Molecular dating of the ‘Gondwanan’ plant family Proteaceae is only partly congruent with the timing of the break-up of Gondwana

Abstract

Aim The flowering plant family Proteaceae is putatively of Gondwanan age, with modern and fossil lineages found on all southern continents. Here we test whether the present distribution of Proteaceae can be explained by vicariance caused by the break-up of Gondwana.
Location Africa, especially southern Africa, Australia, New Zealand, South America, New Caledonia, New Guinea, Southeast Asia, Sulawesi, Tasmania.
Methods We obtained chloroplast DNA sequence data from the rbcL gene, the rbcL-atpB spacer, and the atpB gene from leaf samples of forty-five genera collected from the field and from living collections. We analysed these data using Bayesian phylogenetic and molecular dating methods, with five carefully selected fossil calibration points to obtain age estimates for the nodes within the family.
Results Four of eight trans-continental disjunctions of sister groups within our sample of the Proteaceae post-date the break-up of Gondwana. These involve independent lineages, two with an Africa-Australia disjunction, one with an Africa–South America disjunction, and one with a New Zealand–Australasia disjunction. The date of the radiation of the bird-pollinated Embothriinae corresponds approximately to the hypothesized date of origin of nectar-feeding birds in Australia.
Main conclusions The findings suggest that disjunct distributions in Proteaceae result from both Gondwanan vicariance and transoceanic dispersal. Our results imply that ancestors of some taxa dispersed across oceans rather than rafting with Gondwanan fragments as previously thought. This finding agrees with other studies of Gondwanan plants in dating the divergence of Australian, New Zealand and New Caledonian taxa in the Eocene, consistent with the existence of a shared, ancestral Eocene flora but contrary to a vicariance scenario based on accepted geological knowledge.

Full text

ORIGINAL
ARTICLE
Molecular dating of the ‘Gondwanan’
plant family Proteaceae is only partially
congruent with the timing of the
break-up of Gondwana
Nigel P. Barker
1
*, Peter H. Weston
2
, Frank Rutschmann
3
and Herve
´Sauquet
2,4
INTRODUCTION
Since the scientific revolution in the earth sciences in the
1960s, in which plate tectonics displaced the stable-earth
model as the accepted theoretical framework for historical
geology, the ancient supercontinent of Gondwana has been
invoked as the ancestral distributional area for numerous
groups of flowering plants. As a consequence of this geological
revolution, long-distance dispersal was replaced by vicariance
as a generally accepted null hypothesis to explain disjunct
distributions, and Gondwanan distribution patterns were
reinterpreted as the result of continental drift rather than of
1
Molecular Systematics & Ecology Group,
Department of Botany, Rhodes University,
Grahamstown, 6140, South Africa,
2
Royal
Botanic Gardens and Domain Trust, Mrs
Macquaries Rd, Sydney, NSW 2000, Australia,
3
Institute of Systematic Botany, Zollikerstrasse
107, CH-8008, Zurich, Switzerland,
4
Jodrell
Laboratory, Molecular Systematics Section,
Royal Botanic Gardens, Kew, Richmond TW9
3DS, UK
*Correspondence: Nigel P. Barker, Department
of Botany, Rhodes University, Grahamstown,
6140, South Africa.
E-mail: n.barker@ru.ac.za
ABSTRACT
Aim The flowering plant family Proteaceae is putatively of Gondwanan age, with
modern and fossil lineages found on all southern continents. Here we test
whether the present distribution of Proteaceae can be explained by vicariance
caused by the break-up of Gondwana.
Location Africa, especially southern Africa, Australia, New Zealand, South
America, New Caledonia, New Guinea, Southeast Asia, Sulawesi, Tasmania.
Methods We obtained chloroplast DNA sequence data from the rbcL gene, the
rbcL-atpB spacer, and the atpB gene from leaf samples of forty-five genera
collected from the field and from living collections. We analysed these data using
Bayesian phylogenetic and molecular dating methods, with five carefully selected
fossil calibration points to obtain age estimates for the nodes within the family.
Results Four of eight trans-continental disjunctions of sister groups within our
sample of the Proteaceae post-date the break-up of Gondwana. These involve
independent lineages, two with an Africa-Australia disjunction, one with an
Africa–South America disjunction, and one with a New Zealand–Australasia
disjunction. The date of the radiation of the bird-pollinated Embothriinae
corresponds approximately to the hypothesized date of origin of nectar-feeding
birds in Australia.
Main conclusions The findings suggest that disjunct distributions in Proteaceae
result from both Gondwanan vicariance and transoceanic dispersal. Our results
imply that ancestors of some taxa dispersed across oceans rather than rafting with
Gondwanan fragments as previously thought. This finding agrees with other
studies of Gondwanan plants in dating the divergence of Australian, New Zealand
and New Caledonian taxa in the Eocene, consistent with the existence of a shared,
ancestral Eocene flora but contrary to a vicariance scenario based on accepted
geological knowledge.
Keywords
AtpB, atpB-rbcL spacer, Bayesian dating, dispersal, Gondwana, historical
biogeography, molecular dating, Proteaceae, rbcL, vicariance.
Journal of Biogeography (J. Biogeogr.) (2007) 34, 2012–2027
2012 www.blackwellpublishing.com/jbi ª2007 The Authors
doi:10.1111/j.1365-2699.2007.01749.x Journal compilation ª2007 Blackwell Publishing Ltd

transoceanic dispersal (Raven & Axelrod, 1974; Johnson &
Briggs, 1975; Sanmartı
´n & Ronquist, 2004). Distributions that
are consistent with this ancestral area can be seen in plant
groups such as the Myrtales (Conti et al., 2002; Ladiges et al.,
2003; Sytsma et al., 2004), Annonaceae and Myristicaceae
(Doyle et al., 2004; Richardson et al., 2004; Scharaschkin &
Doyle, 2005), Moraceae (Zerega et al., 2005), Restionaceae
(Linder et al., 2003, 2005), Nothofagus (Swenson et al., 2001;
Cook & Crisp, 2005; Knapp et al., 2005) and Gunnera
(Wanntorp & Wanntorp, 2003). However, of all the modern
flowering plant families that are thought to have existed on
Gondwana, none provides as impressive an array of repeated
transoceanic distributional patterns as the Proteaceae (Johnson
& Briggs, 1975; Weston & Crisp, 1996; see Fig. 1).
Together with their close relatives Platanus and Nelumbo,
Proteaceae form one of the early-diverging lineages in the
eudicot clade of angiosperms (Hoot et al., 1999; Qiu et al.,
2000; Soltis et al., 2000; APG, 2003; Hilu et al., 2003; Kim
et al., 2004), with a fossil record that dates back to the mid-
Cretaceous (Drinnan et al., 1994; Hill et al., 1995; Dettmann &
Jarzen, 1996), when the major Gondwanan continental blocks
were still connected. With 80 genera and over 1700 descendant
species spread over all Gondwana’s current land masses except
Antarctica, the family is thus an ideal group with which to test
hypotheses of the biogeographical history of Gondwana’s flora.
According to the latest intrafamilial classification (Weston &
Barker, 2006), the Proteaceae include at least fourteen taxa
with transoceanic links, each of which offers a potential test of
biogeographical relationships between disjunct Gondwanan
biotas (Fig. 1). Two of these, the genus Lomatia (Fig. 1g), and
the subtribe Embothriinae (Fig. 1h), which have almost
identical distributions, have already been subjected to a
cladistic biogeographical analysis (Weston & Crisp, 1994).
That analysis found statistically significant biogeographical
congruence between the two clades and yielded a general area-
cladogram that was consistent with both geological knowledge
of the sequence of continental break-up and climatological
knowledge of the origin of disjunctions within Australia and
South America. Weston & Crisp (1994) concluded that the
Gondwanan-vicariance hypothesis of proteaceous biogeogra-
phy had been corroborated by their analysis.
A prediction of vicariance biogeography is that clades
showing disjunct, transoceanic distributions must be older
than the ocean basins that fragment their geographic ranges.
For example, taxa occurring in both Australia and Africa must
be more than 105 Myr old, the minimal age of last direct land
contact between these continents through Antarctica and
South America, according to current geological knowledge
(McLoughlin, 2001).
DNA sequence data are increasingly being used in conjunc-
tion with relaxed molecular clock models and fossil calibration
to estimate the age of ancestral lineages, and thus to test
theories of vicariance biogeography. The use of fossils to
calibrate molecular phylogenies involves problems associated
both with the analysis of DNA data and with the identification
and dating of the fossils themselves (Wikstro
¨met al., 2001;
Near & Sanderson, 2004; Forest et al., 2005; Renner, 2005).
Molecular dating methods have now improved to a point at
which problems such as rate heterogeneity can be accommo-
dated to some extent (Magallo
´n, 2004; Rutschmann, 2006), but
in order for such fossil-calibrated studies to provide realistic
dates, the fossils chosen to act as calibration or constraint
points have to be soundly dated and their affinities to modern
taxa well supported (Wikstro
¨met al., 2003; Magallo
´n&
Sanderson, 2005; Rutschmann, 2006).
Using relaxed-clock molecular dating, Wikstro
¨met al.
(2001, 2003) provided a range of estimates for the age of the
flowering plants based on the chloroplast rbcL gene, and dated
the origin of crown eudicots at 147–131 Ma. As noted by those
workers, this age implies a gap in the fossil record, as the
earliest known eudicot fossil dates to c. 125 Ma (Hughes &
McDougall, 1990). The stem age of the Proteaceae was
calculated as 117–108 Myr, but these studies did not include
data from all of the basal lineages of the family, an omission
that may affect the results of molecular dating studies (Linder
& Hardy, 2004). The first multi-gene dating study on
angiosperms by Bell et al. (2005) using Bayesian relaxed clock
(BRC) and penalized likelihood (PL; Sanderson, 2002) meth-
ods retrieved a eudicot stem age of 125–91 Myr, depending on
how fossil-based age constraints were applied in the analyses.
Anderson et al. (2005) produced chronograms from rbcL data
for the basal eudicot lineages using PL and non-parametric rate
smoothing (NPRS; Sanderson, 1997) methods, calibrated using
14 fossils. These analyses estimated the stem age for the
Proteaceae as between 119 Myr (NPRS) and 110 Myr (PL)
when the stem node of the eudicots was fixed at 124 Myr.
In this paper we report a test of the general theory that all
intercontinental biogeographical disjunctions in the Proteaceae
are the result of vicariance (Weston & Crisp, 1996). This
involved estimating the ages of clades showing disjunct,
transoceanic distributions and comparing these estimates with
geological knowledge of the age of the disjunctions. We
conducted Bayesian molecular dating estimates based on a
sample of the proteaceous species for which DNA sequence
data from the chloroplast rbcL, atpB and rbcL-atpB spacer
regions are available, representing approximately half of the
genera. This sampling is a considerable expansion on that of
Anderson et al. (2005), and covers all the subfamilies and
major tribal lineages (Weston & Barker, 2006). The earliest
known fossil that has been attributed to the Proteaceae is a
pollen grain from the mid-Cretaceous, dated at 93 Ma
(Dettmann & Jarzen, 1998). We use this and four other fossils
that display putative synapomorphies of extant clades to
calibrate our analysis.
MATERIALS AND METHODS
Taxon sampling
Species of 45 of the 80 extant genera of Proteaceae, as well as
one species of the outgroup genus, Platanus (Platanaceae),
were included in our study (Table 1). For most genera one
Proteaceae biogeography and the break-up of Gondwana
Journal of Biogeography 34, 2012–2027 2013
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd

species was sampled for all three of the genes analysed, but for
some a different species was used for rbcL sequencing on the
one hand and atpB and the atpB-rbcL spacer on the other (see
Table 1). This occurred because the rbcL data were produced
by one of us (NPB) and augmented with the atpB and atpB-
rbcL spacer data from Hoot & Douglas (1998), with two
ab
cd
ef
gh
Figure 1 Distributional tracks of clades in
the Proteaceae that have disjunct, transoce-
anic distributions (with genera sampled in
this study in brackets). Disjunct areas of
occurrence are filled with black or marked
with black dots, and tracks linking distribu-
tional areas are represented by straight lines.
(a) Tribe Persoonieae (Toronia); (b) tribe
Petrophileae (Petrophile,Aulax); (c) tribe
Proteeae (Protea); (d) subtribes Adenanthi-
nae (Adenanthos) and Leucadendrinae
(Leucadendron); (e) Orites; (f) tribe Helicii-
nae (Helicia,Hollandaea) plus Knightia; (g)
Lomatia; (h) subtribe Embothriinae (Alloxy-
lon,Embothrium,Telopea); (i) subtribe
Stenocarpinae (Stenocarpus); (j) subtribe
Hakeinae (Buckinghamia,Grevillea,Opisthi-
olepis); (k) subtribe Macadamiinae (Brabe-
jum,Macadamia,Panopsis); (l) subtribe
Malagasiinae; (m) subtribe Virotiinae; (n)
subtribe Gevuininae (Cardwellia,Gevuina);
(o) summary of (a) to (n), showing gener-
alized tracks as parallel lines.
N. P. Barker et al.
2014 Journal of Biogeography 34, 2012–2027
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd

notable exceptions. First, we re-sequenced these regions for
Protea, as Hoot & Douglas’s voucher specimen proved to be a
misidentified Leucadendron. Second, the genus Roupala was
excluded from the analysis because the DNA sequences that
Hoot & Douglas (1998: 317) reported for Roupala and Floydia
were identical. This result is incredible, given that these genera
ij
kl
mn
o
Figure 1 continued.
Proteaceae biogeography and the break-up of Gondwana
Journal of Biogeography 34, 2012–2027 2015
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd

Table 1 Voucher details and GenBank numbers for material and taxa sampled for the three regions used in this analysis. Note that voucher details apply to the rbcL sequences only. See Hoot &
Douglas (1998) for details of the other regions. All GenBank numbers in bold indicate sequences generated for this study. Codes associated with voucher numbers are as follows: NSW indicates
herbarium numbers allocated to vouchers housed in the National Herbarium of New South Wales, MEL indicates herbarium numbers allocated to vouchers housed in the National Herbarium of
Melbourne, numbers preceded by a K are garden record numbers from Kirstenbosch Botanic Garden, Cape Town, South Africa. ‘E’ indicates a voucher housed at the Edinburgh Herbarium. GRA
indicates a voucher housed at the Selmar Schonland Herbarium, Grahamstown, South Africa.
Species Voucher (rbcL) Wild source rbcLatpBatpB-rbcL spacer
Platanus occidentalis L. var. glabrata Sarg. NSW 406239 Texas, USA DQ875820 AF528858 AF060755
Toronia toru (A. Cunn.) L.A.S.Johnson & B.G.Briggs NSW 397528 Coromandel Range, New Zealand DQ875823 AF060392 AF060736
Placospermum coriaceum C.T.White & W.D.Francis NSW 398813 Main Coast Range, Queensland DQ875822 AF060391 AF060712
Bellendena montana R.Br. NSW418648 Mt Field, Tasmania DQ875821 AF060390 AF060715
Symphionema montanum R.Br. NSW 397502 Blue Mountains, NSW, Australia DQ875825 AF060394 AF060733
Agastachys odorata R.Br. NSW 397531 McPartlan Pass, Tasmania DQ875824 AF060393 AF060717
Conospermum mitchellii Meisn. NSW 397496 Grampians National Park, Australia DQ875829 AF060398 AF060728
Synaphea spinulosa (Burm.f.) Merr. subsp. major A.S.George NSW 397503 Scaddan, Western Australia DQ875830
Synaphea media A.S.George AF060399 AF060729
Stirlingia latifolia (R.Br.) Steud. NSW 418620 Nedlands, Western Australia DQ875828 AF060397 AF060738
Cenarrhenes nitida Labill. NSW 397495 Melaleuca, Tasmania DQ875827 AF060396 AF060746
Aulax cancellata (L.) Druce Ex cult., Kirstenbosch, South Africa DQ875863
Aulax umbellata (Thunb.) R.Br. AF060732
Petrophile biloba R.Br. NSW 397512 Darling Range, Western Australia DQ875832
Petrophile circinata Kippist ex Meisn. AF060401 AF060735
Leucadendron tinctum I.Williams K12/79 Kirstenbosch Bot. Gdns, South Africa DQ875836
Leucadendron salignum Berg. AF060405 AF060723
Adenanthos sericeus Labill. NSW 202501 Flinders Peninsula, Western Australia DQ875835
Adenanthos obovatus Labill. AF060404 AF060739
Isopogon sphaerocephalus Lindl. NSW 397509 Darling Range, Western Australia DQ875834
Isopogon buxifolius R.Br. AF060403 AF060734
Protea cynaroides (L.) L. N. Barker 2006 (GRA) Ex. cult., Rhodes University, South Africa DQ875837 DQ875866 DQ875867
Beauprea montana (Brongn. & Gris) Virot NSW 368725 New Caledonia DQ875833 AF060402 AF060749
Franklandia fucifolia R.Br. NSW 397493 Fitzgerald River Nat. Pk, W Australia DQ875831 AF060400 AF060721
Eidothea zoexylocarya A.W.Douglas & B.Hyland MEL 2037963 Mt Bartle Frere, Queensland, Australia DQ875826 AF060395 AF060714
Gevuina avellana Molina NSW 397516 Yaquito, Rio Bueno valley, Chile DQ875852
Gevuina bleasdalei (F.Muell.) Sleumer (= Bleasdalea bleasdalei) AF060424 AF060748
Cardwellia sublimis F.Muell. NSW 360127 Herberton Range, Queensland, Australia DQ875851 AF060422 AF060753
Panopsis cinnamomea Pittier NSW 391574 Avila National Park, Venezuela DQ875850
Panopsis feruginea (Meisn.) Pittier AF060421 AF060756
Brabejum stellatifolium L. NSW 403139 Kirstenbosch, South Africa DQ875849 AF060420 AF060757
Macadamia claudiensis C.L.Gross & B.Hyland NSW 397498 Iron Range, Queensland, Australia DQ875848
Macadamia jansenii C.L.Gross & P.H.Weston AF060419 AF060750
Austromuellera trinervia C.T.White NSW 418608 Cooper Creek, Queensland, Australia DQ875865 AF060414 AF060720
Musgravea heterophylla L.S.Sm. NSW 368738 Oliver Creek, Queensland, Australia DQ875844 AF060415 AF060727
N. P. Barker et al.
2016 Journal of Biogeography 34, 2012–2027
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd

Table 1 continued.
Species Voucher (rbcL) Wild source rbcLatpBatpB-rbcL spacer
Banksia ericifolia L.f. NSW 197246 Kurnell, New South Wales, Australia DQ875843
Banksia cuneata A.S.George AF060413 AF060731
Hollandaea riparia B.Hyland NSW 399806 Roaring Meg Creek, Queensland, Australia DQ875854 AF060426 AF060751
Helicia sp. NSW 397522 Ban Khaong, Lao Cai province, Vietnam DQ875853
Helicia australasica F.Muell. AF060425 AF060724
Knightia excelsa R.Br. NSW 383435 New Plymouth, New Zealand DQ875840 AF060422 AF060753
Xylomelum pyriforme (Gaertn.) Knight NSW 399133 Royal National Park, NSW, Australia DQ875847
Xylomelum scottianum (F.Muell.) F.Muell. AF060418 AF060741
Lambertia echinata R.Br. NSW 401668 Cape Le Grand Nat. Pk, Western Australia DQ875846
Lambertia formosa Sm. AF060417 AF060737
Floydia praealta (F.Muell.) L.A.S.Johnson & B.G.Briggs NSW 368734 unknown DQ875845
Orites myrtoidea (Poepp. & Endl.) Benth. & Hook.f. M.Gardner 468 et al. (E) Prq. Nacional de Laguna de La Laja, Chile DQ875842
Orites lancifolia F.Muell. AF060412 AF060718
Neorites kevediana L.S.Sm. NSW 399348 Mt Misery, Queensland, Australia DQ875864 AF060411 AF060716
Sphalmium racemosum (C.T.White) B.G.Briggs,
B.Hyland & L.A.S.Johnson
MEL 2027222 Mt Lewis, Queensland, Australia DQ875839 AF060408 AF060719
Triunia youngiana (C.Moore & F.Muell. ex F.Muell.)
L.A.S.Johnson & B.G.Briggs
NSW 397518 Dorrigo Nat. Pk., NSW, Australia DQ875841
Triunia montana (C.T.White) Foreman AF060410 AF060740
Embothrium coccineum J.R.Forst. & G.Forst. NSW 393588 Cordillera Pelada, Chile DQ875857 AF060429 AF060754
Telopea oreades F.Muell. NSW 239934 Monga State Forest, NSW, Australia DQ875855
Telopea sp. AF060427 AF060758
Alloxylon flammeum P.H. Weston & Crisp NSW 383434 Atherton Tableland, Queensland, Australia DQ875856
Alloxylon wickhamii (W.Hill ex F.Muell.)
P.H. Weston & Crisp
AF060428 AF060752
Stenocarpus trinervis (Montrouz.) Guillaumin NSW 368722 Foret de Mois de Mai, New Caledonia DQ875859
Stenocarpus salignus R.Br. AF060431 AF060743
Lomatia fraxinifolia F.Muell. ex Benth. NSW 393333 Boonjee, Queensland, Australia DQ875858
Lomatia myricoides (C.F.Gaertn.) Domin AF060430 AF060722
Grevillea aquifolium Lindl. NSW 403972 Grampians, Dunkeld, Victoria, Australia DQ875862
Grevillea baileyana McGill. AF060434 AF060747
Buckinghamia ferruginiflora Foreman & B. Hyland NSW 368732 Roaring Meg Creek, Queensland, Australia DQ875861
Buckinghamia celsissima F.Muell. AF060433 AF060742
Opisthiolepis heterophylla L.S.Sm. NSW 368740 Gold Hill, Queensland, Australia DQ875860 AF060432 AF060725
Carnarvonia araliifolia F.Muell. NSW 397032 Gold Hill, Queensland, Australia DQ875838 AF060407 AF060726
Proteaceae biogeography and the break-up of Gondwana
Journal of Biogeography 34, 2012–2027 2017
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd

are morphologically quite dissimilar and occur on different
continents. We have confirmed the identity of one of their
Floydia voucher specimens, but their voucher for Roupala could
not be found. Our molecular data set comprised 46 taxa with
3913 bases, composed of three partitions: atpB (bases 1–1440),
atpB-rbcL spacer (bases 1441–2583), and rbcL (bases
2584–3913). From this data set, we excluded two small regions
because of missing data: bases 1–23 and 1496–1571. Data for the
atpB partition of Aulax were missing. Platanus was used as an
outgroup in the phylogenetic analyses but had to be pruned in
the dating procedure. All new sequences have been deposited in
GenBank under accession numbers DQ875820-DQ87565
(rbcL), DQ875866 and DQ875867 (Protea cynaroides atpB and
atpB-rbcL spacer, respectively). The terminal taxa represented
in our chronogram are genera, which themselves are sometimes
represented by DNA sequences from more than one species.
DNA purification, amplification, sequencing and
alignment
In most cases DNA was extracted and purified using the Cetyl
Trimethyl Ammonium Bromide (CTAB)-diatomite protocol
of Gilmore et al. (1993), but some samples were extracted
using the hot CTAB method of Doyle & Doyle (1987).
The rbcL gene was amplified either as a single unit or in
some cases in overlapping sections. Amplifications were
performed either on a Corbett Research Fts-1s capillary
thermal cycler or on a Hybaid Omn-E thermal cycler. PCR
conditions and thermal cycling parameters were varied in
order to obtain optimum amplification. PCR products were
purified using Wizard DNA Clean-up or PCR-prep systems
(Promega, Madison, WI, USA) as per manufacturers’ proto-
cols. DNA sequencing was carried out using the ABI Prism
TM
DyeDeoxy cycle sequencing system (Foster City, CA, USA),
and both strands of the target sequences were sequenced.
Sequencing reaction products were sent to the Sydney
University and Royal Prince Alfred Molecular Analysis Centre
for electrophoresis on Applied Biosystems ABI 373 and 377
automated sequencers. The rbcL and atpB sequences for Protea
were generated using the primers given by Hoot & Douglas
(1998), and sequenced using an ABI Prism DyeDeoxy cycle
sequencing kit version 3.1 on an ABI 3100 genetic analyzer at
Rhodes University, South Africa. Sequences were manually
edited using sequencher
TM
versions 3.0 and 4.5, exported
into MacClade version 4 (Maddison & Maddison, 2000), and
aligned by eye.
Phylogenetic analyses
Bayesian analyses for topology and branch-length estimation
relied on MrBayes version 3.12 (Huelsenbeck & Ronquist,
2001; Ronquist & Huelsenbeck, 2003). DT-ModSel (Minin
et al., 2003) is a program that evaluates 56 different models
of molecular evolution. It was employed to select the best
model based on the Bayesian information criterion. Topol-
ogy, branch lengths, and model parameter values for the
chosen optimal models were then estimated for each
partition in MrBayes.
Bayesian topology estimation used one cold and three
incrementally heated Markov chain Monte Carlo (MCMC)
chains run for 1.0 ·10
6
cycles, with trees sampled every 100th
generation, each using a random tree as a starting point. For
each data set, MCMC runs were repeated twice. The first 2000
trees were discarded as burn-in. The remaining trees were used
to construct one Bayesian consensus tree with mean branch
lengths and to calculate the posterior probabilities for each
branch. Examination of the logarithmic likelihoods and the
observed consistency between the runs suggested that the used
burn-in periods were sufficiently long.
Bootstrap support values were calculated using the Perl script
BootPHYML version 3.4 (Nylander, 2005). This program first
generates 1000 pseudo-replicates in seqboot (part of phylip
version 3.63; Felsenstein, 2004), then performs a maximum-
likelihood analysis in phyml (Guindon & Gascuel, 2003) for
each replicate under the model of evolution selected above, and
finally computes a 50% majority-rule consensus tree using
consense (also part of the phylip package).
The data were also phylogenetically analysed using maxi-
mum parsimony, as implemented in paup* version 4.0b10.
A heuristic search was completed using default parameters
(stepwise addition of terminals in the order in which they
appear in the data matrix, tree bisection-reconnection (TBR)
branch swapping). Equally parsimonious trees were summar-
ized as a strict consensus tree. Bootstrap support values were
calculated by conducting heuristic searches, using default
parameters, of 1000 pseudo-replicates. The applicability of a
molecular clock was tested using two separate maximum-
likelihood (ML) analyses, first with and then without the
molecular clock enforced. The likelihood ratio test (LRT) was
then used in combination with a chi-squared test to assess
whether or not the molecular clock significantly influenced the
ML analysis (Felsenstein, 1981).
Divergence-time estimation
The Bayesian multidivtime dating method (Thorne et al.,
1998; Thorne & Kishino, 2002) uses a probabilistic model to
describe the change in evolutionary rate over time and uses the
MCMC procedure to derive the posterior distribution of rates
and time. The procedure we followed is divided into three
steps and computer programs, and is described in more detail
in a step-by-step manual by Rutschmann (2005, available on
the multidivtime website, http://statgen.ncsu.edu/thorne/
multidivtime.html).
We ran the Markov chain for at least 1.0 ·10
6
cycles and
collected one sample every 100 cycles, without sampling the first
10
5
cycles (burn-in sector). Each analysis was performed at least
twice using different initial conditions and checking the output
sample files to ensure convergence of the Markov chain.
By using four fossils as minimum-age constraints (ages given
in Table 2 and relevant nodes marked in Fig. 2) and considering
the age of the root to be at least 93 Myr, as supported by the
N. P. Barker et al.
2018 Journal of Biogeography 34, 2012–2027
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd

oldest fossil evidence for the family, we performed a molecular
dating analysis with the software parameters set as follows (in
units of 10 Myr): rttm = 9.3, rttmsd = 3.7, rtrate and
rtratesd = 0.0023, brownmean and brownsd = 0.108, and
bigtime = 13. Using these settings, we effectively constrained
the root node in two ways: first by assigning a minimal
constraint of 93 Myr to this node, and second by flexibly
defining the prior distribution for the age of the root (described
by the parameters rttm and rttmsd). By allowing a standard
deviation (rttmsd) of 37 Myr, we allowed the maximum age to
be 130 Myr, congruent with the commonly accepted age of the
eudicots. The remaining settings were based on commonly
accepted recommendations noted in the step-by-step manual
mentioned above. It must be noted that it is recommended that
multidivtime analyses are bounded from above and below on
the tree. We have fulfilled the requirement for bounding from
below by providing a reasonable prior for the age of the root
(rttm), and the requirement for bounding from above by using
minimum-age constraints for the internal calibration points.
Analyses were performed with and without Aulax, to determine
the effect that the missing data for this taxon had on the analysis.
Fossils used as internal calibration points
In spite of the particularly rich and abundant fossil record
known for Proteaceae (e.g. Hill et al., 1995; Dettmann & Jarzen,
1996, 1998), few fossil taxa are suitable for calibrating or
constraining molecular dating analyses. To serve such a
purpose, a fossil must be appropriately placed on the phylo-
genetic tree based on synapomorphies of extant clades (Mag-
allo
´n, 2004). Because morphological characters are often
limited in fossils, it is also important that such synapomorphies
are chosen among characters with low levels of homoplasy.
Here we have selected some of the few Proteaceae fossils that
meet these conditions, based on our current understanding of
morphological evolution in the group. Except in the particular
case of the root node, each of these fossils has been used as a
minimum-age constraint for the stem node of the clade for
which it shows synapomorphies. All of our internal fossil
constraints are from within the subfamily Grevilleoideae. No
unambiguous fossils are currently known from the remaining
clades of Proteaceae, including the subfamily Proteoideae.
Root node
Crown-group Proteaceae were assumed to be 93 Myr old,
based on the earliest known fossil assigned to the family,
Triorites africaensis Jardine
´& Magloire from the Upper
Cenomanian to Turonian of Senegal and Gabon (Dettmann
& Jarzen, 1998; Ward & Doyle, 1994). Although this fossil
pollen only shows plesiomorphic characters of the family and
lacks distinctive synapomorphies of an internal clade (HS & D.
Cantrill, unpublished data), we decided in this particular case
to use it to calibrate the crown age instead of the stem age of
Proteaceae. This is supported by evidence that stem-group
Platanaceae (the sister lineage to Proteaceae) are at least
15 Myr older (Drinnan et al., 1994), and that crown-group
Proteaceae are at least 82 Myr old (first known record of
Grevilleoideae, the largest clade in the family; Dettmann &
Jarzen, 1996; HS & D. Cantrill, unpublished data). This
provides a minimum age for crown-group Proteaceae. The
origination of the palynological synapomorphies of Proteaceae
in the fossil record is therefore probably closer to the crown
node than to the stem node. For the purpose of this study, it
seemed reasonable to approximate this age with the crown,
with a generous standard deviation of 37 Myr.
Node 1. Stem Embothrium: 35.4 Myr
Fossil pollen grains named Granodiporites nebulosus Stover &
Partridge have been collected from throughout the ‘Upper
N. asperus’to‘Proteacidites tuberculatus’ palynological zones of
the Murray Basin, Australia, dated from 35.4 to 17 Ma.
Granodiporites nebulosus has been identified as a close relative
of Embothrium (MacPhail & Truswell, 1989). This identifica-
tion is sound (as far as any pollen identification can be) because
the grains possess the spinules that are synapomorphic for the
Embothriinae (Feuer, 1990; Weston & Crisp, 1994; HS &
D. Cantrill, unpublished data), and within this subtribe
diporate grains are autapomorphic for Embothrium coccineum
(only Banksieae also have diporate grains in the Proteaceae, and
these look very different from those of Embothrium). This fossil
serves to constrain the age of the Embothrium stem lineage.
Table 2 Age estimates of selected nodes shown in Fig. 1. Esti-
mated ages are given in millions of years (Myr), with the standard
deviation estimates in parentheses. Numbered nodes were used to
calibrate the chronogram; lettered nodes represent clades with
transoceanic disjunctions (see Fig. 1). The age estimates for the
clades given in italics are younger than the ages of the corres-
ponding geological disjunctions.
Node Description Age
Root Proteaceae crown group 118.5 (±8.2)
1 Stem Embothrium
(calibration set to 35.4 Myr);
Australia & Tasmania – S. America
45.8 (±7.0)
2 Stem Embothriinae
(calibration set to 70 Myr)
76.8 (±5.6)
3 Stem Musgraveinae
(calibration set to 35 Myr)
71.2 (±8.9)
4 Stem Banksieae
(calibration set to 65 Myr)
87.9 (±7.0)
A Australia & Islands – S. America 61.2 (±7.5)
B Australia – Australia &
Tasmania & S. America
45.1 (±9.0)
C New Zealand – Australia & SE Asia 45.4 (±9.1)
D Australia – Africa & S. America 53.9 (±10.0)
E Africa – S. America 43.8 (±10.0)
F Australia – South America 51.4 (±10.5
G Australia – Africa 39.7 (±9.3)
H Australia – Africa 64.9 (±11.9)
I New Zealand – Australia 72.3 (±11.8)
Proteaceae biogeography and the break-up of Gondwana
Journal of Biogeography 34, 2012–2027 2019
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd

Node 2. Stem Embothriinae (Alloxylon, Telopea,
Embothrium): 70 Myr
The fossil pollen grain attributed to the species Propylipollis
ambiguus (Stover) Dettmann & Jarzen has been considered as
comparable or nearly identical to pollen of some extant species
of Telopea, and Alloxylon (Dettmann & Jarzen, 1996).
Propylipollis ambiguus possesses the distinctive pollen synapo-
morphy of the subtribe Embothriinae: spinulate exine (Feuer,
1990; Weston & Crisp, 1994; HS & D. Cantrill, unpublished
Figure 2 Chronogram of divergences in the Proteaceae, as determined using four internal calibration points (marked by numbered black
diamonds) set as minimal ages and the mean prior age of the root set to 93 Myr (± 37 Myr; see Methods). Numbers above the branches are
maximum-likelihood bootstrap values; numbers below the branches represent Bayesian clade credibility values (values greater than 0.85 are
shown). Nodes marked in circles correspond to transoceanic disjunctions for which estimated ages match or pre-date continental vicariance
events. Nodes marked in squares correspond to transoceanic disjunctions that post-date continental break-up and thus infer long-distance
dispersal or geodispersal distribution models (see Table 2). The node marked with the star indicates the crown-group radiation of the
subtribe Embothriinae, the Gondwanan bird-pollinated lineage. Dark grey horizontal bars indicate standard deviation estimates of key nodes
discussed in the text. Key tectonic events (McLoughlin, 2001) are indicated below the time line. Key to Tertiary epochs: Pa, Palaeocene; Eo,
Eocene; Ol, Oligocene; Mi, Miocene; Pl , Pliocene. Area codes: Au, Australia; SAm, South America; Af, Africa; Ta, Tasmania; NZ, New
Zealand; NC, New Caledonia; NG, New Guinea; Su, Sulawesi; SE Asia, Southeast Asia.
N. P. Barker et al.
2020 Journal of Biogeography 34, 2012–2027
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd

data). It is thus considered here as a stem species of
Embothriinae. This pollen species has been recovered from
the middle Campanian-Maastrichtian or earliest Danian in the
Otway Basin, from the Eocene elsewhere in Australia (Stover &
Partridge, 1973, 1982; Truswell et al., 1985; Truswell & Owen,
1988; Milne, 1988; Alley & Broadbridge, 1992; Alley &
Beecroft, 1993), and from the Maastrichtian of Argentina
(Baldoni & Askin, 1993). As the Campanian Age spanned from
83 to 71 Ma, and the Maastrichtian Age from 71 to 65 Ma, we
constrained this calibration point at 70 Myr.
Node 3. Stem Musgraveinae (Austromuellera, Musgravea):
35 Myr
The macrofossil Musgraveinanthus alcoensis from the upper
Eocene of Australia (35 Ma; Christophel, 1984) is a well-
preserved fossil inflorescence that is strongly reminiscent of the
morphology of extant members of the subtribe Musgraveinae
(Austromuellera and Musgravea). Detailed cladistic analysis of
macromorphological variation across extant Proteaceae con-
firms this relationship, supported by a combination of
apomorphic characters, including three hypogynous glands
and diporate pollen (PHW & HS, unpublished data). This
fossil serves as a minimum-age constraint for stem Musgra-
veinae.
Node 4. Stem Banksieae (Banksia, Austromuellera,
Musgravea): 65 Myr
Vadala & Drinnan (1998) argue that early Palaeocene leaves
of Banksieaephyllum show the leaf architectural features that
are characteristic of extant Banksieae but that are unknown
elsewhere in the Proteaceae. The diporate pollen grains of the
Banksieae are also synapomorphic for this tribe, and the
earliest fossils of these are dated at the Cretaceous-Palaeocene
boundary (Dettmann & Jarzen, 1998). The age of the stem
lineage of this tribe is thus constrained at 65 Myr on the basis
of these fossils.
RESULTS
The intergeneric relationships reconstructed by our Bayesian
phylogenetic analysis are generally well resolved, and many
nodes receive > 0.95 posterior probability values and good
bootstrap support (Fig. 2). Our Bayesian tree is essentially
congruent with our parsimony bootstrap tree, the topologies
differing only with respect to nodes that received < 50%
bootstrap support in either tree. Moreover, all but one of the
nodes that received > 50% bootstrap support in our analyses
are congruent with groupings in a supertree produced from a
broader sample of genes and taxa (Weston & Barker, 2006).
The one exception to this pattern of congruence concerns the
sister group relationship of Knightia excelsa (New Zealand) to
subtribe Heliciinae (Helicia plus Hollandaea – Australia and
Southeast Asia) on our tree. This relationship has been found
in no other phylogenetic analysis, including Hoot & Douglas’s
(1998) analysis of atpB and rbcL-atpB spacer sequences. Strong
support for a close relationship between Knightia and the
Heliciinae therefore seems to be a feature only of our rbcL data
set, so caution is advised in interpreting the relationships and
age of divergence of Knightia. Where node support is poor, it
does not affect the interpretation of the biogeographical results
presented here, as the nodes and taxa of interest all receive high
levels of support. The effect of the missing atpB data partition
for Aulax was minimal, as results from the analysis that
excluded Aulax were within 1% of the ages found when this
genus was included (results not shown). The analysis including
Aulax is thus presented and discussed here. The likelihood
ratio test strongly rejected the applicability of a molecular
clock. Our analysis provides an age of 118.5 ± 8.2 Myr for the
crown group of the Proteaceae. Table 2 provides the dates for
the key nodes indicated in Fig. 2.
DISCUSSION
Estimated age of the Proteaceae
Our analysis provides an age of 118.5 ± 8.2 Myr for the crown
group of the Proteaceae. This age agrees with those recovered
by Anderson et al. (2005) of 119–110 Myr (stem-group age)
and 96–85 Myr (crown-group age), and by Wikstro
¨met al.
(2003; 117–108 Myr, stem-group age). Unlike these studies,
our analysis included a comprehensive taxonomic sampling of
Proteaceae (including the earliest-diverging lineages) and
internal calibration points, which may explain the slightly
older estimate for the age of the crown.
These estimates are significantly older than the oldest fossil
that has been attributed, with some equivocation, to the
Proteaceae, the 93-Myr-old Cenomanian-Turonian Triorites
africaensis (Dettmann & Jarzen, 1998), used as a minimal
constraint for the age of this node in our analysis. However,
they are not much older than the oldest known platanaceous
fossil (108 Myr; Drinnan et al., 1994), which provides a
minimum age for the divergence of the sister groups Prote-
aceae and Platanaceae. This implies either that the fossil record
of the stem lineage of the Proteaceae is incomplete, or that the
earliest proteaceous fossils have not been recognized as
Proteaceae. Most comparable dating exercises (Wikstro
¨m
et al., 2001, 2003; Anderson et al., 2005; Bell et al., 2005;
Forest et al., 2005; Magallo
´n & Sanderson, 2005), however,
find similar discrepancies between the fossil record and the
estimated ages of taxa, consistent with the fact that the fossil
record provides estimates of only the minimum ages of taxa,
not their actual ages.
Biogeographical implications
Of the ten nodes associated with transoceanic disjunctions, the
age estimates for six (nodes marked A, )1, B, D, F and I in
Fig. 2) are consistent with a Gondwanan vicariance model
because they pre-date the dates of origin of their disjunctions,
known from historical geological research. These include one
Proteaceae biogeography and the break-up of Gondwana
Journal of Biogeography 34, 2012–2027 2021
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd

of the two groups that have already been shown by cladistic
biogeographical analysis to be congruent with a vicariant
history, the subtribe Embothriinae (Weston & Crisp, 1994),
represented in our analysis by the genera Alloxylon,Telopea
and Embothrium (node )1). Lomatia, the other clade inferred
by Weston & Crisp (1994) to be Gondwanan, was not
represented by taxa from both sides of the Pacific Ocean in our
analysis, and so its putatively vicariant history was not directly
tested, although its divergence from its sister group, Steno-
carpus (node A), substantially pre-dates the separation of
Australia from South America. Some of the other clades
inferred by our analysis to pre-date their disjunctions are
represented by few of the relevant species, and more thorough
sampling would allow a more rigorous test of their putative
Gondwanan origins.
Of the nodes associated with transoceanic disjunctions, the
age estimates for four are inconsistent with a Gondwanan
vicariance model (nodes marked C, E, G and H in Fig. 2), as
these considerably post-date major tectonic events associated
with the break-up of Gondwana (Table 2). Three of these
(nodes E, G and H) mark disjunctions involving Africa, two of
which involve an Africa-Australia split.
Node H, representing the well-supported split between
African Aulax and Australian Petrophile, is dated at 64.9 (±
11.9) Ma, some 35 Myr after Africa was isolated from any
other Gondwanan land mass. An even younger disjunction is
shown by Node G – the split between the Australian
Adenanthos and the South African Leucadendron, which is a
representative of the subtribe Leucadendrinae, the major
proteaceous radiation in the Cape flora (Barker et al., 2002;
Weston & Barker, 2006). This split is dated at 39.7 (± 9.3) Ma,
post-dating the isolation of Africa by approximately 66 Myr
(McLoughlin, 2001). Similarly, the divergence of the South
African monotypic Brabejum from the South American
Panopsis, dated at 43.8 (±10) Ma, post-dates the isolation of
Africa from South America by approximately 57 Myr (node E
in Fig. 2). The node immediately basal to this (linking the
Australian genus Macadamia as sister to this pair) pre-dates
the splitting of both Australia and South America from
Antarctica, a biogeographical relationship that is congruent
with the break-up of Gondwana. Thus, if our age estimates are
accurate, the ancestor of Brabejum arrived in Africa as a
consequence of transoceanic dispersal.
Node C marks the divergence of the New Zealand endemic
Knightia excelsa from its Australian/Asian sister clade of Helicia
and Hollandaea 45.4 (± 9.1) Ma. However, this post-dates the
isolation of New Zealand, which started rifting from the rest of
Gondwana about 84 Ma. It should be remembered, however,
that the phylogenetic position, and therefore the inferred date
of divergence of Knightia, should be treated with more caution
than the rest of our results (see above).
Our chronogram can be interpreted in different ways,
depending on the assumptions that one is prepared to accept.
One response would be to argue that, since a fossil provides
an estimate of only the minimum age of a clade, a fossil-
calibrated chronogram logically cannot show that a group is
younger than a geological disjunction (see e.g. Heads, 2005).
This response would immunize vicariance biogeography from
falsification by effectively constraining chronograms accord-
ing to the ages of putative vicariance events. In the case of
our chronogram for the Proteaceae, node G would be
constrained to a minimum of 105 Myr, thus pushing the
crown-group age of the family back substantially. The
obvious difficulty with this alternative is that the complete
absence of any proteaceous fossils before 94 Ma but their
abundance after about 85 Ma requires a general explanation.
We can think of no such explanation and so reject this
alternative.
Another response would be to argue that historical geolo-
gical knowledge concerning the sequence and age of Gondwa-
nan fragmentation is fallible, and that perhaps Africa and New
Zealand separated from the rest of Gondwana much later than
we think, during the Eocene, not the mid to late Cretaceous.
Although there appear to be some grounds for doubting the
commonly accepted timing of the separation of New Zealand
from Australia at 84 Ma (see below), the timing of the
separation of Africa from the rest of Gondwana before 100 Ma
seems very strongly corroborated (McLoughlin, 2001).
Yet another explanation for the discrepancy between our
chronogram and the general theory of Gondwanan vicariance
is that the model of evolution employed in our dating analysis
is fallible. While this possibility cannot be dismissed, it would
imply that the close correlation between the estimated ages of
several taxa showing Australian–South American disjunctions
(nodes marked )1, B, D and F in Fig. 2) and the timing of the
vicariance events that are most likely to have caused them is
merely coincidental.
If our chronogram is interpreted as providing estimates for
the actual ages of taxa (rather than estimates of their minimal
ages), the scientifically unsatisfying alternative of ad hoc appeal
to long-distance dispersal must be invoked, but this alternative
also faces significant difficulties. For the disjunctions separ-
ating Adenanthos from the subtribe Leucadendrinae and
Petrophile from Aulax, dispersal between southern Africa and
what is now Australia must be postulated.
As Adenanthos and the Leucadendrinae are related to a
paraphyletic series of Australian endemics, the most parsimo-
nious explanation for this disjunction is that Leucadendron and
related South African genera evolved from an ancestor that
arrived in Africa by long-distance dispersal from Australia
(node G). However, node G represents a biogeographical
conundrum, as the even more basal placement of the African
genus Protea (together with the unsampled African genus
Faurea, which is widely accepted as the sister genus of Protea;
Barker et al., 2002; Weston & Barker, 2006) suggests that the
more ancient ancestor for this clade as a whole could be
African. The ancestors of Isopogon and later Adenanthos could
thus have been African, dispersing independently twice to
Australia from Africa after Africa split from Gondwana. This is
obviously a less parsimonious explanation than the occurrence
of a single dispersal from Australia to Africa by the ancestor of
Leucadendron and allies, but an eastward (Australia–Africa)
N. P. Barker et al.
2022 Journal of Biogeography 34, 2012–2027
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd

direction of dispersal in many Gondwanan plants has been
inferred (using parsimony methods) to be slightly more
frequent than dispersal from Africa to Australia (Sanmartı
´n
& Ronquist, 2004). All of these inferences rely on at least two
simplifying assumptions: that biogeographical events affecting
individual lineages are independent of one another, and that
the probability (‘cost’) of each event in a class of events
(vicariance, duplication, dispersal, extinction) is identical. The
assumption of independence is clearly false when applied to
contemporaneous extinction events caused by environmental
catastrophes, and is even potentially misleading when applied
to postulated instances of long-distance dispersal, which may
be influenced by generic processes such as the direction and
strength of prevailing air and water currents. The assumption
of equal probability of events within an event class is similarly
simplistic. This example serves as a reminder that inferring the
directionality of dispersal events using parsimony is a specu-
lative exercise, especially when the costs of alternative recon-
structions differ only minimally.
If we accept both these dating results and the accuracy of
geological reconstructions of past distributions of land and sea,
then invoking long-distance dispersal is unavoidable for these
four disjunctions. The only independent, feasible test of this
hypothesis that we can imagine would involve predictions
concerning differences in the means of dispersal between clades
that are predicted to have undergone long-distance dispersal
and those that are predicted not to have. Specifically, one
would expect that putatively dispersed clades would have more
readily mobile fruits than clades for which long-distance
dispersal had not been inferred. However, this expectation is
not corroborated by our knowledge of fruit morphology in the
putatively dispersed taxa, some of which have readily dispersed
fruits, while others have short-lived seeds with no adaptations
for dispersal at all. The tribe Leucadendreae (represented by
Adenanthos and Leucadendron in our analysis), for example, is
characterized by relatively small, light, indehiscent, long-lived
fruits, many of which have structures that facilitate dispersal,
such as long hairs or wings. Dispersal of such fruits by wind
over long distances at least seems plausible (e.g. Renner et al.,
2000; Mun
˜oz et al., 2004), but it is reasonable to ask why there
is no evidence for the similarly mobile, plumose fruits of
Isopogon,Stirlingia and Conospermum ever having dispersed
across the Indian Ocean. The ((Helicia, Hollandaea)Knightia)
clade shows diverse fruit and seed morphology, the dehiscent
fruits of Knightia releasing moderately sized, winged seeds with
some aerodynamic properties. The (Macadamia (Panopsis,
Brabejum)) clade, on the other hand, is characterized by
relatively large, heavy, indehiscent, ballistically dispersed fruits
that quickly lose viability in response to desiccation. However,
the longevity of these seeds in sea water is unknown, and their
potential for dispersal by ocean currents could be readily tested
with immersion experiments using Macadamia seeds. Either
wind or ocean currents carried ancestral propagules of
the ancestor of Brabejum across an ocean gap of at least
1500 km if our dating estimates and geological knowledge are
accurate.
However unlikely it may seem, long-distance dispersal
should be seriously considered as a process that might have
contributed to the origin of these biogeographical patterns.
Similar results from recent studies of other putatively
Gondwanan taxa have seen a resurgence in dispersalist
interpretations (Renner, 2004; de Queiroz, 2005; Won &
Renner, 2006).
Comparisons with other Gondwanan groups
As noted in the Introduction, there are other plant groups that
have Gondwanan distributions, and for which vicariance
explanations have been proposed. Dating studies on some of
these can be compared with our results for the Proteaceae. The
Restionaceae have a long fossil record in southern Africa,
dating back to the late Cretaceous (Linder et al., 2003). This,
and phylogenies that indicate that the southern African lineage
is derived from a basal grade of Australian taxa, suggests that
the family is indeed Gondwanan, with a single African–
Australian vicariant event (Linder et al., 2003). Doyle et al.
(2004) present molecular-based age estimates for the Myris-
ticaceae crown group of 15–21 Myr (method-dependent),
which is in conflict with earlier theories (supported by a
species-level phylogeny suggesting Gondwanan vicariance) that
the family is of Cretaceous origin. The distribution of the
Myristicaceae (found in Africa, Madagascar, Asia and South
America) must thus be a result of transoceanic dispersal.
Richardson et al. (2004) obtained a similar result for the
Rhamnaceae, which, although having a southern continent
distribution, has a stem lineage that is too recent for the group
to have been Gondwanan. Studies on the Annonaceae as a
whole and on the basal genus Anaxagorea in particular suggest
this family to be of Gondwanan origin, but the radiation of
extant taxa is too recent to coincide with tectonic vicariance
events (Doyle et al., 2004; Richardson et al., 2004; Scharasch-
kin & Doyle, 2005).
Studies on Nothofagus, another well-known Gondwanan
genus, also have to resort to dispersal as an explanation. Knapp
et al. (2005) dated the lineages of Nothofagus and found that the
older divergences corresponded to Gondwanan tectonic events,
but inferred that dispersal must have occurred independently at
least twice between New Zealand and Australia, more recently
than 33 (±8) Ma. This date is comparable to that found for our
node G, discussed above. In a second study on Nothofagus,
Cook & Crisp (2005) arrive at similar dates using cpDNA data
(40–14 and 37–13 Ma for the subgenera Lophozoma and
Fucospora, respectively). Renner et al. (2000) dated the diver-
gences in the small family Atherospermataceae (which often co-
occurs with Nothofagus), and found divergence dates for two
disjunct species pairs (Chile–New Zealand and Australia–New
Caledonia) to be dated at 57–33 and 43–25 Ma, respectively
(calibration-dependent). These age estimates correspond well
with the fossil record and suggest that over-water dispersal
occurred in these two instances. However, a more radical
interpretation of the coincident ages of the Australia–New
Zealand–New Caledonian disjunctions is that overland
Proteaceae biogeography and the break-up of Gondwana
Journal of Biogeography 34, 2012–2027 2023
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd

dispersal routes persisted between Australia, New Zealand and
New Caledonia along the Tasman Plateau, Lord Howe Rise and
Norfolk Ridge until much later than is currently thought.
Quilty (2001) presents evidence for a 400-m-high volcanic
island that existed in the Late Eocene east of Tasmania where
the Cascade seamount now occurs, and Herzer et al. (1997)
suggest that Miocene uplift of the Norfolk and Peinga ridges
was sufficient to create islands that could facilitate dispersal. As
noted by McLoughlin (2001, p. 280), however, ‘little geological
data are available for the basement or lower post-break-up
sediments on the Lord Howe Rise’.
In a different region of Gondwana, molecular dating
estimates suggest that Southeast Asian Crypteroniaceae are
likely to have dispersed from Gondwana to the Deccan plate
about 87.25–72.15 Ma. This dispersal occurred across what
would then have been a narrow (200 km) Moc¸ambique
channel while the Deccan plate was rafting along the African
coast (but never in contact with it; Conti et al., 2002;
Rutschmann et al., 2004).
Geodispersal
The possibility of dispersal by ‘island hopping’ as postulated
above represents a third scenario: geodispersal (Lieberman,
2005). This model connects both vicariance and long-distance
dispersal by focusing on the variable strength of the dispersal
barrier over time. If the timeframe is known for when two
lineages with a disjunct distribution diverged, it is possible to
search for compatible dispersal corridors formed by tempo-
rarily reduced barriers that allowed dispersal at that specific
time. In the case of Crypteroniaceae, the Indian Ocean barrier
was temporally reduced by the Deccan plate closely rafting
along the African coast. Similarly, the links between New
Zealand and New Caledonia via the temporarily exposed
Tasman Plateau, Lord Howe Rise and Norfolk Ridge may
have been sufficient to aid in the dispersal of a number of
plant and animal taxa, including those of the Knightia–
Helicia clade. Similarly, hypotheses of now-submerged step-
ping stones between Africa and South America that could
have aided dispersal (as hypothesized for the Panopsis–
Brabejum disjunction, discussed above) have been proposed
(Morley, 2003). The palaeogeography of the South Atlantic
(reviewed by Parrish, 1993) and palynological evidence
(summarized by Morley, 2003) suggest that the now-
submerged Walvis Ridge and Rio Grande Rise (and possibly
the Sierra Leone Ridge) could have been aerial in the past. At
least some areas of these ridges were certainly sufficiently
shallow in places to support red algae (the remains of which
have been found in some drilling sites), and it is thought that
these were finally submerged by the late Eocene (Parrish,
1993), a date that is more recent than the age of the
Panopsis–Brabejum divergence obtained here. However, in the
absence of a detailed, realistic, probabilistic model of
dispersal, and accurate knowledge of the past distribution
of land and sea, ‘geodispersal’ provides just another kind of
narrative ‘explanation’.
Evolution of bird pollination
Our analysis provides an early Eocene date of origin for the
bird-pollinated subtribe Embothriinae (a clade represented by
the genera Alloxylon,Telopea and Embothrium in our chrono-
gram), with a radiation of the crown group in the subtribe
dated to 52.8 (± 6.8) Ma. As several of the morphological
synapomorphies that characterize this clade are adaptations for
bird pollination (Weston & Crisp, 1994), an implication of our
dating results is that nectarivory had evolved in birds at least
by the early Palaeogene. The oldest palaeontological evidence
of nectarivory in birds is a fossil hummingbird from the early
Oligocene of Germany (Mayr, 2004), but molecular dating of
avian clades suggests that the Meliphagoidea, an order
containing the nectar-feeding Meliphagidae (honey-eaters),
radiated in the early Eocene in Australia (Barker et al., 2004), a
timing that is approximately concordant with our estimates for
the diversification of this bird-pollinated group. However, this
result implies the existence of large gaps in the avian fossil
record.
CONCLUDING REMARKS
This dating exercise is the first for the charismatic plant family
Proteaceae as a whole, and raises some interesting and
controversial biogeographical, tectonic and co-evolutionary
questions. Despite an estimated crown-group age for the
Proteaceae that pre-dates the break-up of Gondwana, our
results suggest that the current disjunct distribution pattern of
Proteaceae is a result of both Gondwanan vicariance and more
recent transoceanic dispersal. This conclusion, along with
evidence from other plant groups, conflicts with the previously
accepted model invoking only vicariance and overland disper-
sal. Although still methodologically controversial, molecular
dating studies are proving to be a valuable tool in challenging
commonly accepted wisdom, attracting attention to alternative
biogeographical scenarios, and allowing for fossil evidence to
be re-assessed in an integrated chronological and phylogenetic
framework.
ACKNOWLEDGEMENTS
NPB thanks the Royal Botanic Garden, Sydney, for a
research fellowship in 1996, during which this work was
initiated, and also acknowledges financial support from the
National Research Foundation, South Africa, through
GUN2053645. Seranne Howis is thanked for aligning the
sequence data.
REFERENCES
Alley, N.F. & Beecroft, A. (1993) Foraminiferal and palynolo-
gical evidence from the Pidinga Formation and its bearing
on Eocene sea level events and paleochannel activity, eastern
Eucla Basin, South Australia. Memoir of the Association of
Australasian Palaeontologists,15, 375–393.
N. P. Barker et al.
2024 Journal of Biogeography 34, 2012–2027
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd

Alley, N.F. & Broadbridge, L.M. (1992) Middle Eocene paly-
nofloras from the One Tree Hill area, St Vincent basin,
South Australia. Alcheringa,16, 241–267.
Anderson, C.L., Bremer, K. & Friis, E.M. (2005) Dating phy-
logenetically basal eudicots using rbcL sequences and mul-
tiple fossil reference points. American Journal of Botany,92,
1737–1748.
APG (2003) An update of the Angiosperm Phylogeny Group
classification for the orders and families of flowering plants:
APG II. Botanical Journal of the Linnean Society,141, 399–
436.
Baldoni, A.M. & Askin, R.A. (1993) Palynology of the lower
Lefipa
´n Formation (Upper Cretaceous) of Barranca de los
Perros, Chubut Province, Argentina. Part II. Angiosperm
pollen and discussion. Palynology,17, 241–264.
Barker, N.P., Weston, P.H., Rourke, J.P. & Reeves, G. (2002)
The relationships of the southern African Proteaceae as
elucidated by internal transcribed spacer (ITS) DNA se-
quence data. Kew Bulletin,57, 867–883.
Barker, F.K., Cibois, A., Schikler, P., Feinstein, J. & Cracraft, J.
(2004) Phylogeny and diversification of the largest avian
radiation. Proceedings of the National Academy of Sciences
USA,101, 11040–11045.
Bell, C.D., Soltis, D.E. & Soltis, P.S. (2005) The age of the
angiosperms: a molecular timescale without a clock. Evolu-
tion,59, 1245–1258.
Christophel, D.C. (1984) Early Tertiary Proteaceae: the first
floral evidence for the Musgraveinae. Australian Journal of
Botany,32, 177–186.
Conti, E., Eriksson, T., Scho
¨nenberger, J., Sytsma, K.J. &
Baum, D.A. (2002) Early Tertiary out-of-India dispersal of
Crypteroniaceae: evidence from phylogeny and molecular
dating. Evolution,56, 1931–1942.
Cook, L.G. & Crisp, M.D. (2005) Not so ancient: the extant
crown group of Nothofagus represents a post-Gondwanan
radiation. Proceedings of the Royal Society B: Biological Sci-
ences,272, 2535–2544.
Dettmann, M.E. & Jarzen, D.M. (1996) Pollen of proteaceous-
type from latest Cretaceous sediments, southeastern Aus-
tralia. Alcheringa,20, 103–160.
Dettmann, M.E. & Jarzen, D.M. (1998) The early history of the
Proteaceae in Australia: the pollen record. Australian Sys-
tematic Botany,11, 401–438.
Doyle, J.J. & Doyle, J.L. (1987) A rapid DNA isolation pro-
cedure for small quantities of fresh leaf material. Phyto-
chemical Bulletin,19, 11–15.
Doyle, J.A., Sauquet, H., Scharaschkin, T. & Le Thomas, A.
(2004) Phylogeny, molecular and fossil dating, and bio-
geographic history of Annonaceae and Myristicaceae
(Magnoliales). International Journal of Plant Sciences,165,
S55–S67.
Drinnan, A.N., Crane, P.R. & Hoot, S.B. (1994) Patterns of
floral evolution in the early diversification of non-magnoliid
dicotyledons (eudicots). Plant Systematics and Evolution,8,
93–122.
Felsenstein, J. (1981) Evolutionary trees from DNA sequences:
a maximum likelihood approach. Journal of Molecular
Evolution,17, 368–376.
Felsenstein, J. (2004) PHYLIP (Phylogeny Inference Package).
Department of Genetics, University of Washington, Seattle.
Feuer, S. (1990) Pollen morphology of the Embothrieae
(Proteaceae) II. Embothriinae (Embothrium, Oreocallis,
Telopea). Grana,29, 19–36.
Forest, F., Savolainen, V., Chase, M.W., Lupia, R., Brunneau,
A. & Crane, P.R. (2005) Teasing apart molecular- versus
fossil-based error estimates when dating phylogenetic trees:
a case study in the birch family (Betulaceae). Systematic
Botany,30, 118–133.
Gilmore, S., Weston, P.H. & Thomson, J.A. (1993) A simple,
rapid, inexpensive and widely applicable technique for
purifying plant DNA. Australian Systematic Botany,6, 139–
148.
Guindon, S. & Gascuel, O. (2003) A simple, fast, and accurate
algorithm to estimate large phylogenies by maximum like-
lihood. Systematic Biology,52, 696–704.
Heads, M. (2005) Dating nodes on molecular phylogenies: a
critique of molecular biogeography. Cladistics,21, 62–78.
Herzer, R.H., Chaproniere, G.C.H., Edwards, A.R., Hollis, C.J.,
Pelletier, B., Raine, J.I., Scott, G.H., Stagpoole, V., Strong,
C.P., Symonds, P., Wilson, G.J. & Zhu, H. (1997) Seismic
stratigraphy and structural history of the Reinga basin and
its margins, southern Norfolk Ridge system. New Zealand
Journal of Geology and Geophysics,40, 425–451.
Hill, R.S., Scriven, L.J. & Jordan, G.J. (1995) The fossil record
of Australian Proteaceae. Flora of Australia,16, 21–30.
Hilu, K.W., Borsch, T., Muller, K., Soltis, D.E., Soltis, P.S.,
Savolainen, V., Chase, M.W., Powell, M.P., Alice, L.A.,
Evans, R., Sauquet, H., Neinhuis, C., Slotta, T.A.B., Rohwer,
J.G., Campbell, C.S. & Chatrou, L.W. (2003) Angiosperm
phylogeny based on matK sequence information. American
Journal of Botany,90, 1758–1776.
Hoot, S.B. & Douglas, A.W. (1998) Phylogeny of the Protea-
ceae based on atpb and atpb-rbcl intergenic spacer region
sequences. Australian Systematic Botany,11, 301–320.
Hoot, S.B., Magallo
´n, S. & Crane, P.R. (1999) Phylogeny of
basal eudicots based on three molecular data sets: atpB,
rbcL, and 18S nuclear ribosomal DNA sequences. Annals of
the Missouri Botanical Garden,86, 1–32.
Huelsenbeck, J.P. & Ronquist, F. (2001) MrBayes: Bayesian
inference of phylogeny. Bioinformatics,17, 754–755.
Hughes, N.F. & McDougall, A.B. (1990) Barremian-Aptian
angiospermid pollen records from southern England. Review
of Palaeobotany and Palynology,65, 145–151.
Johnson, L.A.S. & Briggs, B.G. (1975) On the Proteaceae – the
evolution and classification of a southern family. Botanical
Journal of the Linnean Society,70, 83–182.
Kim, S., Soltis, D.E., Soltis, P.S., Zanis, M.J. & Suh, Y. (2004)
Phylogenetic relationships among early-diverging eudicots
based on four genes: were the eudicots ancestrally woody?
Molecular Phylogenetics and Evolution,31, 16–30.
Proteaceae biogeography and the break-up of Gondwana
Journal of Biogeography 34, 2012–2027 2025
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd

Knapp, M., Sto
¨ckler, K., Havell, D., Delsuc, F., Sebastiani, F. &
Lockhart, P.J. (2005) Relaxed molecular clock provides
evidence for long-distance dispersal of Nothofagus (Southern
Beech). PLoS Biology,3, e4.
Ladiges, P.Y., Udovicic, F. & Nelson, G. (2003) Australian
biogeographical connections and the phylogeny of large
genera in the plant family Myrtaceae. Journal of Biogeo-
graphy,30, 989–998.
Lieberman, B.S. (2005) Geobiology and paleobiogeography:
tracking the coevolution of the Earth and its biota. Palaeo-
geography, Palaeoclimatology, Palaeoecology,219, 23–33.
Linder, H.P. & Hardy, C.R. (2004) Evolution of the species-
rich Cape flora. Philosophical Transactions of the Royal So-
ciety B: Biological Sciences,359, 1623–1632.
Linder, H.P., Eldena
¨s, P. & Briggs, B.G. (2003) Contrasting
patterns of radiation in African and Australian Restionaceae.
Evolution,57, 2688–2702.
Linder, H.P., Hardy, C.R. & Rutschmann, F. (2005) Taxon
sampling effects in molecular clock dating: an example from
the African Restionaceae. Molecular Phylogenetics and Evo-
lution,35, 569–582.
MacPhail, M.K. & Truswell, E.M. (1989) Palynostratigraphy of
the central west Murray basin. BMR Journal of Australian
Geology and Geophysics,11, 301–331.
Maddison, W.P. & Maddison, D.R. (2000) MacClade 4: ana-
lysis of phylogeny and character evolution. Sinauer Associates,
Sunderland, MA.
Magallo
´n, S.A. (2004) Dating lineages: molecular and pa-
leontological approaches to the temporal framework of
clades. International Journal of Plant Sciences,165, S7–S21.
Magallo
´n, S.A. & Sanderson, M.J. (2005) Angiosperm diver-
gence times: the effect of genes, codon positions, and time
constraints. Evolution,59, 1653–1670.
Mayr, G. (2004) Old world fossil record of modern-type
hummingbirds. Science,304, 861–864.
McLoughlin, S. (2001) The breakup history of Gondwana and
its impact on pre-Cenozoic floristic provincialism. Austra-
lian Journal of Botany,49, 271–300.
Milne, L.A. (1988) Palynology of a late Eocene lignitic se-
quence from the western margin of the Eucla Basin, Western
Australia. Memoirs of the Association of Australasian Palae-
ontologists,5, 285–310.
Minin, V., Abdo, Z., Joyce, P. & Sullivan, J. (2003) Perfor-
mance-based selection of likelihood models for phylogeny
estimation. Systematic Biology,52, 674–683.
Morley, R.J. (2003) Interplate dispersal paths for megathermal
angiosperms. Perspectives in Plant Ecology, Evolution and
Systematics,6, 5–20.
Mun
˜oz, J., Felicı
´simo, A.M., Cabezas, F., Burgaz, A.R. &
Martı
´nez, I. (2004) Wind as a long-distance dispersal vehicle
in the southern hemisphere. Science,304, 1144–1147.
Near, T.J. & Sanderson, M.J. (2004) Assessing the quality of
molecular divergence time estimates by fossil calibrations
and fossil-based model selection. Philosophical Transactions
of the Royal Society B: Biological Sciences,359, 1477–1483.
Nylander, J. (2005) BootPHYML. School of Computational
Science (SCS), Florida State University, Tallahassee.
Parrish, J.T. (1993) The palaeogeography of the opening
South Atlantic. The Africa–South America connection (ed.
by W. George and R. Lavocat), pp. 8–27. Clarendon Press,
Oxford.
Qiu, Y.L., Lee, J., Bernasconi-Quadroni, F., Soltis, D.E., Soltis,
P.S., Zanis, M., Zimmer, E.A., Chen, Z., Savolainen, V. &
Chase, M.W. (2000) Phylogeny of basal angiosperms: ana-
lyses of five genes from three genomes. International Journal
of Plant Sciences,161, S3–S27.
de Queiroz, A. (2005) The resurrection of oceanic dispersal in
historical biogeography. Trends in Ecology & Evolution,20,
68–73.
Quilty, P.G. (2001) Late Eocene foraminifers and palaeoenvi-
ronment, Cascade Seamount, southwest Pacific Ocean:
implications for seamount subsidence and Australia–Ant-
arctica Eocene connection. Australian Journal of Earth Sci-
ence,48, 633–641.
Raven, P.H. & Axelrod, D.I. (1974) Angiosperm biogeography
and past continental movements. Annals of the Missouri
Botanic Garden,61, 539–673.
Renner, S.S. (2004) Bayesian analysis of combined chloroplast
loci, using multiple calibrations, supports the recent arrival
of Melastomataceae in Africa and Madagascar. American
Journal of Botany,91, 1427–1435.
Renner, S.S. (2005) Relaxed molecular clocks for dating his-
torical plant dispersal events. Trends in Plant Science,10,
550–558.
Renner, C., Foreman, D.B. & Murray, D. (2000) Timing
transantarctic disjunctions in the Atherospermataceae
(Laurales): evidence from coding and noncoding chloroplast
sequences. Systematic Biology,49, 579–591.
Richardson, J.E., Chatrou, L.W., Mols, J.B., Erkens, R.H.J. &
Pirie, M.D. (2004) Historical biogeography of two cosmo-
politan families of flowering plants: Annonaceae and
Rhamnaceae. Philosophical Transactions of the Royal Society
B: Biological Sciences,359, 1495–1508.
Ronquist, F. & Huelsenbeck, J.P. (2003) MrBayes 3: Bayesian
phylogenetic inference under mixed models. Bioinformatics,
19, 1572–1574.
Rutschmann, F. (2005) Bayesian molecular dating using PAML/
multidivtime. A step-by-step manual,version 1.5. http://www.
plant.ch.
Rutschmann, F. (2006) Molecular dating of phylogenetic trees:
a brief review of current methods that estimate divergence
times. Diversity and Distributions,12, 35–48.
Rutschmann, F., Eriksson, T., Scho
¨nenberger, J. & Conti, E.
(2004) Did Crypteroniaceae really disperse out of India?
Molecular dating evidence from rbcL, ndhF, and rpl16 in-
tron sequences. International Journal of Plant Sciences,165,
S69–S83.
Sanderson, M.J. (1997) A nonparametric approach to esti-
mating divergence times in the absence of rate constancy.
Molecular Biology and Evolution,14, 1218–1231.
N. P. Barker et al.
2026 Journal of Biogeography 34, 2012–2027
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd

Sanderson, M.J. (2002) Estimating absolute rates of molecular
evolution and divergence times: a penalized likelihood ap-
proach. Molecular Biology and Evolution,19, 101–109.
Sanmartı
´n, I. & Ronquist, F. (2004) Southern hemisphere
biogeography inferred by event-based models: plant versus
animal patterns. Systematic Biology,53, 216–243.
Scharaschkin, T. & Doyle, J.A. (2005) Phylogeny and historical
biogeography of Anaxagorea (Annonaceae) using mor-
phology and non-coding chloroplast DNA sequence data.
Systematic Botany,30, 712–735.
Soltis, D.E., Soltis, P.S., Chase, M.W., Mort, M.E., Albach,
D.C., Zanis, M., Savolainen, V., Hahn, W.H., Hoot, S.B.,
Fay, M.F., Axtell, M., Swensen, S.M., Prince, L.M., Kress,
W.J., Nixon, K.C. & Farris, J.S. (2000) Angiosperm phylo-
geny inferred from 18S rDNA, rbcL, and atpB sequences.
Botanical Journal of the Linnean Society,133, 381–461.
Stover, L.E. & Partridge, A.D. (1973) Tertiary and Late Cre-
taceous spores and pollen from the Gippsland Basin,
southeastern Australia. Proceedings of the Royal Society of
Victoria,85, 237–286.
Stover, L.E. & Partridge, A.D. (1982) Eocene spore–pollen
from the Werillup Formation, Western Australia. Palynol-
ogy,6, 69–95.
Swenson, U., Hill, R.S. & McLoughlin, S. (2001) Biogeography
of Nothofagus supports the sequence of Gondwana break-
up. Taxon,50, 1025–1041.
Sytsma, K.J., Litt, A., Zjhra, M.L., Pires, J.C., Nepokroeff, M.,
Conti, E., Walker, J. & Wilson, P.G. (2004) Clades, clocks,
and continents: historical and biogeographical analysis of
Myrtaceae, Vochysiaceae, and relatives in the southern
hemisphere. International Journal of Plant Sciences,165, 85–
105.
Thorne, J.L. & Kishino, H. (2002) Divergence time and evo-
lutionary rate estimation with multilocus data. Systematic
Biology,51, 689–702.
Thorne, J.L., Kishino, H. & Painter, I.S. (1998) Estimating the
rate of evolution of the rate of molecular evolution. Mole-
cular Biology and Evolution,15, 1647–1657.
Truswell, E.M. & Owen, J.A. (1988) Eocene pollen from
Bungonia, New South Wales. Association of Australasian
Palaeontologists,5, 259–284.
Truswell, E.M., Sluiter, I.R. & Harris, W.K. (1985) Palynology
of the Oligocene-Miocene sequence in the Oakvale-1 core-
hole, western Murray Basin, South Australia. BMR Journal of
Australian Geology and Geophysics,9, 267–295.
Vadala, A.J. & Drinnan, A.N. (1998) Elaborating the fossil
history of Banksiinae: A new species of Banksieaephyllum
(Proteaceae) from the late Paleocene of New South Wales.
Australian Systematic Botany,11, 439–463.
Wanntorp, L. & Wanntorp, H.E. (2003) The biogeography of
Gunnera L.: vicariance and dispersal. Journal of Biogeo-
graphy,30, 979–987.
Ward, J.V. & Doyle, J.A. (1994) Ultrastructure and relation-
ships of mid-Cretaceous polyforate and triporate pollen
from northern Gondwana. Ultrastructure of fossil spores and
pollen (ed. by M.H. Kurmann and J.A. Doyle), pp. 161–172.
Royal Botanic Gardens, Kew.
Weston, P.H. & Barker, N.P. (2006) A new suprageneric
classification of the Proteaceae, with an annotated checklist
of genera. Telopea,11, 314–344.
Weston, P.H. & Crisp, M.D. (1994) Cladistic biogeography of
Waratahs and their allies (Embothrieae: Proteaceae) across
the Pacific. Australian Systematic Botany,7, 225–249.
Weston, P.H. & Crisp, M.D. (1996) Trans-Pacific biogeo-
graphic patterns in the Proteaceae. The origin and evolution
of Pacific Island biotas, New Guinea to Eastern Polynesia:
patterns and processes (ed. by A. Keast and S.E. Miller), pp.
215–232. SPB Academic Publishing, Amsterdam.
Wikstro
¨m, N., Savolainen, V. & Chase, M.W. (2001) Evolution
of the angiosperms: calibrating the family tree. Proceedings of
the Royal Society B: Biological Sciences,268, 2211–2220.
Wikstro
¨m, N., Savolainen, V. & Chase, M.W. (2003) Angios-
perm divergence times: congruence and incongruence be-
tween fossils and sequence divergence estimates. Telling the
evolutionary time: molecular clocks and the fossil record (ed.
by P.C.J. Donoghue and M.P. Smith), pp. 142–165. Taylor &
Francis, London.
Won, H. & Renner, S.S. (2006) Dating dispersal and radiation
in the Gymnosperm Gnetum (Gnetales) – clock calibration
when outgroup relationships are uncertain. Systematic Bio-
logy,55, 610–622.
Zerega, N.J.C., Clement, W.L., Datwyler, S.L. & Weiblen, G.D.
(2005) Biogeography and divergence times in the mulberry
family (Moraceae). Molecular Phylogenetics and Evolution,
37, 402–416.
BIOSKETCHES
Nigel Barker has a wide range of systematic research
interests, most of which utilize DNA techniques to assess
phylogenetic relationships and biogeographical patterns. Cur-
rent plant groups being studied include the Proteaceae,
Poaceae, Asteraceae and Asphodelaceae.
Peter Weston has a long-standing interest in the Proteaceae
and Orchidaceae, and both he and Nigel Barker are interested
in elucidating the phylogeny and radiation of the Proteaceae,
using both molecular and morphological data.
Frank Rutschmann has specialized expertise in molecular
dating techniques, and has worked on the molecular dating of
a number of groups, including the Crypteroniaceae and
Restionaceae.
Herve
´Sauquet is currently a post-doctoral researcher
investigating the palaeobotanical record of the Proteaceae,
particularly the fossil pollen record.
Editor: Pauline Ladiges
Proteaceae biogeography and the break-up of Gondwana
Journal of Biogeography 34, 2012–2027 2027
ª2007 The Authors. Journal compilation ª2007 Blackwell Publishing Ltd