Content uploaded by Pieter B. Pelser
Author content
All content in this area was uploaded by Pieter B. Pelser on Sep 21, 2017
Content may be subject to copyright.
BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research
libraries, and research funders in the common goal of maximizing access to critical research.
Genetic Diversity and Structure in the Philippine Rafflesia lagascae Complex
(Rafflesiaceae) Inform Its Taxonomic Delimitation and Conservation
Author(s): Pieter B. Pelser, Daniel L. Nickrent, Chrissen E. C. Gemmill, and Julie F. Barcelona
Source: Systematic Botany, 42(3):543-553.
Published By: The American Society of Plant Taxonomists
URL: http://www.bioone.org/doi/full/10.1600/036364417X696186
BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and
environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published
by nonprofit societies, associations, museums, institutions, and presses.
Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of
BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.
Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries
or rights and permissions requests should be directed to the individual publisher as copyright holder.
Systematic Botany (2017), 42(3): pp. 543–553
© Copyright 2017 by the American Society of Plant Taxonomists
DOI 10.1600/036364417X696186
Date of publication August 25, 2017
Genetic Diversity and Structure in the Philippine Rafflesia lagascae Complex
(Rafflesiaceae) inform its Taxonomic Delimitation and Conservation
Pieter B. Pelser,
1,4
Daniel L. Nickrent,
2
Chrissen E. C. Gemmill,
3
and Julie F. Barcelona
1
1
School of Biological Sciences, University of Canterbury, Private Bag 4800, Christchurch 8140, New Zealand
2
Department of Plant Biology, Southern Illinois University Carbondale, 1125 Lincoln Drive, Carbondale,
Illinois 62901-6509, U. S. A.
3
School of Science, University of Waikato, Private Bag 3105, Hamilton 3240, New Zealand
4
Author for correspondence (pieter.pelser@canterbury.ac.nz)
Communicating Editor: Sasa Stefanovic
Abstract—Rafflesia lagascae is a rare endo-holoparasitic species with a disjunct distribution on Luzon Island. It is morphologically very similar to R.
manillana from nearby Samar. This study aims to contribute to the taxonomy and conservation of R. lagascae and R. manillana (i.e. the R. lagascae
complex) by resolving their patterns of genetic diversity and structure. The results of analyses of a microsatellite data set indicate that despite their
frequently extremely small sizes and geographic isolation, Rafflesia populations display moderate genetic diversity and do not show evidence of
pronounced inbreeding. Most populations appear to have limited gene flow among them. Patterns of genetic diversity of staminate and pistillate
Rafflesia flowers growing on the same Tetrastigma host plants indicate that the R. lagascae complex is monoecious and that host plants are regularly
infected by multiple Rafflesia plants. PCoA and Bayesian cluster analyses show that the complex is composed of three genetically isolated taxa. One of
these constitutes R. manillana, supporting the morphology-based hypothesis that it is taxonomically distinct from R. lagascae. The second taxon in this
complex is composed of a morphologically cryptic R. lagascae population from Mt. Labo, which is genetically distinct from all remaining R. lagascae
populations that were studied and that form the third taxon. We recommend that these three taxa are managed as different conservation entities.
Keywords—Microsatellites, parasitic plants, rare species, species complex, taxonomy.
The host species of parasitic plants form essential compo-
nents of their environment (Sandner and Matthies 2017). Their
distribution patterns constitute the maximum potential dis-
tribution area of the parasitic plant species that depend on
them for their existence (Costea and Stefanovi´c 2009) in the
same way that specialists of a rare habitat are confined to areas
where this habitat is found (e.g. Medrano and Herrera 2008;
Williams et al. 2016) and the distribution of elevation spe-
cialists is constrained by topography (e.g. Laurance et al. 2011).
When the prerequisite conditions on which specialist plant
species depend are rare, so are these species. Consequently,
some specialist plants have small population sizes and are
restricted to a single small area (e.g. de Lange 2003) or have a
disjunct distribution pattern (e.g. Medrano and Herrera 2008).
This makes these species vulnerable to inbreeding and genetic
drift as a result of a loss of genetic connectivity between
populations due to habitat fragmentation and degradation
(e.g. Ellstrand and Elam 1993; Schmidt and Jensen 2000;
Stanton et al. 2009; Williams et al. 2016). This might ultimately
result in a loss of genetic diversity and thereby increase the
probability that small and isolated populations go extinct
(Schmidt and Jensen 2000). Indeed, many parasitic plant
species are rare and listed as threatened or endangered
(Marvier and Smith 1997; Sandner and Matthies 2017). Par-
asitic species that are very host specific are especially at risk,
because they are particularly sensitive to changes in host
availability and typically have small distribution areas (Costea
and Stefanovi´c 2009).
Rafflesia R.Br. (Rafflesiaceae) is a Southeast Asian plant
genus of ca. 36 species (Hidayati and Walck 2016), all of which
are obligate endo-holoparasites of Tetrastigma (Miq.) Planch.
(Vitaceae; Meijer 1997; Nais 2001; Pelser et al. 2016). The
Philippines is one of the centers of Rafflesia diversity (Barcelona
et al. 2009) and is home to 13 currently recognized species
(Pelser et al. 2011 onwards; Barcelona et al. 2014; Galindon
et al. 2016), which are all endemic to the country. Rafflesia
lagascae Blanco (1845) is arguably the most common Rafflesia
species in the Philippines and has the largest distributional
area (Barcelona et al. 2009). However, although it can be found
throughout the island of Luzon (Fig. 1), it has a highly disjunct
distribution and is only known from 15 small areas. In some of
these areas, it is only known from one or a few individual host
plants belonging to three species of Tetrastigma (Pelser et al.
2016). Because of the small number of R. lagascae populations
and the extremely small size of many of them, as well as its
narrow host range, this species might be particularly prone to
genetic diversity loss following habitat fragmentation and
degradation. This is potentially a significant conservation
factor, because the Philippine tropical rainforest, to which this
species is confined, has been reduced to an estimated 3–6% of
its original cover (Mittermeier et al. 1998; Ong et al. 2002).
Moreover, remaining tropical forest habitat has become highly
degraded and fragmented as a result of human activities such
as logging, mining, and land conversion for agriculture
(Barcelona et al. 2009). To better understand the conservation
needs of R. lagascae and to improve its management, in-
formation about its patterns of genetic diversity and genetic
structure are needed. These patterns would, for instance,
enable the identification of the populations that have the
lowest genetic diversity and least genetic connectivity with
other populations and that are therefore most at risk of in-
breeding and genetic drift.
At present, nothing is known about patterns of genetic
diversity and structure in any Rafflesia species. Previous mo-
lecular genetic studies of Rafflesia have focused on horizontal
gene transfer from their hosts (Davis and Wurdack 2004;
Nickrent et al. 2004; Xi et al. 2012, 2013), their evolutionary
relationships and diversification (Barkman et al. 2004, 2008;
Nickrent et al. 2004; Davis et al. 2007; Bendiksby et al. 2010),
the molecular regulation of flower development (Nikolov et al.
2013; Ramamoorthy et al. 2013), and the possible loss of their
chloroplast genome (Molina et al. 2014), but none have
addressed research questions at the infraspecific level. Because
of the absence of molecular genetic data for Rafflesia at the
levels of individuals and populations, not only are their pat-
terns of genetic diversity and genetic structure unknown, but
543
other aspects of their biology also remain to be discovered. For
example, most Rafflesia species, including R. lagascae, have
unisexual flowers, but it is not known if they are dioecious or
monoecious (Nais 2001). It is difficult to determine plant
sexuality without genetic data, because of their endoparasitic
nature (Balete et al. 2010). In other words, it is not possible to
determine if staminate and pistillate flowers that emerge from
the same host belong to the same genetically identical indi-
vidual endophyte. Because self-incompatible plants might
be more vulnerable to a loss of genetic diversity resulting
from habitat fragmentation (Stanton et al. 2009), knowing if
R. lagascae is monoecious or dioecious is relevant for its con-
servation management. Similarly, it is presently unclear if
Tetrastigma vines can be host to more than one Rafflesia indi-
vidual. This basic information about Rafflesia will enable more
accurate estimates of population sizes and thereby contribute
to their conservation management.
Population-level genetic data are also needed to inform the
taxonomic delimitation of R. lagascae. This species is morpho-
logically very similar to R. manillana Teschem. (Teschemacher
1844; Fig. 2) and both taxa were considered synonymous
until relatively recently (Madulid et al. 2008; Pelser et al.
2013). Whereas R. lagascae is endemic to Luzon, R. manillana
has only been reported from Basey municipality on Samar
Island. The two allopatric species are of similar sizes (11–24
vs. 11–16 cm diam; Madulid and Agoo 2008; Barcelona et al.
2009) and primarily differ in the color of the floral di-
aphragm and the relative size of the diaphragm aperture
(Pelser et al. 2013; Fig. 2), but it is presently unclear if these
morphological differences are indicative of species level
genetic differences. Rafflesia manillana is currently only known
from a single population and is considered critically endan-
gered by Madulid and Agoo (2008). Therefore, if despite their
morphological differences both taxa are genetically indistinct,
the conservation status of R. manillana would need to be
revised.
In this study, we provide the first examination of the pat-
terns of genetic diversity and genetic structure of R. lagascae
and R. manillana (from here on collectively referred to as the R.
lagascae complex). Specifically, we aim 1) to identify pop-
ulations that have low genetic diversity, 2) to determine if taxa
in the R. lagascae complex are monoecious or dioecious and if
Tetrastigma plants can host more than one Rafflesia individual,
and 3) to resolve the taxonomic delimitation of the R. lagascae
complex.
Materials and Methods
Specimen Sampling—Tissue samples on silica gel and voucher speci-
mens preserved in 70% denatured alcohol were collected from 13 of the 16
areas from which members of the R. lagascae complex have been reported.
For the three unsampled areas (Adams, Ilocos Norte Prov.; Barlig,
Mountain Prov.; Calanasan, Apayao Prov.; Fig. 1), collecting permits or
samples could not be secured in time for this study. Tissue was taken from
at least one flower or flower bud from each infected host plant that was
encountered, with a maximum of six samples taken per site (i.e. one or
more infected host plants in close vicinity of each other and geographically
isolated from other such sites). Nearby host plants were considered distinct
individuals if they were unambiguously spatially separated from each
Fig. 1. Distribution of all currently known populations of R. lagascae
and R. manillana. Black dots indicate populations that were sampled for this
study. Grey dots are populations that could not be sampled. The dotted line
indicates the approximate position of the Guinyangan fault (Besana and
Ando 2005) and marks the area where the Bicol Peninsula connects with
mainland Luzon. Faint dotted lines indicate provincial boundaries.
Fig. 2. Flowers of the three Rafflesia taxa identified in this study. A. R. lagascae s. s. B. Mt. Labo R. lagascae.C.R. manillana.
SYSTEMATIC BOTANY [Volume 42544
other. When possible, damage to Rafflesia plants was minimized by
removing a small amount of tissue from the diaphragm or the perigone
lobes of the flowers, which is not expected to affect their reproductive
function. When flower buds were encountered, we sampled tissue from the
perigone lobes or disk of young flower buds while carefully avoiding
sampling from the cupula area where the tissues of the parasite and host
intertwine. Rafflesia species have a high bud mortality rate (50–90%; Nais
2001 and references therein) and we therefore considered sampling from
young buds as having a lower negative impact on the reproductive success
of plants than sampling from mature buds that have escaped the causes of
early mortality (e.g. nutrient limitation, predation by animals, infection by
wasps; Nais 2001) and are likely to survive until the flowering stage.
Voucher specimens (Appendix 1) are deposited at CAHUP, CANU, PNH,
and SIU (acronyms follow Thiers 2017).
Development of Microsatellite Primers—Contigs from a previously
published Illumina sequence library of R. lagascae obtained from a sample
of this species from Bolos Point (see Molina et al. 2014 for assembly
methodology; NCBI Sequence Read Archive SRX434531, Bioproject
PRJNA235228; Fig. 1) were screened using MSATCOMMANDER 1.0.8-
beta (Faircloth 2008) for microsatellite loci of two or three nucleotides
with a minimum of 15 repeats. This resulted in the identification of 1,254
microsatellite loci. The embedded Primer3 software (Rozen and Skaletsky
1999) subsequently designed M13-tagged universal primers for these using
the following settings: melting temperatures (Tm) of 58–60°C, GC content
of 40–60%, and fragment sizes of 100–450 bp. This resulted in primers for 37
loci, which were screened for amplification success and specificity using
DNA of four R. lagascae specimens. A total of 17 polymorphic loci (Table 1)
were subsequently used for genotyping the additional DNA samples of R.
lagascae and R. manillana available to us. A blastn search using the contig
DNA sequences that contain these 17 loci did not reveal resemblance with
available Vitaceae genomic data in GenBank. We therefore assume that
none of these loci are present in the genome of R. lagascae as a result of
horizontal gene transfer from their host plants.
DNA Extraction and Microsatellite Genotyping—Total genomic DNA
for the microsatellite genotyping analyses was extracted using the Qiagen
DNeasy plant mini kit (Qiagen, Germantown, Maryland) following the
manufacturer’s protocol. Multiplex PCR analyses using up to four primer
combinations per PCR sample were performed in a total volume of 4 ml
using the Qiagen Type-it microsatellite PCR kit. These samples contained
1mlofgenomicDNA(ca.2–60 ng per sample), 2 mlof23PCR master mix ,
0.16 pmol of M13-tagged primer, 0.64 pmol of untagged primer, 2 pmol
of M13 fluorescent-labeled primer (6FAM, NED, PET, or VIC; 50-
GGAAACAGCTATGACCAT-30) and nuclease-free water to volume. The
PCR conditions were as follows: an initial denaturation of 5 min at 95°C
followed by 30 sec at 95°C, 90 sec at 60°C, 30 sec at 72°C, during 28 cycles,
and a final extension of 30 min at 60°C. The PCR products were run on an
ABI 3130xL Genetic Analyzer at the University of Canterbury. Geneious
6.1.7 (Biomatters Ltd, Auckland, New Zealand) was used to determine
fragment lengths.
Data Analyses—We genotyped 143 DNA samples of the R. lagascae
complex obtained from 80 host plants. Microsatellite data of 98 of these
samples were included in our analyses. A total of 45 DNA samples were
excluded, because these had identical genotypes to those of other flowers or
buds obtained from the same host plant and might therefore represent the
same Rafflesia individual.
Exact tests for deviations from Hardy-Weinberg equilibrium and
linkage disequilibrium were performed with GENEPOP v. 4.2 (Raymond
and Rousset 1995) as implemented in GenePop on the web (http://
genepop.curtin.edu.au/) and significance levels were adjusted for multi-
ple tests using the B-Y method FDR at p50.05 (Narum 2006). Only loci
found to display significant linkage disequilibrium across all populations
were considered to be linked. Micro-checker v. 2.2.3 (Van Oosterhout et al.
2004) was used to check for null alleles (95% confidence interval; 1,000
repetitions). Only loci indicated to display null alleles across all pop-
ulations were considered to be truly producing null alleles.
GenAlEx v. 6.5 (Peakall and Smouse 2012) was used to assess the genetic
diversity of each population by calculating the percentage of polymorphic
loci, allelic richness (mean number of alleles and number of effective al-
leles), observed (H
o
) and expected (H
e
) heterozygosity, unbiased hetero-
zygosity (uH
e
), and the inbreeding coefficient. BOTTLENECK 1.2.0.2.
Table 1. Characteristics of microsatellite loci for Rafflesia lagascae and R. manillana. Annealing temperature was 60°C for all loci. Underlined nucleotides
within primers indicate the M13 tag.
Locus Primer sequences (50–30) Repeat motif Allele size range (bp) GenBank accession no.
Man78 F: CCTTGTCACTGCCCATTTCC AG 186–232 KX212099
R: GGAAACAGCTATGACCATCAGTTCATGCGGACCCATC
Man80 F: TCAGGATTCGTGAGCCAAGG AC 269–297 KX212101
R: GGAAACAGCTATGACCATAACACAGGCCGAAGTTGATC
Man109 F: ACGTAGTCATCCATTGAAAGGG AC 376–406 KX212086
R: GGAAACAGCTATGACCATACTTGCCAGCCAGCTTC
Man111 F: GTTGGATTCATCACGTTCATGC AC 398–449 KX212087
R: GGAAACAGCTATGACCATCACCTTCGGCATTCATCCTG
Man120 F: TGTTACTTTGTCTGCCCTTCAC AG 186–206 KX212091
R: GGAAACAGCTATGACCATGTGTATTCCAACGAGCAGG
Man142 F: GGAAACAGCTATGACCATACCCAGGCAAGCGAAGTAC AG 328–346 KX212092
R: TTCATTTGTGAAGAAGACGAGC
Man144 F: GGAAACAGCTATGACCATTTCCTCTTCAGCCAGTCGG AC 179–195 KX212093
R: GTACTCATGAGGTTGTTGGCG
Man166 F: GGAAACAGCTATGACCATGCCCATACATATCCATACACC AC 101–145 KX212094
R: CCCAAGCTCACACAAAGGAG
Man171 F: GGAAACAGCTATGACCATGCCGCCTTCACCATTAATC AAT 238–265 KX212095
R: AGAAGCGAGGTGAAATGATCTC
Man273 F: GCGTGGTTCATTCATGGAGG AC 203–238 KX212096
R: GGAAACAGCTATGACCATAACTTCAGGCCCTTCTCTCG
Man553 F: GGAAACAGCTATGACCATCCACATGCACTCTACCCTC AC 162–197 KX212097
R: TGAGAAAGACTTTGGGAGATGG
Man714 F: GGAAACAGCTATGACCATGTGCGTGCATAACTAACCC AC 220–288 KX212098
R: CATTAGGCTCTGCACACCTTG
Man788 F: GGAAACAGCTATGACCATCCTTCACTTCCACACTACACC AG 334–352 KX212100
R: AGAGATGGGTGGGAAAGGAAG
Man866 F: ATCTACATGAGTCTGTGTGCC AC 147–171 KX212102
R: GGAAACAGCTATGACCATACAGTTACACAGAGACACTTGG
Man1134 F: GGAAACAGCTATGACCATCCTAGACCTTGGTTTGGG AC 439–464 KX212088
R: TTTAGCCTGGGTTTGGAGGG
Man1169 F: CTTTGGTCGAGTAAGGCTAGTC AC 130–176 KX212089
R: GGAAACAGCTATGACCATACCTCAACTTCAATGCGTGC
Man1193 F: GGAAACAGCTATGACCATCCCTCTCCCACTATTTATCGAC AG 331–356 KX212090
R: ACAAGCAAGGAAGATAGACGG
PELSER ET AL.: GENETIC DIVERSITY AND STRUCTURE OF RAFFLESIA LAGASCAE 5452017]
(Cornuet and Luikart 1997) was used to test for recent genetic bottlenecks in
the six Rafflesia populations from which at least nine genetically unique
samples were available. Wilcoxon sign-rank tests were used to determine if
excesses in heterozygosity are statistically significant. Mode-shift tests in
BOTTLENECK were employed to identify shifts in allele frequencies.
Genetic structure was studied by determining the number of private
alleles for each population in GenAlEx. In addition, an analysis of mo-
lecular variance (AMOVA; F
st
, 999 permutations) was carried out in
GenAlEx with a data set from which samples from Mt. Banahaw and Mt.
Mingan were excluded because only one Rafflesia sample was collected
from each of these two areas. Bayesian model-based clustering analyses
were performed with STRUCTURE v. 2.3.4 (Pritchard et al. 2000; Falush
et al. 2003, 2007) using the admixture model and correlated allele fre-
quencies. These analyses were run with K values from 1–20 and with 20
iterations for each K value. Each analysis was run for 200,000 generations of
which the first 100,000 were discarded as burn-in. The STRUCTURE output
was summarized using STRUCTURE HARVESTER (Earl and von Holdt
2012) to determine the value of K that best explains the genetic structure in
the data. For this, both the Evanno et al. (2005) method (K with the highest
value of DK) and the method of Pritchard et al. (2000; K with the highest Pr
(X|K)) were used. CLUMPAK 1.1 (Kopelman et al. 2015) was used for the
summation and graphicalpresentation of theSTRUCTURE results.GenAlEx
was subsequently used for a Principal Coordinate Analysis (PCoA) using a
covariance matrix of co-dominant genotypic pairwise distances between
individual samples with data standardization.
Mantel tests (999 replicates) were used in GenAlEx to test for a corre-
lation between geographic distance and co-dominant genotypic distance
between individuals. These tests were executed with non-transformed
geographic distances as well as with a data set in which these distances
were transformed using the natural logarithm.
Results
Patterns of genetic diversity and genetic structure in the R.
lagascae complex were studied using microsatellite data from
17 loci. These loci yielded a total of 249 different alleles. The
number of alleles per locus across all populations ranged from
5 to 33 (mean 14.59). After B-Y correction (Narum 2006), two of
the 13 populations that were included in our study (Bolos
Point and Mt. Irid) showed a significant (p50.05) deviation
from Hardy-Weinberg proportions because of a deficit of
heterozygotes (Table 2). Overall, however, the observed
heterozygosity (H
o
50.60) was only slightly higher or lower
than estimates of the expected heterozygosity (H
e
50.53,
uH
e
50.64). None of the pairs of loci displayed significant
linkage disequilibrium across all populations and none of
the 17 loci showed evidence of producing null alleles across
all populations.
For 32 of the 80 Tetrastigma host plants of the R. lagascae
complex that were examined, more than one Rafflesia flower or
flower bud was genotyped. Two different Rafflesia genotypes
were obtained from 17 of these hosts (53%), and three different
parasite genotypes were recovered from one host plant. A total
of 57 of the 93 (61%) Rafflesia flowers for which sex could be
determined were staminate. Staminate and pistillate flowers
with identical genotypes were found on four host plants from
the Basey, Bolos Point, Maria Aurora, and Mt. Malinao pop-
ulations. Non-identical Rafflesia genotypes from the same host
have a mean co-dominant genetic distance of 16.5 (n 518;
range 1–32). Rafflesia samples that were obtained from dif-
ferent hosts were all genetically different.
Populations at Bolos Point, Mt. Irid, and Salazar show the
highest levels of genetic diversity as measured by the per-
centage of polymorphic loci, their allelic richness, and het-
erozygosity (Table 2). The inbreeding coefficient is low for the
species complex as a whole (F
IS
5-0.14). It is highest for the
R. manillana population (Basey; F
IS
50.11) and the Mt. Irid
population of R. lagascae (F
IS
50.12). Wilcoxon tests in
Table 2. Genetic diversity and number of private alleles observed at 17 microsatellite loci for 13 populations of the R. lagascae complex. Number of Tetrastigma host plants from which Rafflesia samples were taken,
number of Rafflesia samples, percentage of polymorphic loci (P), allelic richness (Na), number of effective alleles (Ne), number of private alleles (Na(p)), observed heterozygosity (H
o
), expected heterozygosity (H
e
),
unbiased expected heterozygosity (uH
e
), and inbreeding coefficient (F
IS
). SE 5Standard Error. *Populations with a significant deviation from Hardy–Weinberg proportions due to heterozygote deficits after B-Y
correction (Narum 2006).
Population Protected area # Hosts
sampled #Rafflesia
samples P Na (SE) Ne (SE) Na(p) H
o
(SE) H
e
(SE) uH
e
(SE) F
is
(SE)
Aurora Memorial
Natural Park
Aurora Memorial Natural Park 7 9 100% 5.47 (0.67) 4.09 (0.59) 6 0.67 (0.06) 0.67 (0.05) 0.71 (0.05) 20.01 (0.05)
Bolos Point No protective status 11 13 100% 7.82 (0.64) 5.14 (0.52) 21 0.75 (0.05) 0.76 (0.04) 0.79 (0.04) 0.01 (0.05)*
Burgos Pantabangan-Carranglan Watershed Forest Reserve 4 4 94% 4.18 (0.31) 3.19 (0.28) 4 0.58 (0.06) 0.63 (0.05) 0.72 (0.06) 0.05 (0.08)
Maria Aurora No protective status 2 3 88% 3.06 (0.29) 2.60 (0.24) 1 0.73 (0.09) 0.54 (0.06) 0.69 (0.07) 20.35 (0.09)
Mt. Banahaw Mounts Banahaw-San Cristobal Protected Landscape 1 1 65% 1.65 (0.12) 1.65 (0.12) 1 0.65 (0.12) 0.32 (0.06) 0.65 (0.12) 21.00 (0.00)
Mt. Irid No protective status 8 12 100% 6.65 (0.73) 4.48 (0.52) 11 0.66 (0.07) 0.71 (0.05) 0.74 (0.05) 0.12 (0.07)*
Mt. Labo No protective status 20 21 100% 3.94 (0.36) 2.52 (0.25) 6 0.48 (0.05) 0.52 (0.05) 0.54 (0.06) 0.07 (0.04)
Mt. Makiling Mt. Makiling Forest Reserve 4 5 94% 3.82 (0.46) 2.97 (0.43) 8 0.60 (0.08) 0.56 (0.05) 0.63 (0.06) 20.07 (0.09)
Mt. Malinao No protective status 2 2 59% 1.82 (0.29) 1.74 (0.28) 3 0.44 (0.11) 0.32 (0.07) 0.42 (0.10) 20.39 (0.13)
Mt. Mingan No protective status 1 1 65% 1.65 (0.12) 1.65 (0.12) 1 0.65 (0.12) 0.32 (0.06) 0.65 (0.12) 21.00 (0.00)
Mt. Natib Bataan National Park 2 3 82% 2.53 (0.26) 2.21 (0.24) 3 0.45 (0.09) 0.44 (0.06) 0.53 (0.08) 0.01 (0.12)
Salazar Pantabangan-Carranglan Watershed Forest Reserve 8 11 100% 6.53 (0.70) 4.18 (0.48) 7 0.71 (0.06) 0.70 (0.04) 0.74 (0.04) 20.01 (0.07)
Basey Samar Island Natural Park 10 13 82% 3.29 (0.42) 2.36 (0.27) 7 0.43 (0.09) 0.46 (0.07) 0.49 (0.07) 0.13 (0.10)
Total means 80 98 87% 4.03 (0.18) 2.98 (0.12) 6.07 0.60 (0.02) 0.53 (0.02) 0.64 (0.02) 20.14 (0.03)
SYSTEMATIC BOTANY [Volume 42546
BOTTLENECK did not reveal a significant excess of hetero-
zygotes in the six populations that were included in these tests
when a Stepwise Mutation Model was assumed. However,
significant heterozygosity excess was found for the Aurora
Memorial Natural Park (AMNP) and Basey populations under
the Infinite Alleles Model and the Two Phase Model. The Bolos
Point, Mt. Irid, and Mt. Labo populations only displayed
significant heterozygosity excess under an Infinite Alleles
Model. The AMNP population shows evidence of a shift in
allelic frequencies.
All 13 populations of the R. lagascae complex have at least
one private allele (Table 2). The Bolos Point and Mt. Irid
populations have the most private alleles (21 and 11, re-
spectively). Overall genetic differentiation within the R.
lagascae complex is significant (F’
ST
50.651, p,0.001). A total
of 22% of the genetic variation was found among populations
and 9% among individuals. All pairwise F
ST
values are sta-
tistically significant at p50.05 after B-Y correction (Narum
2006), except for the population pairs Burgos-Salazar and
Burgos-Mt. Irid (Table 3). Using the Pritchard et al. (2000) and
Evanno et al. (2005) methods, the results of the STRUCTURE
analysis identified three primary genetic clusters (K 53). One
of these clusters is solely composed of specimens of the Basey
population (R. manillana), the second cluster exclusively
contains R. lagascae specimens from Mt. Labo, and all speci-
mens of the remaining 11 R. lagascae populations form the third
cluster (Fig. 3). The admixture proportions (qvalues) for in-
dividual specimens indicate very little admixture between the
three clusters. The same genetic clusters were also recovered
in a PCoA of pairwise genetic distances (Fig. 4).
To study patterns of genetic structure within R. lagascae s. s.,
STRUCTURE analyses were also performed with a data set
that contains only populations of this group. For this data set,
the Evanno et al. (2005) method favored K 52, whereas the
Pritchard et al. (2000) method preferred K 58. The
STRUCTURE plots of K 52 (Fig. 5A) show that the Bolos
Point population in the northern part of Luzon (Fig. 1) is placed
in the same genetic cluster as populations from the southern
part of mainland Luzon (Mt. Banahaw, Mt. Makiling, Mt.
Natib), and the Bicol peninsula (Mt. Malinao). The remaining
populations of R. lagascae s. s. are found in central Luzon
(AMNP, Burgos, Maria Aurora, Mt. Irid, Mt. Mingan, Salazar)
and show a pattern of admixture between both genetic clusters
(Fig. 5A). The K 58 STRUCTURE plots (Fig. 5B; only the major
clustering pattern presented) show that some nearby pop-
ulations have similar genetic profiles (AMNP, Maria Aurora,
and Mt. Mingan; Burgos and Salazar), that some genetic
clusters are predominantly found in individual populations
(e.g. Mt. Makiling, Mt. Natib, Mt. Irid), and that some pop-
ulations are mostly composed of specimens with high qvalues
for a single genetic cluster (e.g. AMNP, Mt. Makiling,
Mt. Malinao). Overall genetic differentiation within the
R. lagascae s. s. cluster is significant (F’
ST
50.420, p,0.001).
A total of 12% of the genetic variation was found among
populations and 9% among individuals. Within R. lagascae s. s.,
Mantel tests did not reveal a statistically significant positive
correlation at p50.05 between geographic and genetic distances.
Discussion
Because of the size of their flowers, which are among the
largest in the world (Nais 2001), and their intriguing parasitic
life style, Rafflesia species speak to the imagination of biologists
and the general public alike. This fascination is further
heightened by the fact that many Rafflesia species are very
rare. For example, R. schadenbergiana G¨opp. ex Hieron.
(Hieronymus 1885) is currently known from only two living
host plants (Barcelona et al. 2008; Pelser et al. 2016). As such,
Rafflesia are flagships of plant conservation: the giant pandas
of the plant world (Josephson 2000). Yet, to the best of our
knowledge, conservation genetic studies have not previously
been attempted and, consequently, nothing was thus far
known about the patter ns of genetic diversit y and structure of
Rafflesia species. This is not because of a lack of interest by
biologists, but rather because of difficulties typically asso-
ciated with studying rare species with small population sizes.
Even R. lagascae, the most common and widespread species of
Rafflesia in the Philippines, is only known from 15 areas and
few Tetrastigma host plants. For that reason, it is difficult to
attain sample sizes that provide enough statistical power to
conclusively answer population genetic questions. Although
we sampled 13 of the 16 known populations of the R. lagascae
complex, and went through great efforts to locate as many
host plants as possible, many Rafflesia populations appear to
consist of only a single or very few host plants that produce
only a few flowers and, consequently, yielded few samples
for our molecular genetic analyses. For example, from two
populations (Mt. Banahaw and Mt. Mingan) only a single
sample could be obtained (Table 2). This means that, despite
extensive sampling, our results should be interpreted with
care and that our conclusions are tentative. However, al-
though the absolute number of Rafflesia specimens that we
sampled was small, the percentage of individuals sampled in
each population is very high. Therefore, despite of low
sample numbers in some populations, our research resulted
in very accurate estimates of population allele frequencies.
Table 3. Pairwise F
st
values between populations of the R. lagascae complex. *Non-significant pairwise comparisons after B-Y correction (Narum 2006).
Bolos
Point Aurora Memorial
Natural Park Maria
Aurora Salazar Burgos Mt. Irid Mt. Natib Mt.
Makiling Mt.
Malinao Mt.
Labo
R. lagascae Bolos Point
R. lagascae Aurora Memorial Natural Park 0.088
R. lagascae Maria Aurora 0.100 0.053
R. lagascae Salazar 0.077 0.074 0.120
R. lagascae Burgos 0.099 0.087 0.127 0.004*
R. lagascae Mt. Irid 0.066 0.066 0.100 0.042 0.037*
R. lagascae Mt. Natib 0.147 0.209 0.263 0.184 0.192 0.169
R. lagascae Mt. Makiling 0.120 0.186 0.227 0.168 0.182 0.131 0.211
R. lagascae Mt. Malinao 0.254 0.309 0.393 0.209 0.282 0.239 0.385 0.387
R. lagascae Mt. Labo 0.232 0.318 0.373 0.283 0.343 0.275 0.386 0.304 0.413
R. manillana Basey 0.245 0.268 0.338 0.217 0.265 0.214 0.375 0.385 0.381 0.412
PELSER ET AL.: GENETIC DIVERSITY AND STRUCTURE OF RAFFLESIA LAGASCAE 5472017]
Breeding System and Co-infection of Hosts—Genetically
identical Rafflesia samples were exclusively obtained from the
same host plant and Rafflesia samples from the same host plant
that are not genetically identical have a mean co-dominant
genetic distance of 16.5. This indicates that our microsatellite
data set of 17 loci potentially provides enough resolution at the
level of individuals to determine if plants of the R. lagascae
complex are dioecious or monoecious. Under this assumption,
our data indicate that this complex is monoecious, because
genetically identical staminate and pistillate flowers were
found on the same host plants. Previously, however, Rafflesia
was hypothesized to be dioecious (e.g. Olah 1960; Beaman
et al. 1988; Renner and Feil 1993; Bellot and Renner 2013;
Wicaksono et al. 2016; but see Nais 2001 for more cautious
statements). This reproductive strategy was considered more
likely because Rafflesia species are rare and populations are far
apart. Dioecy was therefore regarded as a likely adaptation to
avoid inbreeding depression (Beaman et al. 1988). Two species,
R. baletei Barcelona and Cajano (Barcelona et al. 2006) and R.
verrucosa Balete, Pelser, Nickrent & Barcelona. (2010), how-
ever, have both stamens and a developed ovary on the same
flower, although it is unknown if these are also functionally
bisexual. In addition, non-functional stamens have been re-
ported for some female flowers of R. consueloae Galindon, Ong
and Fernando (2016). If plants of the Rafflesia lagascae complex
and other Rafflesia species are monoecious, it remains to be
determined if they are self-compatible and geitonogamous or
exclusively outcrossing. A geitonogamous breeding system
might benefit small Rafflesia populations, because it increases
the number of available mates, however, it carries with it the
Fig. 3. K 53 STRUCTURE results for the R. lagascae complex. Each bar represents an individual plant and bar colors indicate the proportion of
membership to each genetic cluster (qvalues). Individuals are grouped by population. ANMP 5Aurora Memorial Natural Park.
Fig. 4. The PCoA ordination of the R. lagascae complex using a covariance matrix of co-dominant genotypic pairwise distances between individual
samples with data standardization. The first and second PCoA axes explain 15.5% and 8.6% of the variation respectively. ANMP 5Aurora Memorial
Natural Park.
SYSTEMATIC BOTANY [Volume 42548
potentially negative consequences of inbreeding (Byers and
Meagher 1992; Ellstrand and Elam 1993). Some studies have
suggested that dioecy is overrepresented in parasitic plants
(Renner and Ricklefs 1995; Bellot and Renner 2013), but this
hypothesis rests in part on the assumption that the presence of
pistillate and staminate flowers of endophytic parasites on the
same host belong to different parasite individuals (Bellot and
Renner 2013). If endophytic parasites with unisexual flowers
such as those in Apodanthaceae and Rafflesiaceae are indeed
monoecious, then the incidence of dioecy among all parasites
has been overestimated.
For Tetrastigma host plants from which more than one
Rafflesia sample was taken, genetically different Rafflesia
flowers were recovered from about half. This suggests that
Tetrastigma host plants are regularly home to multiple ge-
netically distinct Rafflesia individuals. Consequently, even
populations composed of a single host plant can maintain
genetic diversity. Potentially, the monoecious breeding system
and co-infection of host plants enable the R. lagascae complex to
maintain viable populations in areas with low host densities.
Genetic Diversity—The numbers of samples taken from
each Rafflesia population can be considered as rough estimates
of relative population sizes, because we sampled Rafflesia from
nearly all encountered infected host plants that produced
flower buds or flowers during our field work. Although
smaller populations of the R. lagascae complex typically dis-
play less genetic diversity than larger populations, as mea-
sured by the percentage of polymorphic loci, allelic richness,
and heterozygosity estimates, they are not genetically de-
pauperate and do not show evidence of marked inbreeding
(Table 2), as might be expected for small and isolated pop-
ulations (Ellstrand and Elam 1993; although see Binks et al.
2015 and references therein). Overall, heterozygosity estimates
are similar to those of other perennial plant species (Nybom
2004). Six populations for which at least nine genetically
unique samples were available were included in analyses to
test for indications of recent genetic bottlenecks. Although we
did not find evidence for a genetic bottleneck in the Salazar
population, there are indications that the other populations
(AMNP, Basey, Bolos Point, Mt. Irid, Mt. Labo) might have
experienced a genetic bottleneck in the recent past. However, if
they indeed went through a genetic bottleneck, it is at present
not known what caused this, because all five populations are
located in areas with relatively intact forests.
Genetic Structure—The results of the genetic structure
analyses indicate low genetic connectivity between most
populations of the R. lagascae complex. All populations have
one or more private alleles (Table 2), most pairwise F
ST
values
and the overall F’
ST
value are statistically significant (Table 3),
and a substantial proportion of the genetic variation in the
complex is found among populations. The results of the
STRUCTURE and PCoA analyses show that there are three
distinct genetic groups that align with R. manillana, the Mt.
Labo R. lagascae population, and the 11 remaining populations
of R. lagascae (Figs. 3, 4). These patterns suggest that these three
groups are genetically isolated from each other.
The distinctiveness of R. manillana as indicated by the
microsatellite data used in this study is supported by mor-
phological differences between this taxon and R. lagascae
(Madulid and Agoo 2008; Madulid et al. 2008; Pelser et al. 2013;
Fig. 2). Except for R. speciosa Barcelona and Fernando (2002),
which is found on the island of Panay and the adjacent
northern part of Negros Island (Barcelona et al. 2009), all
Philippine Rafflesia species are endemic to individual islands in
the archipelago. Rafflesia manillana is endemic to Samar Island
whereas R. lagascae is only known from Luzon Island and their
genetic and morphological differences therefore support the
hypothesis that the sea straits that separate the Philippine
islands form significant barriers to gene flow between Rafflesia
populations. The genetic isolation of R. manillana might be a
contributing factor to its lower genetic diversity relative to
most other populations from which a similar number of
samples was obtained (Table 2).
Also the Mt. Labo R. lagascae population is genetically
isolated from other populations of the complex and likewise
shows relatively low genetic diversity (Table 2). In contrast
with R. manillana, however, the Mt. Labo population does not
appear to be morphologically different from the other R.
lagascae populations (Fig. 2). Its floral characters (e.g. size,
coloration, ornamentation of the diaphragm and perigone
lobes) seem to fall within the range of variation observed in R.
lagascae s. s., but detailed morphometric studies are needed to
determine if the Mt. Labo population is indeed morpholog-
ically indistinct from R. lagascae s. s. Pelser et al. (2016)
Fig. 5. STRUCTURE results for R. lagascae s. s. Each bar represents an individual plant and bar colors indicate the proportion of membership to each
genetic cluster (qvalues). Individuals are grouped by population. A. K 52. B. K 58. Only the major clustering pattern is shown. ANMP 5Aurora Memorial
Natural Park.
PELSER ET AL.: GENETIC DIVERSITY AND STRUCTURE OF RAFFLESIA LAGASCAE 5492017]
identified Tetrastigma loheri Gagnepain (1910) as the host
species of Mt. Labo Rafflesia plants. This species is the most
common host of the R. lagascae complex (Pelser et al. 2016)
and differences in host preference can therefore not explain
the genetic distinctiveness of the Mt. Labo population.
Mt. Labo is located on Bicol Peninsula, which is currently
connected to mainland Luzon by a relatively narrow land
bridge, and at the southeastern end of the distribution area of
R. lagascae (Fig. 1). According to Hall (1998, 2002), the Bicol
Peninsula was part of a cluster of islands that included Samar
and eastern parts of the current Visayas and Mindanao and
that has progressively moved towards mainland Luzon over
the last 45 my. The Bicol Peninsula connected to southeastern
Luzon only ca. 6 mya (Hall 2002). The genetic isolation of the
Mt. Labo population can therefore potentially be explained by
the geographical isolation of the Bicol Peninsula prior to 6 mya.
Molecular dating studies by Bendiksby et al. (2010) suggest
that Rafflesia was present in the Philippines as early as 5.2–9.3
mya. Thus, it is indeed possible that the dynamic geological
history of the Philippines shaped the patterns of genetic
structure of R. lagascae. The Mt. Malinao population, however,
is also located on the Bicol Peninsula (Fig. 1), yet genetically
more similar to populations from mainland Luzon than to the
Mt. Labo population (Fig. 3). If the previously mentioned
geographic hypothesis is correct, then long-distance dispersal
of Rafflesia lagascae s. s. between mainland Luzon to Mt.
Malinao could explain the observed genetic patterns. Zoo-
chory has been proposed as the dispersal method of Rafflesia
seeds (Teijsmann 1856; Justesen 1922; Kuijt 1969; Emmons
et al. 1991; Nais 2001; B¨anziger 2004). Although various ani-
mals have been considered as dispersal agents, observations
by Pelser et al. (2013) suggest seed dispersal by ants. Myr-
mecochory is potentially plesiomorphic in Rafflesiaceae, be-
cause it is also common in Peraceae (Bond and Slingsby 1983;
Lengyel et al. 2009, 2010), which form the sister group of the
clade composed by Euphorbiaceae and Rafflesiaceae (Davis
et al. 2007; Endress et al. 2013). If Rafflesia is indeed myrme-
cochorous, long-distance dispersal is presumably very rare,
and this might explain the high island-endemicity of Rafflesia
in the Philippines. However, other observations and experi-
ments suggest that mammalian frugivores might play a role in
endozoochorous seed dispersal (Emmons et al. 1991; B ¨anziger
2004) and this mode of dispersal might be able to transport
Rafflesia seeds more frequently over greater distances. Alter-
natively, the Mt. Malinao area might be a fragment of main-
land Luzon whose position has been scrambled by tectonic
movement of the Bicol fragment. Although we are not aware of
geological evidence that supports this hypothesis, this part of
the Philippines clearly has a very complex and dynamic
geological history that is presently only poorly known (Hall
2002).
Patterns of genetic structure within R. lagascae s. s. are less
pronounced than those among R. lagascae s. s., Mt. Labo R.
lagascae, and R. manillana, but still indicate low genetic con-
nectivity among most populations (Table 2; Figs. 4, 5).
However, particularly some populations in central Luzon (i.e.
Burgos, Mt. Irid, Salazar) show admixture between different
genetic clusters (Fig. 5) and might therefore have experienced
higher levels of recent gene flow than some of the other
populations (e.g. Mt. Makiling and Mt. Malinao). Although
evidence for isolation by distance in the R. lagascae s. s. cluster
is not statistically significant, pairwise F
ST
values (Table 3) and
the results of Bayesian cluster analyses in STRUCTURE (Fig. 5)
suggest stronger genetic connectivity within two groups of
nearby populations in central Luzon (i.e. among the AMNP,
Maria Aurora, and Mt. Mingan populations and between
Burgos and Salazar) than among other populations. This in-
dicates that nearby populations are perhaps less likely to
become genetically isolated than those that are further away or
that genetic connectivity between populations in central
Luzon is higher than in other parts of the distribution area of R.
lagascae.
Taxonomic Delimitation of the R. lagascae Complex—The
results of our genetic structure analyses inform the taxonomic
delimitation of the R. lagascae complex. In addition to con-
firming that R. manillana is not only morphologically different
from R. lagascae (Fig. 2), but also genetically distinct, we also
demonstrate that a genetically different but morphologically
cryptic population of R. lagascae exists on Mt. Labo. If a species
concept is applied in which genetic distinctiveness or the
absence of interbreeding is considered evidence for recog-
nizing groups of individuals as different species (Coyne and
Orr 2004), then the Mt. Labo population should be described
as a distinct species. We refrain here from formally naming this
species, awaiting the completion of detailed morphological
studies in search of characters in which the Mt. Labo plants are
different from those of R. lagascae s. s.
Conservation Implications—The R. lagascae complex is
known from relatively few and disjunct populations, some of
which are only composed of a few host plants. Despite this, our
results do not provide evidence of pronounced inbreeding and
populations show moderate levels of genetic diversity. This
potentially means that the disjunct distribution of the R.
lagascae complex and the small sizes of some of its populations
are a result of natural differences in host presence and
abundance among areas and that the R. lagascae complex has
adapted to this. In other words, the R. lagascae complex is
possibly naturally rare. It is extremely likely, however, that the
number of populations of the R. lagascae complex has decreased
over the past century and that the geographic distances between
them have increased as a result of widespread habitat de-
struction and degradation in the Philippines (Mittermeier et al.
1998; Ong et al. 2002). Especially in the southern part of its
distribution in Luzon, the natural forest ecosystems in which
R. lagascae is found are confined to small areas separated by
vast areas in which forests have been converted to farmland.
This has undoubtedly resulted in decreased genetic connec-
tivity between the remaining populations.
Our results show substantial differences in allelic compo-
sition and frequencies between populations of the R. lagascae
complex (Table 2). This means that considerable genetic di-
versity might be lost if any of its populations would go extinct
(Schmidt and Jensen 2000). Small or isolated populations
outside of protected areas are most at risk of extinction. These
include the populations from Maria Aurora, Mt. Malinao, and
Mt. Mingan (Table 2). However, also the larger populations
outside of protected areas (Bolos Point, Mt. Irid, and Mt. Labo)
are of conservation concern, because they harbor the highest
levels of genetic diversity and need to be protected from
habitat destruction and degradation to remain centers of ge-
netic diversity.
Because of their pronounced genetic differences and the lack
of gene flow between them, R. lagascae s. s., Mt. Labo R.
lagascae, and R. manillana should be managed as distinct
conservation entities. Mt. Labo R. lagascae and R. manillana are
each only known from a single population, which shows lower
SYSTEMATIC BOTANY [Volume 42550
genetic diversity than similar sized populations of R. lagascae
s. s. (Table 2). Both taxa therefore have a higher risk of losing
genetic diversity and of becoming extinct than R. lagascae s. s.
The habitat of R. manillana is protected as part of the Samar
Island Natural Park and is relatively intact, but the Mt. Labo
area does not enjoy formal protection. The Mt. Labo Rafflesia is
therefore particularly prone to experiencing a reduction in the
size of its single population and should thus be managed as a
critically endangered species (IUCN 2012).
If Rafflesia lagascae s. s. and the Mt. Labo Rafflesia are indeed
morphologically cryptic species, then this poses the question of
whether additional cryptic species have been overlooked in
Rafflesia. If so, this could mean that Rafflesia species of con-
servation concern have remained undetected and that species-
level biodiversity of Rafflesia has been underestimated (e.g.
Sch¨onrogge et al. 2002; Bickford et al. 2007; Shepherd et al.
2015). Future studies into species delimitation using genetic
tools are therefore needed to be able to better inform Rafflesia
conservation management.
Acknowledgments. This publication is dedicated to Danilo S. Balete
(1960-2017), mammologist and discoverer of three new species of Philip-
pine Rafflesia. This project was funded by the Marsden Fund Council from
Government funding administered by the Royal Society of New Zealand
and by the Linnean Society, Percy Sladen Memorial Fund, and the National
Geographic Foundation. We would like to thank the three anonymous
reviewers of this manuscript for their valuable comments. We are grateful
to the National Museum of the Philippines, the Department of Environ-
ment and Natural Resources (DENR) and in particular T. M. S. Lim, Di-
rector of the Biodiversity Management Bureau (BMB) and staff of the
DENR regional and local offices for facilitating the issuance of collecting,
transport, and export permits. Prior Informed Consents were issued by the
Protected Areas and Management Boards (PAMB) of Aurora Memorial
National Park, Bataan Natural Park (Mt. Natib), Pantabangan-Carranglan
Watershed Reserve, Mt. Banahaw-Mt. San Cristobal Protected Area, Samar
Island Natural Park, the Mayors of Buhi, Camarines Sur; Gattaran and
Lal-o, Cagayan; Gabaldon, Nueva Ecija; Labo and San Lorenzo Ruiz,
Camarines Norte; Rodriguez, Rizal, and the Makiling Center for
Mountain Ecosystems. We would also like to thank our NGO partners:
Cagayan Valley Partners in People Development and the Philippine Native
Plants Conservation Society, Inc. For sharing information on distributional
ranges and fieldwork assistance, we thank J. Aliposa, D. S. Balete†,M.O.
Cajano†, J. R. Callado, L. L. Co†, V. Dacumos, A. Diesmos, N. B. Gapas,
J. Sarmiento, and D. Tandang. Numerous field guides/assistants and staff
of DENR and LGU deserve our sincerest gratitude for without their help,
fieldwork would have not been possible. Finally, M. L. Hale and T. E.
Steeves provided helpful feedback on the results of the population genetic
analyses and M. Walters helped with preparing figures for this publication.
Literature Cited
Balete, D. S., P. B. Pelser, D. L. Nickrent, and J. F. Barcelona. 2010. Rafflesia
verrucosa (Rafflesiaceae), a new species of small-flowered Rafflesia
from eastern Mindanao, Philippines. Phytotaxa 10: 49–57.
B¨anziger, H. 2004. Studies on hitherto unknown fruits and seeds of some
Rafflesiaceae, and a method to manually pollinate their flowers for
research and conservation. Linzer Biologische Beitr¨age 36: 1175–1198.
Barcelona, J. F. and E. S. Fernando. 2002. A new species of Rafflesia (Raf-
flesiaceae) from Panay Island, Philippines. Kew Bulletin 57: 647–651.
Barcelona, J. F., M. O. Cajano, and A. S. Hadsall. 2006. Rafflesia baletei,
another new Rafflesia (Rafflesiaceae) from the Philippines. Kew Bulletin
61: 231–237.
Barcelona, J. F., P. B. Pelser, A. M. Tagtag, R. G. Dahonog, and
A. P. Lilangan. 2008. The rediscovery of Rafflesia schadenbergiana
(Rafflesiaceae). Flora Malesiana Bulletin 14: 162–165.
Barcelona, J. F., P. B. Pelser, D. S. Balete, and L. L. Co. 2009. Taxonomy,
ecology, and conservation status of Philippine Rafflesia.Blumea 54:
77–93.
Barcelona, J. F., M. M. E. Manting, R. B. Arbolonio, R. B. Caballero, and
P. B. Pelser. 2014. Rafflesia mixta (Rafflesiaceae), a new species from
Surigao del Norte, Mindanao, Philippines. Phytotaxa 174: 272–278.
Barkman, T. J., S. H. Lim, K. Mat Salleh, and J. Nais. 2004. Mitochondrial
DNA sequences reveal the photosynthetic relatives of Rafflesia,the
world’s largest flower. Proceedings of the National Academy of Sciences
USA 101: 787–792.
Barkman, T. J., M. Bendiksby, S.-H. Lim, K. M. Salleh, J. Nais, D. Madulid,
and T. Schumacher. 2008. Accelerated rates of floral evolution at the
upper size limit for flowers. Current Biology 18: 1508–1513.
Beaman, R. S., P. Decker, and J. Beaman. 1988. Pollination of Rafflesia
(Rafflesiaceae). American Journal of Botany 75: 1148–1162.
Bellot, S. and S. S. Renner. 2013. Pollination and mating systems of
Apodanthaceae and the distribution of reproductive traits in parasitic
angiosperms. American Journal of Botany 100: 1083–1094.
Bendiksby, M., T. Schumacher, G. Gussarova, J. Nais, K. Mat-Salleh,
N. Sofiyanti, D. Madulid, S. A. Smith, and T. Barkman. 2010. Eluci-
dating the evolutionary history of the Southeast Asian, holoparasitic,
giant-flowered Rafflesiaceae: Pliocene vicariance, morphological
convergence and character displacement. Molecular Phylogenetics and
Evolution 57: 620–633.
Besana, G. M. and M. Ando. 2005. The central Philippine Fault Zone:
Location of great earthquakes, slow events, and creep activity. Earth,
Planets, and Space 57: 987–994.
Bickford, D., D. J. Lohman, N. S. Sodhi, P. K. L. Ng, R. Meier, K. Winker,
K. K. Ingram, and I. Das. 2007. Cryptic species as a window on di-
versity and conservation. Trends in Ecology & Evolution 22: 148–155.
Binks, R. M., M. A. Millar, and M. Byrne. 2015. Not all rare species are the
same: Contrasting patterns of genetic diversity and population
structure in two narrow-range endemic sedges. Biological Journal of the
Linnean Society. Linnean Society of London 114: 873–886.
Blanco, F. M. 1845. Flora de Filipinas. Ed. 2. M. Manila: Sanchez.
Bond, W. J. and P. Slingsby. 1983. Seed dispersal by ants in shrublands of
the Cape Province and its evolutionary implications. South African
Journal of Science 79: 231–233.
Byers, D. L. and T. R. Meagher. 1992. Mate availability in small populations
of plant species with homomorphic sporophytic self-incompatibility.
Heredity 68: 353–359.
Cornuet, J. M. and G. Luikart. 1997. Description and power analysis of two
tests for detecting recent population bottlenecks from allele frequency
data. Genetics 144: 2001–2014.
Costea, M. and S. Stefanovi´c. 2009. Cuscuta jepsonii (Convolvulaceae): An
invasive weed or an extinct endemic?. American Journal of Botany 96:
1744–1750.
Coyne, J. A. and H. A. Orr. 2004. Speciation: Reality and concepts. Pp. 9–25
in Speciation, eds. J. A. Coyne and H. A. Orr. Sunderland, Massa-
chussetts: Sinauer Associates, Inc.
Davis, C. C. and K. J. Wurdack. 2004. Host-to-parasite gene transfer in
flowering plants: Phylogenetic evidence from Malpighiales. Science
305: 676–678.
Davis, C. C., M. Latvis, D. L. Nickrent, K. J. Wurdack, and D. A. Baum.
2007. Floral gigantism in Rafflesiaceae. Science 315: 1812.
de Lange, P. J. 2003. Pittosporum serpentinum (de Lange) de Lange
(Pittosporaceae), a new species combination for an ultramafic en-
demic from North Cape, New Zealand. New Zealand Journal of Botany
41: 725–726.
Earl, D. A. and B. M. vonHoldt. 2012. STRUCTURE HARVESTER: A website
and program for visualizing STRUCTURE output and implementing
the Evanno method. Conservation Genetics Resources 4: 359–361.
Ellstrand, N. C. and D. R. Elam. 1993. Population genetic consequences of
small population size: Implications for plant conservation. Annual
Review of Ecology and Systematics 24: 217–242.
Emmons, L. H., J. Nais, and A. Briun. 1991. The fruit and consumers of
Rafflesia keithii (Rafflesiaceae). Biotropica 23: 197–199.
Endress, P. K., C. C. Davis, and M. L. Matthews. 2013. Advances in the
floral structural characterization of the major subclades of Malpigh-
iales, one of the largest orders of flowering plants. Annals of Botany 111:
969–985.
Evanno, G., S. Regnaut, and J. Goudet. 2005. Detecting the number of
clusters of individuals using the software structure: a simulation
study. Molecular Ecology 14: 2611–2620.
Faircloth, B. C. 2008. Msatcommander: Detection of microsatellite repeat
arrays and automated, locus-specific primer design. Molecular Ecology
Resources 8: 92–94.
Falush, D., M. Stephens, and J. K. Pritchard. 2003. Inference of population
structure: Extensions to linked loci and correlated allele frequencies.
Genetics 164: 1567–1587.
PELSER ET AL.: GENETIC DIVERSITY AND STRUCTURE OF RAFFLESIA LAGASCAE 5512017]
Falush, D., M. Stephens, and J. K. Pritchard. 2007. Inference of population
structure using multilocus genotype data: Dominant markers and null
alleles. Molecular Ecology Notes 7: 574–578.
Gagnepain, F. 1910. Tetrastigma (Amp´elidac´ees) nouveaux ou peu connus.
Notululae Systematicae 1: 261–271.
Galindon, J. M. M., P. S. Ong, and E. S. Fernando. 2016. Rafflesia consueloae
(Rafflesiaceae), the smallest among giants: A new species from Luzon
Island, Philippines. PhytoKeys 16: 37–46.
Hall, R. 1998. The plate tectonics of Cenozoic SE Asia and the distribution
of land and sea. Pp. 99–131 in Biogeography and Geological Evolution
of SE Asia, eds. R. Hall and J. D. Holloway. Leiden: Backhuys
Publishers.
Hall, R. 2002. Cenozoic geological and plate tectonic evolution of SE Asia
and the SW Pacific: Computer-based reconstructions, model and
animations. Journal of Asian Earth Sciences 20: 353–431.
Hidayati, S. N. and J. L. Walck. 2016. A review of the biology of Rafflesia:
What do we know and what’s next? Buletin Kebun Raya 19: 67–78.
Hieronymus, G. 1885. Ueber eine neue, von Dr. A. Schadenberg und
O. Koch auf S ¨ud-Mindanao entdeckte Art der Gattung Rafflesia.
Gartenflora 34: 3–7.
IUCN. 2012. IUCN red list categories and criteria: version 3.1. Ed. 2. Gland,
Switzerland and Cambridge, U. K.: IUCN.
Josephson, J. 2000. Going, going, gone? Plant species extinction in the 21st
Century. Environmental Science & Technology 34: 130A–135A.
Justesen, P. T. 1922. Morphological and biological notes on Rafflesia flowers
observed in the highlands of mid-Sumatra. Annales du Jardin Botanique
de Buitenzorg 32: 64–87.
Kopelman, N. M., J. Mayzel, M. Jakobsson, N. A. Rosenberg, and
I. Mayrose. 2015. CLUMPAK: A program for identifying clustering
modes and packaging population structure inferences across K.
Molecular Ecology Resources 15: 1179–1191.
Kuijt, J. 1969. The biology of parasitic flowering plants. Berkeley and Los
Angeles: University of California Press.
Laurance, W. F., D. Carolina Useche, L. P. Shoo, S. K. Herzog, M. Kessler,
F. Escobar, G. Brehm, J. C. Axmacher, I. C. Chen, L. A. G ´amez,
P. Hietz, K. Fiedler, T. Pyrcz, J. Wolf, C. L. Merkord, C. Cardelus,
A. R. Marshall, C. Ah-Peng, G. H. Aplet, M. del Coro Arizmendi,
W. J. Baker, J. Barone, C. A. Br ¨uhl, R. W. Bussmann, D. Cicuzza,
G. Eilu, M. E. Favila, A. Hemp, C. Hemp, J. Homeier, J. Hurtado,
J. Jankowski, G. Katt´an, J. Kluge, T. Kr ¨omer, D. C. Lees, M. Lehnert,
J. T. Longino, J. Lovett, P. H. Martin, B. D. Patterson, R. G. Pearson,
K. S. H. Peh, B. Richardson, M. Richardson, M. J. Samways, F. Senbeta,
T. B. Smith, T. M. A. Utteridge, J. E. Watkins, R. Wilson, S. E. Williams,
and C. D. Thomas. 2011. Global warming, elevational ranges and
the vulnerability of tropical biota. Biological Conservation 144: 548–557.
Lengyel, S., A. D. Gove, A. M. Latimer, J. D. Majer, and R. R. Dunn. 2009.
Ants sow the seeds of global diversification in flowering plants. PLoS
One 4: e5480.
Lengyel, S., A. D. Gove, A. M. Latimer, J. D. Majer, and R. R. Dunn. 2010.
Convergent evolution of seed dispersal by ants, and phylogeny and
biogeography in flowering plants: A global survey. Perspectives in
Plant Ecology, Evolution and Systematics 12: 43–55.
Madulid, D. A. and E. M. G. Agoo. 2008. (2007). On the identity of Rafflesia
manillana Teschem. (Rafflesiaceae). Philippine Scientist 44: 57–70.
Madulid, D. A., I. E. Buot, and E. M. G. Agoo. 2008. (2007). Rafflesia
panchoana (Rafflesiaceae), a new species from Luzon island, Philip-
pines. Acta Manilana 55: 43–47.
Marvier, M. A. and D. L. Smith. 1997. Conservation implications of host use
for rare parasitic plants. Conservation Biology 11: 839–848.
Medrano, M. and C. M. Herrera. 2008. Geographical structuring of genetic
diversity across the whole distribution range of Narcissus
longispathus, a habitat-specialist, Mediterranean narrow endemic.
Annals of Botany 102: 183–194.
Mittermeier, R. A., N. Myers, J. B. Thomsen, G. A. Da Fonseca, and
S. Olivieri. 1998. Biodiversity hotspots and major tropical wilderness
areas: Approaches to setting conservation priorities. Conservation
Biology 12: 516–520.
Meijer, W. 1997. Rafflesiaceae. Flora Malesiana Ser. I 13: 1–42.
Molina, J. E., K. M. Hazzouri, D. L. Nickrent, M. Geisler, R. S. Meyer,
M. M. Pentony, J. M. Flowers, P. B. Pelser, J. F. Barcelona,
S. A. Inovejas, I. Uy, W. Yuan, O. Wilkins, C.-I. Michel, S. LockLear,
G. P. Concepcion, and M. D. Purugganan. 2014. Possible loss of the
chloroplast genome in the parasitic flowering plant Rafflesia lagascae
(Rafflesiaceae). Molecular Biology and Evolution 31: 793–803.
Nais, J. 2001. Rafflesia of the world. Kota Kinabalu: Sabah Parks.
Narum, S. R. 2006. Beyond Bonferroni: Less conservative analyses for
conservation genetics. Conservation Genetics 7: 783–787.
Nickrent, D. L.,A. Blarer, Y.-L. Qiu, R. Vidal-Russell, and F. Anderson. 2004.
Phylogenetic inference in Rafflesiales: The influence of rate heteroge-
neity and horizontal gene transfer. BMC Evolutionary Biology 4: 40.
Nikolov, L. A., P. Endress, M. Sugumaran, S. Sasirat, S. Vessabutr,
E. M. Kramer, and C. C. Davis. 2013. Developmental origins of the
world’s largest flowers, Rafflesiaceae. Proceedings of the National
Academy of Sciences USA 110: 18578–18583.
Nybom, H. 2004. Comparison of different nuclear DNA markers for es-
timating intraspecific genetic diversity in plants. Molecular Ecology 13:
1143–1155.
Olah, L. 1960. Cytological and morphological investigations in Rafflesia
arnoldi R.Br. Bulletin of the Torrey Botanical Club 87: 406–416.
Ong, P. S., L. E. Afuang, and R. Rosell-Ambal. 2002. Philippine biodiversity
conservation priorities: a second iteration of the national biodiversity strategy
and action plan. Quezon City Philippines: DENR-Protected Areas and
Wildlife Bureau, CI-Philippines, UP CIDS and FPE.
Peakall, R. and P. E. Smouse. 2012. GenAlEx 6.5: Genetic analysis in Excel.
Population genetic software for teaching and research—An update.
Bioinformatics 28: 2537–2539.
Pelser, P. B., J. F. Barcelona, and D. L. Nickrent (eds.). 2011. onwards. Co’s
Digital Flora of the Philippines. www.philippineplants.org.
Pelser, P. B., D. L. Nickrent, and J. F. Barcelona. 2016. Untangling a vine and
its parasite: Host specificity of Philippine Rafflesia (Rafflesiaceae).
Taxon 65: 739–758.
Pelser, P. B., D. L. Nickrent, J. R. C. Callado, and J. F. Barcelona. 2013. Mt.
Banahaw reveals: The resurrection and neotypification of the name
Rafflesia lagascae (Rafflesiaceae) and clues to the dispersal of Rafflesia
seeds. Phytotaxa 131: 35–40.
Pritchard, J. K., M. Stephens, and P. Donnelly. 2000. Inference of population
structure using multilocus genotype data. Genetics 155: 945–959.
Ramamoorthy, R., E. E.-K. Phua, S.-H. Lim, H. T.-W. Tan, and P. P. Kumar.
2013. Identification and characterization of RcMADS1,anAGL24
ortholog from the holoparasitic plant Rafflesia cantleyi Solms-Laubach
(Rafflesiaceae). PLoS One 8: e67243.
Raymond, M. and F. Rousset. 1995. GENEPOP (version 1.2): Population
genetics software for exact tests and ecumenicism. The Journal of
Heredity 86: 248–249.
Renner, S. and J. Feil. 1993. Pollinators of tropical dioecious angiosperms.
American Journal of Botany 80: 1100–1107.
Renner, S. and R. E. Ricklefs. 1995. Dioecy and its correlates in the flowering
plants. American Journal of Botany 82: 596–606.
Rozen, S. and H. Skaletsky. 1999. Primer3 on the WWW for general users
and for biologist programmers. Pp. 365–386 in Methods in molecular
biology 132: Bioinformatics methods and protocols, eds. S. Misener and S.
A. Krawetz. Totowa, New Jersey: Humana Press.
Sandner, T. M. and D. Matthies. 2017. Interactions of inbreeding and stress
by poor host quality in a root hemiparasite. Annals of Botany 119:
143–150.
Schmidt, K. and K. Jensen. 2000. Genetic structure and AFLP variation of
remnant populations in the rare plant Pedicularis palustris (Scrophu-
lariaceae) and its relation to population size and reproductive com-
ponents. American Journal of Botany 87: 678–689.
Sch¨onrogge, K., B. Barr, J. C. Wardlaw, E. Napper, M. G. Gardner, J. Breen,
G. W. Elmes, and J. A. Thomas. 2002. When rare species become
endangered: Cryptic speciation in myrmecophilous hoverflies.
Biological Journal of the Linnean Society. Linnean Society of London 75:
291–300.
Shepherd, K. A., K. R. Thiele, J. Sampson, D. Coates, and M. Byrne. 2015. A
rare, new species of Atriplex (Chenopodiaceae) comprising two ge-
netically distinct but morphologically cryptic populations in arid
Western Australia: implications for taxonomy and conservation.
Australian Systematic Botany 28: 234–245.
Stanton, S., O. Honnay, H. Jacquemyn, and I. Rold ´an-Ruiz. 2009. A
comparison of the population genetic structure of parasitic Viscum
album from two landscapes differing in degree of fragmentation. Plant
Systematics and Evolution 281: 161–169.
Teijsmann, J. E. 1856. Nadere bijdrage tot de kennis van de voortteling van
Rafflesia arnoldii R.Br. in ’s Lands Plantentuin te Buitenzorg. Natuur-
kundig Tijdschrift voor Nederlands Indi¨e 12: 277–281.
Teschemacher, J. E. 1844. On a new species of Rafflesia from Manilla. Boston
Journal of Natural History 4: 63–66.
Thiers, B. 2017. [continuously updated]. Index Herbariorum: a global di-
rectory of public herbaria and associated staff. New York Botanical
Garden’s Virtual Herbarium. http://sweetgum.nybg.org/science/ih/.
SYSTEMATIC BOTANY [Volume 42552
VanOosterhout,C.,W.F.Hutchinson,D.P.M.Wills,andP.Shipley.
2004. Micro-checker: Software for identifying and correcting gen-
otyping errors in microsatellite data. Molecular Ecology Notes 4:
535–538.
Wicaksono,A.,S.Mursidawati,L.A.Sukamto,andJ.A.TeixeiradaSilva.
2016. Rafflesia spp.: Propagation and conservation. Planta 244:
289–296.
Williams, E. W., R. Cheung, C. Siegel, M. Howards, J. Fant, and K. Havens.
2016. Persistence of the gypsophile Lepidospartum burgessii (Aster-
aceae) through clonal growth and limited gene flow. Conservation
Genetics 17: 1201–1211.
Xi, Z., R. Bradley, K. Wurdack, K. M. Wong, M. Sugumaran, K. Bomblies,
J. Rest, and C. C. Davis. 2012. Horizontal transfer of expressed genes
in a parasitic flowering plant. BMC Genomics 13: 227.
Xi, Z., Y. Wang, R. K. Bradley, M. Sugumaran, C. J. Marx, J. S. Rest, and
C. C. Davis. 2013. Massive mitochondrial gene transfer in a parasitic
flowering plant clade. PLOS Genetics 9: e1003265.
APPENDIX 1. Voucher information. Herbarium acronyms follow Thiers
(2017).
R. lagascae: Aurora Memorial Natural Park, Aurora Prov., J.R. Callado
382 (CANU, PNH); Bolos Point, Cagayan Prov., J.F. Barcelona 3600 with
J. Payba, R. Echanique & Tabuc (CAHUP, SIU); Burgos, Nueva Ecija Prov.,
J.F. Barcelona 4081 (PNH); Maria Aurora, Aurora Prov., J.F. Barcelona 4033
(PNH); Mt. Banahaw, Quezon Prov., J.F. Barcelona 3819 with P.B. Pelser
(CANU); Mt. Irid, Quezon Prov., J.F. Barcelona 3657 with Joel Sarmiento
(CAHUP, SIU); Mt. Labo. Camarines Norte/Camarines Sur Prov., J.F.
Barcelona 3849 with P.B. Pelser (CANU, PNH); Mt. Makiling, Laguna Prov.,
J.F. Barcelona 3648 with D.L. Nickrent, E. Malinao & R. Breva (CAHUP, SIU);
Mt. Malinao, Camarines Sur Prov., J.F. Barcelona 4046 (CANU); Mt.
Mingan, Aurora Prov., J.F. Barcelona 3867 with P.B. Pelser (PNH); Mt. Natib,
Bataan Prov., F. Barcelona 3563 with L.L. Co (CAHUP, SIU); Salazar, Nueva
Ecija Prov., J.F. Barcelona 4055 (PNH). R. manillana:Basey, Samar Prov., J.F.
Barcelona 3734 et al. (CANU, PNH).
PELSER ET AL.: GENETIC DIVERSITY AND STRUCTURE OF RAFFLESIA LAGASCAE 5532017]
