Polyphyly of Xylosandrus Reitter inferred from nuclear and mitochondrial genes (Coleoptera: Curculionidae: Scolytinae).

Abstract

The Xyleborina ambrosia beetle genus Xylosandrus contains 54 species, several of which are of economic importance. The monophyly of the genus was tested using a data set comprised of multiple gene loci: 28S rDNA; the mitochondrial gene cytochrome oxidase I (COI); and the nuclear genes arginine kinase (ArgK), rudimentary (CAD), and Elongation Factor 1alpha (EF-1alpha). The nuclear protein-coding genes CAD and ArgK were used for the first time in phylogenetics of Scolytinae. Analyses were performed using Parsimony and Bayesian optimality criteria. Our analyses included 43 specimens representing 15 Xylosandrus species and 20 species from Amasa, Anisandrus, Cnestus, Euwallacea and Xyleborus, and two species from the outgroup genus Coccotrypes. All analyses recovered a polyphyletic Xylosandrus. Several species of Xylosandrus were consistently placed in clades with the genera Anisandrus and Cnestus with high support values (100% bootstrap support). Among these, was the economically important invasive species X. mutilatus, which was consistently recovered as part of the "Cnestus" clade. In our analyses, both CAD and ArgK demonstrated phylogenetic utility across varying nodal depths. Despite the selection of genes with signals at complementary phylogenetic depths, the data set used herein did not resolve the phylogeny of Xylosandrus and related genera. Since the taxon sample available for molecular work represents only a fraction of Xylosandrus species, a complete revision that combines molecular and morphological data in a total evidence approach is recommended for the genus.

Full text

Page 1

Polyphyly of Xylosandrus Reitter inferred from nuclear and mitochondrial
genes (Coleoptera: Curculionidae: Scolytinae)
Stephanie A. Dolea,*, Bjarte H. Jordalb, Anthony I. Cognatoa
aDepartment of Entomology, Michigan State University, 243 Natural Science, East Lansing, MI 48824, USA
bThe Natural History Museum, University of Bergen, Mesèplass 3, NO-5007, Bergen, Norway
a r t i c l ei n f o
Article history:
Received 19 March 2009
Revised 12 November 2009
Accepted 12 November 2009
Available online 17 November 2009
Keywords:
Coleoptera
Scolytinae
Scolytidae
Xyleborina
Xyleborini
Ambrosia beetle
Molecular systematics
Phylogenetics
Gene sampling
a b s t r a c t
The Xyleborina ambrosia beetle genus Xylosandrus contains 54 species, several of which are of economic
importance. The monophyly of the genus was tested using a data set comprised of multiple gene loci: 28S
rDNA; the mitochondrial gene cytochrome oxidase I (COI); and the nuclear genes arginine kinase (ArgK),
rudimentary (CAD), and Elongation Factor 1a (EF-1a). The nuclear protein-coding genes CAD and ArgK
were used for the first time in phylogenetics of Scolytinae. Analyses were performed using Parsimony
and Bayesian optimality criteria. Our analyses included 43 specimens representing 15 Xylosandrus species
and 20 species from Amasa, Anisandrus, Cnestus, Euwallacea and Xyleborus, and two species from the out-
group genus Coccotrypes. All analyses recovered a polyphyletic Xylosandrus. Several species of Xylosandrus
were consistently placed in clades with the genera Anisandrus and Cnestus with high support values (100%
bootstrap support). Among these, was the economically important invasive species X. mutilatus, which
was consistently recovered as part of the ‘‘Cnestus” clade. In our analyses, both CAD and ArgK demon-
strated phylogenetic utility across varying nodal depths. Despite the selection of genes with signals at
complementary phylogenetic depths, the data set used herein did not resolve the phylogeny of Xylosan-
drus and related genera. Since the taxon sample available for molecular work represents only a fraction of
Xylosandrus species, a complete revision that combines molecular and morphological data in a total evi-
dence approach is recommended for the genus.
? 2009 Elsevier Inc. All rights reserved.
1. Introduction
Xylosandrus Reitter (1913) is a large genus of xyleborine ambro-
sia beetles with a widespread distribution primarily in tropical, but
also in temperate regions of the world. In their worldwide catalog
of the Scolytinae, Wood and Bright (1992) list 52 species of Xylos-
andrus. Subsequent descriptions, new synonymies, and new com-
binations have brought the present number to 54 species (Bright
and Skidmore, 1997; Saha et al., 2002; Wood, 2007; Dole and Bea-
ver, 2008).
Several Xylosandrus species cause economic losses in nursery
and agricultural settings in their native and introduced ranges. In
Brazil, X. compactus causes losses in several economically impor-
tant host species, including avocado, cacao, coffee, and mango (Oli-
veira, 2008). In North America, three Xylosandrus species (X.
compactus, X. crassiusculus, and X. germanus) have caused ‘‘consid-
erable economic damages” since their introductions (Oliver and
Mannion, 2001). A revised understanding of Xylosandrus generic
boundaries and species diagnosis would aid in the identification
and control of these economically important beetles.
Xylosandrus beetles belong to the weevil subfamily Scolytinae
(Coleoptera: Curculionidae), which is comprised of 26 tribes con-
taining approximately 225 genera and 6000 species worldwide
(Wood and Bright, 1992; Mandelshtam and Beaver, 2003). Due to
the present classification of Scolytinae as a subfamily of the Curcu-
lionidae (Marvaldi, 1997; Kuschel et al., 2000; McKenna et al.,
2009) the previously recognized tribe Xyleborini is herein treated
as the subtribe Xyleborina. Although the Curculionoidea are a rel-
atively young group (152 MYO), discovery of a fossilized scolytine
from Cretaceous amber dates the origin of the subfamily to at least
100 million years ago (Grimaldi and Engel, 2005; Cognato and
Grimaldi, 2009). This long history likely contributes to the diversity
of Scolytines which occupy two ecological groups: phloem-feeding
bark beetles and fungus-feeding ambrosia beetles. The scolytine
subtribe Xyleborina contains approximately 1300 described spe-
cies and constitutes one of the largest radiations of ambrosia bee-
tles (Wood and Bright, 1992; Jordal, 2002). Xyleborina are absent
in the Dominican amber fossil record, suggesting that their radia-
tion began in the Miocene (Bright and Poinar, 1994; Jordal et al.,
2000). Xyleborines are particularly suited for the invasion of new
habitats due to their haplodiploidy mating system, their tendency
to inbreed, and their general lack of host specificity (Beaver, 1989;
Wood and Bright, 1992; Kirkendall, 1993; Normark et al., 1999).
1055-7903/$ - see front matter ? 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.ympev.2009.11.011
* Corresponding author. Fax: +1 415 379 5715.
E-mail address: stephanie@beetlelady.com (S.A. Dole).
Molecular Phylogenetics and Evolution 54 (2010) 773–782
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution
journal homepage: www.elsevier.com/locate/ympev

Page 2

Historically, the classification of the Xyleborina has been chaotic,
with Wood (1986) describing his own classification of the tribe as
‘‘tentative and flawed.” However, recent taxonomic work has begun
to correct this, as the subtribe is finally being studied within a phy-
logeneticcontext(Jordaletal.,2000;Jordal,2002;Hulcretal.,2007).
The genus Xylosandrus currently contains species with highly
variable morphologies, several of whichare similarto those of other
genera, leading to confusion concerning the generic boundaries of
Xylosandrus and its relationship to and distinction from the genera
AmasaLea, Anisandrus Ferrari, andCnestus Sampson.The underlying
cause of this confusion has been the incorrect placement of species
within Xylosandrus, which has blurred the boundaries between it
andothergenera(DoleandBeaver,2008).Commonlyciteddiagnos-
tic characters for Xylosandrus include widely separated procoxae;
very stout, cylindrical bodies; and antennal club that is obliquely
truncatewiththefirstsegmentoftheantennalclubformingacircu-
lar costa and with segments two and three not visible on posterior
face of the club (Bright, 1968; Wood, 1986; Hulcr et al., 2007; Dole
and Beaver, 2008). However, even the frequently cited diagnostic
character of widely separated procoxae is not a synapomorphy for
Xylosandrus and the genus is presently defined by a collection of
homoplastic xyleborine characters (Hulcr et al., 2007).
Both morphological (Dole and Beaver, 2008) and molecular data
suggest that Xylosandrus is not monophyletic (Jordal, 2002). A cla-
distic review of the generic taxonomic characters of Xyleborina
recovered a monophyletic Xylosandrus, but it is important to note
that this study did not include species with morphologies that
deviate from the sensu stricto concept of the genus (Hulcr et al.,
2007). In their review of Australian Xylosandrus, Dole and Beaver
(2008) contributed to the revision of Xylosandrus by defining char-
acters that distinguish it from Cnestus and moving two species
from Xylosandrus to Cnestus. This was only a beginning, given the
narrow geographic perspective of their review. Xylosandrus is in
need of a comprehensive phylogenetic revision that includes other
suspect genera, such as Amasa, Anisandrus, and Cnestus. The preva-
lence of homoplastic morphological characters within Xylosandrus
and related genera necessitates the inclusion of alternative charac-
ters, such as DNA, to help reconstruct the phylogeny of this group.
Molecular studies of scolytine phylogenetics have been based
mainly on mitochondrial genes (16S, COI), the ribosomal-encoding
genes 28S and 18S, and the nuclear protein encoding genes elonga-
tion factor 1a (EF-1a) and Enolase (Normark et al., 1999; Jordal
et al., 2000, 2008; Cognato and Vogler, 2001; Farrell et al., 2001; Jor-
dal,2002;JordalandHewitt,2004).Ofthesegenes,EF-1ahasshown
the greatest potential for resolving the phylogenetic relationship of
Scolytinae (Jordal, 2002, 2007). However, use of these genes alone
has not resolved a Scolytinae phylogeny. Recognizing the underuti-
lized potential of nuclear protein-coding genes, Jordal (2007)
screened multiple gene loci for the phylogenetic reconstruction of
scolytines. The nuclear genes CAD, also known as rudimentary, and
argininekinase(ArgK)werefoundtobethemostpromisingandwere
recommended for use in future phylogenetic studies of Scolytinae.
Here, we use the CAD and ArgK genes for the first time for phy-
logenetic inference in the Scolytinae. Using a data set comprised of
multiple gene loci (28S, ArgK, CAD, COI, and EF-1a) we construct a
phylogenetic hypothesis of the species relationships within Xylos-
andrus, test the monophyly of the genus, and test the relationships
among Xylosandrus and the genera Amasa, Anisandrus, and Cnestus.
2. Materials and methods
2.1. Taxa, DNA sequences, and alignment
We included 43 specimens (Table 1) representing 15 Xylosan-
drus species, 20 species from the genera Amasa, Anisandrus, Cnestus,
Euwallacea, Xyleborus, and two species from the outgroup genus
Coccotrypes. Species from a diverse group of genera were chosen
in order to test the generic limits of Xylosandrus.
DNA was extracted from dissected ethanol preserved thoraces
following protocols of Cognato and Vogler (2001) and the remain-
ing body parts were pinned and vouchered at the A.J. Cook Arthro-
pod Research Collection, Michigan State University. In some cases
DNA was also successfully extracted from dried specimens (X.
monteithi and X. queenslandi). Using the purified DNA, partial gene
regions of mitochondrial cytochrome oxidase I, nuclear ribosomal
28S (D2 and D3 regions), elongation factor-1a, CAD (rudimentary),
and Arginine Kinase (ArgK) were amplified with PCR primers listed
in Table 2. CAD and ArgK sequences were first generated with
degenerate primers designed based on bee sequences (Danforth
et al., 2006; Jordal, 2007). We used these sequences to subse-
quently create new xyleborine primers (Table 2). PCR cocktails
for COI, EF-1a, and 28S followed Cognato and Vogler (2001) and
cocktails for CAD and ArgK followed Jordal (2007). PCR was per-
formed on a thermal cycler (MJ Research, Waltham, MA) under
the following conditions: one cycle for 2 min at 95 ?C (15 min for
Hotstar Taq), 34 cycles for 1 min at 95 ?C, 45 s at 45–55 ?C (see
Table 2 for specific annealing temperatures), 1 min at 72 ?C, and
a final elongation cycle of 5 min at 72 ?C. PCR products were puri-
fied using ExoSAP-IT following the manufacturer’s protocols (USB
Corp., Cleveland, OH). The clean PCR reactions were directly
sequenced using BigDye?Terminator v.1.1 (Applied Biosystems,
Foster City, CA).
Sense and antisense DNA sequences were compiled with the
computer software Sequencer (Ann Arbor, MI), in order to inspect
for ambiguities and create consensus sequences. No sequence
length variation was observed among the protein-coding genes
for the included taxa, however considerable variation occurred in
the 28S sequences. These sequences were aligned by two methods;
manually with reference to a scolytine-specific secondary structure
model (Jordal et al., 2008) and by a multiple progressive pair-wise
alignment with secondary refinement using the computer soft-
ware, MUSCLE v.3.52 (Edgar, 2004). We did not specify stems
and loop regions for the secondary structure alignment. Instead,
we used the scolytid secondary structure model to guide the align-
ment by identifying conserved and length variable regions. For the
other alignment, we used the default settings of the web version of
MUSCLE (http://www.ebi.ac.uk/Tools/muscle/index.html). Nexus
files of both alignments are available at http://www.hisl.ent.msu.
edu/research/publications.php.
2.2. Phylogenetic analyses
We conducted several analyses, which used different optimality
criteria, alignments, and accounted for potential bias in nucleotide
substitution patterns. Parsimony analyses of data aligned by sec-
ondary structure and the MUSCLE software were conducted with
the software PAUP* (Swofford, 2002). For all parsimony analyses,
most-parsimonious reconstructions were obtained by a heuristic
search with 300 random stepwise addition replicates using PAUP*
default settings. Bootstrap values were determined by performing
1000 pseudo-replicates. Bremer support values were calculated by
constructing constraint trees with the software TreeRot (Sorensen,
1996) followed by subsequent analysis with PAUP. For analyses of
both alignments, we considered gap positions as missing data and
as a 5th character state because previous studies have showed that
gap positions are phylogenetically informative (Cognato and Vog-
ler, 2001; Lee, 2001).
Several researchers have proposed direct optimization of nucle-
otide homology during phylogenetic reconstruction given that sta-
tic nucleotide alignments may not accurately reflect positional
homology of nucleotides for all taxa (Kruskal, 1983; Sankoff and
774
S.A. Dole et al./Molecular Phylogenetics and Evolution 54 (2010) 773–782

Page 3

Cedergren, 1983; Wheeler, 1996). We performed direct optimiza-
tion on the data using the program POY ver. 4.0.2881 (Wheeler,
1996; Varón et al., 2008) in order to evaluate the relative signal
within length variable regions. We used the alignment guided by
secondary structure to divide the 28S data into seven input files,
which corresponded to conserved and length variable regions.
The other gene sequences were kept as separate input files. These
files were input into POY and analyzed under a parsimony
optimality criterion using the following tree search commands:
transform(tcm:(1,2),gap_opening:1), build (300), swap (threshold:
5.0), select (), perturb(transform(static_approx),iterations:15,ratchet:
(0.2,3)), select(), fuse(iterations:200,swap()), select(), report (trees:
(total)). An additional analysis considering gap cost of 3 was
conducted with the same tree search strategy as above. Bremer
support values were calculated for the tree recovered in the first
analysis using the following command: calculate_support(bremer,
build(trees:0), swap(tbr, trees:2)). Data and command files are avail-
able at (http://www.hisl.ent.msu.edu/research/publications.php).
We analyzed the data under maximum likelihood using Bayes-
ian estimation of phylogeny as implemented in Mr. Bayes 3.1.2
(Hulsenbeck and Ronquist, 2001) in order to evaluate the effect
of nucleotide substitution models on phylogenetic reconstruction.
We conducted a separate analysis for each of the static alignments.
For both analyses, we partitioned the protein-coding genes by co-
don resulting in 13 partitions thus allowing each to independently
evolve under a general time reversal (GTR) model with a propor-
tion of invariant sites and a gamma distribution. Four Metropo-
lis-Coupled Markov chain Monte Carlo searches (1 cold, 3
heated) were performed twice for 20 ? 106generations each with
sampling every 100th iteration. Approach to stationarity (‘‘burn-
in”) of each search was determined with the graphical interface
of Tracer v. 1.4 (Rambaut and Drummond, 2007) (graph not
shown). All parameters reached stability within 2,000,000 genera-
tions and parameters between runs did not vary. Bayesian poster-
ior probabilities were calculated by a majority-rule consensus of
those trees after the burn-in (for both runs, 180,000 trees).
In order to assess the contribution of support of each gene to the
total support of the tree, the dependency of partitioned Bremer
support (PBS) and nodal distance (the branch length from a given
node to the tip of the tree) were measured for each gene (Baker
et al., 1998; Cognato and Vogler, 2001). In PAUP*, maximum likeli-
hood was used to fit branch lengths to a molecular clock in order to
Table 1
Specimens sequenced, collecting data, and hosts.
SpeciesCollection locality, date and collector Host
Amasa bicostatus (Sampson)
Amasa resectus (Eggers)
Amasa schlichi (Stebbing)
Borneo, Danum Valley, June 2006, Hulcr et al. Coll.
Papua New Guinea, Wannang, March 2006, Hulcr et al. Coll.
Thailand, Kanchanburi Prov., Thong Pha Phoom District, Phu Yae Subdistrict, el. 400,
N14.944? E98.674?. Cognato, Gillogly & Harlin Coll.
Papua New Guinea, Wannang, March 2006, Hulcr et al. Coll.
Borneo, Danum Valley, June 2006, Hulcr et al. Coll.
Madagascar, Réserve Spéciale de Bemarivo, 23.8 km 223? SW Besalampy,
16?55’30”S 044?22’06”E, 19-23 Nov 2002, EF26 at light, B. Fisher
Papua New Guinea, Wannang, March 2006, Hulcr et al. Coll.
Papua New Guinea, Mu, March 2006, Hulcr et al. Coll.
Russia: St. Petersburg. Pushkin. June 5, 2004. M. Mandelshtam Coll.
Thailand, Doi Pui, October 2004, S. Sonthichai Coll.
USA: OH, Hocking Co., Logan, Ethanol trap. May 1, 2004. R. Rabaglia Coll.
USA: MD, Anne Arundel Co. Annapolis. Ethanol trap. April 22, 2005. R. Rabaglia Coll.
Papua New Guinea, Madang Prov., Ohu Village, 2004, Hulcr et al. Coll.
Papua New Guinea, Ohu, March 2006, Hulcr et al. Coll.
USA: Hawaii Oahu. Manoa Falls Trail. N 21.337844, W 157.803558. Ethanol trap August 7,
2007. D. Rubinoff and M. San Jose Coll.
Papua New Guinea, Madang Prov., Ohu Village, 2004, Hulcr et al. Coll.
Papua New Guinea, Mu, March 2006, Hulcr et al. Coll.
Costa Rica, La Selva research station, April 2005, Hulcr Coll.
USA: MD, Anne Arundel Co. Annapolis. Ethanol trap. April 22, 2005. R. Rabaglia Coll.
USA: MD, Anne Arundel Co. Annapolis. Ethanol trap. April 15, 2005. R. Rabaglia Coll.
Papua New Guinea, Mu, March 2006, Hulcr et al. Coll.
Borneo, Danum Valley, June 2006, Hulcr et al. Coll.
Brazil: Espírito Santo Prov. September 26, 2006. P. Sergio Voppi Coll.
Ghana: Ankasa. 2 July 2005. J. Hulcr Coll.
Thailand: Chiangmai Prov. Doi Inthanon. July 28, 2004. A.I. Cognato Coll.
Madagascar: Montagne d’Akirindro 7.6 km 341? NNW Ambinanitelo. 15?17’18”S,
049?32’54”E. March 17-21, 2003. B. Fisher et al. Coll.
USA: MD, Anne Arundel Co. Annapolis. Reared from wood. July 11, 2003. P. Merthel Coll.
USA: North Carolina: Henderson Co. August 8, 2004. P. Merthal Coll.
Thailand: Chiangmai Prov., Mae Rim, June 12, 2005, Dole and Beaver Coll.
USA: MD, Anne Arundel Co. Annapolis. Ethanol trap. April 22, 2005. R. Rabaglia Coll.
USA: Michigan: Ingham Co. East Lansing. Ethanol trap. May 7, 2007 A.I. Cognato Coll.
Borneo, Danum Valley, June 2006, Hulcr et al. Coll.
Thailand: Chumphon Prov., Khao Sok NP, March 24, 2006 Dole et al. Coll.
Australia, Queensland, Palmerston, Henrietta Cr., 550 m, Watchua Falls, 24.I.2000, B. Jordal Coll.
Papua New Guinea, Madang, March 2006, Hulcr et al. Coll.
Ecuador, Napo Prov., Res. Ethnica Waorani, 1 km S. Onkone Gare Camp, Trans. Ent.,
00?39’10”S, 076?26’W, January 2006, T. L. Erwin et al. Coll.
Singapore, Bukit Timah, 50 m, 25-27 Oct 1998, B. H. Jordal Coll.
USA: Mississippi: Oktibbeha Co. 3 mi. W. of Adaton. April 23-26, 2004 T.L Schiefer Coll.
Australia, Queensland, Bunya Mountain NP, 1,100 m, 19.i.2000, B. Jordal & A. Sequeria Coll.
Papua New Guinea, Wannang, March 2006, Hulcr et al. Coll.
Thailand, Chiangmai Prov., Pong Yaeng, 900m a.s.l. July 2005 Dole and Hulcr Coll.
Papua New Guinea, Mu, March 2006, Hulcr et al. Coll.
Borneo, Danum Vallery, June 2006, Hulcr et al. Coll.
Amasa striatotruncatus (Schedl)
Amasa versicolor (Sampson)
Amasa n. sp. 1
Amasa n. sp. 301
Amasa n. sp. 349
Anisandrus dispar (Fabricius)
Anisandrus hirtus (Hagedorn)
Anisandrus obesus (LeConte)
Anisandrus sayi (Hopkins)
Cnestus bimaculatus (Eggers)
Cnestus pseudosuturalis Schedl
Coccotrypes dactyliperda (Fabricius)
Ficus nodosa (twig)
Litsea Lam.
Coccotrypes longior (Eggers)
Euwalllacea russulus (Schedl)
Xyleborus affinis Eichhoff
Xyleborus californicus Wood
Xyleborus pelliculosus Eichhoff
Xyleborus rotundicollis Browne
Xylosandrus ater (Eggers)
Xylosandrus compactus (Eichhoff)
Xylosandrus compactus
Xylosandrus crassiusculus (Motschulsky)
Xylosandrus crassiusculus
Mallotus Lour.
Xylosandrus crassiusculus
Xylosandrus crassiusculus
Xylosandrus discolor (Blandford)
Xylosandrus germanus (Blandford)
Xylosandrus germanus
Xylosandrus improcerus (Sampson)
Xylosandrus mancus (Blandford)
Xylosandrus monteithi Dole & Beaver
Xylosandrus morigerus (Blandford)
Xylosandrus morigerus
Mango
Burseraceae sp. (twig)
Rambutan branch
Xylosandrus morigerus
Xylosandrus mutilatus (Blandford)
Xylosandrus queenslandi Dole & Beaver
Xylosandrus ursa (Eggers)
Xylosandrus ursinus (Hagedorn)
Xylosandrus n. sp.
Xylosandrus n. sp.
Leguminosae sp. (tree)
S.A. Dole et al./Molecular Phylogenetics and Evolution 54 (2010) 773–782
775

Page 4

scale branch lengths throughout the most-parsimonious tree found
in analysis 1 (Table 3). Maximum likelihood estimations were per-
formed using the following settings: general time-reversible mod-
el, basefrequenciesdetermined
invariable sites estimated, among-site rate variation approximated
to a gamma distribution with four rate categories and shape
parameter = 0.5, and molecular clock enforced.
empirically,proportion of
3. Results
3.1. Phylogenetics of Xylosandrus
All analyses recovered a polyphyletic Xylosandrus, including
species from three different genera: Xylosandrus, Anisandrus, and
Cnestus. (Figs. 1 and 2, Table 3). Although tree topologies differed
some, depending on analysis type, alignment parameters, and
treatment of gaps, the following clades were consistently recov-
ered with high support values: ‘‘Anisandrus”, ‘‘Cnestus”, and ((X.
germanus + X. n. sp. Borneo) X. morigerus, X. compactus) (Table 3).
The varied placement of the ‘‘Anisandrus” and ‘‘Cnestus” clades,
along with the placement of several Xylosandrus species within
these clades, was responsible for rendering the genus polyphyletic.
However, the clade containing Xylosandrus morigerus, the type spe-
cies of the genus had bootstrap support of 100% in all analyses for
which bootstrap support was calculated and Bremer supports
ranging from 21 to 28. Likewise, this clade was recovered by the
Bayesian analyses with a posterior probability of 100. The phyloge-
netic placement of ‘‘Anisandrus” and ‘‘Cnestus” changed, depending
on the parameters of the analysis performed, and the various
placements were usually weakly supported (e.g.650% bootstrap
support). While the phylogenetic position of these clades was
weakly supported, the placement of several species of Xylosandrus
within the ‘‘Anisandrus” and ‘‘Cnestus” clades had very strong sup-
port (e.g. 100% bootstrap support). X. ater, X. improcerus and X.
mutilatus were all placed with traditional Cnestus species, with
strongly support (e.g. Fig. 1). The same relationship to Anisandrus
species was found for X. ursa and X. ursinus. Three additional clades
were recovered by multiple analyses with varying levels of
support, but were not consistently present on all of the tree topol-
ogies: ‘‘Cnestus” + X. crassiusculus, X. n. sp. PNG + X. monteithi,
X. mancus + X. discolor (Table 3).
Several Xylosandrus species included in this analysis varied in
their phylogenetic placement: X. crassiusculus, X. discolor, X. man-
cus, X. monteithi, X. queenslandi, and X. n. sp. Papua New Guinea.
These species did not form a larger monophyletic Xylosandrus clade
along with the clade containing the type species, X. morigerus.
However, none of the various placements of these species were
supported by any of the phylogenetic analyses.
3.2. Gene utility
Homoplasy as measured by Consistency Indices (CI) and Reten-
tion Indices (RI) indicated that homoplasy was the lowest for 28S
of all the data partitions (Table 4). However, CAD, ArgK, and EF-
1a, had almost as low CI’s (0.450–0.486) and RI’s (0.631–0.677).
Bremer supports were affected by the alignment method used
Table 2
PCR primers and annealing temperatures used for the amplification of gene sequences.
GenePrimerPrimer sequenceAnnealing Temp. (?C)Ampicon size (bp) Reference
COILCO 1490
HC02198
50-GGTCAACAAATCATAAAGATATTGG-30
50-TAAACTTCAGGGTGACCAAAAAATCA-30
50
50
630
650
Hebert et al. (2003)
Hebert et al. (2003)
28S D2F1
D3R2
3665
4048
50-ACTGTTGGCGACGATGTTCT-30
50-TCTTCGCCCCTATACCC-30
50-AGACAGAGTTCAAGAGTACGTG-30
50-TTGCTCCGTGTTTCAAGACGGG-30
50or 55
50 or 55
50 or 55
50 or 55
500–570
500–570
600–750
600–750
Jordal et al. (2008)
Jordal et al. (2008)
Jordal et al. (2008)
Jordal et al. (2008)
EF-la
ets149
efa754
eflafor 1
eflarev 1
50-ATCGAGAAGTlCGAGAAGGAGGCYCARGAAATGGG-30
50-CCACCAATTTTGTAGACATC-30
50-TACGTAACCATCATTGATGCTYCC-30
50-CCTTCTTTACGTTCAATGGACCATCC-30
48
48
50 or 55
50 or 55
585
585
500
500
Normark et al. (1999)
Normark et al. (1999)
This Study
This Study
ArgK forB2
rev B2
50-GAYTCCGGWATYGGWATCTAYGCTCC-30
50-GTATGYTCMCCRCGRGTACCACG-30
58
58
620
620
This Study
This Study
CAD apCADforB2
apCADrevlmod
50-TGGAARGARGTBGARTACGARGTGGYCG-30
50-GCCATYRCTCBCCTACRCTYTTCAT-30
58
58
740
740
Danforth et al. (2006)
Danforth et al. 2006
Table 3
Support found for selected clades found by different analyses of the five gene dataset. Numbers represent bootstrap/Bremer supports or posterior probabilities for Bayesian
analyses. NA = not applicable or not found in the resulting tree(s).
Analysis type
Number of most-parsimonious trees
1
1
2
3
3
1
4
2
5
2
6
NA
7
NA
Clades
‘‘Anisandrus”
‘‘Cnestus ‘‘
‘‘Cnestus ‘‘ clade + X crassiusculus
‘‘Anisandrus” + ‘‘Cnestus ‘‘
X. n .sp. PNG + X. monteithi
(((X. germanus + X. n. sp. Borneo) X. morigerus) X. compactus)
X. mancus + X. discolor
((X. mancus + X. discolor) X crassiusculus)
((X. mancus + X. discolor) ‘‘Cnestus”)
Xylosandrus
100/32
100/60
<50/5
N/A
97/21
100/24
68/9
N/A
N/A
N/A
100/47
100/68
98/20
N/A
100/26
100/24
<50/18
N/A
N/A
N/A
100/22
100/59
N/A
<50/1
N/A
100/21
65
<50/2
N/A
N/A
100/53
100/53
N/A
N/A
60/7
100/22
<50/74
N/A
<50/74
N/A
N/A/67
N/A/120
N/A
N/A
N/A
N/A/28
N/A/9
N/A
N/A
N/A
100
100
100
N/A
100
100
100
N/A
N/A
N/A
100
100
100
N/A
N/A
100
100
N/A
N/A
N/A
(1) Parsimony/stalk alignment/secondary structure/gaps missing. (2) Parsimony/static alignment/secondary structure/gaps 5th character. (3) Parsimony/static alignmtenl/
MlJSCl /gaps missing. (4) Pasimony/static alignment/MUSCLE/gaps 5th character. (5) Parsimony/dynamic alignment/2 gap cost. (6) Bayesian/static alignment/secondarv
structure/codons + 28s partitions. (7) Bayesian/static aliginment/MUSCLE:/codons + 28s partitions.
776
S.A. Dole et al./Molecular Phylogenetics and Evolution 54 (2010) 773–782

Page 5

and whether gaps were treated as missing data or as a 5th charac-
ter in the analysis. Overall, 28S gave the highest Bremer support
values (Table 5). Static alignment using MUSCLE combined with
gaps treated as missing data had a dramatic effect on 28S Bremer
support values. With these parameters, 28S gave the lowest
Bremer support values, indicating the importance of gaps in the
analysis of the 28S gene. The branch support provided for the var-
ious phylogenies by ArgK was also very high (Table 5). The support
provided by CAD varied depending on the alignment method used
for the 28S data. Alignments informed by secondary structure pro-
duced trees with lower CAD Bremer support values. However,
when the alignment was produced by the software MUSCLE, the
CAD Bremer support values were considerably higher. The compar-
ison of PBS and nodal distances showed no particular trend for
most genes, the only exception being COI which revealed much
higher support values for the most peripheral nodes (Fig. 3). The
phylogenetic information obtained from the majority of the genes
(EF-1a, 28S, ArgK, CAD) occurs throughout the phylogeny, regard-
less of branch length.
Intraspecific distances for EF-1a, ArgK, CAD, and 28S were rela-
tively low, ranging from 0% to 1.18% of sites differing between se-
quences (Table 6). In comparison, the average interspecific
distances observed for the five genes in the dataset ranged from
3.39% to 11.48%. Overall, sequences of COI showed the highest
intraspecific distances, ranging from 0% to 13.16%. In comparison,
the average interspecific distances observed for COI ranged from
14.61% to 17.29%.
4. Discussion
4.1. Analyses
Tree topologies differeddepending on the analytical parameters.
However, the recovery of certain clades was consistent across anal-
yses (Table 3). While the phylogenies illuminate clade membership
andhigher-levelspecies relationships, theyrevealed little aboutthe
relationships among the various genera. Overall, the alignment of
the data and treatment of gap positions as 5th character states
Fig. 1. Most-parsimonious tree found by parsimony analysis of data aligned using 28S secondary structure (Jordal et al., 2008), with gaps treated as missing data. Clade
numbers are given above nodes. Numbers below nodes are bootstrap support values.
S.A. Dole et al./Molecular Phylogenetics and Evolution 54 (2010) 773–782
777

Page 6

resulted in differences in tree topologies. Alignment in the parsi-
mony analyses whether based on secondary structure, MUSCLE, or
direct optimization produced similar trees which differed mostly
in the internal nodes. Alignment using secondary structure verses
MUSCLE made little difference in the outcome of the Bayesian anal-
yses. This is expected because, in the Bayesian analyses, gaps are
considered missing characters and hence contribute no phyloge-
netic information (Ronquist et al., 2005). Gaps provided structure
totheparsimonyphylogenies,producingtreeswithmoreresolution
than the Bayesian trees. However, the Bayesian analysis accounted
forpotentialnucleotidesubstitutionbiases.Therefore,cladesrecov-
ered by both analyses (for example, the ‘‘Anisandrus” clade) were
supported by data that were not compromised by differences in
alignment or nucleotide substitution biases.
4.2. Contributions of data partitions
Commonly used genes, such as 28S, COI, and EF-1a, demon-
strated similar phylogenetic utility as compared with other scoly-
tine studies (Normark et al., 1999; Cognato and Sperling, 2000;
Jordal et al., 2000, 2008; Cognato and Vogler, 2001; Farrell et al.,
2001; Jordal, 2002; Jordal and Hewitt, 2004; Cognato et al., 2005;
Cognato and Sun, 2007; Jordal, 2007). However, all genes included
in this analysis, except COI, showed no significant correlation be-
tween nodal distance and PBS, indicating that they had utility for
resolving multiple levels of the phylogeny (Fig. 3). Typically, the
ribosomal-encoding gene 28S has been shown to resolve relation-
ships within tribes of Scolytinae, but offers little phylogenetic sig-
nal for deeper nodes (Jordal et al., 2008). Here, we found no
Fig. 2. Bayesian tree found by analysis of data aligned using 28S secondary structure (Jordal et al., 2008). Numbers above nodes are Bayesian posterior probabilities.
778
S.A. Dole et al./Molecular Phylogenetics and Evolution 54 (2010) 773–782

Page 7

significant correlation between nodal distance and PBS for 28S, but
these results are not unexpected, given that the relationships being
tested were all bellow the tribal level. Mitochondrial genes, such as
COI, typically do not offer much signal for deeper divergences and
offer better resolution for shallower, species-level nodes (Normark
et al., 1999; Cognato and Sperling, 2000; Cognato and Vogler, 2001;
Farrell et al., 2001; Jordal, 2002, 2007; Jordal and Hewitt, 2004;
Cognato et al., 2005; Cognato and Sun, 2007). For COI, there was
a significant correlation between nodal distance and PBS, with
higher support for shallower nodes, as is predicted for mitochon-
drial genes (Fig. 3). Interestingly, this significance was largely
due to the inclusion of sequences from multiple specimens of
several species. With the removal of the data points representing
X. crassiusculus, X. germanus, and X. morigerus the relationship be-
tween nodal distance and PBS became much less apparent.
In this study we have added new data from the two protein
encoding genes CAD and ArgK, aiming at improved phylogenetic
resolution of the chosen scolytine group. Each of these genes
showed some phylogenetic utility across varying nodal distances
(Fig. 3), but they did not add sufficient information to enable the
resolution of Xylosandrus and related genera. This fits well with
the general performance reported by Wild and Maddison (2008),
who empirically tested the utility of CAD and ArgK along with
other nuclear protein-coding genes in beetle phylogenetics. Both
CAD and ArgK were among the highest performing gene fragments
of the eight tested (Wild and Maddison 2008), quite similar to our
results for Xylosandrus and closely related genera.
4.3. Taxonomic implications and future directions
All analyses performed herein recovered a polyphyletic Xylosan-
drus, suggesting that the genus is in need of taxonomic revision.
This is consistent with previous molecular phylogenies that had
called into question the monophyly of the genus (Jordal, 2002).
Furthermore, the molecular characters utilized by this study have
agreed with previously cited morphological evidence that the cur-
rent classification of the genus contains species belonging to sev-
eral genera (Dole and Beaver, 2008). Three major clades were
recovered with consistently high support: the ‘‘Anisandrus” clade,
‘‘Cnestus” clade, and Xylosandrus sensu stricto clade.
Given that it contains X. morigerus, the type species of Xylosan-
drus, the clade ((X. germanus + X. n. sp. Borneo) X. morigerus, X. com-
pactus) is the highest supported grouping of species belonging to
Xylosandrus sensu stricto resolved by these data. Furthermore, the
four species included in this clade are all morphologically consis-
tent with the strict definition of the genus (Dole and Beaver
2008). Primarily, Xylosandrus sensu stricto can be distinguished
from other xyleborine genera by the synapomorphy of widely sep-
arated procoxae.
The genus Anisandrus was recently resurrected in a phyloge-
netic study of generic taxonomic characters in the Xyleborina (Hul-
cr et al., 2007). Morphological characters distinguishing Anisandrus
from Xylosandrus include contiguous procoxae and lateral protibial
margins with 7–8 socketed teeth. In Xylosandrus sensu stricto the
procoxae are widely separated and the lateral protibial margins
bear 4–5 socketed teeth. The two Xylosandrus species included in
the ‘‘Anisandrus” clade are morphologically consistent with Anisan-
drus. This clade is also supported by a molecular synapomorphy in
28S in the form of a 9–50 base pair insertion comprised almost
exclusively of pyrimidines.
Dole and Beaver (2008) recognized the incorrect placement of
CnestusspecieswithinXylosandrusandmadetentativestepstoward
correcting these taxonomic errors by transferring two species from
Xylosandrus to Cnestus in their review of the Australian species of
Xylosandrus. Morphological characters support the placement of
the three Xylosandrus species within the ‘‘Cnestus” clade. Characters
Table 4
Contribution of data partitions to the character matrix and to the resolution of the most-parsimonious tree found in analysis 1.
Matrix size Variable sitesInformative sites Gaps Contribution to
tree length
Consistency
index (CI)
Retention index (Rl) Mean branch length
EF-1a
C0I
ArgK
CAD
28S
Combined
555
585
594
714
1048
3496
163
257
199
232
318
1169
116
237
165
184
236
938
N/A
N/A
N/A
N/A
473
473
414
2180
613
664
779
4650
0.486
0.205
0.476
0.45
0.614
0.369
0.631
0.339
0.677
0.64
0.769
0.549
5.494
27.542
8.012
8.59
9.964
59.602
Table 5
Partition branch support based on tree found by phylogenetic analysis 1 (see
Materials and methods). Nodes refer to Fig. 1.
Node Gene Partion
EF-1a
COl ArgKCAD 28STotal
1
2
3
4
5
6
7
8
9
7.75
5.75
4
7.4
0
3.8
14
0
?4
3.6
2
2
3
2
3.25
3.6
7
?1
2.33
282
3.8
1.7I
?0.5
6.77
1.67
?0.5
6
3.09
9,05
1.2
0
0
0
2
0
0
0
2
0
3
108.59
?0.5
?1.25
?3
12.1
14
24
7.67
3.67
7.5
?2
?0.5
?0.5
0
?1.57
?2.25
?2
?10.33
?1
2.67
?3.27
7.4
3.79
17
33.46
?1.33
?3
?3
?3.73
?0.2
3.8
1
?3
6
0
?7
?1
7
?4
3
4
103.61
9.25
13
0.67
26.4
1
3.6
10.67
0.67
2
3.4
0.17
3.5
6.33
0.43
7.75
3.4
0
?1
3.67
1.55
6.8
?0.07
5.5
3.92
2.33
1.5
3
1.45
1.25
?0.6
?3
0
0
0
7
19
3
4
0
0
151.53
13.25
8.13
?2
13.3
0
1.4
12
0
2
?4,4
0
1
5.33
?0.57
7.25
?4.4
?1
?1
?1.67
?1.73
?1
1
?0.5
3.85
2
3.5
11
?1.45
?0.3
?1.2
?1
?1
0
?1
0
0
0
6
0
6
72.78
9.25
10.38
1.33
0.8
1
1.2
18.67
0,67
1.5
2.4
0.33
8
13.33
2.71
16
2.4
25.33
5
I8
3.64
7
1.57
5.5
8
4.33
2.5
3
3.64
?5.8
?0.2
5
11
9
0
2
6
?1
1
0
0
20.448
39
36
I
60
16
34
63
5
9
3
2
14
28
10
11
12
13
14
15
16
17
I8
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
3
32
3
21
1
25
3
24
8
27
56
9
4
20
3
4
3
2
7
15
1
2
24
9
9
3
13
Total 641
S.A. Dole et al./Molecular Phylogenetics and Evolution 54 (2010) 773–782
779

Page 8

Fig. 3. Relationships between PBS values and nodal distance (maximum likelihood branch length from a given node to the tip of the tree) for the gene partitions: 28S, COI, and
EF-1a (A), CAD and ArgK (B).
Table 6
COI intra- and interspecific distances expressed as the proportion of sites differing between sequences, with interspecific distances given as the average observed between
species.
EF-1a
COIArgk CAD 28S
Xylosandrus compaclus
Brazil vs. Ghana
Ave Dist Xylosandrus pelliculosus
0.0018
0 0961
0.0769
0.1461
0.0064
0.1033
N/A
0.063
0.0061
0.0355
Xylosandrus crassiusculus
Madagascar vs. Thailand
Madagascar vs. Maryland
Madagascar vs. N. Carolina
Maryland vs. N. Carolina
Maryland vs. Thailand
N. Carolina vs. Thailand
Ave Dist. Xylosandrus pelliculosus
0.0092
0.0031
0.0031
0
0 0121
00118
0.0706
0.0444
0.1094
0.1094
0
0.1009
0.1009
0.1726
0.0026
0.0051
0.0051
0
0.0098
0.0097
0.1104
0.0014
0.0084
0.0084
0
0.007
0.007
0.0676
0
0.0013
0.0014
0
0.0013
0.0013
0.0665
Xylosandrus genuanus
Maryland vs. Michigan
Ave. Disl. Xylosandrus pelliculosus
0
0.0976
0.0034
0.1729
0
0.0918
0
0.0728
0
0.0339
Xylosandrus morigerus
Papua New Guinea vs. Ecuador
Singapore vs. Papua New Guinea
Singapore vs. Ecuador
Ave. Dist Xylosandrus pelliculosus
0.0018
N/A
N/A
0.0932
0.0735
0.0855
0.0974
0.1573
0.0042
0.0106
0.0106
0.0969
0.007
N/A
N/A
0.0742
0.004
0.0016
0.0031
0.0351
780
S.A. Dole et al./Molecular Phylogenetics and Evolution 54 (2010) 773–782

Page 9

distinguishing Cnestus from Xylosandrus sensu stricto include sub-
contiguous procoxae and a four segmented antennal funicle. In
Xylosandrus sensu stricto the antennal funicle is always five seg-
mented. Additionally, in Cnestus the anterior margin of the prono-
tum bears four or fewer asperities, with a coarse pair medially, and
the pronotum is often produced anteriorly. In Xylosandrus sensu
stricto the anterior margin of the pronotum bears six or more asper-
ities of approximately equal size and is never produced anteriorly
(Dole and Cognato, in review). The two Xylosandrus species placed
in the ‘‘Cnestus” clade are morphologically consistent with Cnestus.
OtherXylosandrusspeciesincludedinthisanalysisvariedintheir
phylogenetic placement: X. crassiusculus, X. discolor, X. mancus, X.
monteithi, X. queenslandi, and X. n. sp. Papua New Guinea. Despite
their unresolved positions in the molecular phylogeny, the place-
mentofthesespeciesinXylosandrusmaybecorrect,sincetheirmor-
phologies are consistent with the defining characters of Xylosandrus
sensu stricto. A phylogenetic analysis of morphological data will be
necessary to further test the placement of these species within the
genus. In addition, Xyleborus rotundicollis and Euwallacea russulus
consistently grouped with Xylosandrus sensu lato. This result is sup-
ported by the morphology of these two species, which more closely
resembles Xylosandrus than Euwallacea or Xyleborus. The possible
transferofthesespeciesintoXylosandrusshouldbeexaminedwithin
the context of a morphological study of the genus.
TheuncertainpositionoftheXylosandrus,Anisandrus,andCnestus
clades made it impossible to determine the phylogenetic relation-
ships among these genera within the context of this study. Further-
more,thetaxonsampleavailableformolecularworkrepresentsonly
a fractionof the 54 speciescurrentlyincluded inXylosandrus. Dueto
the unavailability of many species for molecular sequencing, mor-
phological data will be the only way to assess the proper generic
placement of the majority of Xylosandrus species. For instance, sev-
eral species of Xylosandrus have morphologies which suggest that
they should be transferred to Amasa, but these species are so rarely
collected that none were available for DNA sequencing. It is difficult
tountanglewhetherthegenesorlackoftaxacontributedmoretothe
unresolvedphylogeney.However,the additionofbothcouldhelpto
remedythissituation(RokasandCarroll,2005;Edwardsetal.,2007;
Wild and Maddison, 2008).
This study supports the need for a major taxonomic revision of
Xylosandrus. Given the economic importance of Xylosandrus, a revi-
sion should include the development of a morphology-based clas-
sification with clear diagnostic characters and a key to the
worldwide species. Such a revision is currently in review and will
include the five gene data set used here in combination with mor-
phology to provide a more complete phylogenetic classification of
Xylosandrus.
Acknowledgements
We would like to thank Roger Beaver, Don Bright, and Stephen
Wood for their continued advice and valuable guidance throughout
this project and Jiri Hulcr, Aaron Smith, and Sarah Smith for engag-
ing in discussions of molecular systematics. S.A.D. thanks the Cal-
ifornia Academy of Sciences Department of Entomology for the
support of their staff and use of their research facilities. We thank
our colleagues listed in Table 1 for providing specimens. This re-
search was supported by NSF-PEET Grant (DEB-0328920) to A.I.C.
References
Baker, R.H., Yu, X.B., DeSalle, R., 1998. Assessing the relative contribution of
molecular and morphological characters in simultaneous analysis trees. Mol.
Phylogenet. Evol. 9, 427–436.
Beaver, R.A., 1989. Insect–fungus relationships in the bark and ambrosia beetles. In:
Wilding, N., Collins, N.M., Hammond, P.M., Webber, J.F. (Eds.), Insect Fungus
Interactions. Academic Press, London, pp. 121–143.
Bright, D.E., Skidmore, R.E., 1997. A Catalog of Scolytidae and Platypodidae
(Coleoptera), Supplement 1 (1990–1994). NRC Research Press, Ottawa,
Ontario, Canada. 368 pp..
Bright, D.E., 1968. Review of the tribe Xyleborini in America North of Mexico
(Coleoptera: Scolytidae). Can. Entomol. 100, 1288–1323.
Bright, D.E., Poinar, G.O., 1994. Scolytidae and Platypodidae (Coleoptera) from
Dominican Republic amber. Ann. Entomol. Soc. Am. 87, 170–194.
Cognato, A.I., Grimaldi, D., 2009. One hundred million years of morphological
conservation in a bark beetle (Coleoptera: Curculionidae: Scolytinae). Syst.
Entom. 34 (1), 93–100.
Cognato, A.I., Sun, J.H., 2007. DNA based cladograms augment the discovery of a
new Ips species from China (Coleoptera: Curculionidae: Scolytinae). Cladistics
23, 539–551.
Cognato, A.I., Gillette, N.E., Bolaños, R.C., Sperling, F.A.H., 2005. Mitochondrial
phylogeny of pine cone beetles (Scolytinae, Conophthorus) and their affiliation
with geographical area and host. Mol. Phylogenet. Evol. 36, 494–508.
Cognato, A.I., Vogler, A.P., 2001. Exploring data interaction and nucleotide
alignment in a multiple gene analysis of Ips (Coleoptera: Scolytinae). Sys. Biol.
50 (6), 758–780.
Cognato, A.I., Sperling, F.A.H., 2000. Phylogeny of Ips DeGeer species (Coleoptera
Scolytidae) inferred from mitochondrial cytochrome oxidase I DNA sequences.
Mol. Phylogenet. Evol. 14 (3), 445–460.
Danforth, B.N., Fang, J., Sipes, S., 2006. Analysis of family-level relationships in bees
(Hymenoptera: Apiformes) using 28S and two previously unexplored nuclear
genes: CAD and RNA polymerase II. Mol. Phylogenet. Evol. 39, 358–372.
Dole, S.A., Beaver, R.A., 2008. A review of the Australian species of Xylosandrus
Reitter (Coleoptera: Curculionidae: Scolytinae). Coleopt. Bull. 62 (4), 481–492.
Edgar, R.C., 2004. MUSCLE: multiple sequence alignment with high accuracy and
high throughput. Nucleic Acids Res. 32 (5), 1792–1797.
Edwards, S.V., Liu, L., Pearl, D.K., 2007. High-resolution species tress without
concatenation. Proc. Natl. Acad. Sci. USA 104, 5936–5941.
Farrell, B.D., Sequeira, A.S., O’Meara, B.C., Normark, B.B., Chung, J.H., Jordal, B.H.,
2001. The evolution of agriculture in beetles (Curculionidae: Scolytinae and
Platypodinae). Evolution 55 (10), 2011–2027.
Grimaldi, D., Engel, M.S., 2005. Evolution of the Insects. Cambridge University Press,
Cambridge, UK. 755pp..
Hebert, P.D.N., Cywinska, A., Ball, S.L., deWard, J.R., 2003. Biological identifications
through DNA barcodes. Proc. R. Soc. B 270, 313–321.
Hulcr, J., Dole, S.A., Beaver, R.A., Cognato, A.I., 2007. A cladistic review of generic
taxonomic characters in Xyleborina (Coleoptera: Curculionidae: Scolytinae).
Syst. Entom. 32 (3), 568–584.
Hulsenbeck, J.P., Ronquist, F.R., 2001. MrBayes: Bayesian inference of phylogeny.
Biometrics 17, 754–755.
Jordal, B.H., 2007. Reconstructing the phylogeny of Scolytinae and close allies:
major obstacles and prospects for a solution. In: Bentz, B., Cognato, A., Raffa, K.
(Eds.), Proceedings from the Third Workshop on Genetics of Bark Beetles and
Associated Microorganisms. Proc. RMRS-P-45. US. Department of Agriculture,
Forest Service, Rocky Mountain Research Station, Fort Collins, CO, pp. 3–9.
Jordal, B.H., 2002. Elongation Factor 1 resolves the monophyly of the haplodiploid
ambrosia beetles Xyleborini (Coleoptera: Curculionidae). Insect Mol. Biol. 11
(5), 453–465.
Jordal, B., Gillespie, J.J., Cognato, A.I., 2008. Secondary structure alignment and
direct optimization of 28S rDNA sequences provide limited phylogenetic
resolution in bark and ambrosia beetles (Curculionidae: Scolytinae). Zool. Scr.
37 (1), 43–56.
Jordal, B.H., Hewitt, G.M., 2004. The origin and radiation of Macroneasian beetles
breeding in Euphorbia: the relative importance of multiple data partitions and
population sampling. Syst. Biol. 53 (5), 711–734.
Jordal, B.H., Normark, B.J., Farrell, B.D., 2000. Evolutionary radiation of an inbreeding
haplodiploid beetle lineage (Curculionidae: Scolytinae). Biol. J. Linn. Soc. 71,
483–499.
Kirkendall, L.R., 1993. Ecology and evolution of biased sex ratios in bark and
ambrosia beetles (Scolytidae). In: Wrensch, D.L., Ebbert, M.A. (Eds.), Evolution
and Diversity of Sex Ratio: Insects and Mites. Chapman and Hall, New York, pp.
235–345.
Kruskal, J.B., 1983. An overview of sequence comparison. In: Sankoff, D., Kruskal, J.B.
(Eds.), Time Warps, String Edits, and Macromolecules: The Theory and Practice
of Sequence Comparison. Addison-Wesley, Reading, Massachusetts, pp. 1–45.
Kuschel, G., Leschen, R.A.B., Zimmermann, E.C., 2000. Platypodidae under scrutiny.
Invertebr. Taxon. 14, 771–805.
Lee, M.S.Y., 2001. Unalignable sequences and molecular evolution. Trends Ecol. Evol.
16, 681–685.
Marvaldi, A.E., 1997. Higher level phylogeny of Curculionidae (Coleoptera:
Curculionoidea) based mainly on larval characters, with special reference to
broad-nosed weevils. Cladistics 13, 285–312.
Mandelshtam, M.Y.,Beaver,R.A., 2003.
Amphiscolytini—a new tribe of Scolytidae (Coleoptera) for Dacryophthorus
capensis Schedl. Zootaxa 298, 1–8.
McKenna, D.D., Sequeira, A.S., Marvaldi, A.E., Farrell, B.D., 2009. Temporal lags and
overlap in the diversification of weevils and flowering plants. Proc. Natl. Acad.
Sci. USA 17, 7083–7088.
Normark, B.B., Jordal, B.H., Farrell, B.D., 1999. Origin of a haplodiploidy beetle
lineage. Proc. R. Soc. Lond. B 266, 2253–2259.
Oliveira, C.M., 2008. First record of Xylosandrus compactus (Eichhoff) (Coleoptera:
Curculionidae: Scolytinae) on soursop, Annona muricata L. (Annonanceae) in
Brazil, with a list of host plants. Coleopt. Bull. 62 (1), 45–48.
Amphiscolytus—anew genus,and
S.A. Dole et al./Molecular Phylogenetics and Evolution 54 (2010) 773–782
781

Page 10

Oliver, J.B., Mannion, C.M., 2001. Ambrosia beetle (Coleoptera: Scolytidae) species
attacking chestnut and captured in ethanol-baited traps in middle Tennessee.
Popul. Ecol. 30 (5), 909–918.
Rambaut.A.,Drummond,A.J.2007.
beast.bio.ed.ac.uk/Tracer>.
Reitter, E., 1913. Bestimmungs-tabelle der Borkenkafer (Scolytidae) aus Europa
und den angrenzenden Landren. Wiener Entomologische Zeitung 32 (Beiheft),
1–116.
Rokas, A., Carroll, S.B., 2005. More genes or more taxa? The relative contribution of
gene number and taxon number to phylogenetic accuracy. Mol. Biol. Evol. 22,
1337–1344.
Ronquist, F., Huelsenbeck, J.P., van der Mark, P., 2005. MrBayes 3.1. Available from:
<http://mrbayes.csit.fsu.edu/index.php>.
Saha, N., Maiti, P.K., Chakraborti, S., 2002. On some species of Xylosandrus
Reitter (Scolytidae: Coleoptera) from the Sub-Himalayan West Bengal with
description of a new species. Record of the Zoological Survey of India 91
(1), 9–27.
Sankoff, D., Cedergren, R.J., 1983. Simultaneous comparison of three or more
sequences related by a tree. In: Sankoff, D., Kruskal, J.B. (Eds.), Time Warps,
Tracerv1.4.Availablefrom<http://
String Edits, and Macromolecules: The Theory and Practice of Sequence
Comparison. Addison-Wesley, Reading, Massachusetts, pp. 253–263.
Sorensen, M.D., 1996. Treerot. University of Michigan, Ann Arbor, MI.
Swofford, D.L., 2002. PAUP*: Phylogenetic Analysis Using Parsimony (*and Other
Methods), 4.0b10. Sinauer Associates, Sunderland, Massachusetts.
Varón, A., Vinh, L.S., Bomash, I., Wheeler, W.C., 2008. POY 4.0.2881. American
Museum of Natural History, New York.
Wheeler, W.C., 1996. Optimization alignment: the end of multiple sequence
alignment in phylogenetics? Cladistics 12, 1–9.
Wild, A.L., Maddison, D.R., 2008. Evaluating nuclear protein-coding genes for
phylogenetic utility in beetles. Mol. Phylogenet. Evol. 48, 877–891.
Wood, S.L., 2007. The Bark and Ambrosia Beetles of South America (Coleoptera:
Scolytidae). Monte Bean Life Science Museum, Brigham Young University,
Provo, 900 pp. + 230 plates.
Wood, S.L., 1986. A reclassification of the genera of Scolytidae (Coleoptera). Gt.
Basin Nat. Mem. 10, 126.
Wood, S.L., Bright, D.E., 1992. A catalog of Scolytidae and Platypodidae (Coleoptera):
Part 2. Taxonomic Index, Volumes A and B. Gt. Basin Nat. Mem. 13.
782
S.A. Dole et al./Molecular Phylogenetics and Evolution 54 (2010) 773–782