Isolated with persistence or dynamically connected? Genetic patterns in a

Diversity and Distributions
A Journal of Conservation Biogeography
Diversity and Distributions, (Diversity Distrib.) (2014) 1–15
BIODIVERSITY
RESEARCH
Isolated with persistence or dynamically
connected? Genetic patterns in a
common granite outcrop endemic
S.-L. Tapper1, M. Byrne1,2*, C. J. Yates1, G. Keppel3,4, S. D. Hopper5,
K. Van Niel6, A. G. T. Schut3, L. Mucina2,3 and G. W. Wardell-Johnson3
1
Science and Conservation Division,
Department of Parks and Wildlife, Locked
Bag 104, Bentley Delivery Centre, WA 6983,
Australia, 2School of Plant Biology, The
University of Western Australia, 35 Stirling
Highway, Crawley, WA 6009, Australia,
3
Curtin Institute for Biodiversity and
Climate, Curtin University, GPO Box
U1987, Perth, WA 6845, Australia, 4School
of Natural and Built Environments and
Barbara Hardy Institute, University of South
Australia, GPO Box 2471, Adelaide, SA
5001, Australia, 5Centre of Excellence in
Natural Resource Management and School
of Plant Biology, The University of Western
Australia, Foreshore House, Proudlove
Parade, Albany, WA 6330, Australia,
6
School of Earth and Environment, The
University of Western Australia, 35 Stirling
Highway, Crawley, WA 6009, Australia
ABSTRACT
Aim Granite outcrops are prominent throughout the world and harbour many
endemic species. Their topographic complexity and range of environments have
led to the hypothesis that they act as refugia facilitating the persistence of species through climate change. We evaluate this hypothesis by investigating the
phylogeographic patterns in a common granite endemic shrub.
Location Granite outcrops of the Southwest Australian Floristic Region.
Methods Chloroplast haplotypes of 89 Kunzea pulchella individuals from 16
granite outcrops were determined from sequences of three chloroplast intergenic
spacer regions. Phylogenetic reconstruction and divergence dating was inferred
using Bayesian and Parsimony analyses and phylogenetic relationships between
haplotypes were examined in relation to geographic distributions. Nuclear diversity and differentiation of populations were assessed through analysis of 11
nuclear microsatellite loci across 384 individuals from the 16 granite outcrops.
Results Kunzea pulchella exhibited low haplotype and allelic diversity within
outcrops and high levels of divergence among outcrops, indicating an ancient
restriction to specific outcrops with genetic drift as the main driver of evolution. Two divergent lineages were revealed in the chloroplast phylogeny dating
to the Pliocene and potentially reflecting the initial impact of increased aridity
prior to isolation on individual outcrops.
Main conclusions Rather than uncovering the typical pattern for Pleistocene
*Correspondence: Margaret Byrne, Science
and Conservation Division, Department of
Parks and Wildlife, Locked Bag 104, Bentley
Delivery Centre, WA 6983, Australia.
E-mail: [email protected]
refugia with contraction to, and expansion from particular granite outcrops, we
observed persistence, prolonged isolation and divergence of populations. We
suggest the persistence of K. pulchella on multiple outcrops through a period of
considerable climatic change may be a result of broad climatic tolerances or
contraction and expansion dynamics operating at microrefugial scales within
outcrops. Our observations of low haplotype and allelic diversity within populations of K. pulchella provide some support for the latter. The enduring nature
of K. pulchella and evolutionary potential of populations on individual outcrops accentuates the value of these environments for biodiversity conservation
planning in a changing climate.
Keywords
Chloroplast divergence, evolutionary history, nuclear diversity, phylogeography, Pleistocene refugia, terrestrial islands.
Refugia are habitats that species retreat to, persist in and
potentially expand from under changing environmental
conditions (Keppel et al., 2012). Often a refugium will be a
place providing environmental diversity and stability as
regional environments change (Keppel et al., 2012). These
places have intrinsic conservation value because they
ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
DOI: 10.1111/ddi.12185
http://wileyonlinelibrary.com/journal/ddi
INTRODUCTION
1
S.-L. Tapper et al.
facilitate the local persistence of species and genotypes when
regional conditions are unsuitable, foster evolutionary processes that may lead to diversification, and may serve as
depositories of rare species or genotypes (Ashcroft et al.,
2012).
The role of refugia in facilitating the persistence and diversification of species during Pleistocene climate changes has
been examined broadly in the literature (e.g. Soltis et al.,
1997; Taberlet et al., 1998; Sch€
onswetter et al., 2002;
Anthony et al., 2007). Most studies have focused on the glaciated landscapes of the Northern Hemisphere (Hewitt,
2004), but an increasing number of studies have considered
species responses in regions that remained unaffected by ice
sheets, yet experiencing profound climate changes (Beheregaray, 2008; Keppel et al., 2012). In these landscapes, refugia
are often associated with complex topography, such as
mountain ranges and deep valleys that are buffered from
environmental extremes and provide some climate stability
(Garrick, 2011).
South-west Western Australia, classified biogeographically
as the South-West Australian Floristic Region (SWAFR;
Hopper & Gioia, 2004), is one of five regions characterized
by mediterranean-type climate and is recognized as a global
biodiversity hotspot (Myres et al., 2000). Unlike the other
mediterranean-climate regions, SWAFR has low relief and,
with the exception of the Stirling Range that reaches 1109 m
in altitude, there is limited scope for contraction and expansion along elevational gradients as regional climates change
(Yates et al., 2010). Yet, some phylogeographic patterns
emerging from recent studies show that as climate became
drier during the Pleistocene glacial phases, species contracted
to and persisted in localized refugia (Byrne, 2008). The characteristics of these places remain cryptic and warrant further
investigation (Ashcroft, 2010; Ashcroft & Gollan, 2013).
Granite outcrops are prominent topographical features in
the otherwise flat landscape of the inland SWAFR. They are
of great geological age (Archaean), with evidence for persistence as continuously exposed terrestrial landscape features
from the mid-Cretaceous (Twidale & Bourne, 1998). While
the rock surface itself is generally very arid, granite outcrops
can act as water-harvesting sites by channelling run-off water
to crevices and margins, as well as gnammas (rock pools),
creating seasonally water-rich microhabitats (York Main,
1997). The association of these features with putative relictual species and early branching lineages in molecular phylogenies (e.g. Hsiao et al., 1998; Byrne et al., 2001; Fay et al.,
2001; Saarela et al., 2007) has led to the hypothesis that
granite outcrops in the SWAFR may have been refugia during periods of increased aridity through the Pleistocene and
earlier periods (Hopper et al., 1997; Byrne, 2008). Some outcrops may be more likely to act as refugia than others based
on their topography. For example, larger outcrops harvest a
greater amount of rainfall, providing a higher level of moisture to crevices, gnammas and fringes (Laing & Hauck, 1997;
York Main, 1997). Outcrops with a greater number of cracks,
fissures, or gnammas may provide greater habitat than more
2
uniform, smooth rock faces. If granite outcrops have acted
as refugia during the Pleistocene arid periods they may prove
essential for the survival of some species during increased
aridity predicted in the SWAFR as a consequence of anthropogenic climate change (Bates et al., 2008).
The role of granite outcrops as refugia for species that
occupy the intervening matrix is difficult to test because
much of this landscape has been cleared for agriculture.
However, there are many species endemic to granite outcrop
habitats, and these may show evidence of contraction to, and
expansion from, particular outcrops if local extinction and
recolonization have occurred through Pleistocene climate
cycles. To date, phylogeographic research has failed to provide clear evidence for species contracting to and expanding
from specific granite outcrops. The few phylogeographical
and genetic analyses of granite endemics in the SWAFR have
uncovered patterns of persistence, prolonged isolation and
divergence (Sampson et al., 1988; Yates et al., 2007; Byrne &
Hopper, 2008; Levy et al., 2012), rather than the contraction/local extinction and recolonization patterns common to
many Pleistocene refugia. This pattern of localized persistence may be a result of the rarity of the plant species studied to date. Rare species might be less informative than
widespread species in identifying signals of range dynamics,
as they may be restricted to certain granite outcrops because
of attributes of their ecology, morphology or reproductive
biology that reduce their ability to expand from local areas
of persistence. Studies of widespread species across a broad
range of environmental conditions and a range of outcrops
would enhance the power to detect signals of expansion and
contraction from particular rock complexes.
Kunzea pulchella Lindl. (Myrtaceae) is a bird-pollinated
evergreen shrub endemic to granite outcrops of the SWAFR
and adjacent Eremean province. The species grows in the
abundant small cracks or crevices in the granite surface
where rainfall run-off is often channelled creating a moist
microhabitat. Kunzea pulchella exhibits similar life-history
traits to a previously studied rare species that is restricted to
a small number of outcrops, Eucalyptus caesia; however,
K. pulchella is common and geographically widespread. York
Main (1997) described granite outcrops as a collective ecosystem linked by the aeolianosphere, and interactions
between granite outcrops and the atmosphere show thermal
drafts above granite outcrops lift air to considerable heights
(Szarzynski, 2000). The widespread distribution and small
light seeds of K. pulchella suggests that dispersal between
rocks could occur through wind dynamics, and that refugial
dynamics at the collective ecosystem scale might be observed.
Thus, the species may harbour a signature of contraction
and expansion that would provide evidence for particular
outcrops acting as refugia within the granite network.
Phylogeographic analyses can identify the presence of refugia by revealing genetic signatures of historical processes in
extant species (Hewitt, 2000, 2004). Refugia are characterized
by high diversity, and where species have expanded from
refugia, a genetic signature of low diversity among popula-
Diversity and Distributions, 1–15, ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
Isolation and persistence in Kunzea pulchella
tions is expected (Hewitt, 2000, 2004; Byrne, 2008), although
expansion signals may be complex due to long distance colonization ahead of an expansion front (Ibrahim et al., 1996)
or fixation of low frequency alleles driven by ‘allele surfing’
(Excoffier & Ray, 2008; Excoffier et al., 2009). We undertook
a phylogeographic analysis of K. pulchella to investigate
genetic signatures of expansion and contraction that would
provide evidence for specific granite outcrops acting as Pleistocene refugia. Specifically, we hypothesized that granite outcrops acting as refugia would have high haplotype and allelic
diversity, and outcrops that had been recolonized would
show low haplotype and allelic diversity with patterns of geographic structure among adjacent outcrops. We also investigated the extent to which rock size influenced genetic
variation and hypothesized that larger granite outcrops with
more fissures and a greater variety of environments would
show signals of being refugia.
METHODS
Collections and DNA extractions
Leaf material was collected from 24 individuals from each of
16 K. pulchella populations on granite outcrops in the centre
of the species range (Fig. 1a). Approximately, 40 mg of dried
leaf material was ground to a fine powder using a Tissue
Lyser (Retsch, D-Haan, Germany) and DNA extracted via a
cetyltrimethyl ammonium bromide method (Tapper et al.,
2013).
Chloroplast DNA sequencing and analysis
Polymerase chain reaction (PCR) and sequencing trials
were conducted on two samples across six non-coding
chloroplast DNA (cpDNA) regions that have been shown
to be useful for phylogeographic studies in Australian
plants (Byrne & Hankinson, 2012). The psbA–trnH, psbD–
trnT and trnS–trnG intergenic spacer regions were selected
for further analysis based on sequence quality and nucleotide diversity. Sequencing of these three regions was conducted for six individuals from each of the 16 sampled
populations of K. pulchella, as well as samples of K. baxteri, sister to K. pulchella, and K. preissiana, also within
the Salisia subgenus (de Lange et al., 2010), for use as
outgroups. PCR amplification was completed in 50 lL volumes with the following reagent proportions: 40 ng template DNA, 10 lL of 59 PCR buffer (50 mM KCl, 20 mM
Tris-HCl pH 8.4, 0.2 mM dNTP), 0.1 lM each primer,
either 3 mM (trnH–psbA) or 1.5 mM (psbD–trnT, trnS–
trnG) MgCl2 and 0.5 unit Taq polymerase (Life Technologies, Melbourne). The PCR cycling protocol involved initial denaturation at 80 °C for 5 min, 30 cycles of 95 °C
for 1 min, 50 °C for 1 min ramping by 0.3 °C per second
to 65 °C, 65 °C for 4 min, and final extension at 65 °C
for 5 min. Amplification products were cleaned using a
polyethanol glycol precipitation method and sequencing
Diversity and Distributions, 1–15, ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
reactions were completed through the Macrogen Inc. EZseq service (Seoul, South Korea).
DNA sequence chromatograms for all three regions were
edited for miscalls in SEQUENCHER 5.0 (Genecodes Corp., MI,
USA). Sequence alignment was initially carried out using
ClustalW in MEGA 5.05 (Tamura et al., 2007) with manual
alignment using MESQUITE where large indels prevented parsimonious alignment (Maddison & Maddison, 2007).
Sequences of all three loci were concatenated in MESQUITE. A
19 base pair inversion was uncovered in the psbA-trnH
region. To prevent the overestimation of substitution events
and incorrect phylogeny reconstruction, one inversion configuration was replaced with its reverse complement. A single
transversion was then added to the end of the sequence to
code for the inversion, as outlined by Whitlock et al. (2010).
Potentially informative indels were coded in SEQSTATE 1.4.1
(M€
uller, 2005) under the Simmons & Ochoterena (2000)
simple coding scheme.
Tajima’s D (Tajima, 1989) and Ramos-Onsins & Rozas
(2002) R2 were calculated in DNASP 5.10.01 (Librado & Rozas, 2009) across all samples and within each of the two
divergent clades to test for neutrality and population growth
or decline. Identification of haplotypes and calculation of
nucleotide and haplotype diversity was undertaken with
DNASP. One representative of each haplotype was used in
analyses thereafter. Parsimony analysis with PAUP* 4.0B10
(Swofford, 2003) used 1000 random heuristic search replicates with the Tree Bisection–Reconnection branch swapping
algorithm and bootstrapping was conducted using heuristic
searches with simple stepwise addition and 1000 bootstrap
replicates. The level of homoplasy was assessed using the
Consistency Index (CI) and Retention Index (RI). A medianjoining maximum parsimony network of the 18 chloroplast
haplotypes was drawn in NETWORK 4.6.1.1 (Bandelt et al.,
1999) with epsilon set to 0 and indels treated as binary characters as coded in SEQSTATE.
Molecular dating and topology was simultaneously estimated via a relaxed clock Bayesian analysis in BEAST 1.7.4
(Drummond & Rambaut, 2007). The binary indel partition
was omitted from the analysis because it does not meet the
assumption of a clockwise mutation rate. A coalescent tree
prior was chosen as the most appropriate for this type of
dataset following the findings of Drummond et al. (2002).
An uncorrelated lognormal relaxed clock model was used for
the rate variation among branches (Drummond et al., 2006).
Divergence time estimates were constrained by placing a calibration at the root using the mean (4.34 Myr) and 95% CI
(1.84–7.61) of the time since most recent common ancestor
(TMRCA) of K. pulchella and K. baxteri acquired from an
unpublished dataset (A. Thornhill, pers. comm.) that was
calibrated using the divergence dates from Thornhill et al.
(2012) based on pollen fossil data. The GTR model with
gamma distributed rate variation was the best-fit model
for Bayesian analysis as determined in JMODELTEST 0.1.1
(Guindon & Gascuel, 2003; Posada, 2008). Markov Chain
Monte Carlo was executed with two chains for 1.0 9 107
3
S.-L. Tapper et al.
(a)
(b)
generations with trees and parameters sampled every 1000
generations. Convergence of parameters were assessed by
examining trace shape and ensuring all effective sample sizes
were > 200 using TRACER software (Rambaut & Drummond,
2007). TREEANNOTATOR (Drummond & Rambaut, 2007) was
used to identify the maximum clade credibility tree.
Nuclear analysis
Primers and protocols previously developed by Tapper et al.
(2013) were used to amplify eleven microsatellite loci
(Kun03, 06, 07, 09, 10, 12, 13, 14. 16, 18, 20) for 24 individuals from each of 16 sampled populations. An Applied Biosystems 3730 DNA Analyser (Applied Biosystems) was used
to visualize the amplification products, and genotypes were
scored using GENEMAPPERTM 4.0 analysis software (Applied
Biosystems). Polymerase Chain Reaction and genotyping was
repeated for 5% of samples across all 11 loci to quantify
scoring error in PEDANT 1.0 (Johnson & Haydon, 2007).
4
Figure 1 (a) Locations in Western
Australia of 16 Kunzea pulchella
populations sampled on granite outcrops
for this study and geographic
distribution of chloroplast DNA
haplotypes found in populations. Dark
gray shading on inset map represents the
natural distribution of K. pulchella (b)
Median-joining network for chloroplast
haplotypes identified in K. pulchella. The
size of the portions of pie charts
represent the number of individuals of
that haplotype. Colour-coding of
haplotypes in the map correspond to
those in the median-joining haplotype
network. Circle sizes in the network are
relative to haplotype frequency and
branch lengths indicate the number of
mutations between haplotypes.
The mean number of alleles per locus, the average allelic
richness across all loci, and Weir & Cockerham’s (1984)
measure of FIS were calculated for all 16 populations with
the GENEPOP 4.0.1 web interface (Raymond & Rousset, 1995).
Deviations from Hardy–Weinberg equilibrium were assessed
by exact tests using GENEPOP 4.0.1. Linkage Disequilibrium
between each pair of loci was analysed in each population
using the Log-likelihood statistic and significance values were
corrected using a sequential Bonferroni correction (Holm,
1979) to minimize the effect of Type I errors. Null allele frequencies (A0) and large allele drop-out were estimated using
FREENA (Chapuis & Estoup, 2007). Observed and expected
heterozygosity were calculated in GENALEX 6.41 (Peakall &
Smouse, 2006). BOTTLENECK 1.2.02 (Piry et al., 1999) was used
to test for the evidence of recent reductions in effective population size. The Wilcoxon’s signed-rank test was used to
identify significant heterozygosity excess (Piry et al., 1999).
Mode-shift tests were used to detect shifts in allele frequencies characteristic of more recent bottlenecks.
Diversity and Distributions, 1–15, ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
Isolation and persistence in Kunzea pulchella
Bayesian analysis with the assumption of independent alleles
was implemented in STRUCTURE 2.3.3 (Pritchard et al., 2000) to
estimate the affinity of individuals to genetically homogenous
groups (K) using the admixture ancestry model. Burnin length
was set to 100,000 followed by 100,000 Markov Chain Monte
Carlo repetitions with 10 simulations for each proposed value
of K (K = 1–15). Results were uploaded to STRUCTURE HARVESTER 0.6.92 (Earl & vonHolt, 2012) to identify the most probable value of K. Clustering patterns were visualized in DISTRUCT
1.1 (Rosenberg, 2004) after aligning all 10 runs for the optimum K with CLUMPP 1.1.2 (Jakobsson & Rosenberg, 2007) to
obtain a similarity coefficient (h′).
A principal coordinates analysis of Nei’s unbiased genetic
distance was conducted using GENALEX 6.41 (Peakall &
Smouse, 2006). The Carvalli-Sforza & Edwards’ (1967)
Chord genetic distance was used to generate a majority-rule
consensus neighbour-joining tree from 1000 bootstrap replicates in PHYLIP 3.69 (Felsenstein, 1989). Pairwise population
FST and Dest were calculated with 999 permutations in GENALEX 6.5 (Excoffier et al., 2005). Calculation of global FST, partitioning of genetic variance by analysis of molecular
variance (AMOVA), and tests for isolation by distance (IBD)
were determined in GENALEX 6.41.
Outcrop size was measured by creating polygons outlining
each outcrop based on satellite imagery in Googleearth and
imported into ArcMap v9.2 (ESRI, 2009). The polygons were
drawn around the base, where a visual transition of slope
was observed and were therefore to some extent, arbitrary.
Outcrop size was not normally distributed and square-root
and log transformations improved the distributions sufficiently to investigate the relationship between outcrop size
and genetic diversity using Spearman’s Rank correlation (rs)
in Statistica 6 (Statsoft 2001) (Table 1).
RESULTS
Chloroplast sequence dataset
The cpDNA sequence data set revealed 18 different K. pulchella haplotypes in 89 samples (Table 2). Outgroup samples
included one K. baxteri and three K. priessiana haplotypes
(Table 2). Seven samples of K. pulchella (four from Talgomine Rock 2, two from Marshall Rock, one from Sandford
Rocks) consistently produced poor quality sequence data in
at least one of the three regions and thus were omitted from
analyses. Bayesian analysis of data from individual regions
produced generally concordant data sets (see Fig. S1 in Supporting Information), and the sequences from each region
were combined. Aligned sequences (including 47 binary
characters representing indels) were 2549 bp in length. The
data set encompassed 185 variable sites of which 157 were
parsimony informative. Parsimony analysis resulted in 82
equally parsimonious trees of 209 steps (CI = 0.90,
RI = 0.99). Parsimony bootstrap support and posterior probability scores for the combined data set are presented in
Fig. 2. Nucleotide and haplotype diversity across all samples
of K. pulchella were 0.005 and 0.868, respectively. Neutrality
tests were non-significant for both lineages for Tajima’s D
and R2 (Table 3).
Bayesian and parsimony analysis split the samples into
two divergent lineages with strong support (PP = 1.0,
BS = 80). The northern clade contained haplotypes from
Geeraning and Elachbutting Hill (Fig. 2). The southern
clade, consisting of haplotypes from all remaining outcrops,
showed little geographic structure and multiple, poorly supported, higher level branches (Fig. 2). The same phylogeographic patterns were evident in the median-joining
Table 1 Population names and location information for populations of Kunzea pulchella and outgroup samples of K. baxteri and
K. preissiana.
Population
Locality details
Latitude/Longitude
Billycatting Hill
Bushfire Rock
Chiddarcooping
Eaglestone Rock
Elachbutting Hill
Geeraning
Graham Rock
The Humps
Jilakin Rock
Marshal Rock
Murray Rock
Sandford Rocks
Talgomine Rock 1
Talgomine Rock 2
Wave Rock
Yeerakine Rock
Kunzea baxteri
Kunzea preissiana
8.9 km NNE of Kununoppin
45.5 km E of Hyden
Chiddarcooping Nature Reserve
12.8 km ENE of Nungarin
Elachbutting Nature Reserve
Geeraning Nature Reserve
Graham Rock Nature Reserve
17 km NE of Hyden
~1 km W of Jilakin Lake
5.5 km SE of Bencubbin
31 km NE of Hyden
Sandford Rocks Nature Reserve
19.2 km W of Nungarin
20.5 km W of Nungarin
Wave Rock Reserve
8.7 km SE of Kondinin
Cape Arid National Park
Fitzgerald River National Park
31°02′
32°26′
30°54′
31°08′
30°35′
30°31′
32°27′
32°18′
32°39′
30°50′
32°18′
31°13′
31°12′
31°12′
32°26′
32°33′
33°53′
33°46′
Diversity and Distributions, 1–15, ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
29.51″S
32.67″S
21.02″S
07.88″S
36.12″S
31.56″S
36.00″S
56.02″S
24.67″S
33.56″S
59.82″S
54.76″S
48.45″S
59.21″S
36.18″S
37.53″S
59.30″S
36.15″S
117°57′
119°20′
118°39′
118°13′
118°36′
118°35′
119°03′
118°57′
118°19′
117°54′
119°09′
118°45′
118°18′
118°18′
118°53′
118°19′
123°18′
119°39′
Outcrop size (km2)
38.99″E
50.85″E
24.87″E
26.81″E
39.97″E
57.95″E
16.50″E
30.68″E
36.39″E
18.90″E
02.35″E
26.74″E
00.09″E
34.33″E
29.21″E
10.07″E
45.70″E
18.10″E
7.57
4.62
314.97
8.62
6.67
4.67
2.54
9.80
6.30
1.19
26.21
35.59
2.89
0.34
6.78
1.72
–
–
5
S.-L. Tapper et al.
Table 2 List of haplotypes identified through chloroplast DNA sequence analysis of K. pulchella, K. preissiana and K. baxteri.
Haplotypes were defined based on sequences from three intergenic spacer regions; psbD-trnT, psbA-trnH, trnS-trnG. GenBank Accession
Numbers are listed for each region and haplotype.
Haplotype
code
GenBank Accession Number
psbD-trnT
psbA-trnH
trnS-trnG
Population (No. individuals)
H1
H2
H3
H4
H5
H6
H7
H8
H9
H10
H11
H12
H13
H14
H15
H16
JX417102
JX417113
JX417115
JX417103
JX417104
JX417105
JX417108
JX417109
JX417107
JX417110
JX417111
JX417112
JX417114
JX417114
JX417117
JX417118
JX417092
JX417092
JX417092
JX417092
JX417093
JX417094
JX417096
JX417096
JX417095
JX417097
JX417097
JX417098
JX417092
JX417097
JX417101
JX417095
JX417122
JX417122
JX417122
JX417123
JX417124
JX417124
JX417126
KC894607
JX417125
JX417123
JX417123
JX417123
JX417123
JX417123
JX417124
JX417125
H17
H18
JX417119
JX417112
JX417095
JX417092
JX417125
JX417123
K.
K.
K.
K.
JX417121
JX417121
JX417121
JX417120
JX417087
JX417089
JX417090
JX417091
JX417130
JX417130
JX417130
JX417128
Billycatting Hill (6)
Marshal Rock (3)
Marshal Rock (1)
Bushfire Rock (6)
Chiddarcooping (5)
Chiddarcooping (1)
Elachbutting Hill (6)
Geeraning (6)
Eaglestone Rock (6)
Graham Rock (6)
The Humps (6)
Jilakin Rock (6)
Murray Rock (1)
Murray Rock (5)
Sandford Rocks (5)
Talgomine Rock 1 (5)
Talgomine Rock 2 (2)
Talgomine Rock 1 (1)
Wave Rock (6)
Yeerakine Rock (6)
K. preissiana (2)
K. preissiana (1)
K. preissiana (2)
K. baxteri (3)
preissiana1
preissiana2
preissiana3
baxteri1
haplotype network, with Elachbutting Hill and Geeraning
separated from the remaining outcrops on a long branch
(Fig. 1b). Only one haplotype was identified on each outcrop
except for Chiddarcooping, Marshal Rock, Murray Rock and
Talgomine Rock 1, where two haplotypes were identified
(Table 2). The TRMCA of the two major lineages was estimated to be 3.47 Myr (95% CI: 2.21–4.73 Myr).
Nuclear microsatellite dataset
Across the 11 microsatellite loci amplified in 384 K. pulchella
individuals, 2.6% of data were missing as a result of nonamplification at a locus or unscorable peaks. No single locus
had more that 4.7% missing data. The test for scoring accuracy produced an error rate of < 8% for all 11 loci. The total
number of alleles per locus ranged from six to 18. All loci
were polymorphic in all populations, and 12 populations
exhibited unique alleles.
The highest level of diversity was found at Sandford Rocks,
and the lowest diversity was detected at The Humps. The average number of alleles per locus in a population was 5.42
( 0.17) and ranged from 4.00 to 6.82. Observed heterozygosity (Ho) ranged from 0.43 to 0.66, averaging 0.58 ( 0.02)
(Table 4). Expected heterozygosity (He) ranged from 0.55 on
The Humps to 0.71 on Chiddarcooping and Sandford Rocks,
averaging 0.64 ( 0.01) (Table 4). The average inbreeding
6
coefficient (FIS) was 0.101 ( 0.02), ranging from 0.033 at
Bushfire Rock to 0.243 at Yeerakine Rock.
Tests for Hardy–Weinberg equilibrium (HWE) (P < 0.05)
following sequential Bonferroni correction revealed significant deviation in only 35 out of 176 locus-population combinations, with the majority showing heterozygote deficiency.
The Graham Rock population showed deviation from HWE
at six loci, three exhibited a heterozygote deficiency and
three exhibited an excess. The majority (83%) of the significant tests for deviation from HWE involved three loci
(Kun03, Kun09 and Kun13), and null alleles were also
detected at these three loci as a result of heterozygote deficiency; however, null alleles appeared to have only a small
effect on FST estimates as pairwise and single-locus estimates
were very similar with and without correction for null alleles
(data not shown). Global FST across all 16 populations was
0.130 (CI: 0.11–0.15) using original genotypes and 0.122 (CI:
0.11–0.14) when corrected using the method of Chapuis &
Estoup (2007). Significant linkage disequilibrium was
detected in 52 of 1056 tests after sequential Bonferroni
correction (P < 4.2 9 10 5). No loci pair showed significant
linkage disequilibrium across the majority of populations,
and most disequilibrium (92% of the significant locus combinations) was found within three populations, Bushfire
Rock (25%), Graham Rock (52%) and Billycatting Hill
(15%).
Diversity and Distributions, 1–15, ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
Isolation and persistence in Kunzea pulchella
Figure 2 Maximum-clade-credibility
chronogram obtained through Bayesian
phylogenetic analysis of three chloroplast
DNA intergenic spacer regions (psbDtrnT, psbA-trnH, trnS-trnG) and
calibrated using a known root age.
Branch lengths are scaled according to
time with nodes representing median age
estimates and numbers in brackets
describing 95% confidence intervals for
node ages. Posterior probabilities (> 0.8)/
Parsimony Bootstrap Support (> 80) are
shown below branches.
Table 3 Haplotype and nucleotide diversity and neutrality tests. Statistics calculated from chloroplast DNA sequence data from three
combined regions psbD-trnT, psbA-trnH, trnS-trnG of n number of samples. All estimates for Tajima’s D and Ramos-Onsins & Rozas R2
were non-significant (P > 0.05).
K. pulchella
Northern lineage
Southern lineage
n
Haplotype diversity
Nucleotide diversity
Tajima’s D
Ramos-Onsins & Rozas R2
89
12
77
0.868
0.545
0.880
0.0054
0.0005
0.0028
0.589
1.486
0.659
0.116
0.272
0.122
A bottleneck was detected in the Bushfire Rock population
through the Wilcoxon test for significant heterozygosity
excess (P < 0.05) and was consistent for both the stepwise
mutation and the infinite allele models. The Wave Rock population exhibited a shifted distribution of rare and common
alleles, indicating that the population has also undergone a
recent bottleneck.
Bayesian analysis of the complete data set with STRUCTURE identified three genetic clusters (K = 3) that was further supported by the similarity coefficient of the runs
(h′ = 0.956); analysis with removal of the three loci that
showed some deviation from HWE gave similar results. Two
clusters corresponded with populations found in the south
of the sampling range with some admixture in Yeerakine
Rock and The Humps (Fig. 3). The third cluster consistently
associated the populations to the north, but some admixture
with the other two clusters was evident in all but Eaglestone
Rock and the two most northern populations, Elachbutting
Hill and Geeraning, that were identified as a divergent lineage in the chloroplast phylogeny (Fig. 3). The principal coordinate analysis explained 48.82% of the variation using two
axes and 66.31% using three axes (Fig. 4). It showed the
northern and southern populations generally occupying sepa-
Diversity and Distributions, 1–15, ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
rate ordination space but no clear geographical pattern to
the population relationships. A neighbour-joining tree of
Chord genetic distance at the population level also revealed
no clear geographical pattern (see Fig. S2 in supporting
information).
All pairwise FST and Dest estimates among populations
were significant (P < 0.01). Both pairwise FST and Dest indicated the greatest differentiation was between Wave Rock
towards the south and the most northern sampled population at Geeraning (Table 5). The least differentiation was evident between The Humps and Jilakin Rock based on FST,
and between Eaglestone Rock and Talgomine Rock 1 based
on Dest (Table 5). Global FST (a measure of population subdivision) was 0.130 ( 0.012). There was a significant positive correlation between FST and geographic distance
(r = 0.398 and P = 0.0001) across all pairwise population
comparisons supporting an isolation by distance model. An
AMOVA for pairwise ΦST illustrated that most of the total
variance (80%) was explained by differences within populations while 20% of the variation was explained by differences
among populations.
Granite outcrop area was not significantly correlated (P >
0.05) with average number of alleles per locus (rs = 0.36),
7
S.-L. Tapper et al.
Figure 3 Genetic ancestry of 16
K. pulchella populations based on
analysis of 11 nuclear microsatellite loci
in STRUCTURE 2.3.3. Each thin column
represents an individual which is divided
into three shaded segments proportional
to the individual’s genome belonging to
each cluster. qA, dark grey; qB, light grey;
qC, white. Results are the optimal
alignment of 10 replicates.
(a)
(b)
Figure 4 Principal Coordinates Analysis of genetic distances between 16 populations of K. pulchella based on nuclear microsatellites
displayed across (a) two axes and (b) three axes.
observed heterozygosity (rs = 0.12), expected heterozygosity
(rs = 0.41) or inbreeding coefficient (rs = 0.24).
DISCUSSION
The phylogeographic patterns found in this study of K. pulchella are not indicative of dynamics involving contraction
to, and expansion from, particular granite outcrops. Rather,
K. pulchella exhibits a genetic signature consistent with a history of prolonged isolation and persistence on specific granite outcrops throughout Pleistocene climatic oscillations.
While there are caveats to the rule (Ibrahim et al., 1996;
Excoffier & Ray, 2008), species that have expanded from
refugia typically exhibit low diversity and a small number of
common widespread haplotypes as well as a geographic
direction to structure within genetically divergent lineages
(Hewitt, 2000, 2004; Byrne, 2008). Kunzea pulchella displays
a different genetic signature; haplotypes are highly divergent
and all but two are unique to single granite outcrops. Sampled
populations had only one or two chloroplast haplotypes and
8
the genetic distance between granite outcrops was not correlated
with geographic distance except for outcrops < 1 km apart.
Despite a widespread current distribution, the phylogeographic history of the populations of K. pulchella studied
here is consistent with that of previously studied rare granite
endemics, such as Eucalyptus caesia (Moran & Hopper, 1983;
Byrne & Hopper, 2008), Verticordia staminosa (Yates et al.,
2007), Eucalyptus crucis (Sampson et al., 1988) and the lizard
species Ctenophorus ornatus (Levy et al., 2012), which all
show a history of isolation on single or geographically close
granite outcrops and divergence driven by genetic drift. Our
results indicate that the apparent absence of particular outcrops acting as refugia within the granite network may be a
feature of granite species generally and not an idiosyncratic
feature of the rare species studied to date.
Isolation and persistence
Populations of K. pulchella show genetic signals of high
differentiation and low haplotype diversity that are indicative
Diversity and Distributions, 1–15, ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
Isolation and persistence in Kunzea pulchella
Table 4 Genetic diversity measures for all sampled populations of K. pulchella based on analysis of 11 microsatellite loci.
Population
n
A′
Billycatting Hill
Chiddarcooping
Eaglestone Rock
Elachbutting Hill
Geeraning
Marshal Rock
Sandford Rocks
Talgomine Rock 1
Talgomine Rock 2
Bushfire Rock
The Humps
Graham Rock
Jilakin Rock
Murray Rock
Wave Rock
Yeerakine Rock
Mean SE
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
4.73
6.36
5.91
5.91
5.91
5.91
6.82
5.91
4.91
4.46
4.00
5.91
5.18
5.64
4.46
4.64
5.42
Ho
0.72
0.68
0.60
0.84
0.80
0.60
0.41
0.67
0.58
0.39
0.43
0.71
0.54
0.77
0.46
0.61
0.17
0.48
0.60
0.60
0.62
0.61
0.64
0.65
0.64
0.60
0.66
0.52
0.53
0.54
0.57
0.51
0.43
0.58
He
0.09
0.05
0.06
0.07
0.08
0.06
0.09
0.05
0.06
0.08
0.07
0.07
0.08
0.07
0.09
0.08
0.02
0.58
0.71
0.69
0.63
0.67
0.64
0.71
0.69
0.61
0.65
0.55
0.65
0.63
0.66
0.56
0.56
0.64
FIS
0.05
0.02
0.04
0.06
0.05
0.04
0.03
0.03
0.05
0.03
0.07
0.06
0.05
0.04
0.05
0.06
0.01
0.192
0.164
0.120
0.032
0.119
0.009
0.101
0.067
0.033
0.033
0.031
0.164
0.123
0.145
0.104
0.243
0.101
0.11
0.06
0.07
0.05
0.08
0.06
0.11
0.06
0.04
0.12
0.09
0.08
0.10
0.07
0.11
0.09
0.02
n, sample size; A′, average number of alleles per loci; Ho, observed heterozygosity; He, expected heterozgosity; FIS, inbreeding cefficient.
of prolonged genetic isolation on granite outcrops as
opposed to contraction to specific granite outcrops and
recolonization of adjacent outcrops within the granite
network. A high level of divergence between populations was
evident in both chloroplast and nuclear genomes, with all
but two chloroplast haplotypes population-specific and a
high global population differentiation in the nuclear genome.
Aside from the divergence of two ancient lineages, there was
little evidence for consistent geographic patterns of genetic
structure among populations. At a local scale, there was
some similarity among some groups of populations, but
these were not consistent between the chloroplast and
nuclear data sets, as would be expected if similarity was a
result of recent expansion. Patterns of genetic differentiation
suggest that long-term isolation of populations on granite
outcrops has driven differing patterns of divergence.
In addition to high population differentiation, chloroplast
data sets revealed low diversity within K. pulchella populations with no more than two chloroplast haplotypes found
in any population. In contrast, nuclear analysis showed
diversity within populations, with notably higher diversity in
K. pulchella than other granite endemic species of the Myrtaceae family in Western Australia, E. caesia (Byrne & Hopper,
2008) and V. staminosa (Yates et al., 2007). Lower population diversity in E. caesia and V. staminosa may be a result
of their rarity and historically small effective population size.
Comparison of diversity with other non-granite endemic
Myrtaceae species analysed using microsatellites revealed
diversity in K. pulchella was the same as the common Calothamnus quadrifidus (Byrne et al., 2007) and lower than most
eucalypts (Ho = 0.56–0.91, data not shown), with only three
other species lower, the rare Eucalyptus curtisii (Ho = 0.47,
Smith et al., 2003) and Metrosideros boninensis (Ho = 0.37,
Kaneko et al., 2007), and the common Eugenia dysenterica
(Ho = 0.46, Zucchi et al., 2003).
Diversity and Distributions, 1–15, ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
Although higher than rare species, the nuclear diversity
of K. pulchella populations is generally lower than other
widespread Myrtaceae species, noting that these species are
mostly eucalypts and not granite endemics. Low diversity is
commonly a result of population bottlenecks, low levels of
gene flow or small population size (Amos & Harwood,
1998). While the low diversity in the chloroplast genome
may be a result of ancient bottlenecks, there was little evidence of more recent bottlenecks in the nuclear genome.
The low nuclear diversity in K. pulchella in comparison
with other widespread but more generalist species may be a
result of limited gene flow and small effective population
size. Kunzea pulchella is bird pollinated and some gene flow
via pollen dispersal may be expected given evidence for
extensive pollen dispersal in another bird pollinated species
in a fragmented landscape (Byrne et al., 2007). The admixture identified by STRUCTURE analysis among geographically close populations, such as Wave Rock and The
Humps, is likely due to pollen dispersal. Admixture among
geographically distant populations is more likely to be a
result of drift than gene flow by pollen dispersal. Genetic
distance between populations was independent of geographic distance beyond 1 km, and there was significant
differentiation among all populations, suggesting that gene
flow has been highly restricted. Low levels of gene flow
coupled with small effective population size may have
increased genetic drift, obscuring any signal of localized
common ancestry. Thus, isolation and population size are
likely to be the major force in maintenance of lower diversity in K. pulchella populations in relation to other Myrtaceae species.
Granite outcrops that are large in size are expected to
have greater water-harvesting capacity and habitat variety,
enabling them to support larger populations and therefore
be most suitable to act as Pleistocene refugia. However, in
9
10
Billycatting
Hill
Bushfire Rock
Chiddarcooping
Eaglestone
Rock
Elachbutting
Hill
Geeraning
Graham Rock
The Humps
Jilakin
Marshal Rock
Murray Rock
Sandford
Rocks
Talgomine
Rock 1
Talgomine
Rock 2
Wave Rock
Yeerakine
Rock
0.271
–
0.051
–
0.067
0.091
0.093
0.091
0.112
0.080
0.088
0.089
0.071
0.076
0.065
0.098
0.112
0.102
0.092
0.073
0.085
0.053
0.097
0.097
0.082
0.101
0.070
0.078
0.073
0.076
0.077
0.117
0.086
0.106
0.067
0.056
0.049
0.051
0.100
0.057
0.051
0.083
0.061
0.051
0.058
0.250
0.296
–
Chiddarcooping
Bushfire
Rock
Billycatting
Hill
0.111
0.092
0.052
0.034
0.053
0.109
0.070
0.074
0.090
0.067
0.063
0.068
0.376
0.206
–
0.291
Eaglestone
Rock
0.117
0.101
0.063
0.066
0.135
0.133
0.063
0.066
0.074
0.126
0.100
0.079
0.372
–
0.063
0.101
0.108
0.102
0.102
–
0.113
0.090
0.094
0.094
0.087
0.068
0.059
0.110
0.088
0.086
0.066
0.079
0.054
0.331
0.195
–
0.355
0.352
0.375
0.252
Graham
Rock
0.361
0.197
0.198
0.326
Geeraning
0.337
0.209
0.248
0.136
Elachbutting
Hill
0.085
0.063
0.078
0.052
0.353
0.172
–
0.040
0.087
0.049
0.053
0.311
0.291
0.213
0.269
0.251
The
Humps
0.102
0.066
0.085
0.069
0.350
0.296
0.106
–
0.092
0.052
0.057
0.289
0.313
0.176
0.277
0.313
Jilakin
0.116
0.124
0.107
0.060
0.361
0.328
0.315
0.319
–
0.074
0.073
0.213
0.327
0.339
0.360
0.200
Marshal
Rock
0.112
0.083
0.097
0.062
0.339
0.320
0.154
0.159
0.262
–
0.052
0.276
0.255
0.237
0.257
0.239
Murray
Rock
0.096
0.068
0.085
0.062
0.288
0.364
0.197
0.207
0.292
0.195
–
0.194
0.321
0.217
0.275
0.253
Sandford
Rocks
0.237
–
–
0.094
0.095
0.080
0.132
0.116
0.346
0.412
–
0.056
0.257
0.324
0.186
0.464
0.336
0.169
0.173
0.395
0.248
0.217
0.301
0.320
0.214
0.305
0.217
Yeerakine
Rock
–
0.476
0.172
0.256
0.305
0.370
0.366
0.343
0.366
0.364
0.386
0.391
0.324
Wave
Rock
0.203
0.282
0.257
0.274
0.375
0.342
0.331
0.187
0.346
0.192
0.166
0.213
Talgomine
Rock 2
0.263
0.253
0.185
0.251
0.219
0.237
0.269
0.241
0.248
0.199
0.111
0.252
Talgomine
Rock 1
Table 5 Genetic differentiation between population pairs of K. pulchella. Numbers below the diagonal describe pairwise FST while those above diagonal describe pairwise Dest. All values
were significant (P < 0.01).
S.-L. Tapper et al.
Diversity and Distributions, 1–15, ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
Isolation and persistence in Kunzea pulchella
this study, outcrop size and nuclear diversity were not
correlated, suggesting that larger outcrops have not
necessarily maintained larger effective population sizes than
smaller outcrops. Higher levels of diversity on particular
granite outcrops may be more closely correlated with other
environmental characteristics for example, habitat heterogeneity, or the presence of symbionts (Vellend & Geber,
2005).
It could be interpreted from the current widespread distribution of K. pulchella that there has been effective seed dispersal across the landscape within and between suitable
granite outcrops. The genetic signal of isolation revealed in
this study is not consistent with this hypothesis. As seen in
other granite endemics of the SWAFR (Sampson et al., 1988;
Yates et al., 2007; Byrne & Hopper, 2008; Levy et al., 2012),
K. pulchella shows no evidence of gene flow through seed
dispersal between outcrops. Therefore, the current distribution is unlikely to have arisen through recent expansion and
colonization of granite outcrops, but rather through persistence of populations that have remained isolated for
prolonged periods of time.
Divergent lineages
Analysis of chloroplast diversity revealed two divergent lineages in the sampled K. pulchella populations, with one
genetic lineage occurring at Elachbutting Hill and Geeraning,
the two northern most sampled populations, and the other
present in all remaining sampled populations. The estimated
divergence time between these lineages dates to the late Pliocene/early Pleistocene (3.47 Myr BP, 95% CI: 2.21–4.73 Myr)
coinciding with increasing aridity (Fujioka et al., 2005). The
separation of chloroplast diversity into two highly divergent
lineages is a common finding in genetic analyses of widespread
generalist species in the SWAFR (e.g. Byrne et al., 2002, 2003;
Byrne & Hines, 2004; Wheeler & Byrne, 2006; H. Nistelberger,
N. Gibson, B. Macdonald, S.-L. Tapper & M. Byrne, unpublished data), generally reflecting a correlation with major aridification in the mid-Pleistocene resulting from change in
oscillation patterns during Milankovich cycles. The divergence
among populations based on molecular dating in this study of
K. pulchella indicates a similar correlation with major aridification in the mid-Pleistocene, but the divergence of northern
and southern lineages in K. pulchella appears much older than
that in other species, indicating influence of events in the late
Pliocene/early Pleistocene in this species that is both widespread and confined to a specific habitat with a patchy distribution. Pliocene divergence among lineages has also been
observed in invertebrate short range endemics confined to
specific habitats in the more southern areas of the SWAFR
(Cooper et al., 2011; Rix & Harvey, 2012).
Refugia for granite endemics
After extensive review of the literature, Keppel et al. (2012)
defined refugia as habitats that components of biodiversity
Diversity and Distributions, 1–15, ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
retreat to, persist in and can potentially expand from under
changing environmental conditions. This definition is easily
understood in the context of species that occur across continuous environmental gradients where responses to changing climatic conditions drive contraction and expansion
dynamics in species’ ranges. It raises difficulties when evaluating responses in species endemic to patchily distributed
topographically complex environments, such as granite outcrops, where persistence may be facilitated through localized
contraction and expansion from in situ refugia.
We hypothesized that, for a widespread species endemic to
granite outcrops, we might see the classic signature of contraction and expansion indicating a refugial function involving recent range dynamics for particular outcrops. Our study
of K. pulchella has not revealed such a signal. Rather, it has
shown evidence for prolonged persistence and isolation of
the species on individual outcrops across the granite network. There are two plausible explanations, which may act
in concert, for persistence through climatic change. First,
wide physiological tolerances and specialist traits for the
granite environment made the species resistant to fluctuating
arid and more mesic climatic conditions. Second, during
periods of unfavourable conditions the species contracted to
favourable habitats (localized refugia) within outcrops with
subsequent expansion of the population when conditions
became more favourable. These explanations require further
investigation in K. pulchella. Our study did not sample at the
subpopulation scale needed to detect a genetic signature of
expansion and contraction within outcrops, but the low levels of both haplotype and allelic diversity suggest K. pulchella
populations have been through substantial, but not recent,
bottlenecks. Moreover, recent work by Schut et al. (2014)
provides further evidence for potential change in population
distribution within outcrops under changing climate conditions. They found a very strong relationship between annual
precipitation, soil depth and vegetation structure on granite
rocks across a rainfall gradient, which, when modelled and
mapped onto granite outcrops, showed contraction of more
mesic vegetation types to favourable localized refugia under
more arid conditions.
Whereas Pleistocene climatic oscillations would be expected
to have driven some change in the species’ range during arid
periods, it appears that K. pulchella persisted in small populations, with possible localized contraction and expansion within
outcrops, rather than local extirpation and recolonization of
outcrops across the granite network. This isolation on multiple
outcrops through a period of considerable climatic turbulence
indicates that most outcrops provide sufficient heterogeneity
and refugial opportunities within the outcrop to facilitate persistence of populations. In this sense, all outcrops may be
important localized refugia for impending anthropogenic climate change. Our study also illustrates the evolutionary potential of populations on individual outcrops in both rare and
common species and accentuates the value of these environments for biodiversity conservation planning in a changing
climate.
11
S.-L. Tapper et al.
ACKNOWLEDGEMENTS
This study was funded by Australian Research Council Linkage
grant, LP0990914. We thank Lee Bornman and Douglas Wardell-Johnson for field assistance, and Andrew Thornhill for
providing divergence dates from his unpublished dataset and
guidance in the use of BEAST software for dating the phylogeny.
REFERENCES
Amos, W. & Harwood, J. (1998) Factors affecting levels of
genetic diversity in natural populations. Philosophical
Transactions of the Royal Society of London Series B, Biological Sciences, 353, 177–186.
Anthony, N.M., Johnson-Bawe, M., Jeffery, K., Clifford, S.L.,
Abernethy, K.A., Tutin, C.E., Lahm, S.A., White, L.J.T.,
Utley, J.F., Wickings, E.J. & Bruford, M.W. (2007) The role
of Pleistocene refugia and rivers in shaping gorilla genetic
diversity in central Africa. Proceedings of the National Academy of Sciences USA, 104, 20432–20436.
Ashcroft, M.B. (2010) Identifying refugia from climate
change. Journal of Biogeography, 37, 1407–1413.
Ashcroft, M.B. & Gollan, J.R. (2013) Moisture, thermal inertia, and the spatial distributions of near-surface soil and air
temperatures: understanding factors that promote microrefugia. Agricultural and Forest Meteorology, 176, 77–89.
Ashcroft, M.B., Gollan, J.R., Warton, D.I. & Ramp, D. (2012)
A novel approach to quantify and locate potential microrefugia using topoclimate, climate stability, and isolation from
the matrix. Global Change Biology, 18, 1866–1879.
Bandelt, H.-J., Forster, P. & R€
ohl, A. (1999) Median-joining
networks for inferring intraspecific phylogenies. Molecular
Biology and Evolution, 16, 37–48.
Bates, B.C., Hope, P., Ryan, B., Smith, I. & Charles, S.
(2008) Key findings of the Indian Ocean climate initiative
and their impact on policy development in Australia. Climatic Change, 89, 339–354.
Beheregaray, L.B. (2008) Twenty years of phylogeography:
the state of the field and the challenges for the Southern
Hemisphere. Molecular Ecology, 17, 3754–3774.
Byrne, M. (2008) Evidence for multiple refugia at different
time scales during the Pleistocene climatic oscillations in
southern Australia inferred from phylogeography. Quaternary Science Reviews, 27, 2576–2585.
Byrne, M. & Hankinson, M. (2012) Testing the variability of
chloroplast sequences for plant phylogeography. Australian
Journal of Botany, 60, 569–574.
Byrne, M. & Hines, B. (2004) Phylogeographical analysis of
cpDNA variation in Eucalyptus loxophleba (Myrtaceae).
Australian Journal of Botany, 52, 459–470.
Byrne, M. & Hopper, S.D. (2008) Granite outcrops as
ancient islands in old landscapes: evidence from the phylogeography and population genetics of Eucalyptus caesia
(Myrtaceae) in Western Australia. Biological Journal of the
Linnaean Society, 93, 177–188.
12
Byrne, M., Tischler, G., McComb, J. & Coates, D. (2001)
Phylogenetic relationship between two rare Acacia species
and their wide spread relatives. Conservation Genetics, 2,
157–166.
Byrne, M., Macdonald, B. & Coates, D.J. (2002) Phylogeographical patterns in chloroplast DNA variation within
Acacia accuminata (Leguminosae: Mimosoideae) complex
in Western Australia. Journal of Evolutionary Biology, 15,
576–587.
Byrne, M., Macdonald, B. & Brand, J. (2003) Phylogeography
and divergence in the chloroplast genome of Western Australian Sandalwood (Santalum spicatum). Heredity, 91,
389–395.
Byrne, M., Elliott, C.P., Yates, C. & Coates, D.J. (2007)
Extensive pollen dispersal in a bird-pollinated shrub, Calothamnus quadrifidus, in a fragmented landscape. Molecular
Ecology, 16, 1303–1314.
Carvalli-Sforza, L.L. & Edwards, A.W.F. (1967) Phylogenetic
analysis: models and estimation procedures. American Journal of Human Genetics, 19, 233–257.
Chapuis, M.P. & Estoup, A. (2007) Microsatellite null alleles
and estimation of population differentiation. Molecular
Biology and Evolution, 24, 621–631.
Cooper, S.J.B., Harvey, M.S., Saint, K.M. & Main, B.Y.
(2011) Deep phylogeographic stucturing of populations of
the trapdoor spider Morggridgea tingle (Migidae) from
southwestern Autralia: evidence for long-term refugia
within refugia. Molecular Ecology, 20, 3219–3236.
Drummond, A.J. & Rambaut, A. (2007) BEAST: bayesian
evolutionary analysis by sampling trees. BMC Evolutionary
Biology, 7, 214.
Drummond, A.J., Nicholls, G.K., Rodrigo, A.G. & Solomon,
W. (2002) Estimating mutation parameters, population
history and genealogy simultaneously from temporally
spaced sequence data. Genetics, 161, 1307–1320.
Drummond, A.J., Ho, S.Y.W., Phillips, M.J. & Rambaut, A.
(2006) Relaxed phylogenetics and dating with confidence.
PLoS Biology, 4, 699–710.
Earl, D.A. & vonHolt, B.M. (2012) STRUCTURE HARVESTER: a website and program for visualizing STRUCTURE output and implementing the Evanno method.
Conservation Genetics Resources, 4, 359–361.
ESRI (Environmental Systems Resource Institute) (2009)
ArcMap 9.2. ESRI, Redlands, CA.
Excoffier, L. & Ray, N. (2008) Surfing during population
expansions promotes genetic revolutions and structuration.
Trends in Ecology and Evolution, 23, 347–351.
Excoffier, L.G., Laval, L.G. & Schneider, S. (2005) Arlequin ver
3.0: an integrated software package for population genetics
data analysis. Evolutionary Bioinformatics Online, 1, 47–50.
Excoffier, L., Foll, M. & Petit, R. (2009) Genetic consequences of range expansions. Annual Review of Ecology,
Evolution and Systematics, 40, 481–501.
Fay, M.F., Lledo, M.D., Richardson, J.E., Rye, B.L. &
Hopper, S.D. (2001) Molecular data confirm the affinities
Diversity and Distributions, 1–15, ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
Isolation and persistence in Kunzea pulchella
of the south-west Australian endemic Granitites with
Alphitonia (Rhamnaceae). Kew Bulletin, 56, 669–675.
Felsenstein, J. (1989) PHYLIP – Phylogeny inference package
(Version 3.2). Cladistics: The International Journal of the
Willi Hennig Society, 5, 164–166.
Fujioka, T., Chappell, J., Honda, M., Yatsevich, I., Fifield, K.
& Fabel, D. (2005) Global cooling initiated in stoney
deserts in central Australia 2-4 Ma, dated by cosmogenic
21
Ne-10Be. Geology, 12, 993–996.
Garrick, R.C. (2011) Montane refuges and topographic complexity generate and maintain invertebrate biodiversity:
recurring themes across space and time. Journal of Insect
Conservation, 15, 469–478.
Guindon, S. & Gascuel, O. (2003) A simple, fast, and accurate algorithm to estimate large phylogenies by maximum
likelihood. Systematic Biology, 52, 696–704.
Hewitt, G.M. (2000) The genetic legacy of the Quaternary
ice ages. Nature, 405, 907–913.
Hewitt, G.M. (2004) Genetic consequences of climatic oscillations in the Quaternary. Philosophical Transactions of the
Royal Society of London Series B, Biological Sciences, 359,
183–195.
Holm, S. (1979) A simple sequential rejective multiple test
procedure. Scandinavian Journal of Statistics, 6, 65–70.
Hopper, S.D. & Gioia, P. (2004) The southwest Australian
floristic region: evolution and conservation of a global hot
spot of biodiversity. Annual Review of Ecology, Evolution
and Systematics, 35, 623–650.
Hopper, S.D., Brown, A.P. & Marchant, N.G. (1997) Plants
of Western Australian granite outcrops. Journal of the Royal
Society of Western Australia, 80, 141–158.
Hsiao, C., Jacobs, S.W.L., Barker, N.P. & Chatterton, N.J.
(1998) A molecular phylogeny of the subfamily Arundinoideae (Poaceae) based on sequences of rDNA. Australian
Systematic Botany, 11, 41–52.
Ibrahim, K.M., Nichols, R.A. & Hewitt, G.M. (1996) Spatial
patterns of genetic variation generated by different
forms of dispersal during range expansion. Heredity, 77,
282–291.
Jakobsson, M. & Rosenberg, N.A. (2007) CLUMPP: a cluster
matching and permutation program for dealing with label
switching and multimodality in analysis of population
structure. Bioinformatics, 23, 1801–1806.
Johnson, P.C.D. & Haydon, D.T. (2007) Software for quantifying and simulating microsatellite genotyping error. Bioinformatics and Biology Insights, 1, 71–75.
Kaneko, S., Isagi, Y. & Nobushima, F. (2007) Development
of microsatellite markers for Metrosideros boninensis
(Myrtaceae), an endangered endemic plant species from
the Bonin Islands, Japan. Conservation Genetics, 8, 753–
755.
Keppel, G., Van Niel, K., Wardell-Johnson, G., Yates, C.J.,
Byrne, M., Mucina, L., Schut, A.G.T., Hopper, S.D. &
Franklin, S.E. (2012) Refugia: identifying and understanding safe havens for biodiversity under climate change.
Global Ecology and Biogeography, 21, 393–404.
Diversity and Distributions, 1–15, ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
Laing, I.A.F. & Hauck, E.J. (1997) Water harvesting from
granite outcrops in Western Australia. Journal of the Royal
Society of Western Australia, 80, 181–184.
de Lange, P.J., Smissen, R.D., Wagstaff, S.J., Keeling, D.J.,
Murray, B.G. & Toelken, H.R. (2010) A molecular phylogeny and infrageneric classification for Kunzea (Myrtaceae)
inferred from rDNA ITS and ETS sequences. Australian
Systematic Botany, 23, 309–319.
Levy, E., Kennington, W.J., Tomkins, J.L. & LeBas, N.R.
(2012) Phylogeography and population genetic structure of
the Ornate Dragon Lizard, Ctenophorus ornatus. PLoS
ONE, 7, e47351.
Librado, P. & Rozas, J. (2009) DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics, 25, 1451–1452.
Maddison, W.P. & Maddison, D.R. (2007) Mesquite: a modular system for evolutionary analysis. Version 2.75 http://
mesquiteproject.org.
Moran, G.F. & Hopper, S.D. (1983) Genetic diversity and
the insular population structure of the rare granite rock
species, Eucalyptus caesia Benth. Australian Journal of Botany, 31, 161–172.
M€
uller, K. (2005) SeqState – primer design and sequence statistics for phylogenetic DNA data sets. Applied Bioinformatics, 4, 65–69.
Myres, N., Mittermeier, R.A., Mittermeier, C.G., da Fonesca,
G.A.B. & Kent, J. (2000) Biodiversity hotspots for conservation priorities. Nature, 403, 853–858.
Peakall, R. & Smouse, P.E. (2006) GENALEX 6: genetic
analysis in Excel. Population genetic software for teaching
and research. Molecular Ecology Notes, 6, 288–295.
Piry, S., Luikart, G. & Cornuet, J.-M. (1999) BOTTLENECK:
a computer program for detecting recent reductions in the
effective population size using allele frequency data. Journal
of Heredity, 90, 502–503.
Posada, D. (2008) jModelTest: phylogenetic model averaging.
Molecular Biology and Evolution, 25, 1253–1256.
Pritchard, J.K., Stephens, M. & Donnelly, P. (2000) Inference
of population structure using multilocus genotype data.
Genetics, 155, 945–959.
Rambaut, A. & Drummond, A.J. (2007) Tracer v1.4, Available at: http://beast.bio.ed.ac.uk/Tracer.
Ramos-Onsins, S.E. & Rozas, J. (2002) Statistical properties
of new neutrality tests against population growth. Molecular Biology and Evolution, 19, 2092–2100.
Raymond, M. & Rousset, F. (1995) GENEPOP (version 1.2):
population genetics software for exact tests and ecumenicism. Journal Heredity, 86, 248–249.
Rix, M.G. & Harvey, M.S. (2012) Phylogeny and historical
biogeography of ancient assassin spiders (Araneae: Archaeidae) in the Australian mesic zone: evidence for Miocene
speciation within Tertiary refugia. Molecular Phylogenetics
and Evolution, 62, 375–396.
Rosenberg, N.A. (2004) DISTRUCT: a program for the
graphical display of population structure. Molecular Ecology
Notes, 4, 137–138.
13
S.-L. Tapper et al.
Saarela, J.M., Rai, H.S., Doyle, J.A., Endress, P.K., Mathews, S., Marchant, A.D., Briggs, B.G. & Graham, S.W.
(2007) Hydatellaceae identified as a new branch near the
base of the angiosperm phylogenetic tree. Nature, 446,
312–315.
Sampson, J.F., Hopper, S.D. & James, S.H. (1988) Genetic
diversity and the conservation of Eucalyptus crucis Maiden.
Australian Journal of Botany, 36, 447–460.
Sch€
onswetter, P., Tribsch, A., Barfuss, M. & Niklfeld, H.
(2002) Several Pleistocene refugia detected in the high
alpine plant Phyteuma globulariifolium Sternb. & Hoppe
(Campanulaceae) in the European Alps. Molecular Ecology,
11, 2637–2647.
Schut, A.G.T., Wardell-Johnson, G.W., Keppel, G., Baran,
I., Franklin, S.E., Hopper, S.D., van Niel, K., Mucina, L.
& Byrne, M. (2014) Rapid characterization of vegetation
structure to predict refugia and climate change impacts
across a global biodiversity hotspot. PLoS ONE, 9,
e82778.
Simmons, M. & Ochoterena, H. (2000) Gaps as characters in
sequence-based phylogenetic analyses. Systematic Biology,
49, 369–381.
Smith, S., Hughes, J. & Wardell-Johnson, G. (2003) High
population differentiation and extensive clonality in a rare
mallee eucalypt: Eucalyptus curtisii. Conservation Genetics,
4, 289–300.
Soltis, D.E., Gitzendanner, M.A., Strenge, D.D. & Soltis, P.S.
(1997) Chloroplast DNA intraspecific phylogeography of
plants from the Pacific Northwest of North America. Plant
Systematics and Evolution, 206, 353–373.
Statsoft (2001) Statistica. A data analysis software system,
Release 6.
Swofford, D.L. (2003) PAUP*: phylogenetic analysis using
parsimony. Version 4.0. Sinauer Association, Sunderland,
MA.
Szarzynski, J. (2000) Xeric islands: environmental conditions
on inselbergs. Inselbergs: biotic diversity of isolated rock outcrops in tropical and temperate regions, Vol. 146, (ed. by S.
Porembski and W. Barthlott), pp. 37–48. Springer Verlag,
Berlin, DE.
Taberlet, P., Fumagalli, L., Wust-Saucy, A.-G. & Cosson,
J.-F. (1998) Comparative phylogeography and postglacial
colonization routes in Europe. Molecular Ecology, 7, 453–464.
Tajima, F. (1989) Statistical method for testing the neutral
mutation hypothesis by DNA polymorphism. Genetics, 123,
585–595.
Tamura, K., Dudley, J., Nei, M. & Kumar, S. (2007) MEGA4:
molecular evolutionary genetics analysis (MEGA) software
version 4.0. Molecular Biology and Evolution, 24, 1596–1599.
Tapper, S.-L., Byrne, M. & Gardner, M. (2013) Characterisation of microsatellite markers for the granite endemic Kunzea pulchella (Lindl.) A.S.George (Myrtaceae) identified
using next generation sequencing. Conservation Genetics
Resources, 5, 529–531.
14
Thornhill, A.H., Popple, L.W., Carter, R.J., Ho, S.Y.W. &
Crisp, M.D. (2012) Are pollen fossils useful for calibrating
relaxed molecular clock dating of phylogenies? A comparative study using Myrtaceae. Molecular Phylogenetics and
Evolution, 63, 15–27.
Twidale, C.R. & Bourne, J.A. (1998) Origin and age of bornhardts, southwest Western Australia. Australian Journal of
Earth Sciences, 45, 903–914.
Vellend, M. & Geber, M.A. (2005) Connections between
species diversity and genetic diversity. Ecology Letters, 8,
767–781.
Weir, B.S. & Cockerham, C.C. (1984) Estimating F-statistics
for the analysis of population structure. Evolution, 38,
1358–1370.
Wheeler, M.A. & Byrne, M. (2006) Congruence between phylogeographic patterns in cpDNA variation in Eucalyptus
marginata (Myrtaceae) and geomorphology of the Darling
Plateau, south-west of Western Australia. Australian Journal of Botany, 54, 17–26.
Whitlock, B.A., Hale, A.M. & Groff, P.A. (2010) Intraspecific
inversions pose a challenge for the trnH-psbA plant DNA
barcode. PLoS ONE, 7, e11533.
Yates, C., Ladd, P., Coates, D. & McArthur, S. (2007)
Hierarchies of cause: understanding rarity in an endemic
shrub Verticordia staminosa (Myrtaceae) with a highly
restricted distribution. Australian Journal of Botany, 55,
194–205.
Yates, C., Elith, J., Latimer, A.M., Le Maitre, D., Midgley,
G.F., Schurr, F.M. & West, A.G. (2010) Projecting climate
change impacts on species distribution in megadiverse
South African Cape and Southwest Australian Floristic
Regions: opportunities and challenges. Austral Ecology, 35,
374–391.
York Main, B. (1997) Granite outcrops: a collective ecosystem. Journal of the Royal Society of Western Australia, 80,
113–122.
Zucchi, M.I., Brondani, R.P.V., Pinheiro, J.B., Chaves, L.J.,
Coelho, A.S.G. & Vencovsky, R. (2003) Genetic structure
and gene flow in Eugenia dysenterica DC in the Brazilian
Cerrado utilizing SSR markers. Genetics and Molecular Biology, 26, 449–457.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the
online version of this article:
Figure S1 Bayesian phylogenetic analysis of three chloroplast
DNA intergenic spacer regions (a) psbD-trnT, (b) psbA-trnH,
(c) trnS-trnG.
Figure S2 Neighbour-joining tree of Chord genetic distance
between populations of Kunzea pulchella obtained from
analysis of 11 nuclear microsatellite loci.
Diversity and Distributions, 1–15, ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
Isolation and persistence in Kunzea pulchella
BIOSKETCH
The authors are engaged in a multidisciplinary study on the role
of granite outcrops as refugia under future climate change.
the study, G.K. and G.W.J. collected samples, S.-L.T. collected
the data, S.-L.T. and M.B. analysed the data and led the writing.
Editor: Jeremy Austin
Author contributions: All authors conceived the broad ideas
and contributed to the writing, M.B., C.Y. and G.W.J. designed
Diversity and Distributions, 1–15, ª 2014 State of Western Australia.
Diversity and Distributions ª 2014 John Wiley & Sons Ltd
15