Content uploaded by Norris H Williams
Author content
All content in this area was uploaded by Norris H Williams on Jan 08, 2018
Content may be subject to copyright.
ORIGINAL
ARTICLE
Orchid historical biogeography,
diversification, Antarctica and the
paradox of orchid dispersal
Thomas J. Givnish
1
*, Daniel Spalink
1
, Mercedes Ames
1
, Stephanie P. Lyon
1
,
Steven J. Hunter
1
, Alejandro Zuluaga
1,2
, Alfonso Doucette
1
,
Giovanny Giraldo Caro
1
, James McDaniel
1
, Mark A. Clements
3
,
Mary T. K. Arroyo
4
, Lorena Endara
5
, Ricardo Kriebel
1
, Norris H. Williams
5
and Kenneth M. Cameron
1
1
Department of Botany, University of
Wisconsin-Madison, Madison, WI 53706,
USA,
2
Departamento de Biolog
ıa,
Universidad del Valle, Cali, Colombia,
3
Centre for Australian National Biodiversity
Research, Canberra, ACT 2601, Australia,
4
Institute of Ecology and Biodiversity,
Facultad de Ciencias, Universidad de Chile,
Santiago, Chile,
5
Department of Biology,
University of Florida, Gainesville, FL 32611,
USA
*Correspondence: Thomas J. Givnish,
Department of Botany, University of
Wisconsin-Madison, 430 Lincoln Drive,
Madison, WI 53706, USA.
E-mail: givnish@wisc.edu
ABSTRACT
Aim Orchidaceae is the most species-rich angiosperm family and has one of
the broadest distributions. Until now, the lack of a well-resolved phylogeny has
prevented analyses of orchid historical biogeography. In this study, we use such
a phylogeny to estimate the geographical spread of orchids, evaluate the impor-
tance of different regions in their diversification and assess the role of long-dis-
tance dispersal (LDD) in generating orchid diversity.
Location Global.
Methods Analyses use a phylogeny including species representing all five orchid
subfamilies and almost all tribes and subtribes, calibrated against 17 angiosperm
fossils. We estimated historical biogeography and assessed the importance of dif-
ferent regions for rates of speciation, extinction and net species diversification.
We evaluated the impact of particular LDD events on orchid diversity by asking
how many species evolved in the new range subsequent to those events.
Results Orchids appear to have arisen in Australia 112 Ma (95% higher prob-
ability distribution: 102.0–120.0 Ma), then spread to the Neotropics via Antarc-
tica by 90 Ma (HPD: 79.7–99.5 Ma), when all three continents were in close
contact and apostasioids split from the ancestor of all other orchids. Ancestors
of vanilloids, cypripedioids and orchidoids+epidendroids appear to have origi-
nated in the Neotropics 84–64 Ma. Repeated long- and short-distance dispersal
occurred through orchid history: stochastic mapping identified a mean total of
74 LDD events or 0.8 Ma
1
. Across orchid history, Southeast Asia was the
most important source and maximally accelerated net diversification; across
epidendroids, the Neotropics maximally accelerated diversification.
Main conclusions Our analysis provides the first biogeographical history of
the orchids, implicating Australia, the Neotropics and Antarctica in their ori-
gin. LDD and life in the Neotropics –especially the Andes –had profound
effects on their spread and diversification; >97% of all orchid species are
restricted to individual continents.
Keywords
Asparagales, BioGeoBEARS, BiSSE, long-distance dispersal, Neotropics,
Southeast Asia
INTRODUCTION
Orchids are the largest family of angiosperms, with roughly
880 genera and 27,800 species. They comprise ~8% of all
vascular plants, grow in almost all terrestrial habitats except
the driest deserts and are native to all continents except
Antarctica (Pridgeon et al., 1999–2014; The Plant List, 2015).
Givnish et al. (2015) recently showed that the drivers of
ª2016 John Wiley & Sons Ltd http://wileyonlinelibrary.com/journal/jbi 1905
doi:10.1111/jbi.12854
Journal of Biogeography (J. Biogeogr.) (2016) 43, 1905–1916
extraordinary orchid diversity include the evolution of
pollinia, epiphytic habit, CAM photosynthesis, pollination
via Lepidoptera, euglossine bees, and deceit, and life in
extensive tropical cordilleras such as the Andes and New
Guinea Highlands. The defining characteristics of orchids –
minute seeds, germination aided by fungal symbionts, and
floral column of fused male and female parts –appear not
to have accelerated orchid speciation themselves but to have
interacted in several cases with the traits just mentioned to
generate exceptional levels of orchid diversity.
Across angiosperms, species richness is greater in families
with broader geographical ranges and latitudinal extents
(Ricklefs & Renner, 1994). Indeed, the area of ecozones or
continents occupied explains 50% of the variance in log-
transformed species richness in a phylogenetically structured
analysis of 409 angiosperm families (Vamosi & Vamosi,
2011). There remains the central question of whether large
areas or ecological volumes occupied cause large numbers of
species or vice versa (Ricklefs & Renner, 1994, 2000; Dodd
et al., 1999; Givnish et al., 2014). But given that orchids
occur on all continents save Antarctica, and have one of the
widest latitudinal ranges of any plant family –from 72°N for
Corallorhiza trifida in the Canadian Arctic Archipelago to
55°S for Chloraea, Codonorchis and Gavilea in Tierra del
Fuego (Pridgeon et al., 1999–2014) –it seems plausible that
high orchid diversity may partly reflect their broad
distribution.
This raises what we term the ‘paradox of orchid dispersal’.
On one hand, the broad distribution of orchids (and their
great diversity) might partly reflect their excellent dispersal,
conferred by the dust-like seeds of almost all species. On the
other hand, frequent long-distance seed dispersal should
work against differentiation within species and thus, ulti-
mately, speciation (Givnish, 2010). Dressler (1981) suggested
that only 34 orchid genera had been able to cross tropical
oceans and establish disjunct distributions on continents
long isolated from each other (i.e. South America, Africa and
Southeast Asia). Such disjunctions are seen within genera or
pairs of closely related genera in all orchid subfamilies except
Apostasioideae, suggesting that they may often have arisen
early enough for continental drift to affect present-day distri-
butions (Dressler, 1981; Chase, 2001; but see Guo et al.,
2012). Such inferences are not based, however, on actual
calculations tied to the ages of specific fossils. Only three
orchid species reached the Hawaiian Islands by natural
means (Wagner et al., 1990), suggesting limits imposed by
seed dispersal or by missing pollinators or fungal symbionts.
Dozens of closely related species of Teaguiea appear to have
speciated within a few kilometres of each other in Andean
Ecuador (Jost, 2004), which also seems consistent with
short-distance dispersal of orchid seeds or their mutualists
(Givnish et al., 2015).
Where did the orchids originate? To what extent do rela-
tionships above the generic level reflect intercontinental dis-
persal or vicariance events? Does the remarkable diversity of
orchids reflect their ability to disperse among continents, or
is that ability limited? Where did the key plant traits that
appear to have accelerated net rates of orchid diversification
arise?
Answers to these questions have been blocked by the lack
of a well-resolved, strongly supported backbone phylogeny
for the orchids. Givnish et al. (2015) recently derived such a
phylogeny using a phylogenomics approach, calibrated
against time using 17 angiosperm fossils. Here, we use this
tree to estimate the historical biogeography of Orchidaceae,
identify its area of origin, assess the roles of vicariance and
intercontinental dispersal, and locate the origin of epi-
phytism, a key trait that appears to have accelerated orchid
diversification (see also Chomicki et al., 2015). We test
whether occurrence on different continents had a significant
effect on rates of speciation, extinction and net species diver-
sification. Finally, we assess the extent by which long-dis-
tance dispersal (LDD) may have increased orchid diversity
by asking how many species evolved in new ranges after
LDD events. These across-family biogeographical analyses
complement those that have been conducted on several smal-
ler groups of orchids (Gravendeel et al., 2004; Bouetard
et al., 2010; Smidt et al., 2011; Guo et al., 2012, 2015; Dueck
et al., 2014; Freudenstein & Chase, 2015).
MATERIALS AND METHODS
Phylogenetics and tree calibration
Our study uses the supermatrix tree of Givnish et al. (2015),
based on sequences of 75 plastid genes for 39 orchid species
and 96 angiosperm outgroups, and three plastid genes for
another 162 orchid species. This analysis includes placehold-
ers for all five subfamilies, 18 of 19 tribes, and 40 of 43 sub-
tribes recognized by Chase et al. (2003), collectively
representing 99.6% of all orchid species. We continue using
the tribal and subtribal nomenclature of Chase et al. (2003)
rather than its recent re-arrangement by Chase et al. (2015),
to make our results regarding biogeography directly compa-
rable to those in our earlier article devoted to plant traits
(Givnish et al., 2015).
Givnish et al. (2015) calibrated the single maximum-likeli-
hood tree resulting from the supermatrix analysis against the
ages of 17 angiosperm fossils using beast 1.80 (Drummond
et al., 2012) and branch lengths based on atpB, psaB and
rbcL sequences. Dates and 95% highest posterior densities
(hereafter, HPD) were calculated for each node. Here, we
replace our sampling of Cypripedioideae with the more
extensive set of taxa studied by Guo et al. (2012). We were
unable to incorporate these directly in our dating analyses,
as the only locus shared between the two datasets was rbcL.
Instead, we replicated the BEAST analysis of Guo et al. and
grafted the resulting time-calibrated phylogeny of cypri-
pedioids onto ours. To refine modelling of geographical
diversification in tribe Vanilleae, we grafted 12 Vanilla tips
onto our tree, matching the topology and divergence times
of representatives of the lettered clades in groups band cof
Journal of Biogeography 43, 1905–1916
ª2016 John Wiley & Sons Ltd
1906
T. J. Givnish et al.
Bouetard et al. (2010), and used their results to graft Pogonia
japonica as sister to Pogonia minor and Cleistesopsis as sister
to Pogonia. We included placeholders for a total of 173
orchid genera (21.7% of those tabulated by Chase et al.
(2003).
Historical biogeography
We estimated ancestral areas on the modified BEAST
chronogram using the maximum-likelihood approach imple-
mented in BioGeoBEARS (Matzke, 2014). BioGeoBEARS
incorporates a founder-event parameter for ‘jump specia-
tion’, which allows cladogenetic dispersal outside parental
areas. To test the effect of the J-parameter on ancestral-area
estimation, we conducted two independent runs using DEC
(dispersal–extinction–cladogenesis analysis) and DEC+J; two
using DIVA and DIVA+J; and two using BayArea and
BayArea+J. A priori, we considered DEC superior to the
other models because it permitted speciation via the biologi-
cally relevant mechanisms of widespread vicariance and sub-
set sympatry. We conducted likelihood ratio tests based on
AICc scores on nested models to assess goodness of overall
fits.
All tips were coded as present/absent in seven geographical
regions (North America, Neotropics, Eurasia, Africa, South-
east Asia, Australia, Pacific) based on generic distributions
acquired from Pridgeon et al. (1999–2014), eMonocot Team
(2015) and WCSP (2015). Regional boundaries are discussed
in Appendix S1.Representatives of individual genera were
coded based on the entire distribution of that genus, except
for species of Cypripedioideae and Vanilloideae grafted onto
the phylogeny to represent lineages from different regions. In
this placeholder analysis, terminal branch lengths to the taxa
included are assumed to be accurate proxies for branch
lengths to all species within genera, subtribes or tribes; devia-
tions from this assumption are negligible at the scale of the
entire orchid phylogeny. Ancestral-area estimations included
representatives of Iridaceae and all astelid families (Astelia-
ceae, Blandfordiaceae, Boryaceae, Hypoxidaceae, Lanari-
aceae); the clade composed of the latter five families root at
the node immediately above Orchidaceae in order
Asparagales.
Relative dispersal probabilities among regions were con-
strained based on their emergent areas (particularly for Paci-
fic islands) and widths of water barriers between areas
during five time slices (0–2, 2–30, 30–60, 60–90 and 90–
150 Ma; see Table S1 of Appendix S1). Dispersal probabili-
ties between Australia and South America take into account
their connection via then-warmer Antarctica during the two
oldest periods (Givnish et al., 2015). We considered move-
ment between regions to be (1) long-distance dispersal
(LDD) if it occurred over a substantial distance of water at
the time of dispersal; (2) short-distance dispersal (SDD)
across the short distances of water that separate Southeast
Asia and Australia at Wallace’s Line, or intermittently sepa-
rate Eurasia and North America at Beringia; and (3) simple
geographical spread if regions were contiguous at the time of
dispersal (e.g. Southeast Asia and Eurasia, or North America
and the Neotropics after these areas came into near contact
by the Middle Miocene [Montes et al., 2012]). Distributions
were considered a result of vicariance if ancestors had spread
across formerly contiguous regions and then became extinct
in an intervening region (e.g. Antarctica).
Stochastic mapping
We performed biogeographical stochastic mapping to esti-
mate the number of dispersal events between each region,
based on the dispersal and extinction parameters of our
model, presence/absence of the terminal taxa in each region,
and our ancestral-area estimation using BioGeoBEARS. Fol-
lowing Matzke (2014), we calculated the average numbers of
anagenetic range expansions (e.g. A–>A+B), extinctions (e.g.
A+B–>A), range switches (A–>B), cladogenetic range
expansions involving sympatry or vicariance, and jump-dis-
persal events (e.g. A–>A, B) from 50 stochastic maps to esti-
mate the total number and directionality such events during
orchid evolution. Although the number of placeholder taxa
(~250) is small relative to the total number of species, the
excellent representation at the subtribal level combined with
many subtribes being restricted to a single region reduces the
noise expected in this analysis. Conversely, distributional
variation among species within genera or subtribes unaccom-
panied by dense sampling within those groups is likely to
increase the variation detected via stochastic mapping.
Geographical correlates of rates of speciation,
extinction and net diversification
We explored correlations between distribution in different
regions and apparent rates of speciation, extinction and net
species diversification using the BiSSE model (Maddison
et al., 2007) implemented in the R package Diversitree
(FitzJohn et al., 2009). BiSSE cannot calculate likelihoods on
unresolved tips representing more than 190 taxa, so the
numbers of species at all tips were down-weighted by a fac-
tor of 25, with small clades rounded up to 1 (Givnish et al.,
2014, 2015).
To test whether presence in a region affected diversifica-
tion, we estimated the proportion of species present in each
region for each of the taxonomic units comprising the tips
of a pruned chronogram restricted to the set of subfamilies,
tribes and subtribes employed by Givnish et al. (2015). We
conducted analyses for orchids as a whole and for Epiden-
droideae. Data on the numbers of species in each taxonomic
unit were drawn from Chase et al. (2003); data on the frac-
tion of species distributed within and outside each region
were derived from Pridgeon et al. (1999–2014), eMonocot
Team (2015) and WCSP (2015). Where necessary, tallies of
species occurrences for individual genera in each of the
regions were down-weighted by the ratio of currently tabu-
lated species to the total fide Chase et al. (2003). This
Journal of Biogeography 43, 1905–1916
ª2016 John Wiley & Sons Ltd
1907
Orchid biogeography and diversification
approach sidesteps issues raised by species, subtribes or tribes
occurring in multiple regions by focusing, one region at a
time, on the apparent effects of species presence versus
absence in that region. This simplification seems reasonable,
given that many of our taxonomic units and the great major-
ity of individual species are restricted to single regions.
For each region, an unconstrained model for diversifica-
tion was compared to models where speciation (k), extinc-
tion (l) and character-state transition rates (q) were
individually constrained (k
0
=k
1
,l
0
=l
1
,q
01
=q
10
). Likeli-
hoods of constrained models were compared to the uncon-
strained model, and significance of likelihood scores assessed
using ANOVA. Net rates of diversification were calculated as
kl. For each region, we measured the advantage in net
diversification per million years within lineages conferred by
a character state as f=exp((k
1
l
1
)(k
0
l
0
)) 1
(Givnish et al., 2015). We used q
10
as a measure of the
extent to which a region served as a source of dispersal per-
taxon present; q
01
measured the extent to which it served as
a target for dispersal from other regions per-taxon present in
those regions.
Some authors have criticized BiSSE analyses as being sen-
sitive to small sample sizes and extreme tip bias (Davis
et al., 2013) or correlations with characters not studied
(Rabosky & Goldberg, 2015). Even after down-weighting by
a factor of 25, we have 249 tips in the main analysis –close
to the 300 recommended by Davis et al. Our epidendroid
analysis, however, has only 99 tips, so inferences from it
must be viewed with caution. Three of our seven regions
(North America, Eurasia, Pacific) have <10% of the taxa
(the threshold identified by Davis et al.), but none have
especially high rates of speciation or diversification, so the
reduced precision of rate inference for these would likely not
be of great consequence (see Results). We have previously
summarized the morphological and ecological correlates of
high rates of orchid speciation and diversification (Givnish
et al., 2015) and discussed their wide distribution in the
tropics, so confounding factors are unlikely to be an issue
here.
Effects of intercontinental movements on orchid
diversity
We calculated the number of species resulting from each of
several selected LDD events across tropical oceans, based on
the position of such events in the pruned chronogram and
the number of species found per region in each tip. We
studied (1) all LDD events, (2) the LDD event preceding the
rise of the upper epidendroids and (3) the LDD events pre-
ceding the origins of the pleurothallid alliance, of Cymbideae
minus Cymbidineae and of Angraecinae and Aerangidinae.
We also identified the region associated with the origin of
epiphytism at the base of the upper epidendroids –inferred
by Givnish et al. (2015) –and whether the spread of epi-
phytism required additional LDD events.
RESULTS
Models including jump speciation provided a significantly
better fit than those without (Table S2), so we used them to
estimate historical biogeography. There were very few differ-
ences in ancestral age estimation using DEC+J, DIVA+J and
BayArea+J, and so here we report the findings based on the
biologically more reasonable DEC+J model. All dates
described below are stem ages unless otherwise indicated.
This model places the most likely origin of orchids in Aus-
tralia 112 million years ago (Ma) (HPD: 102.0–120.0 Ma)
(Fig. 1), when Africa and India/Madagascar had already sep-
arated from Antarctica and Australia, but Australia and
South America remained connected via Antarctica. By the
time that apostasioids diverged from other orchids 90 Ma
(HPD: 79.7–99.5 Ma), the most likely distribution of the
orchid crown was Australia+Neotropics, with the common
ancestors of the vanilloids, cypripedioids and orchidoids+epi-
dendroids then originating most likely in the Neotropics
roughly 84, 76 and 64 Ma, respectively (HPDs: 74.4–92.9,
64.6–87.4 and 54.8–73.7 Ma) (Fig. 1). Stem orchids arose in
Australia under both DIVA+J and BayArea+J as well, with
the initial orchids in Australia and the Neotropics under
DIVA+J and in the Neotropics alone in BayArea+J. Under
Dec+J, apostasioids reached Southeast Asia sometime during
the last 43 Ma (HPD: 24.2–65.0 Ma) (Fig. 1), presumably
via collision of the Australian and Eurasian Plates and short-
distance dispersal.
Within vanilloids, tribe Pogonieae apparently underwent
long-distance dispersal from the Neotropics to North Amer-
ica ca. 44 Ma (HPD: 33.5–54.9 Ma), spawning Cleistesiopsis,
Isotria and Pogonia, with the latter moving to Eurasia within
the last 11 Myr (HPD: 9.1–26.2 Ma) (Fig. 1). Tribe Vanilleae
originated in the Neotropics and underwent long-distance
dispersal across the Pacific 64–59 Ma (HPDs: 55.1–73.4 Ma,
47.9–69.0 Ma) to New Caledonia, leading to Clematepis-
tephium and Eriaxis.Vanilla originated in the Neotropics ca.
61 Ma (HPD: 51.6–70.0 Ma), underwent LDD to Africa
26–18 Ma, and then moved from Africa via LDD into the
Indian Ocean 13 Ma, and from Africa to the Caribbean
12–4 Ma (last several dates based on grafted branches)
(Fig. 1). The remaining clade (Pseudovanilla through Cyr-
tosia) appears to have originated in Australia or Southeast
Asia 61 Ma (HPD: 51.6–70.0 Ma) after LDD from the
Neotropics, and then spread into Southeast Asia and further
overland into East Asia in some Cyrtosia.Pseudovanilla
appears to have undergone LDD to Pohnpei and Fiji in the
Pacific 31–6 Ma (HPDs: 22.3–58.5 Ma, 2.1–11.2 Ma).
Subfamily Cypripedioideae and the genera Cypripedium,
Selenipedium, Phragmipedium and Mexipedium appear to
have arisen in the Neotropics roughly 76, 31, 28, 23 and
21 Ma, respectively (HPDs: 64.6–87.4, 30.2–43.0, 18.0–39.2,
14.1–33.3 and 12.4–31.5 Ma) (Fig. 1). Cypripedium spread
into Eurasia (and rarely, Southeast Asia) and then back into
North America several times. Paphiopedilum appears to have
Journal of Biogeography 43, 1905–1916
ª2016 John Wiley & Sons Ltd
1908
T. J. Givnish et al.
100 80 60 40 20 0
Millions of years ago
Chloraea 3
Chloraea 2
Chloraea 1
Megastylis glandulosa
Pterostylis 2
Pterostylis 1
Pachyplectron 2
Pachyplectron 1
Gonatostylis
Pristiglottis
Goodyera
Kreodanthus
Erythrodes
Platythelys
Prescottia 2
Prescottia 1
Stenoptera
Altensteinia
Gomphichis
Porphyrostachys
Aa
Myrosmodes
Pterichis
Cranichis
Ponthieva
Baskervilla
Cyclopogon
Spiranthes 2
Spiranthes 1
Stenorrhynchos
Sarcoglottis
Sauroglossum
Pelexia
Palmorchis
Cephalanthera
Epipactis
Listera
Neottia
Elleanthus
Sobralia
Xerorchis
Monophyllorchis
Triphora
Nervilia
Tropidia
Corymborkis
Coelogyne
Dendrochilum
Arundina
Bletilla
Thunia
Glomera
Calopogon
Arethusa
Eleorchis
Liparis lilifolia
Oberonia
Liparis viridiflora
Dendrobium
Cadetia
Bulbophyllum
Epigeneium
Phreatia
Eria
Trichotosia
Podochilus
Appendicula
Acanthephippium
Spathoglottis
Collabium
Nephelaphyllum
Phaius
Calanthe
Tipularia
Calypso
Dactylostalix
Govenia
Cremastra
Aplectrum
Oreorchis
Corallorhiza
Coelia
Earina 2
Earina 1
Bletia
Chysis
Arpophyllum
Encyclia
Epidendrum
Cattleya
Meiracyllium
Isochilus
Dilomilis
Restrepia
Masdevalia
Pleurothallis
Polystachya
Bromheadia
Amesiella
Cleisostoma
Neofinetia
Phalaenopsis
Aeranthes
Angraecum
Aerangis
Diaphananthe
Grammatophyllum
Cymbidium
Cyrtopodium
Galeandra
Catasetum
Dressleria
Dipodium
Eulophia
Stellilabium
Oncidium
Eriopsis
Zygopetalum
Cryptarrhena
Huntleya
Dichaea
Acineta
Kegeliella
Houlletia
Stanhopea
Coryanthes
Lycomormium
Lycaste
Neomoorea
Xylobium
Maxillaria
Cryptocentrum
SAP
NTE
NTESAP
TAP
TAP
a
b
c
d
e
f
g
h
i
j
k
l
n
m
o
p
q
r
s
t
u
v
w
x
y
z
aa
ab
ac
ad
ae
af
6
7
8
9
10
11
12
13
1
2
3
4
5
E
O
North America (N)
Neotropics (T)
Eurasia (E)
Africa (F)
Southeast Asia (S)
Australia (A)
Pacific (P)
NTEFSAPNTEFSAP
Figure 1 Diagram depicting estimation of orchid historical biogeography using BioGeoBEARS. Pie diagrams at each node denote the
geographical regions (or combinations thereof) inferred to have been occupied by ancestral taxa. Wedge colour indicates ancestral region (see
inset), wedge width, the probability of that region or combination of region. Combination of regions are indicated by hatching of colours or
by lettering; white wedges indicate the sum of origins in regions (or regional combinations) with individual probabilities <15%. Distributions
of genera and individual species are indicated by coloured boxes. Vertical lines with numbers and letters indicate subtribes, tribes and
subfamilies. Complete names for these lineages are provided in Figure S2, together with the pie diagrams on the shoulders of each node.
Journal of Biogeography 43, 1905–1916
ª2016 John Wiley & Sons Ltd
1909
Orchid biogeography and diversification
colonized Southeast Asia from the Neotropics 46 Ma (from
grafted branch), most likely involving long-distance dispersal
via boreotropical migration through Beringia.
Orchidoids arose via range expansion from the Neotropics
to include Africa, together with Australia and/or the Pacific,
with Orchideae spreading from Africa into Eurasia and
ultimately North America and Japan. Codonorchis remained
in South America, where it today occurs in the far south.
Diurideae spread to Australia and then to Southeast Asia,
New Zealand and New Caledonia. Cranichidae underwent a
major diversification in the Neotropics, with dispersal to
North America in Cranichis and Spiranthes, and dispersal to
100 80 60 40 20 0
Millions of years a
g
o
Iridaceae
Boryaceae
Blandfordiaceae
Astelia
Neoastelia
Milligania
Lanariaceae
Pauridia
Rhodohypoxis
Hypoxis
Apostasia
Neuwiedia
Duckeella
Cleistes rosea
Isotria medeoloides
Isotria verticillata
Cleistesiopsis
Pogonia ophioglossoides
Pogonia japonica
Pogonia minor
Epistephium 2
Epistephium 1
Eriaxis
Clematepistephium
Pseudovanilla 2
Pseudovanilla 1
Erythrorchis 2
Erythrorchis 1
Lecanorchis
Galeola
Cyrtosia 2
Cyrtosia 1
V. inodora
V. mexicana
V. planifolia
V. palmarum
V. africana
V. crenulata
V. albida
V. aphylla
V. imperialis
V. humblottii
V. phalaenopsis
V. barbellata
Vanilla dilloniana
Selenipedium
Mexipedium
Ph. besseae
Ph. caricinum
Ph. longifolium
Ph. exstaminodium
Phragmipedium lindleyanum
Pa. vietnamense
Pa. delenatii
Pa. bellatulum
Pa. wardii
Pa. hirsutissimum
Pa. primulinum
Pa. dianthum
Paphiopedilum adductum
Cy. molle
Cy. irapeanum
Cy. debile
Cy. subtropicum
Cy. acaule
Cy. palangshanense
Cy. fasciculatum
Cy. bardolphianum
Cy. margaritaceum
Cy. flavum
Cy. passerinum
Cy. japonicum
Cy. californicum
Cy. candidum
Cy. tibeticum
Cypripedium farreri
Disperis
Corycium
Disa
Satyrium
Cynorkis
Habenaria
Orchis
Ophrys
Dactylorhiza
Platanthera
Galearis
Codonorchis
Acianthus
Corybas
Adenochilus
Eriochilus
Caladenia
Glossodia
Cyanicula
Microtis
Cryptostylis
Coilochilus
Orthoceras
Diuris
Rhizanthella
Aporostylis
Thelymitra
Calochilus
Lyperanthus
Rimacola
Leporella
Megastylis latissima
Chiloglottis
Megastylis rara
SAP
NES
TFA
TFP TFAP
TAP
TAP
NTFESA
TA
TA
TA
ag
ah
ai
aj
ak
al
am
an
ao
ap
14
15
16
17
18
O
C
V
A
NTEFSAPNTEFSAPNTEFSAP
Figure 1 Continued.
Journal of Biogeography 43, 1905–1916
ª2016 John Wiley & Sons Ltd
1910
T. J. Givnish et al.
Australia and the Pacific in Pachyplectron, Pterostylis and
relatives (Fig. 1).
Epidendroids appear to have arisen in the Neotropics ca.
64 Ma (HPD: 54.8–73.7 Ma), with subsequent dispersal to
Eurasia, Southeast Asia and North America in Neottieae.
Dispersal to Southeast Asia occurred in the vast clade of epi-
dendroids sister to Sobralieae +Triphoreae, including Tropi-
deae and Nervileae (Fig. 1). Two long-distance dispersals
back to the Neotropics from Southeast Asia appear to have
occurred: in the large clade allied with the pleurothallids
(Coelia-Agrostrophyllinae-Bletiinae-Laeliinae-Pleurothallidinae-
Ponerinae) roughly 32 Ma (28.4–34.2 Ma), with LDD to the
Pacific in Agrostophyllinae (Earina)30Ma(HPD:26.7–
35.1Ma); and in Cymbideae minus Cymbidinae, between 31
and 24 Ma (HPDs: 26.5–34.9 and 17.5–26.7 Ma), with dis-
persal in Cymbidinae to Australia, the Pacific basin, and Eur-
asia in the last 17 Ma (HPD: 10.3–24.0 Ma) (Fig. 1).
Calypsoeae appears to have undergone LDD from Southeast
Asia to North America 32 Ma (HPD: 28.4–35.9 Ma) and
then spread back to Eurasia and Southeast Asia at least twice.
Arethuseae appears to have undergone LDD from Southeast
Asia from North America 15 Ma (HPD: 8.5–18.4 Ma) and
then dispersed back to Southeast Asia to form Eleorchis.
Aerangidinae and Angraecinae of Vandeae appear to have
arisen in Africa after LDD from Southeast Asia 21 Ma (HPD:
17.5–24.0 Ma). Secondary LDDs from Southeast Asia to
Australia and the Pacific Basin appear to have occurred in
Collabinae, Podochileae, Dendrobieae and Phreatia (Fig. 1).
The main difference between the DEC+JandBAYAREA+J
estimations is that movements to the Neotropics in Cym-
bideae and Epidendreae are from Southeast Asia in the for-
mer and from Australia and Southeast Asia in the latter
(Fig. S1 of Appendix 1).
Stochastic mapping
BioGeoBEARS inferred an average total of 15.0 jump-disper-
sal events and 121 anagenetic dispersal events; there were
also 225.9 cladogenetic events involving extinction but not
dispersal (Tables S3–S5 of Appendix S1). Sixteen of 69
jump-dispersal scenarios had a mean occurrence ≥0.3, total-
ling 10.3 of 15.0 events; 51 of 312 anagenetic scenarios had a
mean occurrence ≥0.3, totalling 121 of 138.7 events. Of the
more common jump-dispersal events, 8.4 (82%) involved
LDD, 1.5 (14%) involved SDD across Wallace’s Line or Ber-
ingia and 0.44 (4%) involved simple spread over contiguous
areas (Table S4). Of the more common anagenetic dispersal
events, 65.9 (55%) involved LDD, 40.3 (34%) involved SDD
and 12.6 (11%) involved simple geographical spread. Our
analysis, thus, identified at least 74.3 LDD events in the his-
tory of orchid evolution (Table S5). Most anagenetic migra-
tions were between adjoining regions.
Effects of geographical distribution on
diversification
Different regions had different effects on speciation, extinc-
tion and net species diversification (Tables 1, S6). Across
orchid history, presence in Southeast Asia led to the highest
advantage in net diversification rate relative to other regions
per million years (f=15.4%), with the second highest
advantage in Australia (f=7.7%); the remaining regions
had only small advantages or disadvantages in diversification.
Speciation and extinction rates were significantly lower in
Southeast Asia and Australia than elsewhere. In general,
diversification rates were not correlated significantly with k
1
or l
1
, but –as expected –were highly correlated with
Table 1 Apparent rates of speciation (k), extinction (l) and character-state transition (q) associated with presence in particular regions for
orchids as a whole and for epidendroids. States with significantly higher rates are indicated by *(P<0.05), **(P<0.01) and ***(P<0.001).
D
1
=(k
1
l
1
) measures the diversification rate associated with a particular distribution; f=exp((k
1
l
1
)(k
0
l
0
)) 1measuresthe
advantage in net diversification (r=kl) per million years within lineages conferred by that distribution.
k
1
k
0
l
1
l
0
D
1
fq
01
q
10
Orchidaceae
Southeast Asia 0.159 0.751*** 4.4 910
5
0.735*** 0.159 15.4% 0.0014 0.117***
Australia 0.119 0.662*** 2.5 910
6
0.616*** 0.119 7.7% 0.0015 0.0791***
Neotropics 0.471 0.311 0.388 0.260 0.083 3.2% 0.0075*0.0006
Africa 0.280 0.503 0.236 0.454 0.044 0.2% 0.0037 0.0309*
Eurasia 0.049 0.406*** 1.5 910
5
0.339*** 0.049 1.8% 0.0017 0.0467***
North America 0.143 0.440*0.111 0.378 0.032 3.0% 0.0057*0.0007
Pacific 7.0 910
6
0.477*** 2.0 910
5
0.422*** 0.000 5.6% 0.0047 0.0414**
Epidendroideae
Neotropics 0.186 0.316*** 4.9 910
6
0.003*** 0.186 11.5% 0.0033*** 2.6 910
8
Australia 0.126 0.372*3.2 910
7
0.255*0.126 9.1% 0.0007 0.0000
Southeast Asia 0.129 0.425*5.3 910
6
0.310 0.129 1.5% 8.1 910
4
0.0510***
Eurasia 0.042 0.281*** 7.0 910
6
0.153 0.042 1.5% 0.0410*0.0015
Africa 0.193 0.464 0.086 0.381 0.107 0.2% 0.0028 0.0320*
Pacific 0.014 0.439*** 2.5 910
8
0.345*** 0.014 7.7% 0.0048 0.0444
North America 0.297 0.392*** 0.756** 0.297 0.459 42.6% 0.0058 1.9 910
5
Journal of Biogeography 43, 1905–1916
ª2016 John Wiley & Sons Ltd
1911
Orchid biogeography and diversification
(k
1
k
0
)(r=0.988***) and (l
1
l
0
)(r=0.904***).
Southeast Asia was also the strongest per-taxon source area,
reflecting the proximity of Eurasia to New Guinea and off-
shore islands. Australia was the second strongest source.
Across epidendroids, presence in the Neotropics led to the
highest diversification rate relative to other regions
(f=11.5%) (Table 1). Southeast Asia was again the stron-
gest souce; Eurasia was the strongest target. North America
had by far the greatest disadvantage in diversification
(f=42.6% Myr
1
), presumably because it is the least
tropical of the regions and epidendroids are mostly tropical
in distribution. Speciation and extinction rates were not cor-
related significantly with relative diversification rates, or rela-
tive advantage in extinction rates (l
1
l
0
), but speciation
rates were highly correlated with (k
1
k
0
)(r=0.988***).
Southeast Asia was both the strongest per-taxon source and
target during epidendroid evolution (Table 1).
Effects of dispersal on diversification at large spatial
scales
The products of all LDD events following the vicariant spread
of the family to the Neotropics and Australia include all orch-
ids save ca. 675 species, or 2.7% of all orchids tabulated by
Chase et al. (2003). Just one LDD event, from the Neotropics
to Southeast Asia roughly 48 Ma (HPD: 40.5–56.9 Ma) –in
the common ancestor of Nervilieae, Tropidieae and the upper
epidendroids (Fig. 1) –precedes diversification of 79.4% of all
orchid species within the new range. A later LDD event, from
Southeast Asia to the Neotropics roughly 31 Ma (25.0–
38.9 Ma) –in the common ancestor of Bletiinae, Laeliinae,
Pleurothallidinae and Ponerinae –accounts for 5857 species,
or 23.5% of all orchid species. The LDD event in the same
direction that led to Cymbideae minus Cymbidinae 24 Ma
(HPD: 20.1–28.1 Ma) accounts for 14.9% of all orchids; the
LDD event from Southeast Asia to Africa 21 Ma (HPD: 16.1–
26.5 Ma) that led to Angraecinae and Aerangidae generated
3.1% of all orchids.
Overall, 160 orchid species among the clades tabulated
occur in North America, 11,448 in the Neotropics, 595 in
Eurasia, 2583 in Africa, 5444 in Southeast Asia, 4405 in Aus-
tralia and 777 in the Pacific. Given the total number of
orchid species in these clades, this implies that only 2.6% of
individual orchid species occur in more than one region.
Epiphytism appears to have arisen in the upper epidendroids
in Southeast Asia, with the spread of epiphytic orchids
throughout the tropics requiring several subsequent LDD
events.
DISCUSSION
Orchids appear to have arisen in Australia 112 Ma (see
Results for HPDs of this and all other phylogeny-based dates
mentioned below) and then migrated to South America via
Antarctica by 90 Ma (Fig. 1). Between 100 and 48 Ma, global
temperatures reached exceptional highs driven by rising
atmospheric CO
2
levels; Antarctica had a summer maximum
temperature of ca. 20°C and supported tropical to subtropi-
cal vegetation (Pross et al., 2012). A growing list of monocot
genera and families are known to have a disjunct distribution
in Australia and South America, represented by extant or
fossil taxa, supporting an Antarctic connection (Conran
et al., 2015). The origin of Liliales in Australia (Givnish
et al., 2016) and the origin of the commelinid orders in
South America –then connected to Antarctica and thus Aus-
tralia (Givnish et al., 1999) –bracket the origin of the orch-
ids and reinforce the conclusion that orchids arose in
Australia. Fossil pollen from Antarctica from 52 Ma indicates
the presence of palms and near-tropical rain forests even at
that late date (Pross et al., 2012). It is ironic that Antarctica
–the one continent where orchids currently are not found –
appears to have played a key role in early orchid evolution.
Dressler (1981) argued that, of the 33 orchid genera with
disjunct distributions on continents separated by tropical
oceans (South America, Africa and Southeast Asia), only five
might be old enough to reflect vicariance based on continen-
tal drift. However, our data indicate that these ‘old’ disjunc-
tions –in Vanilla, Corymborkis, Palmorchis-Diceratostele,
Epistephium-Clematepistephium and Tropidia –are in fact
too recent to reflect continental drift and instead seem to be
products of long-distance dispersal. Dressler (1981) recog-
nized the distinctiveness of the orchid floras on each of the
three tropical continental regions and suspected that they
evolved after they had separated from each other. Our data
support a more complex scenario, with multiple introduc-
tions to each region. Dressler realized that exchanges between
North America and Eurasia could have occurred frequently
during the Cenozoic as a result of intermittent land bridges
across Beringia and the North Atlantic. Based on early, time-
calibrated broad-scale phylogenies for angiosperms, Chase
(2001) concluded that orchids likely arose around 110 Ma
and might have evolved on Gondwana. However, he argued
that trying to identify the continent on which they arose was
futile, given how close the continents were to each other
then.
Our biogeographical estimation suggests that 15 transocea-
nic LDD events resulted in the origin of orchid tribes or
smaller groups restricted or nearly so to individual conti-
nents, and occurred long after the breakup of Gondwana.
These events were (1–2) from Southeast Asia to Neotropics
twice, forming Cymbideae minus Cymbidinae and the pleu-
rothallid alliance; (3) from Southeast Asia to Africa, forming
Angraecinae +Aerangidae; (4) from Southeast Asia to North
America, forming large parts of Calypsoeae; (5–6) from
Neotropics to Southeast Asia twice, forming Neottieae and
the upper epidendroids +Tropideae +Nervilieae; (7–9) from
Neotropics to Africa forming Orchideae, to Australia form-
ing Diurideae, and to the Pacific forming Pterostylidinae and
part of Goodyerinae; (10–11) from the Pacific to Neotropics
twice, forming part of Goodyerinae and Spiranthinae; (12)
from Neotropics to Africa, forming part of Vanilla; (13)
from Africa back to the Caribbean, forming more of Vanilla;
Journal of Biogeography 43, 1905–1916
ª2016 John Wiley & Sons Ltd
1912
T. J. Givnish et al.
(14) from Neotropics to Southeast Asia, forming part of
Vanilleae; and (15) from Neotropics to the Pacific, forming
Clematepistephium +Eriaxis (Fig. 1). The movement of apos-
tasioids from Australia to New Guinea and thence across
Wallace’s Line to Southeast Asia presumably occurred after
the Australian, Eurasian and Pacific plates began to collide
25 Ma, causing the uplift of New Guinea and nearby islands
(Pigram & Davies, 1987).
Over time, relatively low rates of dispersal –and/or disper-
sal over relatively short distances –should maximize specia-
tion and net rates of diversification. At the scale of our
sampling, orchids underwent about 74 transoceanic LDD
events in 90 Ma, or 0.8 events Ma
1
–not a very high rate,
less than ca. 1.3 events Ma
1
for Dryopteris over the past
25 Ma (Sessa et al., 2012), and far less if expressed in events
species
1
Ma
1
.
Integrated across orchid history, Southeast Asia had the
highest net diversification rates and was the strongest source
area (Table 1). The geological history of Southeast Asia is
unusually complex, reflecting early collisions of Gondwanan
fragments with the Eurasian plate and later collisions of the
latter with the Australian, Pacific and Philippine plates (Hall,
2009). By 80 Ma, Gondwanan fragments had sutured to Eur-
asia, forming western Malesia, much of which remained
above water after 65 Ma. In the middle Eocene, rifting of the
Makassar Strait formed a deep channel that persisted
through the Cenozoic and separated the Australian plate
from emergent parts of the Sunda Shelf (Hall, 2009).
Numerous north–south mountain ranges (e.g. Tenasserim,
Kayah-Karen, Luang Prabang) mark mainland Southeast Asia
and grade into the Hengduan and Himalayan massifs to the
west; all were uplifted by continuing collision of the Indian
plate with Eurasia over the past 50 Myr and provide exten-
sive venues for speciation (Yu et al., 2015). Importantly,
these mountains and the volcanoes of insular Southeast Asia
are most likely where epiphytism evolved in epidendroids
and then spread around the world. Repeated fluctuations in
sea level and climate fragmented rain-forest areas during the
Pleistocene and probably fostered extensive speciation in
insular and mainland Southeast Asia (Guo et al., 2012, 2015;
Thomas et al., 2012). Southeast Asia is one the world’s lead-
ing hotspots for plant diversity, with 42,000 species of vascu-
lar plants (Brooks et al., 2006). Integrated biogeographical
studies of 29 plant and animal clades clearly separate South-
east Asia west of Wallace’s Line from Australia, New Guinea
and Pacific islands (Turner et al., 2001). Southeast Asia’s
position would have facilitated orchid dispersal into the
adjoining Australian region and warmer parts of contiguous
Eurasia.
Across all orchid history, Australia is second in net diversi-
fication rate and the second strongest source area for migra-
tion (Table 1). Most orchids in this region are restricted to
New Guinea, Sulawesi and nearby islands, all of which have
complex histories. Western Sulawesi rafted from Borneo in
the middle Eocene; eastern Sulawesi, the Moluccas, the lesser
Sunda islands and New Guinea resulted from collision of the
Australian, Eurasian and Pacific plates 25 Ma (Pigram &
Davies, 1987; Hall, 2009). Much of this area emerged only
during the late Miocene, but southeastern Sulawesi and smal-
ler, short-lived islands emerged by 20 Ma (Hall, 2009) and
presumably facilitated short-distance dispersal of apostasioids
from the Australian mainland (see Results). The New Guinea
orogeny began 10 Ma, with over 30 smaller terranes assem-
bled on the northern periphery of the oncoming Australian
plate (Hall, 2009; Baldwin et al., 2012). The great area,
height and topographic complexity of the New Guinea High-
lands favoured the evolution of large numbers of orchids
over the last 10 Myr (De Vogel & Schuiteman, 2001;
Schuiteman et al., 2010; Givnish et al., 2015). New Guinea’s
position, in turn, would have facilitated dispersal into South-
east Asia and tall, wet islands of the southwest Pacific.
It is surprising that the Neotropics showed such a low
advantage (3.2% Myr
1
) relative to other regions in net
diversification, given that the fastest speciating groups of
orchids (viz., Pleurothallidinae, Laeliinae and close relatives
[Givnish et al., 2015]) are found in tropical America. How-
ever, the Neotropics were also the cradle of several ancient,
low-diversity lineages (viz., Vanilloideae, Cypripedioideae)
with very low rates of diversification (Givnish et al., 2015).
When we restrict attention to more recent history and focus
on epidendroids (crown group ca. 48 Ma), the Neotropics
are indeed characterized by the highest rates of diversifica-
tion (f=11.5% in Table 1) and speciation relative to other
regions (k
1
k
0
=0.186). These unusually high rates may
be tied to the uplift of the northern Andes over the past
15 Myr and increased access to Central America from South
America over the same period (Givnish et al., 2014). Both
areas are geographically extensive and climatically and topo-
graphically complex. The Andes especially have been shown
to foster rapid speciation in bromeliads and orchids (Givnish
et al., 2014, 2015) and many other groups (Hughes et al.,
2013), and they are the leading hotspot for plant biodiversity
worldwide (Brooks et al., 2006). Two exceptionally diverse
groups –the pleurothallid alliance and Cymbideae minus
Cymbidinae –are largely restricted to the northern Andes
and Central America. Detailed phylogenies and biogeographi-
cal estimations at finer geographical scales of these two
clades could be used to test these hypotheses. Both should
show a substantial increase in speciation and the narrowness
of species ranges starting 15 Ma.
Finally, the relatively low rate of long-distance dispersal by
orchids worldwide appears to have played a key role in their
diversification. The products of all LDD events following the
vicariant origin of orchids in Australia and the Neotropics
include 97.3% of present-day orchid species. This suggests
that access provided by LDD to other continental areas –
and the possibilities for speciation, range expansions and
avoidance of extinction afforded by their landforms, geo-
graphical extent, pollinators and mycorrhizal partners –
increased orchid diversity by 36-fold, directly and indirectly.
LDD from the Neotropics to Southeast Asia at the base of
the upper epidendroids +Sobralieae +Triphoreae alone
Journal of Biogeography 43, 1905–1916
ª2016 John Wiley & Sons Ltd
1913
Orchid biogeography and diversification
appears to have multiplied orchid diversity fivefold; LDD
from Southeast Asia to the Neotropics 32 Ma to form the
pleurothallid alliance alone appears to have generated nearly
one-quarter of present-day species. Yet limited dispersal of
dust-like seeds in tropical mountains is at least one possible
driver of high local diversity (Givnish et al., 2015). Thus,
while pollinia, epiphytism, life in extensive tropical cordil-
leras and specialization on particular groups of pollinators all
accelerated orchid diversification (Givnish et al., 2015), infre-
quent long-distance dispersal also played an important role
by permitting occasional access to new regions and pools of
mutualists. Geographically, the three significant accelerations
of net species diversification (Givnish et al., 2015) corre-
spond to (1) LDD from the Neotropics to Africa, Australia
and the Pacific basin; (2) LDD from the Neotropics to
Southeast Asia, followed by invasion of higher elevations and
the evolution of epiphytism; and (3) invasion of the north-
ern Andes by the pleurothallid alliance. We stand by our
original interpretation of these accelerations as reflecting the
origins of (1) pollinia, (2) epiphytism and (3) invasion of an
extensive tropical cordillera, but here we identify where these
key morphological and ecological shifts occurred.
ACKNOWLEDGEMENT
This research was supported by the NSF grant DEB-0830836
to T.J.G. under the Assembling the Tree of Life Program.
REFERENCES
Baldwin, S.L., Fitzgerald, P.G. & Webb, L.E. (2012) Tectonics
of the New Guinea region. Annual Review of Earth and
Planetary Science,40, 495–520.
Bouetard, A., LeFeuvre, P., Gigant, R., Bory, S., Pignal, M.,
Besse, P. & Grisoni, M. (2010) Evidence of transoceanic
dispersion of the genus Vanilla based on plastid DNA
phylogenetic analysis. Molecular Phylogenetics and Evolu-
tion,55, 621–630.
Brooks, T.M., Mittermeier, R.A., da Fonseca, G.A.B., Gerlach,
J., Hoffmann, M., Lamoreux, J.F., Mittermeier, C.G.,
Pilgrim, J.D. & Rodrigues, A.S.L. (2006) Global biodiver-
sity conservation priorities. Science,313,58–61.
Chase, M.W. (2001) The origin and biogeography of Orchi-
daceae. Genera Orchidacearum, Vol. 2 (ed. by A.M. Prid-
geon, P.J. Cribb, M.W. Chase and F. Rasmussen), pp. 1–5.
Oxford University Press, Oxford.
Chase, M.W., Cameron, K.M., Barrett, R.L. & Freudenstein,
J.V. (2003) DNA data and Orchidaceae systematics: a new
phylogenetic classification. Orchid Conservation (ed. by
K.W. Dixon, S.P. Kell, R.L. Barrett and P.J. Cribb), pp.
69–89. Natural History Publications, Kota Kinabalu,
Sabah.
Chase, M.W., Cameron, K.M., Freudenstein, J.V., Pridgeon,
A.M., Salazar, G., van den Berg, C. & Schuiteman, A.
(2015) An updated classification of Orchidaceae. Botanical
Journal of the Linnean Society,177, 151–174.
Chomicki, G., Bidel, L.P., Ming, F., Coiro, M., Zhang, X.,
Wang, Y., Baissac, Y., Jay-Allemand, C. & Renner, S.S.
(2015) The velamen protects photosynthetic orchid roots
against UV-B damage, and a large dated phylogeny implies
multiple gains and losses of this function during the Ceno-
zoic. New Phytologist,205, 1330–1341.
Conran, J.G., Bannister, J.M., Lee, D.E., Carpenter, R.J., Ken-
nedy, E.M., Reichgelt, T. & Fordyce, R.E. (2015) An
update of monocot macrofossil data from New Zealand
and Australia. Botanical Journal of the Linnean Society,
178, 394–420.
Davis, M.P., Midford, P.E. & Maddison, W. (2013) Exploring
power and parameter estimation of the BiSSE method for
analyzing species diversification. BMC Evolutionary Biology,
13, 38.
De Vogel, E. & Schuiteman, A. (2001) Flora Malesiana: orch-
ids of New Guinea, Volume 1: illustrated checklist and gen-
era. National Herbarium of The Netherlands, Leiden.
Dodd, M.E., Silvertown, J. & Chase, M.W. (1999)
Phylogenetic analysis of trait evolution and species
diversity variation among angiosperm families. Evolution,
53, 732–744.
Dressler, R.L. (1981) Orchids: natural history and classifica-
tion. Harvard University Press, Cambridge, MA.
Drummond, A.J., Suchard, M.A., Xie, D. & Rambaut, A.
(2012) Bayesian phylogenetics with BEAUti and the
BEAST 1.7. Molecular Biology and Evolution,29, 1969–
1973.
Dueck, L.A., Aygoren, D. & Cameron, K.M. (2014) A molec-
ular framework for understanding the phylogeny of Spi-
ranthes (Orchidaceae), a cosmopolitan genus with a North
American center of diversity. American Journal of Botany,
101, 1551–1571.
FitzJohn, R.G., Maddison, W.P. & Otto, S.P. (2009) Estimat-
ing trait-dependent speciation and extinction rates from
incompletely resolved phylogenies. Systematic Biology,58,
595–611.
Freudenstein, J.V. & Chase, M.W. (2015) Phylogenetic rela-
tionships in Epidendroideae (Orchidaceae), one of the
great flowering plant radiations: progressive specialization
and diversification. Annals of Botany,115, 665–681.
Givnish, T.J. (2010) Ecology of plant speciation. Taxon,59,
1326–1366.
Givnish, T.J., Evans, T.M., Pires, J.C. & Sytsma, K.J. (1999)
Polyphyly and convergent morphological evolution in
Commelinales and Commelinidae: evidence from rbcL
sequence data. Molecular Phylogenetics and Evolution,12,
360–385.
Givnish, T.J., Barfuss, M.H.J., Van Ee, B., Riina, R., Schulte,
K., Horres, R., Gonsiska, P.A., Jabaily, R.S., Crayn, D.M.,
Smith, J.A.C., Winter, K., Brown, G.K., Evans, T.M., Holst,
B.K., Luther, H.E., Till, W., Zizka, G., Berry, P.E. &
Sytsma, K.J. (2014) Adaptive radiation, correlated and
contingent evolution, and determinants of net species
diversification in Bromeliaceae. Molecular Phylogenetics
and Evolution,71,55–78.
Journal of Biogeography 43, 1905–1916
ª2016 John Wiley & Sons Ltd
1914
T. J. Givnish et al.
Givnish, T.J., Spalink, D., Ames, M., Lyon, S.P., Hunter, S.J.,
Zuluaga, A., Clements, M.A., Arroyo, M.T.K., Leebens-
Mack, J., Endara, L., Kriebel, R., Neubig, K.M., Whitten,
W.M., Williams, N.H. & Cameron, K.M. (2015) Orchid
phylogenomics and multiple drivers of extraordinary
diversification. Proceedings of the Royal Society of London,
Series B,282, 171–180.
Givnish, T.J., Zuluaga, A., Lam, V.K.Y., Gomez, M.S., Iles,
W.J.D., Spalink, D., Moeller, J.R., Lyon, S.P., Briggs, B.G.,
Zomlefer, W.B. & Graham, S.W. (2016) Plastome phy-
logeny and historical biogeography of the monocot order
Liliales: out of Australia and through Antarctica. Cladistics,
32. (Early View –doi:10.1111/cla.12153).
Gravendeel, B., Smithson, A., Slik, F.J.W. & Schuieman, A.
(2004) Epiphytism and pollinator specialization: drivers
for orchid diversity? Philosophical Transactions of the Royal
Society of London, Series B: Biological Sciences,359,
1523–1535.
Guo, Y.Y., Luo, Y.B., Liu, Z.J. & Wang, X.Q. (2012) Evolu-
tion and biogeography of the slipper orchids: Eocene
vicariance of the conduplicate genera in the Old and New
World tropics. PLoS ONE,7, e38788.
Guo, Y.Y., Luo, Y.B., Liu, Z.J. & Wang, X.Q. (2015)
Reticulate evolution and sea-level fluctuations together
drove species diversification of slipper orchids (Paphiope-
dilum) in South-East Asia. Molecular Ecology,24,
2838–2855.
Hall, R. (2009) Southeast Asia’s changing palaeogeography.
Blumea,54, 148–161.
Hughes, C.E., Pennington, R.T. & Antonelli, A. (2013)
Neotropical plant evolution: assembling the big picture.
Botanical Journal of the Linnean Society,171,1–18.
Jost, L. (2004) Explosive local radiation of the genus Tea-
gueia (Orchidaceae) in the Upper Pastaza watershed of
Ecuador. Lyonia,7,42–47.
Maddison, W.P., Midford, P.E. & Otto, S.P. (2007) Estimat-
ing a binary character’s effect on speciation and extinction.
Systematic Biology,56, 701–710.
Matzke, N.J. (2014) Model selection in historical
biogeography reveals that founder-event speciation is a
crucial process in island clades. Systematic Biology,63,
951–970.
eMonocot Team (2015) eMonocot. http://e-monocot.org/
Montes, C., Cardona, A., McFadden, R., Mor
on, S.E., Silva,
C.A., Restrepo-Moreno, S., Ram
ırez, D.A., Hoyos, N., Wil-
son, J., Farris, D., Bayona, G.A., Jaramillo, C.A., Valencia,
V., Bryan, J. & Flores, J.A. (2012) Evidence for middle
Eocene and younger land emergence in central Panama:
implications for Isthmus closure. Geological Society of
America Bulletin,124, 780–799.
Pigram, C.J. & Davies, H.L. (1987) Terranes and the accre-
tion history of the New Guinea orogen. BMR Journal of
Australian Geology and Geophysics,10, 193–211.
Pridgeon, A.M., Cribb, J.P., Chase, W.M. & Rasmussen, F.
(1999–2014) Genera Orchidacearum, Volumes 1–6. Oxford
University Press, Oxford.
Pross, J., Contreras, L., Bijl, P.K., Greenwood, D.R., Bohaty,
S.M., Schouten, S., Bendle, J.A., Rohl, U., Tauxe, L., Raine,
J.I., Huck, C.E., van de Flierdt, T., Jamieson, S.S.R., Stick-
ley, C.E., van de Schootbrugge, B., Escutia, C. & Brinkhuis,
H. (2012) Persistent near-tropical warmth on the Antarctic
continent during the early Eocene epoch. Nature,488,
73–77.
Rabosky, D.L. & Goldberg, E.E. (2015) Model inadequacy
and mistaken inferences of trait-dependent speciation. Sys-
tematic Biology,64, 340–355.
Ricklefs, R.E. & Renner, S.S. (1994) Species richness within
families of flowering plants. Evolution,48, 1619–1636.
Ricklefs, R.E. & Renner, S.S. (2000) Evolutionary flexibility
and flowering plant familial diversity: a comment on
Dodd, Silvertown, and Chase. Evolution,54, 1061–1065.
Schuiteman, A., Vermuelen, J.J. & De Vogel, E. (2010) Flora
Malesiana: orchids of New Guinea, Volume 6: Genus Bulbo-
phyllum. National Herbarium of The Netherlands, Leiden.
Sessa, E.B., Zimmer, E.A. & Givnish, T.J. (2012) Phylogeny,
divergence times and historical biogeography of New
World Dryopteris (Dryopteridaceae). American Journal of
Botany,99,1–21.
Smidt, E.C., Borba, E.L., Gravendeel, B., Gischer, G.A. & van
den Berg, C. (2011) Molecular phylogeny of the Neotropi-
cal sections of Bulbophyllum (Orchidaceae) using nuclear
and plastid markers. Taxon,60, 1050–1064.
The Plant List (2015) http://www.theplantlist.org/
Thomas, D.C., Hughes, M., Phutthai, T., Ardi, W.H., Rajb-
handary, S., Rubite, R., Twyford, A.D. & Richardson, J.E.
(2012) West to east dispersal and subsequent rapid diversi-
fication of the mega-diverse Begonia (Begoniaceae) in the
Malesian archipelago. Journal of Biogeography,39,98–113.
Turner, H., Hovencamp, P. & van Welzen, P.C. (2001) Bio-
geography of Southeast Asia and the West Pacific. Journal
of Biogeography,28, 217–230.
Vamosi, J.C. & Vamosi, S.M. (2011) Factors influencing
diversification in angiosperms: at the crossroads of intrin-
sic and extrinsic traits. American Journal of Botany,98,
460–471.
Wagner, W.L., Herbst, D.R. & Sohmer, S.H. (1990) Manual
of the flowering plants of Hawai‘i. University of Hawaii
Press, Honolulu.
WCSP (2015). World Checklist of Selected Plant Families.
http://apps.kew.org/wcsp/
Yu, W.B., Liu, M.L., Wang, H., Mill, R.R., Ree, R.H., Yang,
J.B. & Li, D.Z. (2015) Towards a comprehensive phylogeny
of the large temperate genus Pedicularis (Orobanchaceae),
with an emphasis on species from the Himalaya-Hengduan
Mountains. BMC Plant Biology,15, 176.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the
online version of this article:
Appendix S1 Additional methods, tables and figure.
Journal of Biogeography 43, 1905–1916
ª2016 John Wiley & Sons Ltd
1915
Orchid biogeography and diversification
BIOSKETCH
This article is the product of an international collaboration among specialists in orchid systematics, phylogenetics, ecology and
biogeography.
Author contributions: T.J.G. and D.S. designed the study and conducted the analyses; T.J.G., S.P.L, S.J.H, A.D., G.G.C. and J.M.
compiled distributional data; T.J.G. wrote the first draft; all authors except A.D., G.G.C. and J.M. produced the time-calibrated
phylogeny.
Editor: Alexandre Antonelli
Journal of Biogeography 43, 1905–1916
ª2016 John Wiley & Sons Ltd
1916
T. J. Givnish et al.
