Journal of Biogeography (J. Biogeogr.) (2012) 39, 1427–1438
ORIGINAL
ARTICLE
Radiation following long-distance
dispersal: the contributions of time,
opportunity and diaspore morphology
in Sicyos (Cucurbitaceae)
Patrizia Sebastian1, Hanno Schaefer2, Rafael Lira3, Ian R. H. Telford4 and
Susanne S. Renner1*
1
Systematic Botany and Mycology, University
of Munich (LMU), Menzinger Strasse 67,
80638 Munich, Germany, 2Organismic and
Evolutionary Biology, Harvard University,
Cambridge, MA 02138, USA, 3Facultad de
Estudios Superiores Iztacala, Universidad
Nacional Autónoma de México, Tlalnepantla,
C.P. 54090, Mexico, 4School of Environmental
and Rural Science, University of New England,
Armidale, NSW 2351, Australia
ABSTRACT
Aim To infer the most plausible explanations for the presence of 14 species of the
Neotropical cucurbit genus Sicyos on the Hawaiian Islands, two on the Galápagos
Islands, two in Australia, and one in New Zealand.
Location Neotropics, the Hawaiian and Galápagos archipelagos, Australia and
New Zealand.
Methods We tested long-problematic generic boundaries in the tribe Sicyoeae
and reconstructed the history of Sicyos using plastid and nuclear DNA sequences
from 87 species (many with multiple accessions) representing the group’s generic
and geographic diversity. Maximum likelihood and Bayesian approaches were
used to infer relationships, divergence times, biogeographic history and ancestral
traits.
Results Thirteen smaller genera, including Sechium, are embedded in Sicyos,
which when re-circumscribed as a monophyletic group comprises 75 species. The
14 Hawaiian species of Sicyos descended from a single ancestor that arrived
c. 3 million years ago (Ma), Galápagos was reached twice at c. 4.5 and 1 Ma, the
species in Australia descended from a Neotropical ancestor (c. 2 Ma), and New
Zealand was reached from Australia. Time since arrival thus does not correlate
with Sicyos species numbers on the two archipelagos.
*Correspondence: Susanne S. Renner,
Systematic Botany and Mycology, University of
Munich (LMU), Menzinger Strasse 67, 80638
Munich, Germany.
E-mail: renner@lrz.uni-muenchen.de
Main conclusions A plausible mechanism for the four trans-Pacific dispersal
events is adherence to birds of the tiny hard fruit with retrorsely barbed spines
found in those lineages that underwent long-distance migrations. The Hawaiian
clade has lost these spines, resulting in a lower dispersal ability compared with the
Galápagos and Australian lineages, and perhaps favouring allopatric speciation.
Keywords
Australia, bird dispersal, diversification, Galápagos, Hawaiian radiation, Miocene, New Zealand, trans-Pacific dispersal.
INTRODUCTION
The geographic origin and speed of diversification of flowering
plant clades occurring on the archipelagos of the Pacific Ocean
have attracted much recent attention (e.g. Wright et al., 2000;
Cronk et al., 2005; Harbaugh & Baldwin, 2007; Clark et al.,
2008; Harbaugh et al., 2009; Keppel et al., 2009). Especially
striking are radiations on the Hawaiian Islands, including the
ª 2012 Blackwell Publishing Ltd
Lobelioideae with 126 species, Cyrtandra (Gesneriaceae) with
59, Melicope/Platydesma (Rutaceae) with 52, and eight more
genus-level clades each with ‡ 19 species (Baldwin & Wagner,
2010). Most Hawaiian lineages are younger than five million
years (Myr) old (Lindqvist & Albert, 2002; Price & Clague,
2002; Clark et al., 2009; Havran et al., 2009; Willyard et al.,
2011), and, compared with other angiosperm rates so far
reported, some of them have diversified at a higher rate
http://wileyonlinelibrary.com/journal/jbi
doi:10.1111/j.1365-2699.2012.02695.x
1427
P. Sebastian et al.
(Baldwin & Sanderson, 1998). The Galápagos Islands by
contrast have produced few plant radiations, the largest among
them the daisy genus Scalesia with 15 species going back to a
common ancestor living 1.9–6.2 million years ago (Ma)
(Schilling et al., 1994) and Varronia with four species going
back 1.12–4.5 Myr (Weeks et al., 2010). The difference in the
number and size of plant radiations on the two archipelagos
could reflect the time available for diversification, ecological
opportunity, and the propensity of particular clades to form
isolated populations prone to interruption of gene flow. Other
possible explanations include different extinction effects (e.g.
resulting from different palaeoclimatic histories) or taxonomic
bias (different species concepts applied on different archipelagos; Carine & Schaefer, 2010; Schaefer et al., 2011).
Clades occurring on both the Hawaiian and Galápagos
archipelagos should in principle allow the contribution of time
and ecological opportunity to be disentangled from cladespecific propensities to form new species. The latter may
correlate with dispersal ability, mating system, ease of hybridization, and speed of karyotype rearrangements (see also Price
& Wagner, 2004). The indigenous floras of the Hawaiian and
Galápagos archipelagos share only 24 genera, with 13 having
endemic species on both archipelagos (Table 1). Among them
Table 1 Shared native angiosperm genera of the Hawaiian and
Galápagos Islands. Genera in bold have endemic species on both
archipelagos (data from Wiggins & Porter, 1971; Wagner et al.,
1990; Carr, 2006; Bungartz et al., 2009).
Hawaii
Galápagos
Genus
Family
Endemic Native Endemic Native
Abutilon
Acacia
Amaranthus
Cordia/Varronia
Cuscuta
Dodonaea
Chamaesyce
Gossypium
Heliotropium
Ipomaea
Lobelia
Lycium
Peperomia
Phyllanthus
Phytolacca
Pilea
Pisonia
Plantago
Plumbago
Portulaca
Psychotria
Sesuvium
Sicyos
Solanum
Waltheria
Malvaceae
Mimosaceae
Amaranthaceae
Boraginaceae
Cuscutaceae
Sapindaceae
Euphorbiaceae
Malvaceae
Boraginaceae
Convolvulaceae
Campanulaceae
Solanaceae
Piperaceae
Euphorbiaceae
Phytolaccaceae
Urticaceae
Nyctaginaceae
Plantaginaceae
Plumbaginaceae
Portulacaceae
Rubiaceae
Aizoaceae
Cucurbitaceae
Solanaceae
Sterculiaceae
3
2
1
0
1
0
15
1
0
1
13
0
23
1
1
0
2
3
0
3
11
0
14
3
0
1428
1
0
0
1
0
1
0
0
2
4
0
1
2
0
0
1
3
0
1
1
0
1
0
1
1
1
0
4
4
2
1
8
2
1
2–3
0
1
4
0
0
1
1
1
0
1
2
1
2
1
0
0
3
1
2
0
0
0
0
3
3
1
0
2
1
1
1
0
0
2
0
0
1
0
1
1
is Sicyos, the focal clade of this study. Sicyos is a genus in the
Cucurbitaceae that has 14 endemic species on the Hawaiian
Islands (Wagner & Shannon, 1999), two species on the
Galápagos Islands (Sebastian et al., 2010a), two in Australia,
one in New Zealand, and between 41 and 56 species in the
Americas, depending on the taxonomic concept applied:
several small genera have been included in Sicyos or segregated
from it based mostly on fruit characters. Species of Sicyos are
climbing or trailing annual or perennial vines that often occur
in disturbed habitats. All Sicyos species are monoecious, with
male and female flowers on each individual; the flowers are
diurnal and depend on wasps and short-tongued bees for
pollination (LaBerge & Hurd, 1965; Fig. 1a) because automatic
selfing is precluded by their unisexuality. Fruit and seed
morphology in the Sicyos alliance is exceptionally variable, and
traits such as fleshy or hard fruits, with smooth surfaces or
surfaces bearing barbed or hooked spines (Fig. 1e–i), are likely
to influence dispersal.
The Sicyos clade, including the segregate genera, is especially
diverse in Mexico, where several new species have been
discovered in the recent past (Lira, 1994; Lira & Rodrı́guezArévalo, 1999; Rodrı́guez-Arévalo & Lira, 2001; Rodrı́guezArévalo, 2003; Rodrı́guez-Arévalo et al., 2004, 2005). Sicyos is
the name-giving taxon of the Sicyoeae, a tribe with 265 species
in perhaps a dozen genera (Schaefer & Renner, 2011a,b).
Family-wide molecular phylogenies relying on plastid and
recently also nuclear data suggest that Sicyoeae are monophyletic (Kocyan et al., 2007; Schaefer et al., 2009; Schaefer &
Renner, 2011b). However, none sampled more than a few
species of Sicyos.
Here we use Sicyos to study whether time since arrival,
ecological opportunity, or different fruit morphologies more
plausibly explain the different species numbers produced in
regions reached by long-distance dispersal, namely Hawaii (14
species), Galápagos (2), Australia (2) and New Zealand (1). A
single medium-sized clade having produced species in all these
areas provides a rare opportunity to disentangle the relative
effects of age, traits and ecological opportunity on diversification.
MATERIALS AND METHODS
Taxon sampling, DNA sequencing, alignment and
phylogenetic analysis
We used 112 accessions representing 87 species of Sicyoeae,
including the type species of all relevant generic names so as to
be able to decide the taxonomic fate of the previously
segregated genera Anomalosicyos Gentry (7 species, Central to
South America), Cladocarpa (H. St John) H. St John (20
species, Hawaii), Costarica L.D. Gómez (1 species, Costa Rica),
Microsechium Naudin (2–4 species, Mexico and Guatemala),
Parasicyos Dieterle (2 species, Guatemala), Pterosicyos Brandegee (1 species, Mexico and Guatemala), Sarx H. St John
(2 species, Hawaii), Sechiopsis Naudin (5 species, Mexico
and Guatemala), Sechium P. Browne (5 species, Mexico),
Journal of Biogeography 39, 1427–1438
ª 2012 Blackwell Publishing Ltd
Sicyos evolution in Hawaii, Galápagos and Australia/New Zealand
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
Figure 1 Habitats, flowers, and armed or unarmed fruits of Sicyos species from Hawaii, Australia and the continental mainland: (a)
S. angulatus visited by Vespula germanica (USA), (b) S. pachycarpus (Hawaii), (c) S. maximowiczii growing in a colony of great frigatebird
(Fregata minor), (d) S. undara (Australia), (e) S. pachycarpus (Hawaii), (f) S. weberbaueri (Peru), (g) S. australis (Australia), (h) S. acarieanthus (Peru), (i) Microsechium ruderale (Guatemala). Scale bar = 1 cm. Photographs by: Forest & Kim Starr (a, e), T. Rau
(b; Carr, 2006), H. Schaefer (b), P. Sebastian (d), M. Weigend (f, h), A. Lyne (g; APII) and M. Nee (i).
Sicyocarya (A. Gray) H. St John (25 species, Hawaii),
Sicyocaulis Wiggins (1 species, Galápagos Islands), Sicyosperma
A. Gray (1 species, Mexico and Arizona) and Skottsbergiliana
H. St John (2 species, Hawaii). No previous taxonomic fusions
or segregations were based on molecular data. We were able to
sample all species known from the Galápagos Islands, Australia
and New Zealand, and 13 of the 14 known from the Hawaiian
Islands. The missing Hawaiian species, Sicyos semitonsus, is
close to S. herbstii, S. hispidus and S. maximowiczii, judging
from the shared hairy fruit protuberances, but may actually be
a hybrid (Telford, 1990; Starr & Martz, 1999). The type
specimen of Costarica hamata was unavailable for sequencing,
and instead we used material collected at the type locality on
the slopes of the Irazú Volcano in Costa Rica. Appendix S1 in
the Supporting Information lists all included species with their
authors, geographic origin of the sequenced sample, voucher
deposition and GenBank accession numbers. A total of 420
chloroplast and 98 nuclear sequences were newly generated for
this study and have been submitted to GenBank (accession
numbers JN560179–JN560696).
Genomic DNA was isolated from herbarium specimens or
from silica-dried leaves, using the NucleoSpin plant kit
(Machery-Nagel, Düren, Germany). Polymerase chain reaction
(PCR) protocols and primers were the same as in Sebastian
et al. (2010b). The plastid DNA regions sequenced were the
Journal of Biogeography 39, 1427–1438
ª 2012 Blackwell Publishing Ltd
trnL intron, the adjacent trnL–trnF intergenic spacer, the
rpl20–rps12 intergenic spacer, the trnS–trnG intergenic spacer,
the psbA–trnH intergenic spacer and the rbcL gene; the nuclear
region sequenced was the complete internal transcribed spacer
(ITS) region of the ribosomal DNA. For amplification of the
psbA–trnH intergenic spacer, we used the primers listed in
Volz & Renner (2009). The PCR products were purified with
the PCR Wizard clean-up kit (Promega GmbH, Mannheim,
Germany) or ExoSap (Fermentas, St Leon-Rot, Germany).
Cycle sequencing was performed with the BigDye Terminator
cycle sequencing kit on an ABI Prism 3100 Avant automated
sequencer (Applied Biosystems, Foster City, CA). Sequencing
primers were the same as those used for DNA amplification.
The ITS region yielded single bands and unambiguous base
calls, and we therefore refrained from cloning. Sequence
assembly of forward and reverse strands was carried out with
Sequencher 4.7 (Gene Codes, Ann Arbor, MI), and sequences
were aligned by eye using MacClade 4.08 (Maddison &
Maddison, 2003).
The aligned plastid DNA matrix comprised 4527 nucleotides, and the aligned ITS matrix 872 nucleotides. In eight
cases, we combined plastid and nuclear sequences from
different samples (Appendix S1). Maximum likelihood (ML)
analyses and ML bootstrap searches (using 500 replicates) were
carried out using RAxML 7.2.8 (Stamatakis, 2006). RAxML
1429
P. Sebastian et al.
searches relied on the GTR + C model, with model parameters
estimated over the duration of specified runs. Analyses of the
separate plastid and nuclear datasets (gene trees not shown)
with one exception yielded congruent tree topologies, with
differences restricted to tip nodes with low statistical support
(bootstrap < 75%); the sole difference concerned Microsechium gonzalo-palomae (see Results).
Molecular clock analyses and diversification rates
To obtain age ranges for the nodes of biogeographic interest
we used Bayesian time estimation and a relaxed clock
uncorrelated-rates model as implemented in beast 1.6.1
(Drummond & Rambaut, 2007). Species with nearly identical
sequences were excluded from the dating analysis to reduce
stochastic error and rate heterogeneity, yielding an alignment
of 81 species, including the early-diverging Sicyoeae Luffa
aegyptiaca, Nothoalsomitra suberosa and Trichosanthes ovigera
for rooting purposes (Schaefer & Renner, 2011b). The pollen
Hexacolpites echinatus from the Oligocene (33.9 to 23 Ma) of
Cameroon (Salard-Cheboldaeff, 1978) is the oldest known
hexacolpate echinate Sicyoeae-type pollen and was used as a
calibration point. The most conservative assignment of this
pollen is to the split between Linnaeosicyos with 4-colporate
reticulate pollen (Schaefer et al., 2008a) and the remaining
New World Sicyoeae with 4–16 colpate/colporate and mostly
echinate pollen. To cover the uncertainty in the pollen age, we
applied a normally distributed prior probability distribution of
28.5 Ma ± 6 Myr to this node, as there is little justification for
weighting the probability towards the minimum bound of the
stratum in which the pollen was found (Ho & Phillips, 2009).
The root of the Sicyoeae was constrained to 37 Ma ± 3 Myr
(again with a normal prior distribution) based on the age
found for this node in the family-wide analysis by Schaefer
et al. (2009). All beast runs used a Yule tree prior and the
GTR + C model with six rate categories; Monte Carlo Markov
chains (MCMC) were run for 20 million generations, sampling
every 1000th generation. Mixing of the chains and convergence
were checked using Tracer 1.5 (Rambaut & Drummond,
2007); of the 20,001 posterior trees, the first 5000 were
discarded as burn-in based on inspection of the Tracer files.
Final trees were edited in FigTree 1.3.1 (Rambaut, 2006).
We modelled diversification as a time-homogeneous birth/
death process, with a net diversification rate k and a relative
extinction rate j = 0 or j = 0.9 (Magallón & Sanderson,
2001). Rates were calculated using the rate.estimate algorithm
implemented in the R package geiger 1.3.1 (Harmon et al.,
2008).
Ancestral area reconstruction
To infer the biogeographic history of Sicyoeae we used
Bayesian MCMC searches in beast and the continuous-time
Markov chain (CTMC) biogeographic reconstruction approach proposed by Lemey et al. (2009). The advantage of
this approach is that it incorporates branch length information
1430
as well as uncertainty in the tree topology. Priors for migration
rates used a C distribution for the relative rate parameter
(shape parameter = 1.0) and an exponential distribution
(mean = 1.0) for the geosite model parameter as recommended by Lemey et al. (2009). The CTMC phylogeographic
model assumes that ancestral ranges are limited to single
regions, making it particularly relevant for clades in which
dispersal plays a larger role than vicariance, as is the case for
oceanic island archipelagos. The geographic regions coded
were: (1) North American plate, (2) Caribbean plate, (3) South
American plate, (4) Hawaii, (5) Galápagos, and (6) Australia/
New Zealand.
Evolution of habitat occupation and fruit armature
To assess whether habitat diversity in a region is correlated
with diversification we categorized habitats into (1) shrublands
and coastal vegetation, (2) tropical deciduous forest, (3)
tropical evergreen forest, (4) tropical wet montane or cloud
forest, (5) Pinus–Quercus forest, and (6) dry montane forest,
and then coded each species for its preferred habitat(s). Data
on habitat preferences came from specimen labels and
taxonomic and floristic treatments (Macbride, 1960; Wiggins
& Porter, 1971; Wagner et al., 1990; Jeffrey & Trujillo, 1992).
Preferences were plotted on a ML phylogeny for the same 81
taxa as used for the molecular clock dating.
To infer ancestral states of fruit armature in Sicyos, we used
ML as implemented in Mesquite 2.74 (Maddison & Maddison, 2009), employing the Markov k-state one-parameter
model (Lewis, 2001). The coded character states were: (1)
armed, (2) unarmed, (3) winged, and (4) variable withinspecies (this was relevant for Sicyos edulis; Lira et al., 1999).
Transition parameters were estimated on the 81-taxon ML
phylogeny.
RESULTS
Phylogenetic relationships of the Sicyoeae
The ML phylogeny for the Sicyoeae (Fig. 2) shows that almost
all species of Anomalosicyos, Cladocarpa, Costarica, Microsechium (only the type species, M. ruderale), Parasicyos, Pterosicyos,
Sarx, Sechiopsis, Sechium, Sicyocarya, Sicyocaulis, Skottsbergiliana and Sicyosperma are embedded among species of Sicyos, a
clade that itself has 100% bootstrap support. In addition, all
the segregate genera that had more than one species (Anomalosicyos, Microsechium, Parasicyos, Sechiopsis, Sechium) turn
out to be polyphyletic, and Frantzia, which traditionally was
seen as close to Sechium, instead constitutes a genetically
distant lineage (Fig. 2). From now on, we focus on the
monophyletic genus Sicyos as circumscribed in Fig. 2, that is,
including all the former segregates. The Hawaiian species of
Sicyos form a robustly supported monophyletic group, and this
is also supported by a 6-bp deletion in their trnL intron. The
single New Zealand species groups with the two Australian
species. By contrast, the two species on the Galápagos
Journal of Biogeography 39, 1427–1438
ª 2012 Blackwell Publishing Ltd
Sicyos evolution in Hawaii, Galápagos and Australia/New Zealand
Sicyos albus (Sarx) Hawai’i *
95 Sicyos herbstii Kaua’i
84 Sicyos hispidus Hawai’i
Sicyos maximowiczii (Cladocarpa) Lehua*
86
Sicyos lanceoloideus O’ahu
Sicyos macrophyllus (Sicyocarya) Hawai’i*
Sicyos lasiocephalus (Skottsbergiliana) Hawai’i*
Sicyos cucumerinus Maui
82
Sicyos pachycarpus Maui
Sicyos waimanaloensis Moloka’i
Sicyos anunu Hawai’i
Sicyos hillebrandii Hawai’i
Sicyos erostratus Moloka’i
Sicyos collinus Mexico
Sicyos microphyllus Mexico
Sicyos angulatus USA*
75 Sicyos davilae Mexico
Sicyos longisepalus Mexico
98 Sicyos weberbaueri Peru
Sicyos baderoa Chile
Sicyos villosus Galapagos
Sicyos debilis Peru
97 Sicyos montanus Bolivia
Sicyos warmingii Argentina
97 Sicyos mcvaughii Mexico
95 Sicyos sertuliferus Mexico
Sicyos polyacanthus Argentina
Sechium mexicanum Mexico
100 Microsechium ruderale Guatemala *
Sicyos lirae Mexico
Sicyos macrocarpus Peru
Sicyocaulis pentagonus Galapagos *
86 Sicyos acarieanthus Peru
100 Sicyos palmatilobus Ecuador
100
100
Sicyos quinquelobatus Brazil
Sicyos barbatus (Anomalosicyos) Mexico*
Sicyos undara Australia
81 Sicyos mawhai New Zealand
Sicyos australis Australia
94
Sicyos laciniatus Mexico
Sicyos ampelophyllus USA (New Mexico)
Sicyos malvifolius Bolivia
Sicyos peninsularis Mexico
94
Sicyos longisetosus Ecuador
Sicyos triqueter (Sechiopsis) Mexico *
100
Sechiopsis tetraptera Mexico
Sicyos guatemalensis Mexico
100
Sicyos glaber USA (Texas)
77 Sechium chinantlense Mexico
100 Sicyos edulis (Sechium) Panama
*
100
Sechium compositum Mexico
98
Sechium hintonii Mexico
Sicyos motozintlensis Mexico
100
90
Sicyos chiriquensis (Costarica) Costa Rica*
89
Sechiopsis diptera Mexico
100 80
Sechiopsis laciniatus (Pterosicyos) Mexico*
Sicyos andreanus Ecuador
Sicyos dieterleae Mexico
98
Sicyos parviflorus Mexico
100
100
80
Sicyosperma gracile Mexico*
100 Parasicyos maculatus Guatemala*
95 Sicyos bulbosus Mexico
Sicyos galeottii Mexico
Parasicyos dieterleae Mexico
Frantzia venosa Costa Rica
100
Frantzia talamancensis Costa Rica
Frantzia panamensis Panama
98 Frantzia pittieri Costa Rica *
Frantzia villosa Costa Rica
Frantzia tacaco Costa Rica
100
Echinopepon bigelovii (Brandegea) USA (Arizona) *
100
Echinopepon insularis Mexico
Echinopepon paniculatus Mexico
88
Echinopepon arachoideus (Apatzingania) Mexico*
Echinocystis lobata USA*
78
100 Marah macrocarpus USA (California)
Marah fabaceus USA (California)
Cyclanthera australis (Pseudocyclanthera) Paraguay *
100
100
Cyclanthera carthagenensis Guatemala
99
Cyclanthera brachystachya unknown
Hanburia mexicana Mexico*
100
Hanburia oerstedii Costa Rica
Linnaeosicyos amara Dominican Republic *
SICYOS
100
99
98
Trichosanthes ovigera Japan
Luffa aegyptiaca Asia*
Nothoalsomitra suberosa Australia*
Figure 2 Maximum likelihood phylogram for 86 species of Sicyoeae (excluding Microsechium gonzalo-palomae; see Results) based on 5399
aligned nucleotides of plastid and nuclear sequences analysed under a GTR + C model. The tree is rooted on Nothoalsomitra. Values at nodes
give likelihood bootstrap support ‡ 75% based on 500 replicates. Boxes around clades and arrows in the inset mark the four long-distance
dispersals to: Hawaii (blue), Galápagos (yellow) and Australia/New Zealand (pink). Stars indicate type species of currently or formerly
accepted genera. Inset: Geographic origins of the sequenced plant material. Circles, Sicyos; stars, other Sicyoeae.
archipelago result from independent dispersals to the islands
(Fig. 2).
The Mexican species Microsechium gonzalo-palomae is the
only species placed differently in the plastid and nuclear (ITS)
gene trees: based on its plastid sequences it clusters with the
Frantzia clade, but based on its nuclear sequences it belongs in
Sicyos. Two ITS sequences from duplicates of one of the two
Journal of Biogeography 39, 1427–1438
ª 2012 Blackwell Publishing Ltd
existing herbarium collections of this species showed 10
nucleotide differences but nevertheless clustered together,
suggesting multiple coexisting ITS copies, such as would be
expected following hybridization. We excluded this species
from our further combined analyses, as investigating the
nature of this species or hybrid population will require
fieldwork to collect more material.
1431
P. Sebastian et al.
Divergence times, direction of dispersal
and diversification rates
Sicyos originated in North America (probably Mexico, see inset
in Fig. 2) during the early Miocene, 23.6–15.1 Ma (Fig. 3). The
common ancestor of the Hawaiian radiation is inferred to have
diverged from a North American (Mexican) lineage 5.5–
1.9 Ma and to have given rise to the extant Hawaiian species
around 3 (4.1–1.3) Ma (Fig. 3). The Galápagos species Sicyos
villosus is part of a clade occurring in Ecuador, Peru, Chile,
Bolivia, Argentina and southern Brazil from which it diverged
about 4.5 (6.4–2.8) Ma. The other Galápagos species, Sicyocaulis pentagonus, is nested in a separate clade among species
from Ecuador, Peru and Brazil from which it diverged 1 (1.5–
0.08) Ma, so the ancestral areas of both Galápagos species were
probably in South America, possibly in adjacent mainland
Ecuador (Fig. 3). The Australia/New Zealand clade is sister to
species from the south-western United States, Mexico and
Bolivia, and diverged from a North American ancestor 5.6–
1.7 Ma. The Australian Sicyos undara is about 1 (3.6–0.5) Myr
old, and its close relatives, S. australis and S. mawhai, the latter
endemic to New Zealand, are about 0.7 (1.6–0.1) Myr old. The
two new species, S. mawhai and S. undara, differ from
S. australis in fruit morphology, number of flowers per
inflorescence, and flower size and peduncle length (Telford
et al., 2012).
The net diversification rate (k) of the Sicyos crown group in
the Hawaiian Islands is k = 0.47–1.45 species Myr)1, assuming
no extinction (j = 0), or k = 0.18–0.58 species Myr)1 if
extinction rates are high (j = 0.9; Table 2, which also summarizes the characteristics of the four trans-Pacific disjunctions). In the Australian/New Zealand clade, the diversification
rate is k = 0.11–0.81 species Myr)1 (j = 0) or k = 0.03–0.24
species Myr)1 (j = 0.9). Of course, it is zero for the two singlespecies Galápagos lineages.
Habitat diversity and fruit morphology
Habitat preferences in Sicyos are variable even within very
young clades and notably so in the Hawaiian clade (Fig. 1
and Appendix S2). We did not code disturbed versus
undisturbed habitats; however, Sicyos species often grow in
disturbed sites. Some of the widespread species, such as
Sicyos angulatus and S. polyacanthus, occur in a wide range
of habitats in their native ranges and are locally invasive in
the Old World. The Hawaiian species are found in coastal
areas, in arid or moist shrublands, or in openings in rain
forest up to 2000 m elevation. Sicyocaulis pentagonus on the
Galápagos is known only from a few collections in moist
forests of the Scalesia zone between 130–200 and 400–
550 m. The other Galápagos endemic, Sicyos villosus, is
known from a single collection made by Charles Darwin on
Floreana, which lacks habitat details. Habitats on Floreana,
which has a maximum elevation of 640 m a.s.l., could have
been coastal vegetation, arid shrublands, and/or moist
Scalesia forest.
1432
Of the two Australian species, Sicyos australis is widespread
from northern Queensland to Tasmania, where it occurs in
deciduous vine thickets, eucalypt forest, and in montane and
near-coastal habitats; it has also been collected on New
Zealand’s North Island, Lord Howe (now extinct) and Norfolk
Islands. The second Australian species, S. undara, is known
only from the Undara Volcanic National Park in Queensland,
where it grows in clay loam in boulder gullies of collapsed lava
tubes in tropical deciduous forest. The New Zealand endemic,
S. mawhai, is restricted to islands adjacent to the North Island
and the Kermadec Islands, where it occurs in scrubs or forest
margins of near-coastal sites.
Ancestral state reconstruction suggests that fruits armed
with spines or hooks and winged fruits evolved several times
(Appendix S3). Both Galápagos species as well as the Australian and New Zealand species have spiny fruits and are derived
from relatives with such fruits (Fig. 1g and Appendix S3). By
contrast, the Hawaiian clade lost the spines and instead has
smooth fruits (Fig. 1e) or fruits with stubby, hairy protuberances (Sicyos maximowiczii, S. hispidus, S. herbstii and
S. semitonsus).
DISCUSSION
Here we use a medium-sized clade (the genus Sicyos, which
comprises 75 species in its new monophyletic circumscription)
in order to study whether ecological opportunity or time
available for diversification more plausibly explain the strikingly different species numbers in four regions reached by
long-distance dispersal. Our dense species sampling allowed
identification of the geographic origins of the disjunct species
on Hawaii, the Galápagos, Australia and New Zealand. The
distances from North America to Hawaii (3800 km) and from
the South American mainland (Ecuador) to Galápagos
(930 km) should favour plant arrival on the Galápagos, but
hardly make it so frequent as to retard speciation. Indeed,
based on our biogeographic reconstruction, Hawaii and
Australia/New Zealand were reached a single time, while
Galápagos was reached twice. The finding that more distant
archipelagos have fewer arrivals and larger radiations fits with
findings from archipelagic birds (Ricklefs & Bermingham,
2007).
We estimate that the genus Sicyos is about 19 Myr old, the
Hawaiian radiation occurred 3 Ma, the Australian species are
2 Myr old, the New Zealand species 0.7 Myr old, and the two
Galápagos species Sicyos villosus and Sicyocaulis pentagonus are
4.5 and 1 Myr old, respectively (for error margins see Table 2
and Fig. 3). Time per se therefore cannot explain the different
species numbers in the four areas reached by long-distance
dispersal. Instead, the species build-up on the Hawaiian Islands
clearly exceeded that in the other regions, although the
Hawaiian diversification rate of Sicyos is not exceptional
compared with other plant radiations (Valente et al., 2010).
Morphologically, the Hawaiian Sicyos species are distinct from
each other (Wagner & Shannon, 1999), although their genetic
divergence is low (Fig. 2), a combination also found in other
Journal of Biogeography 39, 1427–1438
ª 2012 Blackwell Publishing Ltd
Sicyos evolution in Hawaii, Galápagos and Australia/New Zealand
Oligocene
Eocene
QuatPliocene ernary
Miocene
0.4
7
0.7
1.0
1.4
4
2.6
North American plate
0.8
0
3.6
Caribbean plate
1.2
4.7
1.1
1.4
4
South American plate
2.3
Hawaii
5.5
1.5
3.6
4.2
Galápagos
6.0
Australia/New Zealand
4.5
1.7
1 7
7.1
1.2
3.0
0.7
0.9
9
1
.1
1.1
7
1.7
9.0
5.4
7.8
1.3
9.9
1.9
8
2.8
3.5
0.7
13.1
1.9
5.8
3.9
14.8
6.7
4.5
*
1.0
16.6
2
1.9
4.9
9.4
5.2
11.4
17.7
12.3
9.7
6.5
10.3
19.2
14.3
Hexacolpites echinatus
(Salard-Cheboldaeff 1978)
0.7
2.4
0.5
3.6
4
3
4.3
5
5
5.5
22.4
7 4
7.4
13.8
4.5
15.5
10.9
*
19
9.9
19.9
2
1.4
29.2
9.4
14.9
20.8
8.9
31.9
1.8
14.2
33.9
35.6
*
1
40
35
30
25
20
15
10
5
Sicyos lasiocephalus
Sicyos anunu
Sicyos pachycarpus
Sicyos lanceoloideus
Sicyos albus
Sicyos hispidus
Sicyos maximowiczii
Sicyos herbstii
Sicyos collinus
Sicyos angulatus
Sicyos davilae
Sicyos longisepalus
Sicyos microphyllus
Sicyos warmingii
Sicyos montanus
Sicyos debilis
Sicyos weberbaueri
Sicyos baderoa
Sicyos villosus
Sicyos sertuliferus
Sicyos mcvaughii
Sicyos polyacanthus
Sechium mexicanum
Sicyocaulis pentagonus
Sicyos acarieanthus
Sicyos palmatilobus
Sicyos macrocarpus
Sicyos quinquelobatus
Sicyos barbatus
Microsechium ruderale
Sicyos lirae
Sicyos ampelophyllus
Sicyos laciniatus
Sicyos malvifolius
Sicyos australis
Sicyos mawhai
Sicyos undara
Sicyos longisetosus
Sicyos peninsularis
Sicyos triqueter
Sechiopsis tetraptera
Sicyos guatemalensis
Sicyos glaber
Sechium chinantlense
Sicyos edulis
Sechium compositum
Sechium hintonii
Sicyos chiriquensis
Sicyos motozintlensis
Sechiopsis laciniatus
Sechiopsis diptera
Sicyos andreanus
Sicyos parviflorus
Sicyosperma gracile
Sicyos dieterleae
Sicyos bulbosus
Sicyos galeottii
Parasicyos maculatus
Parasicyos dieterleae
Frantzia pittieri
Frantzia villosa
Frantzia tacaco
Frantzia panamensis
Frantzia talamancensis
Frantzia venosa
Echinopepon bigelovii
Echinopepon insularis
Echinopepon arachoideus
Echinopepon paniculatus
Cyclanthera carthagenensis
Cyclanthera australis
Cyclanthera brachystachya
Hanburia oerstedii
Hanburia mexicana
Marah fabaceus
Marah macrocarpus
Echinocystis lobata
Linnaeosicyos amara
Trichosanthes ovigera
Luffa aegyptiaca
Nothoalsomitra suberosa
0 age in Ma
Figure 3 Chronogram and ancestral area reconstruction for Sicyos and related Sicyoeae obtained under a Bayesian relaxed clock and a
continuous-time Markov chain biogeographic reconstruction applied to the data set used for Fig. 2 but excluding five Hawaiian species with
almost identical sequences. Bars at nodes indicate the 95% confidence intervals around the estimated times. Numbers above branches give
the node age, stars mark the calibration nodes, and arrows the four long-distance dispersals discussed in the text. Branch colour indicates
character states for ancestral areas with the highest posterior probability (all ‡ 0.92, except for a few outgroups), with the coding explained in
the inset.
Journal of Biogeography 39, 1427–1438
ª 2012 Blackwell Publishing Ltd
1433
P. Sebastian et al.
Table 2 Characteristics of the four Sicyos disjunctions compared here.
Characteristic
Hawaii
Australia/New Zealand
Galápagos 1
(Sicyos villosus)
Galápagos 2
(Sicyocaulis pentagonus)
Number of extant species
Distance from likely region of origin [km]
Relative habitat diversity
Dispersal ability (diaspore morphology)
Stem age [Ma]
Crown age [Ma]
Diversification rate (k = 0) [species Myr)1]
Diversification rate (k = 0.9) [species Myr)1]
14
3800
High
Low
5.5–1.9
4.1–1.3
0.47–1.45
0.18–0.58
3
6000
High
High
5.6–1.7
3.6–0.5
0.11–0.81
0.03–0.24
1
930
Low
High
6.4–2.8
NA
0
0
1
930
Low
High
1.5–0.08
NA
0
0
NA, not applicable.
Hawaiian radiations (Baldwin & Robichaux, 1995; Lindqvist
et al., 2003).
As time for speciation can be discounted as a strong
explanation for the numbers of species in the four areas
reached following long-distance dispersal, it might be extrinsic
factors that make the Hawaiian Islands especially conducive to
Sicyos speciation. Available area per se may influence diversification in island archipelagos (Ricklefs & Bermingham, 2007;
Price & Wagner, 2011). The land area of the Hawaiian
archipelago (16,300 km2) is more than twice that of the
Galápagos (8000 km2). This could explain the larger species
number on Hawaii compared with the Galápagos but certainly
not that only three species evolved in Australia and New
Zealand. However, extrapolation from land surface to number
of suitable habitats is difficult (see also Price & Wagner, 2011).
The Hawaiian archipelago harbours many types of plant
formations (Mueller-Dombois & Fosberg, 1998), and species
of Sicyos occupy most of them (Appendix S2). Especially
relevant may be the extreme breadth of rainfall regimes on
Hawaii. High islands, such as Kaua’i and Hawai’i, receive on
their windward slopes as much as 12 m annual precipitation,
whereas their leeward slopes experience warm-season droughts
and as little as 250 mm annual precipitation, in common with
some of the low islands. The Hawaiian soils derived from
volcanism have accordingly experienced very different weathering (Cuddihy, 1989).
The Galápagos archipelago by comparison has fewer climate
and vegetation types. Aridification there set in c. 3 Ma, while
before that conditions were warmer and moister (Wara et al.,
2005; Federov et al., 2006), potentially supporting rain forest
vegetation. Today, the climate of the Galápagos Islands is
characterized by highly variable convective rainfall during the
hot season and by a prolonged cool season (June to
December), with only little orographic rainfall on the higher
windward sides of the islands and frequent droughts at lower
elevations with arid-adapted vegetation types (Mueller-Dombois & Fosberg, 1998). The overall more arid climate on
Galápagos compared with Hawaii, combined with fewer
habitat types on smaller islands, may explain the absence of
any large plant radiations on Galápagos. Among the 13
angiosperm genera with endemic species in both archipelagos
1434
(Table 1) the Hawaiian genera usually have more species, and
radiations of more than four species are almost entirely
restricted to Hawaii.
An even broader range of climatic and edaphic conditions in
the Australian region than in the Hawaiian Islands, however,
did not result in a radiation of Sicyos, although further
collecting and sequencing efforts may well reveal additional
Australian species. The limited diversification of Australian
Sicyos resembles the situation in the remaining Cucurbitaceae
on that continent. These go back to some 20 independent
arrivals from the Malesian region that occurred at different
times over the past 35 Myr, with none currently having more
than four living descendant species (Schaefer et al., 2008b,
2009). The largest cucurbit radiations in Australia are Austrobryonia, with four species in the arid and semi-arid regions of
the continent (Schaefer et al., 2008b), and Cucumis, with two
radiations of two/three species in tropical savanna habitats
(Sebastian et al., 2010b; Telford et al., 2011). The reasons for
the limited cucurbit diversification in Australia remain
unclear.
In New Zealand, Cucurbitaceae had no endemic species
before the discovery of Sicyos mawhai. This is surprising
because Cucurbitaceae are successful transoceanic dispersers
(Schaefer et al., 2009; Duchen & Renner, 2010; Schaefer &
Renner, 2010), and the distance of 2100 km between Australia
and New Zealand has been overcome by many other disjunctly
distributed Australia/New Zealand clades (Pole, 1994). Nor is
Sicyos the only angiosperm genus disjunctly distributed
between the Americas and the Australian/New Zealand region:
Californian species of Lepidium (Brassicaceae) are phylogenetically closest to Australian/New Zealand species, which has
been explained by transoceanic dispersals (Mummenhoff et al.,
2004).
Seabirds can act as dispersal agents across the Pacific Ocean
(Falla, 1960; Carlquist, 1967) because at least some of them,
such as storm petrels, shearwaters and frigate birds, nest on the
ground or in burrows in coastal vegetation where they may
contact fruiting plants [Marks & Leasure, 1992; Starr & Martz,
1999; our Fig. 1c shows a great frigatebird (Fregata minor)
nesting among Sicyos maximowiczii on Laysan Island, Hawaii].
Some petrel populations migrate between Mexico, Central and
Journal of Biogeography 39, 1427–1438
ª 2012 Blackwell Publishing Ltd
Sicyos evolution in Hawaii, Galápagos and Australia/New Zealand
South America, the Galápagos and Hawaii, and one race of the
white-faced storm-petrel (Pelagodroma marina) migrates
across the Pacific between New Zealand and the Humboldt
Current (off the coast of Peru), then west past the Galápagos
Islands (Tomkins, 1982). Species of Siycos have fleshy fruits
(fresh up to c. 10 cm long) or hard fruits (5–10 mm long) that
are smooth or barbed with hooked spines (Fig. 1 and
Appendix S3). As the spiny Sicyos fruits are presented in
multi-seeded infructescences, one contact with a bird can lead
to several seeds being transported. Fruit morphology has been
evolutionarily labile, and even fairly large wings have arisen
several times (Appendix S3), the latter surprisingly not linked
to any long-distance dispersal events. All lineages or species
involved in long-distance dispersal to Hawaii, the Galápagos,
Australia or New Zealand have spiny fruits. The spines,
however, are readily lost; in Sicyos edulis, natural populations
can contain individuals with spiny or smooth fruits (Lira et al.,
1999). Spines were also lost in the Hawaiian clade, where fruits
are unarmed or in four species retain stubby outgrowths
(Telford, 1990). Loss or reduction of dispersal ability is well
documented in other insular plant and animal species
(Carlquist, 1965, 1966a,b, 1974), the prime example being
Bidens, which on the Hawaiian Islands lost the barbed awns
responsible for long-distance dispersal in the mainland species
(Carlquist, 1966a, 1967). Selection for loss of dispersibility
should be strong because the majority of propagules that are
dispersed away from islands will be lost at sea. In Hawaiian
Sicyos, such limited dispersal ability could have promoted the
isolation of populations and thus allopatric speciation in the
diverse habitats and species build-up.
CONCLUSIONS
Sicyos, a clade of 75 species once the names in the segregate
genera are transferred to the genus Sicyos, includes four transPacific disjunctions. Small-scale habitat diversity and morphological adaptations (loss of spines leading to reduced dispersal
ability) are the most plausible factors that could account for
the significantly higher and more rapid accumulation of
species on the Hawaiian Islands compared with the Galápagos,
Australia or New Zealand.
ACKNOWLEDGEMENTS
For material we thank W. Wagner (United States National
Herbarium, Washington, DC), B. Hammel (Missouri Botanical Garden) and M. Nee (New York Botanical Garden). We
also thank C. Hughes, W. Wagner and an anonymous referee
for helpful comments on our manuscript.
REFERENCES
Baldwin, B.G. & Robichaux, R.H. (1995) Historical biogeography and ecology of the Hawaiian silversword alliance
(Asteraceae): new molecular phylogenetic perspectives.
Hawaiian biogeography: evolution on a hot spot archipelago
Journal of Biogeography 39, 1427–1438
ª 2012 Blackwell Publishing Ltd
(ed. by W.L. Wagner and V.A. Funk), pp. 259–287.
Smithsonian Institution Press, Washington, DC.
Baldwin, B.G. & Sanderson, M.J. (1998) Age and rate of
diversification of the Hawaiian silversword alliance (Compositae). Proceedings of the National Academy of Sciences
USA, 95, 9402–9406.
Baldwin, B.G. & Wagner, W.L. (2010) Hawaiian angiosperm
radiations of North American origin. Annals of Botany, 105,
849–879.
Bungartz, F., Herrera, H.W., Jaramillo, P., Tirado, N., Jı́menezUzcategui, G., Ruiz, D., Guézou, A. & Ziemmeck, F. (2009)
Charles Darwin Foundation Galápagos species checklist.
Charles Darwin Foundation, Puerto Ayora, Galápagos.
Available at: http://www.darwinfoundation.org/datazone/
checklists/ (accessed 10 July 2011).
Carine, M.A. & Schaefer, H. (2010) The Azores diversity
enigma: why are there so few Azorean endemic flowering
plants and why are they so widespread? Journal of Biogeography, 37, 77–89.
Carlquist, S. (1965) Island life. Natural History Press, New
York.
Carlquist, S. (1966a) The biota of long-distance dispersal. II.
Loss of dispersibility in Pacific Compositae. Evolution, 20,
433–455.
Carlquist, S. (1966b) The biota of long-distance dispersal. III.
Loss of dispersibility in the Hawaiian flora. Brittonia, 20, 30–
48.
Carlquist, S. (1967) The biota of long-distance dispersal. V.
Plant dispersal to Pacific islands. Bulletin of the Torrey
Botanical Club, 94, 129–162.
Carlquist, S. (1974) Island biology. Columbia University Press,
New York.
Carr, G. (2006) Hawaiian native plant genera. University of
Hawaii Botany Department, Manoa, Hawaii. Available at:
http://www.botany.hawaii.edu/faculty/carr/natives.htm
(accessed 10 July 2011).
Clark, J.R., Ree, R.H., Alfaro, M.E., King, M.G., Wagner,
W.L. & Roalson, E.H. (2008) A comparative study in
ancestral range reconstruction methods: retracing the
uncertain histories of insular lineages. Systematic Biology,
57, 693–707.
Clark, J.R., Wagner, W.L. & Roalson, E.H. (2009) Patterns of
diversification and ancestral range reconstruction in the
southeast Asian–Pacific angiosperm lineage Cyrtandra
(Gesneriaceae). Molecular Phylogenetics and Evolution, 53,
982–994.
Cronk, Q.C.B., Kiehn, M., Wagner, W.L. & Smith, J.F. (2005)
Evolution of Cyrtandra (Gesneriaceae) in the Pacific Ocean:
the origin of a supertramp clade. American Journal of Botany, 92, 1017–1024.
Cuddihy, L.W. (1989) Vegetation zones of the Hawaiian Islands.
Conservation biology in Hawai’i (ed. by C.P. Stone and D.B.
Stone), pp. 27–37. University of Hawaii Press, Honolulu.
Drummond, A.J. & Rambaut, A. (2007) BEAST: Bayesian
evolutionary analysis by sampling trees. BMC Evolutionary
Biology, 7, 214.
1435
P. Sebastian et al.
Duchen, P. & Renner, S.S. (2010) The evolution of Cayaponia
(Cucurbitaceae): repeated shifts from bat to bee pollination
and long-distance dispersal to Africa 2–5 million years ago.
American Journal of Botany, 97, 1129–1141.
Falla, R.A. (1960) Birds as dispersal agents. Proceedings of the
Royal Society B: Biological Sciences, 152, 655–659.
Federov, A.V., Dekens, P.S., McCarthy, M., Ravelo, A.C.,
deMenocal, P.B., Barreiro, M., Pacanowski, R.C. & Philander, S.G. (2006) The Pliocene paradox (mechanisms for a
permanent El Niño). Science, 312, 1485–1489.
Harbaugh, D.T. & Baldwin, B.G. (2007) Phylogeny and biogeography of the sandalwoods (Santalum, Santalaceae):
repeated dispersal throughout the Pacific. American Journal
of Botany, 94, 1028–1040.
Harbaugh, D.T., Wagner, W.L., Allan, G.J. & Zimmer, E.A.
(2009) The Hawaiian Archipelago is a stepping stone for
dispersal in the Pacific: an example from the plant genus
Melicope (Rutaceae). Journal of Biogeography, 36, 230–241.
Harmon, L.J., Weir, J.T., Brock, C., Glor, R.E. & Challenger,
W. (2008) GEIGER: investigating evolutionary radiations.
Bioinformatics, 24, 129–131.
Havran, J.C., Sytsma, K.J. & Ballard, H.E., Jr (2009) Evolutionary relationships, inter-island biogeography and
molecular evolution in the Hawaiian violets (Viola: Violaceae). American Journal of Botany, 96, 2087–2099.
Ho, S.Y.W. & Phillips, M.J. (2009) Accounting for calibration
uncertainty in phylogenetic estimation of evolutionary
divergence times. Systematic Biology, 58, 367–380.
Jeffrey, C. & Trujillo, B. (1992) Cucurbitaceae. Flora of Venezuela, Vol. V, Part 1 (ed. by G. Morillo), pp. 11–202. Fondo
Editorial Acta Cientifica Venezolana, Caracas.
Keppel, G., Lowe, A.J. & Possingham, H.P. (2009) Changing
perspectives on the biogeography of the tropical South
Pacific: influences of dispersal, vicariance and extinction.
Journal of Biogeography, 36, 1035–1054.
Kocyan, A., Zhang, L.-B., Schaefer, H. & Renner, S.S. (2007) A
multi-locus chloroplast phylogeny for the Cucurbitaceae
and its implications for character evolution and classification. Molecular Phylogenetics and Evolution, 44, 553–577.
LaBerge, W.E. & Hurd, P.D. (1965) A new subgenus and
species of matinal Andrena from the flowers of Sicyos
(Cucurbitaceae) in Mexico. Pan-Pacific Entomologist, 41,
186–193.
Lemey, P., Rambaut, A., Drummond, A.J. & Suchard, M.A.
(2009) Bayesian phylogeography finds its roots. PLoS
Computational Biology, 5, e1000520.
Lewis, P.O. (2001) A likelihood approach to estimating phylogeny from discrete morphological character data. Systematic Biology, 50, 913–925.
Lindqvist, C. & Albert, V.A. (2002) Origin of the Hawaiian
endemic mints within North American Stachys (Lamiaceae).
American Journal of Botany, 89, 1709–1724.
Lindqvist, C., Motley, T.J., Jeffrey, J.J. & Albert, V.A. (2003)
Cladogenesis and reticulation in the Hawaiian endemic
mints (Lamiaceae). Cladistics, 19, 480–495.
1436
Lira, R. (1994) Especie nueva de Microsechium (Cucurbitaceae,
tribu Sicyoeae, subtribu Sicyinae) del estado de Oaxaca,
México. Anales del Instituto de Biologı́a de la Universidad
Nacional Autónoma de México, Serie Botánica, 65, 73–81.
Lira, R. & Rodrı́guez-Arévalo, I. (1999) Nuevas especies de los
géneros Cyclanthera Schrad. y Sicyos L. (Sicyoeae, Cucurbitaceae) para la flora de México. Acta Botanica Mexicana,
48, 11–19.
Lira, R., Castrejón, J., Zamudio, S. & Rojas-Zenteno, C. (1999)
Propuesta de ubicación taxonómica para los chayotes silvestres (Sechium edule, Cucurbitaceae) de México. Acta
Botánica Mexicana, 49, 47–61.
Macbride, F.J. (1960) Flora of Peru, Botanical Series of the
Field Museum of Natural History, Vol. XIII, Part V, No. 2.
Chicago Natural History Museum Press, Chicago.
Maddison, D.R. & Maddison, W.P. (2003) MacClade, version
4.0.8. Sinauer Associates, Sunderland, MA.
Maddison, W.P. & Maddison, D.R. (2009) MESQUITE: a
modular system for evolutionary analysis, version 2.72.
Available at: http://mesquiteproject.org/mesquite/mesquite.
html (accessed 10 July 2011).
Magallón, S. & Sanderson, M.J. (2001) Absolute diversification
rates in angiosperm clades. Evolution, 55, 1762–1780.
Marks, J.S. & Leasure, S.M. (1992) Breeding biology of Tristram’s storm-petrel on Laysan Island. Wilson Bulletin, 104,
719–731.
Mueller-Dombois, D. & Fosberg, F.R. (1998) Vegetation of the
tropical Pacific islands. Springer, Berlin.
Mummenhoff, K., Linder, P., Friesen, N., Bowman, J.L., Lee,
J.-Y. & Franzke, A. (2004) Molecular evidence for bicontinental hybridogenous genomic constitution in Lepidium
sensu stricto (Brassicaceae) species from Australia and New
Zealand. American Journal of Botany, 91, 254–261.
Pole, M. (1994) The New Zealand flora – entirely long-distance
dispersal? Journal of Biogeography, 21, 625–635.
Price, J.P. & Clague, D.A. (2002) How old is the Hawaiian
biota? Geology and phylogeny suggest recent divergence.
Proceedings of the Royal Society B: Biological Sciences, 269,
2429–2435.
Price, J.P. & Wagner, W.L. (2004) Speciation in Hawaiian
angiosperm lineages: cause, consequence, and mode. Evolution, 58, 2185–2200.
Price, J.P. & Wagner, W.L. (2011) A phylogenetic basis for
species–area relationships among three Pacific Island floras.
American Journal of Botany, 98, 449–459.
Rambaut, A. (2006) FigTree v. 1.3.1. Available at: http://tree.
bio.ed.ac.uk/software/figtree/ (accessed August 2011).
Rambaut, A. & Drummond, A.J. (2007) Tracer 1.5. Available
at: http://beast.bio.ed.ac.uk/Tracer (accessed August 2011).
Ricklefs, R.E. & Bermingham, E. (2007) The causes of evolutionary radiations in archipelagoes: passerine birds in the
Lesser Antilles. The American Naturalist, 169, 285–297.
Rodrı́guez-Arévalo, I. (2003) A new species of Sicyos (Cucurbitaceae, Sicyoeae, Sicyinae) from Mexico and Guatemala.
Brittonia, 55, 69–72.
Journal of Biogeography 39, 1427–1438
ª 2012 Blackwell Publishing Ltd
Sicyos evolution in Hawaii, Galápagos and Australia/New Zealand
Rodrı́guez-Arévalo, I. & Lira, R. (2001) Nueva especie del
género Sicyos L. (Cucurbitaceae) para la República
Mexicana. Boletin de la Sociedad Botanica de México, 68, 81–
84.
Rodrı́guez-Arévalo, I., Lira, R. & Dávila, P. (2004) Two new
species of Sicyos L. (Cucurbitaceae) from Guerrero and
Oaxaca, Mexico. Botanical Journal of the Linnean Society,
145, 373–378.
Rodrı́guez-Arévalo, I., Lira, R. & Calzada, I. (2005) A new
species of Sicyos L. (Cucurbitaceae) from Oaxaca, Mexico.
Brittonia, 57, 43–46.
Salard-Cheboldaeff, M. (1978) Sur la palynoflore Maestrichtienne et Tertiaire du bassin sédimentaire littoral du
Cameroun. Pollen et Spores, 20, 215–260.
Schaefer, H. & Renner, S.S. (2010) A gift from the New World?
The West African crop Cucumeropsis mannii and the
American Posadaea sphaerocarpa (Cucurbitaceae) are the
same species. Systematic Botany, 35, 534–540.
Schaefer, H. & Renner, S.S. (2011a) Cucurbitaceae. Families
and genera of vascular plants, Vol. 10 (ed. by K. Kubitzki),
pp. 112–174. Springer Verlag, Berlin.
Schaefer, H. & Renner, S.S. (2011b) Phylogenetic relationships
in the order Cucurbitales and a new classification of the
gourd family (Cucurbitaceae). Taxon, 60, 122–138.
Schaefer, H., Kocyan, A. & Renner, S.S. (2008a) Linnaeosicyos
(Cucurbitaceae), a new genus for the Caribbean sister
group of the New World Sicyeae. Systematic Botany, 33,
349–355.
Schaefer, H., Telford, I.R.H. & Renner, S.S. (2008b) Austrobryonia (Cucurbitaceae), a new Australian endemic genus,
is the closest living relative to the Eurasian and Mediterranean Bryonia and Ecballium. Systematic Botany, 33, 125–
132.
Schaefer, H., Heibl, C. & Renner, S.S. (2009) Gourds afloat: a
dated phylogeny reveals an Asian origin of the gourd family
(Cucurbitaceae) and numerous oversea dispersal events.
Proceedings of the Royal Society B: Biological Sciences, 276,
843–851.
Schaefer, H., Moura, M., Maciel, M.G.B., Silva, L., Rumsey, F.J.
& Carine, M.A. (2011) The Linnean shortfall in oceanic
island biogeography: a case study in the Azores. Journal of
Biogeography, 38, 1345–1355.
Schilling, E.E., Panero, J.L. & Eliasson, U.H. (1994) Evidence
from chloroplast DNA restriction site analysis on the relationships of Scalesia (Asteraceae: Heliantheae). American
Journal of Botany, 8, 248–254.
Sebastian, P., Schaefer, H. & Renner, S.S. (2010a) Darwin’s
Galápagos gourd: providing new insights 175 years after his
visit. Journal of Biogeography, 37, 975–980.
Sebastian, P., Schaefer, H., Telford, I.R.H. & Renner, S.S.
(2010b) Cucumber (Cucumis sativus) and melon (C. melo)
have numerous wild relatives in Asia and Australia,
and the sister species of melon is from Australia. Proceedings of the National Academy of Sciences USA, 107,
14269–14273.
Journal of Biogeography 39, 1427–1438
ª 2012 Blackwell Publishing Ltd
Stamatakis, A. (2006) RAxML-VI-HPC: maximum likelihoodbased phylogenetic analyses with thousands of taxa and
mixed models. Bioinformatics, 22, 2688–2690.
Starr, F. & Martz, K. (1999) S.S. Midway Expedition – May 21,
1999–June 16, 1999. Trip Report. United States Fish and
Wildlife Service. Available at: http://www.hear.org/starr/
publications/ssmidway.pdf (accessed August 2011).
Telford, I.R.H. (1990) Cucurbitaceae. Manual of the flowering
plants of Hawai’i (ed. by W.L. Wagner, D.R. Herbst and S.H.
Sohmer), pp. 568–581. University of Hawaii Press & Bishop
Museum Press, Honolulu.
Telford, I.R.H., Sebastian, P., Bruhl, J.J. & Renner, S.S. (2011)
Cucumis (Cucurbitaceae) in Australia and eastern Melanesia, including newly recognized species and the sister species
to C. melo. Systematic Botany, 36, 376–389.
Telford, I.R.H., Sebastian, P., de Lange, P.J., Bruhl, J.J. &
Renner, S.S. (2012) Morphological and molecular data reveal
three rather than one species of Sicyos (Cucurbitaceae) in
Australia, New Zealand and the Islands of the South West
Pacific. Australian Systematic Botany, doi: 10.1071/SB11032.
Tomkins, R.J. (1982) Migration of white-faced storm-petrels
Pelagodroma marina in the South Pacific and the status of
the Kermadec subspecies. Emu, 84, 32–35.
Valente, L.M., Savolainen, V. & Vargas, P. (2010) Unparalleled
rates of species diversification in Europe. Proceedings of the
Royal Society B: Biological Sciences, 277, 1489–1496.
Volz, S.M. & Renner, S.S. (2009) Phylogeography of the
ancient Eurasian medicinal plant genus Bryonia (Cucurbitaceae) inferred from nuclear and chloroplast sequences.
Taxon, 58, 550–560.
Wagner, W.L. & Shannon, R.K. (1999) Nomenclator of
Hawaiian Sicyos (Cucurbitaceae). Novon, 9, 441–447.
Wagner, W.L., Herbst, D.R. & Sohmer, S.H. (1990) Manual of
the flowering plants of Hawai’i. University of Hawaii Press &
Bishop Museum Press, Honolulu.
Wara, M.W., Ravelo, A.C. & Delaney, M.L. (2005) Permanent
El Niño-like conditions during the Pliocene warm period.
Science, 309, 758.
Weeks, A., Baird, K.E. & McMullen, C.K. (2010) Origin
and evolution of endemic Galápagos Varronia species
(Cordiaceae). Molecular Phylogenetics and Evolution, 57,
948–954.
Wiggins, I.L. & Porter, D.M. (1971) Flora of the Galápagos
Islands. Stanford University Press, Stanford, CA.
Willyard, A., Wallace, L.E., Wagner, W.L., Weller, S.G., Sakai,
A.K. & Nepokroeff, M. (2011) Estimating the species tree for
Hawaiian Schiedea (Caryophyllaceae) from multiple loci in
the presence of reticulate evolution. Molecular Phylogenetics
and Evolution, 60, 29–48.
Wright, S.D., Yong, C.G., Dawson, J.W., Whittaker, D.J. &
Gardner, R.C. (2000) Riding the ice age El Niño? Pacific
biogeography and evolution of Metrosideros subg. Metrosideros (Myrtaceae) inferred from nuclear ribosomal DNA.
Proceedings of the National Academy of Sciences USA, 97,
4118–4123.
1437
P. Sebastian et al.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the
online version of this article:
Appendix S1 List of Sicyoeae species with their authors,
geographic origin of the sequenced sample, voucher deposition, and GenBank accession numbers for all sequences
included in this study.
Appendix S2 Habitat types of Sicyoeae species plotted on an
86-species maximum likelihood phylogeny.
Appendix S3 Evolution of armed fruits and unarmed fruits
in the Sicyoeae inferred on an 86-species maximum likelihood
phylogeny under maximum likelihood optimization.
As a service to our authors and readers, this journal provides
supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online
delivery, but are not copy-edited or typeset. Technical support
issues arising from supporting information (other than
missing files) should be addressed to the authors.
1438
BIOSKETCH
Patrizia Sebastian is a PhD student in the laboratory of
Susanne Renner at the University of Munich (LMU), where
she is working on the evolution of Cucumis, Sicyos and other
Cucurbitaceae.
The main research interests of the authors are the evolution,
biogeography and systematics of plants.
Author contributions: P.S. and S.R. designed the research; P.S.
collected and analysed the data; H.S., I.T. and R.L. selected and
supplied plant materials; P.S., H.S. and S.R. wrote the paper.
Editor: José Marı́a Fernández-Palacios
Journal of Biogeography 39, 1427–1438
ª 2012 Blackwell Publishing Ltd