
lished by Kim & Choi (2011) which comprise 21
unique haplotypes. In addition, we obtained from Kim
& Choi (2011), and from some additional samples that
we collected, a total of 16 haplotypes from six
additional species in the genus Typha, and ﬁve
haplotypes from ﬁve Sparganium spp., for use as
outgroups in our phylogenetic analyses (see below).
The genus Sparganium was chosen as an outgroup
because it is the closest sister group to the genus Typha
(Bremer, 2000). Sample locations are provided in
Figs. 1,2, and 3, and Table S1. For the newly acquired
sequences, a leaf tip of approximately 2 cm from each
sampled plant was dried in a bag of silica beads (Fisher
Scientiﬁc) and stored at -20°C on return to the lab.
Dried plant material was ground in a mixer mill
(Retsch MM300; Retsch, Newtown, Pennsylvania,
USA), and genomic DNA was extracted from 10 to
30 mg dried tissue per plant using the E.Z.N.A spin
Fig. 1 Map of T. angustifolia indicating the sampling sites
represented by black circles and the contemporary distribution
areas represented in dark grey. Distribution information from
WCSP (2012), Kim & Choi (2011), and Ciotir et al. (2013).
Solid lines with arrows identify approximate pathway of
intercontinental movement of naturalized plants, and small
dashes indicate the approximate provenance of samples
obtained from garden centres in Canada (see text and
Table S1 for details)
Hydrobiologia (2016) 768:137–150 139
column DNA plant mini kits (Omega Bio-Tek,
Georgia, USA), eluted into a ﬁnal volume of 100 ll.
From each collected or purchased sample, three
chloroplast DNA (cpDNA) regions were ampliﬁed:
trnL–trnF(trnL gene, trnL intron, and trnL–Finter-
genic spacer), trnC–petN (intergenic spacer), and
psbM–trnD (intergenic spacer). The trnL–trnF region
was ampliﬁed using the primers ‘c’ and ‘f’ described in
Taberlet et al. (1991). The trnC–petN and psbM–trnD
regions were ampliﬁed using primers developed by
Kim & Choi (2011). Each PCR reaction included 19
Taq reaction buffer (UBI Life Sciences), 2 mM
, 0.2 mM dNTPs, 20 lM of each primer, 2.5U
HP Taq (UBI Life Sciences), 1.2 ll (100%) of BSA, and
approximately 10 ng DNA in a total volume of 25 ll.
PCR ampliﬁcations were performed in a Mastercycler
epgradient thermal cycler (Eppendorf) (Table 1), and
PCR products were puriﬁed by incubating with 10U
exonuclease I and 2U shrimp alkaline phosphatase
(Fermentas International, Inc.) for 15 min at 15°C
followed by 15 min at 80°C. Puriﬁed products were
sequenced in both directions using a BigDye Termina-
tor v3.1 Cycle Sequencing Kit (Applied Biosystems,
Foster City, California, USA). Sequences were run on a
DNA Analyser 3730xl ABI (Applied Biosystems).
The plants acquired from nurseries that were
identiﬁed as either T. angustifolia or T. latifolia were
additionally genotyped at microsatellite loci TA3,
TA5, TA8, and TA20 (Tsyusko-Omeltchenko et al.,
2003), which have alleles that are speciﬁc to either T.
latifolia or T. angustifolia (Snow et al., 2010; Kirk
et al., 2011) and can therefore be used to discriminate
T. angustifolia, T. latifolia, and T. 9glauca; the latter
is the invasive hybrid of T. angustifolia and T. latifolia
that dominates the Great Lakes and St. Lawrence
Seaway regions of North America (Olson et al., 2009;
Travis et al., 2010; Freeland et al., 2013).
Sequence alignments and data analysis
De novo sequences were assembled and edited in
Seqman (DNASTAR LaserGene 5, Madison, WI,
Fig. 2 Map of T. domingensis indicating the sampling sites
represented by black circles and the contemporary distribution
areas represented in dark grey.Dotted lines show potential
pathways of long-distance dispersal, although direction of
movements could not be determined (see text and Table S1
for details). Distribution information from WCSP (2012) and
Kim & Choi (2011)
140 Hydrobiologia (2016) 768:137–150
USA) and aligned with the previously published Typha
and Sparganium sequences (Table S1). All sequences
were imported into BioEdit Software (Hall, 1999).
Each region was aligned using CLUSTALW (Thomp-
son et al., 1994) with subsequent manual adjustments,
based on 55 sequences of each chloroplast region from
T. angustifolia,22fromT. domingensis, and 63 from T.
latifolia (Table S1). The three cpDNA regions were
then appended in a single 3147 bp alignment.
Duplicate sequences were removed, leaving 10 haplo-
types from T. angustifolia (three novel, and seven from
Kim & Choi, 2011), 15 haplotypes from T. domingen-
sis (11 novel, and four from Kim& Choi, 2011), and 38
haplotypes from T. latifolia (28 novel, and 10 from
Kim & Choi, 2011). Phylogenetic analyses were
performed on these haplotypes plus an additional 35
haplotypes from T. capensis Rohrb., T. elephantine
Roxb., T. laxmannii Lepech., T. minima Lepech.,T.
Fig. 3 Map of T. latifolia indicating the sampling sites
represented by black circles and the contemporary distribution
areas represented in dark grey.Solid lines with arrows identify
approximate pathway of intercontinental movements of
naturalized plants, and small dashes indicate the approximate
provenance of samples obtained from garden centres in Canada
(see text and Table S1 for details). Distribution information
from WCSP (2012), Kim & Choi (2011), and Ciotir et al. (2013)
Table 1 PCR conditions used for amplifying the three cpDNA regions
Region PCR conditions
trnL–trnF94°C/120 s 94°C/60 s 55°C/60 s 72°C/120 s 72°C/420 s 32
trnC–petN95°C/240 s 95°C/60 s 59.5°C/60 s 72°C/90 s 72°C/420 s 40
psbM–trnD94°C/240 s 94°C/60 s 58°C/60 s 72°C/90 s 72°C/600 s 40
Hydrobiologia (2016) 768:137–150 141
shuttleworthii Koch & Sond., T. orientalis C.Presl.,
and the outgroup Sparganium (Table S1).
Bayesian inference was used to reconstruct the
phylogeny of Typha spp. in MRBAYES ver. 3.2
(Huelsenbeck & Ronquist, 2001). Three Bayesian runs
were generated with independent partitions for trnL–
trnF, trnC–petN, and psbM–trnD according to their
alignment lengths and models. The default prior and
likelihood settings were used for all parameters except
for the nucleotide substitution models which were set to
GTR ?I for the trnL–trnF and psbM–trnD regions
(with alignment lengths of 1006 bp and 1129 bp,
respectively), and GTR ?G for the trnC–petN region
(alignment length of 1012 bp). Each analysis included
77 sequences, of which ﬁve represented outgroups.
Models of nucleotide sequence substitution were
generated by MrModeltest ver. 2.3 based on all variable
sites and the good ﬁt of Akaike Information Criterion
(AIC) (Nylander, 2004). Substitution model parameters
and rates of substitution were allowed to vary across
partitions using ratepr =variable and the ‘unlink’
command. MCMC chain analyses were run for up to
50,000,000 iterations, sampling trees once every 1,000
iterations. Convergence of the four Markov chains and
assessment of ‘burn-in’ values were determined by
examining the average standard deviation of split
frequencies and by plotting the likelihood values
against the number of generations on a linear regression
graph. Using a burn-in value of 10% of the total number
of trees, we generated the consensus tree and posterior
probability values. Multiple output ﬁles were examined
to assess the convergence between samples in Tracer
ver.1.5. (Rambaut & Drummond, 2008). The consensus
trees were visualized in FigTree ver.1.3.1. (Drummond
& Rambaut, 2007).
Concatenated alignments of all sequences (i.e.
duplicates were returned to the dataset) of each of T.
angustifolia,T. domingensis, and T. latifolia were used
to generate parsimony networks in TCS ver. 1.21
(Clement et al., 2000). The analyses were set with 95%
statistical parsimony and gaps as the ﬁfth state
(Templeton & Singh, 1993).
The Bayesian 50% majority rule consensus tree
reconstructed T. angustifolia and T. domingensis as
paraphyletic or polyphyletic species, and T. latifolia as
a monophyletic species (Fig. 4). Typha angustifolia is
divided into one clade that comprises multiple lin-
eages, plus one divergent haplotype (E4/N1/P3) that
was found in both Europe and North America. The
former has two clusters, the ﬁrst of which includes
three Asian, one North American, and two purchased
(from Canadian nurseries) haplotypes, plus one haplo-
type that is dispersed across multiple continents (Asia,
Europe, North America, and Australia). The second
cluster comprises two European haplotypes and one
Asian haplotype. The divergent haplotype of T.
angustifolia (E4/N1/P3) groups with T. capensis,T.
elephantina, and T. domingensis, with T. capensis as its
nearest neighbour. These relationships are supported in
all analyses by high posterior probability values (pp:
0.71–1). In T. domingensis, ﬁve basal lineages repre-
sented by two North American haplotypes (N2, N3),
one African (Af2), and two Asian (A2, A3) haplotypes
are intermediate to T. elephantina and the remaining T.
domingensis sequences, which form a single clade
comprising haplotypes from Asia, Africa, Europe, and
North America. The T. domingensis topology is
supported by high posterior probability values (pp:
0.93–0.99). One haplotype is shared between Europe
and Australia (E4/Au). For T. latifolia, the phylogeny
recovered three broad clusters: one from North Amer-
ica, one from Asia and eastern Europe, and the third
primarily from western Europe (Fig. 4). All relation-
ships within T. latifolia were supported by high
posterior probability values (pp: 0.99–1). The relative
extent of divergence among haplotypes inferred from
the phylogenetic tree was largely reﬂected in each of
the haplotype networks (Figs. 5,6,7).
In both the phylogenetic tree and the haplotype
networks, samples purchased from nurseries in
Canada have the haplotype preﬁx ‘P’ (Figs. 4,5,6,
7). With the exception of three plants identiﬁed as T.
angustifolia that turned out to be T. latifolia, each
purchased plant had a haplotype sequence that agreed
with the retailer’s taxonomic identiﬁcation; however,
the microsatellite data identiﬁed ten of the plants
labelled as T. angustifolia as the hybrid T. 9glauca.
Purchased samples of T. angustifolia comprised
haplotypes from the two divergent genetic lineages,
and the purchased T. latifolia plant represented an
Asian lineage. The purchased T. minima matched a
haplotype previously identiﬁed from Europe, and the
purchased T. laxmannii matched a haplotype previ-
ously identiﬁed in both Europe and Asia (Table S1).
142 Hydrobiologia (2016) 768:137–150
Fig. 4 The 50% majority rule consensus tree from Bayesian
analyses based on combined cpDNA sequences of Typha and
Sparganium species. Numbers on nodes represent the posterior
probability support values. Haplotypes are labelled with letters
that indicate provenance: AAsia, EEurope, Af Africa, Au
Australia, NNorth America, Ppurchased samples (purchased
from plant nurseries in Canada). Numbers in brackets represent
the numbers of individuals from each location. T. latifolia
haplotypes E4, E5, and E6 are from eastern Europe, and E1, E2,
E3, E7, E8, and E9 are from western Europe. See Table S1 for
more detailed information on locations
Hydrobiologia (2016) 768:137–150 143
Although globally distributed species are among those
most likely to become invasive due to their high
dispersal and colonizing capacity (Nikulina et al.,
2007), global movements are often underestimated
due to cryptic introductions (Lavoie et al., 1999;
Geller et al., 2010). In this study, we found evidence
that the designation of native and non-native Typha
taxa in North America is complicated by cryptic
introductions of non-native lineages and, in the case of
T. angustifolia, by taxonomic uncertainty. Below, we
will discuss our ﬁndings and the associated implica-
tions for each species.
Our results clearly show that T. angustifolia is
polyphyletic, with one divergent haplotype (E4/N1/
P3) more closely related to T. domingensis,T. capensis,
and T. elephantina than to its conspeciﬁc lineages. This
polyphyly could potentially be explained by inter-
speciﬁc hybridization (either ancient or recent) with T.
capensis and subsequent introgression of the chloro-
plast genome; previous examples of introgression
following ancient hybridization have been found in
diverse taxa (e.g. Gross & Rieseberg, 2005; Klymus
et al., 2010). Alternatively, haplotype E4/N1/P3 may
belong to a cryptic species that closely resembles T.
Fig. 5 TCS network for T. angustifolia. Haplotype labels:
AAsia, EEurope, Af Africa, Au Australia, NNorth America,
Ppurchased samples (purchased from plant nurseries in Canada).
Haplotypes are coloured according to their actual or inferred
origin: white North America, light grey Asia, dark grey Europe.
Haplotypes present in multiple continents are marked with a thick
line.Sizesofcircles are proportional to haplotype frequencies.
See Table S1 for more detailed information on locations
Fig. 6 TCS network for T.
labels: AAsia, EEurope, Af
Africa, Au Australia,
(purchased from plant
nurseries in Canada).
Haplotypes are coloured
according to their actual or
inferred origin: white North
America, light grey Asia,
dark grey Europe.
Haplotypes present in
multiple continents are
marked with a thick line.
Sizes of circles are
proportional to haplotype
frequencies. See Table S1
for more detailed
information on locations
144 Hydrobiologia (2016) 768:137–150
angustifolia. Polyphyly and cryptic species have often
been identiﬁed from molecular data in taxa with
cosmopolitan distributions (e.g. Bickford et al., 2007;
Lemmon et al., 2007; Nikulina et al., 2007). Further
investigation is necessary before we can explain this
highly divergent haplotype.
The remaining T. angustifolia haplotypes comprise
Asian, European, and North American haplotypes. A
previous study based on microsatellite genotypes
suggested that Typha angustifolia was introduced to
North America from Europe (Ciotir et al., 2013), and
the current study agrees with this conclusion because
we found three haplotypes in North America (N2, E4/
N1/P3, A1/E1/N3/Au) that are sufﬁciently divergent
to represent at least one introduction of a non-native
haplotype (Fig. 5), and possibly multiple introduc-
tions. Because it is embedded within the predomi-
nantly Asian cluster (Fig. 5), one of the T. angustifolia
haplotypes introduced into North America (A1/E1/
N3/Au) seems also to have cryptically moved from
Asia to Europe, where T. angustifolia is considered
native (Geze, 1912; Cirujano, 2002). That same
Fig. 7 TCS network for T. latifolia. Haplotype labels: AAsia,
EEurope, Af Africa, Au Australia, NNorth America,
Ppurchased samples (purchased from plant nurseries in
Canada). Haplotypes E4, E5, and E6 were from eastern Europe
(east of longitude 13°220.127.116.1100), while haplotypes E1, E2, E3,
E7, E8, and E9 were from western Europe (west of longitude
13°01.39.1500). Haplotypes are coloured according to their
actual or inferred origin: white North America, light grey Asia,
dark grey Europe. Haplotypes present in multiple continents are
marked with a thick line. Sizes of circles are proportional to
haplotype frequencies. See Table S1 for more detailed
information on locations
Hydrobiologia (2016) 768:137–150 145
haplotype (A1/E1/N3/Au) was also blatantly intro-
duced into Australia, where T. angustifolia was ﬁrst
recorded in 1943 (AVH, 2013).
Formally differentiating between the contributions
of processes such as dispersal, vicariance, and lineage
sorting to the current distributions of haplotypes is
challenging in species with complex evolutionary
histories (e.g. Kodandaramaiah, 2010). Typha angus-
tifolia should be considered one example of such a
species because it has repeatedly experienced both
vicariance and dispersal throughout its history and
also hybridizes with multiple congeneric species (see
Kim & Choi, 2011, and references therein). In
addition, samples for widespread species such as T.
angustifolia are unlikely to represent the entire
species’ distribution. Despite these complications,
incomplete lineage sorting must be considered as a
possible explanation for haplotypes sharing between
continents. Nevertheless, we do not believe that
incomplete lineage sorting can explain the patterns
that we found in T. angustifolia, partly because this
study is based on haplotypes that reﬂect [3000 bp of
the chloroplast genome, which evolves relatively
rapidly in Typha (Guisinger et al., 2010); this high
resolution, combined with the relatively rapid lineage
sorting that occurs in plastid versus nuclear DNA
markers, means that identical cpDNA haplotypes
shared a relatively recent common ancestor. Further-
more, intercontinental dispersal of T. angustifolia is
unequivocal in locations from which it was previously
absent, including Australia (AVH, 2013). Finally, as
noted above, a previous study used microsatellite
genotypes to conclude that T. angustifolia had dis-
persed from Europe to North America (Ciotir et al.,
2013). Therefore, although constraints imposed by our
study species leave us unable to unequivocally
demonstrate that lineage sorting was complete prior
to more recent episodes of intercontinental dispersal,
several lines of evidence collectively suggest that
incomplete lineage sorting cannot be the sole expla-
nation for our conclusion of long-distance dispersal in
Our data therefore reﬂect at least one introduction of
a non-native T. angustifolia lineage into North Amer-
ica. In addition, we found three non-native T. angus-
tifolia haplotypes in plants obtained from garden
centres, and one of these haplotypes is found in natural
populations. The remaining two nursery haplotypes,
which are of Asian provenance, have not been found in
the wild, but are very closely related to the widespread
non-native A1/E1/N3/Au haplotype that is also found
in North America, and more exhaustive sampling in
North America is required before we can conclude that
these lineages have not been naturalized at any
location. We cannot exclude the possibility that garden
centres obtained their plants from natural populations,
although this seems much less likely for haplotypes
that have not been found in the wild. Regardless, it is
notable that we had a T. angustifolia plant delivered to
us in Ontario through an online order from a supplier in
British Columbia, a transaction that led to a dispersal of
a non-native T. angustifolia haplotype across a distance
of approximately 4500 km. Less dramatically, a recent
survey of garden owners in Ontario identiﬁed a mean
Euclidean distance of *40 km between sites of
purchased plants (including invasive species) and their
destinations, but there were also rare long-distance
dispersal events of up to 500 km between purchase
location and destination (Marson et al., 2009). Com-
mercial suppliers are therefore facilitating the dispersal
of non-native lineages regardless of whether they were
the initial source of the introductions. Finally, the sale
by nurseries of the hybrid T. 9glauca, labelled as T.
angustifolia, could also help with the spread of the
hybrid which is dominant in the Great Lakes region
(Travis et al., 2010; Kirk et al., 2011; Freeland et al.,
Typha domingensis is paraphyletic, with several North
American, Asian, and African haplotypes (Af2, A2,
A3, N2, N3) marginally more closely related to T.
elephantina than to the remaining T. domingensis
haplotypes, which form a monophyletic group
(Fig. 4). The proximity of the paraphyletic clades to
one another suggests that paraphyly in this case may
be explained by incomplete lineage sorting following
speciation. The parsimony network reveals three
lineages in North America that have a fairly high
degree of relatedness (N2, N3, N4), and which
therefore likely shared a recent common provenance,
although it is unclear whether these three haplotypes
were more likely to have dispersed from Africa (their
nearest neighbour) to North America or vice versa. A
third lineage that is also present in North America (N1)
shows substantial divergence from the other two North
American lineages, which suggests that at least one
146 Hydrobiologia (2016) 768:137–150
non-native lineage of T. domingensis may have been
introduced into North America. There also seems to
have been at least one dispersal event of T. domingensis
between Europe and Australia (E4/Au) (Fig. 2). Typha
domingensis is introduced and invasive in Hawaii
(HISC, 2013), and invasive in Florida (Zedler &
Kercher, 2004), southern California (Beare & Zedler,
1987), and Costa Rica (Osland et al., 2011). Cryptic
intercontinental dispersal of T. domingensis lineages to
North America could help explain the invasiveness of
putatively native populations, although the lack of
continental clustering for this species, combined with
paraphyly at the species level (Fig. 4), means that we
cannot rule out a potential inﬂuence of incomplete
lineage sorting on the patterns that we found. We are
therefore more cautious about inferring intercontinen-
tal dispersal in T. domingensis, and conclude that
although dispersal has likely occurred to some extent
(e.g. between Europe and Australia, based on an
identical haplotype found at geographically distant
sites), additional markers should be employed in future
studies to further investigate the phylogeographic
history of this species.
The T. latifolia haplotypes identiﬁed a single mono-
phyletic group with continental phylogeographic struc-
turing: the three subclades represent North America,
western Europe, and a combination of eastern Europe
and Asia (Fig. 4). The continental phylogeographic
structuring is also evident in the network, which shows
a distinct separation between two clusters: Asia and
eastern Europe versus North America and western
Europe (Fig. 7). The phylogenetic tree suggests that
haplotype E9/N1 was introduced into North America
from Europe, as it is the only North American haplotype
within a European cluster on the phylogenetic tree
(Fig. 4). In addition, the network suggests that haplo-
type E13 was introduced into Europe from North
America, a conclusion that is supported by an earlier
study that used microsatellite data to suggest that T.
latifolia colonized Europe from North America (Ciotir
et al., 2013); these lines of evidence collectively
identify bidirectional intercontinental dispersal of T.
latifolia between North America and Europe. Eastern
European haplotypes share relatively recent ancestry
with Asian haplotypes, reﬂecting dual colonization of
Europe by T. latifolia: once from Asia into eastern
Europe, and once from North America into western
Europe (Fig. 7). In addition, North America is the most
plausible source of Australian T. latifolia, where it is
considered an introduced and invasive species (Smith,
2000; Parsons & Cuthbertson, 2001; ISSG, 2006).
Incomplete lineage sorting is unlikely to explain
inferred patterns of intercontinental dispersal in T.
latifolia because of the same arguments that were
applied to T. angustifolia, plus a strong pattern of
continental phylogeographic structuring (Figs. 4,7).
We also determined that a T. latifolia plant
purchased from a Canadian retailer was an Asian
lineage, which again raises questions about the
potential role that nurseries may play in facilitating
the long-distance dispersal of non-native lineages, and
also about the authenticity of some ‘native plant’
claims by retailers. This non-native lineage has not
been found in any natural North American T. latifolia
populations, and the likelihood of its naturalization is
difﬁcult to evaluate, particularly when we do not know
how long it has been marketed by North American
nurseries; however, there are many precedents of
Asian invasive species and lineages becoming inva-
sive in North America (e.g. Reichard & White, 2001).
T. latifolia has increased its geographic distribution in
North America in recent decades (Shih & Finklestein,
2008), and if increasingly dominant behaviour is noted
within this species, researchers should monitor for the
potential involvement of non-native lineages, partic-
ularly those (such as the Asian lineage we recovered)
that can be plausibly introduced through garden
centres and water gardens.
In conclusion, our data strongly suggest cryptic
intercontinental dispersal of genetic lineages in both T.
angustifolia and T. latifolia, which has resulted in the
introduction of non-native lineages into North Amer-
ica. Our conclusions regarding T. domingensis are
more circumspect because of the potential inﬂuence of
incomplete lineage sorting, which should be further
investigated in future studies using additional markers.
We have also reiterated the known fact that commer-
cial retailers play a role in the global dispersal of plants
(e.g. Reichard & White, 2001; Mack & Lonsdale,
2007; Martin & Coetzee, 2011), both within and
between continents; further investigation into nurs-
eries as potential conduits of non-native Typha
lineages should include commercial samples from
multiple countries and continents. Additionally, our
data revealed that T. angustifolia is rendered
Hydrobiologia (2016) 768:137–150 147
polyphyletic by a highly divergent lineage, and further
investigation will be required in order to determine
whether the domination of wetlands by T. angustifolia
and its hybrid T. xglauca in North America involves
one or both of the divergent, polyphyletic T. angus-
tifolia lineages. Wetlands are particularly susceptible
to biological invasions, including invasions by species
in the genus Typha (Mills et al., 1993; Galatowitsch
et al., 1999; Brinson & Malvarez, 2002; Zedler &
Kercher, 2004). Hybrids and novel intraspeciﬁc
admixtures have been repeatedly implicated in bio-
logical invasions (Schierenbeck & Ellstrand, 2009),
including invasions by the hybrid Typha 9glauca
(Freeland et al., 2013), and future work should use
variable nuclear markers to infer speciﬁc pathways of
introductions and investigate the potential contribu-
tions of non-native lineages to regional patterns of
invasion by Typha spp. in North America.
Acknowledgments We are very grateful to Serena Caplins,
Jennifer Coughlan, Sarah Dungan, Heather Kirk, Jonathan
Mitchley Nicole Vachon, Morgan Wehtje, Walter Wehtje, Stan
Yavno, and Chris Yesson for collecting cattail samples during
their ﬁeld trips in North America, Europe, and Africa. We are
grateful to Dr. Jonathan Mitchley who provided shelter,
transportation, and logistics during the ﬁeldtrip in the UK and
Europe. Special thanks to Dr. Chris Yesson for his contribution to
the global map distributions, and for insightfulbiogeographic and
phylogenetic advice. Many thanks to the Jack Matthews
international scholarship at Trent University for covering the
costs of the European ﬁeld trip. This study was also funded by a
Natural Sciences and Engineering Council (NSERC) Discovery
Grant to JF, and Trent University.
AVH: Australia’s Virtual Herbarium. Available at: http://avh.
chah.org.au. Accessed on August 12, 2013.
Barnard, J., 1882. Economic value of the aquatic plant Typha
latifolia. Papers & Proceedings of the Royal Society of
Tasmania 17: 163–167.
Beare, P. A. & J. B. Zedler, 1987. Cattail invasion and persis-
tence in a coastal salt-marsh – the role of salinity reduction.
Estuaries 10: 165–170.
Bickford, D., D. J. Lohman, N. S. Sodhi, P. K. L. Ng, R. Meier,
K. Winker, K. K. Ingram & I. Das, 2007. Cryptic species as
a window on diversity and conservation. Trends in Ecology
and Evolution 22: 148–155.
Briggs, B.G., 1987. Flora of Australia Online eds. Available
Accessed June 15, 2013.
Brinson, M. M. & A. I. Malvarez, 2002. Temperate freshwater
wetlands: types, status, and threats. Environmental Con-
servation 29: 115–133.
Champion P. D. C. & J. S. Clayton, 2001. Border control for
potential aquatic weeds. Stage 2. Weed risk assessment. In:
(ed. Department of Conservation, Wellington, New Zeal-
and) 185: 1–30. Department of Conservation Science
Publications, Science & Research Unit.
Ciotir, C., H. Kirk, J. Row & J. R. Freeland, 2013. Interconti-
nental dispersal of Typha angustifolia and T. latifolia
between Europe and North America has implications for
Typha invasions. Biological Invasions 15: 1377–1390.
Cirujano, S., 2002. Typhaceae. In Bolibar, S. C. (ed), Flora
Iberica. Real Jardı
´nico, Vol. 18. CSIC, Madrid.
Clement, M., D. Posada & K. A. Crandall, 2000. TCS: a com-
puter program to estimate gene genealogies. Molecular
Ecology 9: 1657–1659.
Dlugosch, K. M. & I. M. Parker, 2008. Founding events in
species invasions: genetic variation, adaptive evolution,
and the role of multiple introductions. Molecular Ecology
Drummond, A. J. & A. Rambaut, 2007. BEAST: Bayesian
evolutionary analysis by sampling trees. BMC Evolution-
ary Biology 7: 8.
Freeland, J. R., C. Ciotir & H. Kirk, 2013. Regional differences
in the abundance of native, introduced, and hybrid Typha
spp. in northeastern North America inﬂuence wetland
invasions. Biological Invasions 15: 2651–2665.
Galatowitsch, S. M., N. O. Anderson & P. D. Ascher, 1999.
Invasiveness in wetland plants in temperate North Amer-
ica. Wetlands 19: 733–755.
Geller, J. B., J. A. Darling & J. T. Carlton, 2010. Genetic per-
spectives on marine biological invasions. Annual Reviews
in Marine Science 2: 367–393.
Geze, J. B., 1912. Etudes botanique et agronomique sur les
Typha et quelques autre plantes palustres. Socie
yme D’Imprimerie de Villefranche-de-Rouergue.
Grace, J. B. & J. S. Harrison, 1986. The biology of Canadian
Weeds. 73. Typha latifolia L., Typha angustifolia L. and
Typha 9glauca Godr. Canadian Journal of Plant Sciences
Gross, B. L. & L. H. Rieseberg, 2005. The ecological genetics of
homoploid hybrid speciation. Journal of Heredity 96:
Guisinger, M. M., T. W. Chumley, J. V. Kuehl, J. L. Boore, R.
K. Jansen, 2010. Implications of the plastid genome
sequence of Typha (Typhaceae, Poales) for understanding
genome evolution in Poaceae. Journal of Molecular Evo-
lution 70: 149–156.
Gupta, A. K., 2013. Typha domingensis. In: IUCN 2013. IUCN
Red List of Threatened Species version 2013.1. www.
iucnredlist.org. Accessed on August 12, 2013.
Hall, T. A., 1999. Bioedit: a user friendly biological sequence
alignment editor and analysis program for Windows
95/98NT. Nucleic Acids Symposium Series 41: 95–98.
Hawaii Invasive Species Council (HISC) Available at: http://
sed on August 12, 2013.
Hickman, J. C., 1993. The Jepson Manual. Higher Plants of
California. University of California Press, Berkeley.
Huelsenbeck, J. P. & F. Ronquist, 2001. MRBAYES: Bayesian
inferenceof phylogenetic trees.Bioinformatics 17: 754–755.
Hussner, A., 2012. Alien aquatic plant species in European
countries. Weed Research 52: 297–306.
148 Hydrobiologia (2016) 768:137–150
ISSG: Global Invasive Species Database, 2006. Typha latifolia.
Available at: http://www.issg.org/database/species/ecology.
Kim, C. & H. K. Choi, 2011. Molecular systematics and char-
acter evolution of Typha (Typhaceae) inferred from
nuclear and plastid DNA sequence data. Taxon 60:
Kirk, H., C. Connolly & J. R. Freeland, 2011. Molecular genetic
data reveal hybridization between Typha angustifolia and
Typha latifolia across a broad spatial scale in eastern North
America. Aquatic Botany 95: 189–193.
Klymus, K. E., S. C. Humfeld, V. T. Marshall, D. Cannatella &
H. C. Gerhardt, 2010. Molecular patterns of differentiation
in canyon treefrogs (Hyla arenicolor): evidence for intro-
gressive hybridization with the Arizona treefrog
(H.wrightorum) and correlations with advertisement call
differences. Journal of Evolutionary Biology 23:
Kolbe, J. J., R. E. Glor, L. R. G. Schettino, A. C. Lara, A. Larson
& J. B. Losos, 2004. Genetic variation increases during
biological invasion by a Cuban lizard. Nature 431:
Kodandaramaiah, U., 2010. Use of dispersal–vicariance analy-
sis in biogeography – a critique. Journal of Biogeography
Kun, S. & D. Simpson, 2010. Flora of China, Typhaceae, 23,
161-163. Available at: http://ﬂora.huh.harvard.edu/china/
Accessed on January 07, 2014.
Lavergne, S. & J. Molofsky, 2007. Increased genetic variation
and evolutionary potential drive the success of an invasive
grass. Proceedings of the National Academy of Sciences
Lavoie, D. M., L. D. Smith & G. M. Ruiz, 1999. The potential
for intracoastal transfer of non-indigenous species in the
ballast water of ships. Estuarine, Coastal and Shelf Science
Lemmon, E. M., A. R. Lemmon, J. T. Collins, J. A. Lee-Yaw &
D. C. Cannatella, 2007. Phylogeny-based delimitation of
species boundaries and contact zones in the trilling chorus
frogs (Pseudacris). Molecular Phylogenetics and Evolu-
tion 44: 1068–1082.
Mack, R. N. & W. M. Lonsdale, 2007. Humans as global plant
dispersers: getting more than we bargained for. Bioscience
Marson, D., B. Cudmore, D.A.R. Drake & N.E. Mandrak, 2009.
Summary of a survey of water garden owners in Canada.
Canadian manuscript report of ﬁsheries and aquatic sci-
ences 2906: vi, 23p.
Martin, G. D. & J. A. Coetzee, 2011. Pet stores, aquarists and the
internet trade as modes of introduction and spread of
invasive macrophytes in South Africa. Water SA 37:
Miao, S. L., 2004. Rhizome growth and nutrient resorption:
mechanisms underlying the replacement of two clonal
species in Florida Everglades. Aquatic Botany 78:
Mills, E. L., J. H. Leach, J. T. Carlton & C. L. Secor, 1993.
Exotic species in the Great-Lakes – a history of biotic crises
and anthropogenic introductions. Journal of Great Lakes
Research 19: 1–54.
Newman, S., J. Schuette, J. B. Grace, K. Rutchey, T. Fontaine,
K. R. Reddy & M. Pietrucha, 1998. Factors inﬂuencing
cattail abundance in the northern Everglades. Aquatic
Botany 60: 265–280.
Nikulina, E. A., R. Hanel & P. Schafer, 2007. Cryptic speciation
and paraphyly in the cosmopolitan bryozoan Electra pilosa
– Impact of the Tethys closing on species evolution.
Molecular Phylogenetics and Evolution 45: 765–776.
Nylander, J. A., 2004. MrModelTest, version 2.3. Evolutionary
Biology Centre, Uppsala University, Uppsala, Sweden.
Available at: https://github.com/nylander/MrModeltest2.
Olson, A., J. Paul, J. R. Freeland, 2009. Habitat preferences of
cattail species and hybrids (Typha spp.) in eastern Cana-
da. Aquatic Botany 91: 67–70.
Osland, M. J., E. Gonzalez & C. J. Richardson, 2011. Restoring
diversity after cattail expansion: disturbance, resilience,
and seasonality in a tropical dry wetland. Ecological
Applications 21: 715–728.
Paczkowska, G., 1994. Typha domingensis in Western Aus-
tralian Herbarium (1998–). FloraBase—the Western Aus-
tralian Flora. Department of Parks and Wildlife eds.
Available at: http://ﬂorabase.dpaw.wa.gov.au/.
Parsons, W. T. & E. G., Cuthbertson, 2001 Family Typhaceae.
Noxious Weeds of Australia. CSIRO Publishing, Victoria.
Rambaut, A. & A. J. Drummond, 2008. Tracer version 1.5.
Available at: http://tree.bio.ed.ac.uk/software/tracer/.
Reichard, S. H. & P. White, 2001. Horticulture as a pathway of
invasive plant introductions in the United States. BioS-
cience 51: 103–113.
Roman, J. & J. A. Darling, 2007. Paradox lost: genetic diversity
and the success of aquatic invasions. Trends in Ecology and
Evolution 22: 454–464.
Sawada, M., A. E. Viau, K. Gajewski, 2003. The biogeography
of aquatic macrophytes in North America since the last
glacial maximum. Journal of Biogeography 30: 999–1017.
Schierenbeck, K. A. & N. C. Ellstrand, 2009. Hybridization and
the evolution of invasiveness in plants and other organisms.
Biological Invasions 11: 1093–1105.
Shih, J. G. & S. A. Finklestein, 2008. Range dynamics and
invasive tendencies in Typha Latifolia and Typha angus-
tifolia in eastern North America derived from herbarium
and pollen records. Wetlands 28: 1–16.
Simberloff, D., 2009. The role of propagule pressure in bio-
logical invasions. Annual Review of Ecology and Sys-
tematics 40: 81–102.
Smith, S. G., 2000. Typhaceae A. L. Jussieu. – In: Flora of North
America. Editorial Committee (eds). New York and
Oxford. Online version, Available at: http://www.eﬂoras.
sed on August 10, 2013.
Snow, A. A., S. E. Travis, R. Wildova
`, et al., 2010. Species-
speciﬁc SSR alleles for studies of hybrid cattails (Typha
latifolia 9T. angustifolia; Typhaceae) in North America.
American Journal of Botany 97: 2061–2067.
Stevens, M. & C. Hoag, 2000. USDA-NRCS Plant Guide,
Narrow leaf cattail, Typha angustifolia L., National Plant
Data Center and Idaho Plant Materials Center. Available
Accessed on June 25, 2013.
Stevens, M. & C. Hoag, 2006. USDA – NRCS Plant Guide,
Southern cattail, Typha domingensis Pers., National Plant
Hydrobiologia (2016) 768:137–150 149
Data Center and Idaho Plant Materials Center. Available at:
sed on August 11, 2013.
Taberlet, P., L. Gielly, G. Pautou & J. Bouvet, 1991. Universal
primers for ampliﬁcation of 3 noncoding regions of
chloroplast DNA. Plant Molecular Biology 17: 1105–1109.
Templeton, A. R. & C. F. Singh, 1993. A cladistics analysis of
phenotypic associations with haplotypes inferred from
restriction endonuclease mapping.4. Nested analyses with
cladogram uncertainty and recombination. Genetics 134:
Thompson, J. D., D. G. Higgins & T. J. Gibson, 1994. Clustal-W
– Improving the sensitivity of progressive multiple
sequence alignment through sequence weighting, position-
speciﬁc gap penalties and weight matrix choice. Nucleic
Acids Research 22: 4673–4680.
Travis, S. E., J. E. Marburger, S. Windels & B. Kubatova, 2010.
Hybridization dynamics of invasive cattail (Typhaceae)
stands in the western Great Lakes Region of North
America: a molecular analysis. Journal of Ecology 98:
Tsyusko-Omeltchenko, O. V., N. A. Schable, M. H. Smith, T.
C. Glenn, 2003. Microsatellite loci isolated from nar-
rowleaved cattail Typha angustifolia. Molecular Ecology
Notes 3: 535–538.
USDA-ARS, 2005. National Genetic Resources Program.
Typha latifolia L. Germplasm Resources Information
Network – (GRIN) (Online Database). National Germ-
plasm Resources Laboratory, Beltsville, Maryland.
Available at: http://www.ars-grin.gov2/cgi-bin/npgs/html/
taxon.pl?101629. Accessed on January 09, 2014.
USDA-NRCS, 2014. The PLANTS Database. Baton Rouge,
USA: National Plant Data Center. Available at: http://
nativity,http://plants.usda.gov/. Accessed on January 1st,
Vibrans, H., 2004. Malezas de Me
´xico (on-line resource).
Available at: http://www.conabio.gob.mx/malezasde
Accessed on June 15, 2013.
WCSP, 2012. World Checklist of Selected Plant Families.
Facilitated by the Royal Botanic Gardens, Kew. Available
at: http://apps.kew.org/wcsp/ Retrieved http://apps.kew.
6F3EC6ED30DAA4. Accessed on September 5, 2012.
Zapfe, L. & J. R. Freeland, 2015. Heterosis in invasive F1 cattail
hybrids (Typha 9glauca). Aquatic Botany 125: 44–47.
Zedler, J. B. & S. Kercher, 2004. Causes and consequences of
invasive plants in wetlands: opportunities, opportunists, and
outcomes. Critical Reviews in Plant Science 23: 431–452.
150 Hydrobiologia (2016) 768:137–150