Content uploaded by Seunggwan Shin
Author content
All content in this area was uploaded by Seunggwan Shin on Feb 05, 2018
Content may be subject to copyright.
Phylogenomic Data Yield New and Robust Insights into the
Phylogeny and Evolution of Weevils
Seunggwan Shin,*
†,1
Dave J. Clarke,
†,1
Alan R. Lemmon,
2
Emily Moriarty Lemmon,
3
Alexander L. Aitken,
1
Stephanie Haddad,
1
Brian D. Farrell,
4
Adriana E. Marvaldi,
5
Rolf G. Oberprieler,
6
and Duane D. McKenna
1
1
Department of Biological Sciences, University of Memphis, Memphis, TN
2
Department of Scientific Computing, Florida State University, Tallahassee, FL
3
Department of Biological Science, Florida State University, Tallahassee, FL
4
Museum of Comparative Zoology, Harvard University, Cambridge, MA
5
CONICET, Divisi
on Entomolog
ıa, Universidad Nacional de La Plata, La Plata, Buenos Aires, Argentina
6
CSIRO, Australian National Insect Collection, Canberra, ACT, Australia
†
These authors contributed equally to this work.
*Corresponding author: E-mail: sciaridae1@gmail.com.
Associate editor: Claudia Russo
Abstract
The phylogeny and evolution of weevils (the beetle superfamily Curculionoidea) has been extensively studied, but many
relationships, especially in the large family Curculionidae (true weevils; >50,000 species), remain uncertain. We used
phylogenomic methods to obtain DNA sequences from 522 protein-coding genes for representatives of all families of
weevils and all subfamilies of Curculionidae. Most of our phylogenomic results had strong statistical support, and the
inferred relationships were generally congruent with those reported in previous studies, but with some interesting
exceptions. Notably, the backbone relationships of the weevil phylogeny were consistently strongly supported, and
the former Nemonychidae (pine flower snout beetles) were polyphyletic, with the subfamily Cimberidinae (here elevated
to Cimberididae) placed as sister group of all other weevils. The clade comprising the sister families Brentidae (straight-
snouted weevils) and Curculionidae was maximally supported and the composition of both families was firmly estab-
lished. The contributions of substitution modeling, codon usage and/or mutational bias to differences between trees
reconstructed from amino acid and nucleotide sequences were explored. A reconstructed timetree for weevils is con-
sistent with a Mesozoic radiation of gymnosperm-associated taxa to form most extant families and diversification of
Curculionidae alongside flowering plants—first monocots, then other groups—beginning in the Cretaceous.
Key words: Curculionoidea, Curculionidae, chronogram, exon, hybrid enrichment, phylogenetics.
Introduction
The beetle superfamily Curculionoidea Latreille (weevils) con-
tains approximately 62,000 described extant species in more
than 5,800 genera (Kuschel 1995;Oberprieler et al. 2007),
making it one of the most species-rich radiations of metazo-
ans. The apparent success of weevils has been ascribed to
coevolution with plants, especially flowering plants (e.g.,
Farrell 1998;McKenna et al. 2009), and the development of
a specialized “oviposition rostrum” (e.g., Anderson 1993,
1995)—a purported key innovation in which the female
uses her rostrum/mouthparts to drill an oviposition site
deep inside the host plant. However, weevil diversity is likely
attributable to a “cascade of evolutionary innovations”
(Oberprieler et al. 2007), at least some of which facilitated
specialized trophic interactions with plants, and also with
fungi, as several weevil lineages are mycetophagous or feed
on plant substrates modified by fungi (Holloway 1982;
Zimmerman 1994;Marvaldi et al. 2002). Together, these inno-
vations are proposed to have enhanced weevil speciation
rates and/or reduced extinction rates, promoting lineage
accumulation (e.g., McKenna et al. 2015). Achieving a stable
higher-level classification and a robust phylogenetic backbone
for the superfamily Curculionoidea promises to facilitate both
the exploration of factors underlying the apparent evolution-
ary success of weevils and the evolutionary dynamics of their
intimate associations with plants (e.g. Farrell 1998;
Oberprieler et al. 2007;McKenna et al. 2009). It will also
facilitate the predictive power expected of classifications
rooted in robust phylogenetics and thereby provide the
much needed resources for biosecurity, conservation and
pest control applications that is currently lacking for this
largest of phytophagous insect radiations.
Most recent classifications recognize seven major lineages
of weevils (Anthribidae, Attelabidae, Belidae, Brentidae,
Caridae, Curculionidae, and Nemonychidae: e.g., Oberprieler
et al. 2007;Oberprieler, Anderson, et al. 2014). These have
been variously sampled and supported as monophyletic
groups in recent morphological and molecular phylogenetic
studies (e.g., Marvaldi and Morrone 2000;Marvaldi et al. 2002;
Hunt et al. 2007;McKenna et al. 2009,2015;McKenna 2011;
Article
ßThe Author(s) 2018. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution.
All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com
Mol. Biol. Evol. doi:10.1093/molbev/msx324 Advance Access publication December 26, 2017 1
Downloaded from https://academic.oup.com/mbe/advance-article-abstract/doi/10.1093/molbev/msx324/4765916
by University of Memphis - Library, sciaridae1@gmail.com
on 05 February 2018
Haran et al. 2013;Bocak et al. 2014;Gillett et al. 2014;Gunter
et al. 2016). Curculionidae, with approximately 51,000 de-
scribed species in more than 4,600 genera (Oberprieler
2014c), is the largest family of weevils and the second-
largest family of metazoans. Subfamily concepts in
Curculionidae remain highly tentative and controversial,
though some natural groupings have been suggested by stud-
ies of adult morphology (Morimoto 1962;Kuschel 1971;
Thompson 1992;Alonso-Zarazaga 2007) and phylogenetic
studies employing morphological and/or molecular data
(Kuschel 1995;Marvaldi and Morrone 2000;Marvaldi et al.
2002,2009;Morimoto and Kojima 2006;Hundsdo¨rfer et al.
2009;McKenna et al. 2009,2015;Jordal et al. 2011;Haran
et al. 2013;Gillett et al. 2014;Gunter et al. 2016).
The phylogeny of the superfamily Curculionoidea has been
reconstructed using molecular data from one or a small num-
ber of genes obtained via traditional polymerase chain reac-
tion and Sanger sequencing (e.g., Wink et al. 1997;Marvaldi
et al. 2002,2009;Hundsdo¨rfer et al. 2009;McKenna et al. 2009,
2015;Jordal et al. 2011;Haran et al. 2013;Bocak et al. 2014;
Gillett et al. 2014;Gunter et al. 2016)andusingnext-
generation-sequencing methods to generate data from mito-
chondrial genomes (Haran et al. 2013;Gillett et al. 2014).
Nonetheless, several family-level relationships and most
subfamily-level relationships in Curculionidae remain uncer-
tain due to weak statistical support and/or limited taxon
sampling in all studies to date. McKenna et al. (2009) under-
took the only molecular phylogenetic study that has sampled
all families and subfamilies of weevils. However, in common
with most other studies, nodal support values were moderate
to low or lacking for nearly all relationships resolved by
McKenna et al. (2009), thus leaving the phylogenetic place-
ment, monophyly and morphological definition of most
higher weevil taxa uncertain.
One traditional means of increasing resolving power in
molecular phylogenetic data sets is to sample more loci
(e.g., Niehuis et al. 2012;Misof et al. 2014;Peters et al.
2017). Leache and Rannala (2011) demonstrated that hun-
dreds of nuclear loci measuring 1kb(perlocus)inlength
may be needed to resolve difficult nodes resulting from pre-
sumably rapid radiations and/or recent divergences (Leache
and Rannala 2011;Prum et al. 2015). However, until recently
there were no widely available cost- or time-efficient
approaches for increasing the number of loci included in
molecular phylogenetic studies of nonmodel taxa (Lemmon
EM and Lemmon AR 2013). Phylogenomic approaches have
now become available that allow for the generation of DNA
sequence data from large numbers of known/targeted loci
from nonmodel taxa (e.g., see Lemmon EM and Lemmon AR
2013;Misof et al. 2014). These approaches have been widely
used in the studies of vertebrates but have only recently been
employed in the studies of insects (Young et al. 2016;
Breinholtetal.2018;Haddad et al. 2018).
This study was designed to reconstruct family-level rela-
tionships across Curculionoidea and subfamily-level relation-
ships in Curculionidae using phylogenomic data comprised of
DNA sequences (separate analyses of nucleotide [NT] and
amino acid [AA] data) from >500 1:1 orthologous nuclear
genes (see Materials and Methods) obtained via anchored
hybrid enrichment (AHE) from an exemplar set of weevil
species and outgroups. AHE is a highly efficient and scalable
method for generating high-throughput DNA sequence data.
It allows for the capture and amplification of target DNA
using specially designed probes (Lemmon et al. 2012)and
can efficiently and selectively harvest specific desired parts
of the genome (here 941 exons). We used the results of our
phylogenetic analyses along with information from the fossil
record to estimate chronograms for weevils using several dif-
ferent calibration schemes and compared and contrasted the
results with those from previous studies.
New Approaches
In this study we report results from use of the first AHE probe
set for Coleoptera (beetles), the most species-rich order of
metazoans. The AHE probes were designed for broad utility
across Coleoptera as well as the other neuropteroid insect
orders (Strepsiptera, Megaloptera, Neuroptera and
Raphidioptera [McKenna 2014,2016]) (Haddad et al. 2018;
Materials and Methods). Probe sequences are available via
Dryad (accession number doi: 10.5061/dryad.v0b7v). It is dif-
ficult to translate AHE sequence data using existing analytical
pipelines for hybrid enrichment data because many of the
assembled targeted exons include flanking intron sequences
of varying lengths (Breinholt et al. 2018). Our AHE analytical
pipeline uses protein-based orthology searches to identify
coding regions in the AHE data and translate these regions
into AA sequences for orthology assessment (see Materials
and Methods).
Results and Discussion
New Insights into the Phylogeny and Evolution of
Weevils
Most of the nodes in our phylogenies had maximal statistical
support (100% maximum likelihood [ML] bootstrap support
[MLBS], 1.0 Bayesian posterior probability [PP]), demonstrat-
ing the utility of our probes for generating nuclear DNA se-
quence data useful in resolving both deep and shallow
divergences. Furthermore, most major weevil groupings
were consistently supported regardless of data set (AA or
NT) or method of phylogenetic inference (ML or Bayesian
[BI]), and under various modeling and partitioning schemes
(fig. 1;supplementary table S2,Supplementary Material on-
line). Phylogenetic trees resulting from separate concatenated
analyses of each codon position (C1, C2, C3: supplementary
figs. S7, S8, and S9,respectively,Supplementary Material on-
line), a concatenated analysis of combined first and second
codons (C12: supplementary fig. S10,Supplementary Material
online), and both AA- and NT-based binned and weighted
coalescent species tree analyses were also largely congruent
(supplementary table S2,Supplementary Material online).
One exception was the discordant placement of
Cimberididae as sister group of the clade
(Nemonychidae þAnthribidae) in the ASTRAL NT analysis
(with 91% support). However, this relationship was otherwise
found and only weakly supported in analyses of third-codon
Shin et al. .doi:10.1093/molbev/msx324 MBE
2
Downloaded from https://academic.oup.com/mbe/advance-article-abstract/doi/10.1093/molbev/msx324/4765916
by University of Memphis - Library, sciaridae1@gmail.com
on 05 February 2018
FIG.1.Chronogram for weevils (superfamily Curculionoidea) showing RelTime estimates of lineage divergence times based on the partitioned ML
analysis of amino acid data (supplementary fig. S6,Supplementary Material online). Subfamily (family for chrysomeloid outgroups) and genus
names are given for all ingroup (weevil) terminals. The three taxon names in bold indicate (elevated) ranks different from the classification of
Oberprieler, Anderson et al. (2014) in Leschen and Beutel (2014). The outgroups Aethina and Cucujus (both Cucujoidea) are not shown (following
Kumar et al. 2016;Mello et al. 2017). Stars indicate nodes for which all trees from primary analyses (supplementary figs. S1–S6,Supplementary
Material online) have support values greater than 97% MLBS and 1.0 PP; diamonds indicate nodes with support values greater than 97% MLBS and
1.0 PP in trees resulting from the primary AA analyses (supplementary figs. S2, S4, and S6,Supplementary Material online); triangles indicate nodes
with support values greater than 97% MLBS and 1.0 PP in trees resulting from the primary NT analyses (supplementary figs. S1, S3, and S5,
Supplementary Material online). The origin of angiosperms is indicated, as estimated from fossils (132–141 Ma [Brenner 1996]) and molecules
(140–180 Ma [Bell et al. 2005]). CUR, Curculionoidea; CHR, Chrysomeloidea; inc.sed., incertae sedis (where a taxon’s broader relationships are
Phylogenomic Data Yield Insights into Weevil Evolution .doi:10.1093/molbev/msx324 MBE
3
Downloaded from https://academic.oup.com/mbe/advance-article-abstract/doi/10.1093/molbev/msx324/4765916
by University of Memphis - Library, sciaridae1@gmail.com
on 05 February 2018
positions (C3; supplementary fig. S9,40%MLBS;supplemen-
tary table S2,Supplementary Material online), which also
supported a relationship between Belidae and Attelabidae
(same as in the partitioned NT analysis; supplementary fig.
S5,Supplementary Material online). Specifically, our analyses
robustly supported a “phylogenetic backbone” for weevils—a
series of early-diverging relationships among families.
Curculionoidea and five of its seven family-level lineages
were maximally supported monophyletic groups across all
of our analyses. Notably, Cimberididae (formerly a subfamily
of Nemonychidae; pine flower snout beetles) were placed as
the sister group of all other weevils, a relationship previously
suggested by Haran et al. (2013) based on data from mito-
chondrial genomes (Supplementary Material online). The
morphologically plesiomorphic and largely conifer-feeding
pine flower snout beetles therefore do not form a natural
group as currently defined, warranting a narrowing of the
limits of the family and recognition of Cimberididae as a dis-
tinct family-level lineage (Crowson, 1985 previously suggested
the family-group name Cimberidae [sic] for Cimberis and
Rhinorhynchinae). As Nemonychidae are widely regarded as
retaining many morphologically “ancestral” character states
and have the oldest fossil history among extant weevils
(Kuschel 1983,2003;Oberprieler RG and Oberprieler SK
2012;Kuschel and Leschen 2011), our results suggest a
more complex evolutionary history for the many fossil nem-
onychids. Similar to some other recent molecular phyloge-
netic studies (e.g., Marvaldi et al. 2002,2009;McKenna et al.
2009), our analyses maximally support a monophyletic group
comprising fungus weevils (family Anthribidae) and its place-
ment as sister group of the other Nemonychidae sampled.
Our analyses robustly support placement of the conifer-
associated family Caridae as sister group of the predominantly
angiosperm-feeding clade comprising straight-snouted and
true weevils (Brentidae and Curculionidae, respectively).
Previous studies, with the exception of McKenna et al.
(2015), have either placed this isolated family in different
positions (Gunter et al. 2016) or found only weak to moder-
ate support for it among analyses (Supplementary Material
online). The placement of Caridae in our trees is also consis-
tent with results from analyses of morphological data (e.g.,
Marvaldi and Morrone 2000;Marvaldi et al. 2002;
Supplementary Material online).
Monophyly of the species-rich clade comprising the sister
families Brentidae and Curculionidae was maximally (and
ubiquitously) supported (fig. 1;supplementary figs. S1–S6
and table S2,Supplementary Material online); support for
this relationship in previous molecular studies was inconsis-
tent and/or those studies were inconclusive about the defi-
nition and limits of the two families, especially Brentidae. The
true weevils (Curculionidae) were strongly supported as a
monophyletic group in our study regardless of data set,
partitioning scheme or analysis method. This therefore firmly
establishes the limits of the family by demonstrating, for
example, inclusion of the palm and pinhole borer weevils
(Dryophthorinae and Platypodinae), while definitively exclud-
ing other groups. Notable among the latter are the morpho-
logically enigmatic subfamilies Microcerinae and Ithycerinae
(monotypic; New York weevil), which were here placed
among the straight-snouted weevils. Brentidae in the present
sense (Oberprieler 2000,2014a;Oberprieler et al. 2007)were
maximally supported as a monophyletic group in all of our
analyses. This is notable because there is currently no strong
morphological evidence for its monophyly (Oberprieler
2014a) and because Ithycerinae and Microcerinae were
both placed in Curculionidae by McKenna et al. (2009) and
have had different systematic placements in the past
(reviewed by Oberprieler 2014b).
Most lower-level relationships found within the five
broadly sampled weevil families were maximally supported
(with the exception of relationships in the “CCCMS clade” of
Curculionidae [Conoderinae, Cossoninae, Curculioninae,
Molytinae, Scolytinae]; fig. 1,supplementary figs. S1–S6 and
table S2,Supplementary Material online). Lower-level rela-
tionships found by ASTRAL analyses (binned and weighted),
such as those within the “CEGH clade” of Curculionidae
(Cyclominae, Entiminae, Gonipterini, Hyperinae) and the
CCCMS clade, were either insufficiently or no better resolved
than those found by analyses of concatenated data. This is
not a surprising result given the deep time scale (Jurassic
origin) over which weevils have diversified (fig. 1,supplemen-
tary table S4,Supplementary Material online; and see below)
and also given that individual genes in AHE data sets may
have relatively low phylogenetic signal (e.g., Prum et al. 2015).
Our results therefore establish a robust phylogenetic
“backbone” for Curculionidae, and both the relationships
and divergence times we found are consistent with the pro-
posals by Marvaldi et al. (2002),Oberprieler et al. (2007) and
McKenna et al. (2009) (see also Farrell 1998)thattheancestral
diversification of Curculionidae occurred in association with
monocotyledonous angiosperm host plants and that the
family subsequently diversified onto other plant groups, in-
cluding other angiosperms (fig. 1;supplementary figs. S18–
S25,Supplementary Material online).
Consistent with other phylogenetic studies (Marvaldi 1997;
McKenna et al. 2009;Haran et al. 2013;Gillett et al. 2014;
Gunter et al. 2016), the hypothesized ancestrally monocot-
associated subfamilies Brachycerinae and Dryophthorinae
(palm weevils) and the wood-boring Platypodinae (pinhole
borers) form early-diverging groups of true weevils, the first
of these subfamilies a polyphyletic grade but the others
monophyletic clades, subtending a major lineage within the
family, here informally called the “higher Curculionidae”
(Curculionidae sensu Thompson, 1992; “higher weevils”). The
FIG.1.Continued
unknown or undefined). Weevil photos are courtesy of Udo Schmidt (used with permission). Left column, top to bottom: Anthribus albinus,
Involvulus caeruleus,Cerobates (Cerobates) sexsulcatus,Odoiporus longicollis,Hypera nigrirostris,Gronops lunatus,Scolytus scolytus,Cossonus
parallelepipedus,Cryptorhynchus lapathi. Right column, top to bottom: Cimberis attelaboides,Attelabus nitens,Apion rubens,Tanysphyrus lemnae,
Platypus cylindrus,Eupholus cuvieri,Mononychus punctumalbum,Curculio glandium,Magdalis duplicata.
Shin et al. .doi:10.1093/molbev/msx324 MBE
4
Downloaded from https://academic.oup.com/mbe/advance-article-abstract/doi/10.1093/molbev/msx324/4765916
by University of Memphis - Library, sciaridae1@gmail.com
on 05 February 2018
relationships among subgroups of Brachycerinae and the
nested placement of palm and pinhole borer weevils were
strongly supported (fig. 1;supplementary table S2,
Supplementary Material online). This has important implica-
tions for the classification and interpretation of morphological
evolution of weevils because although the majority of species
in the subfamilies Brachycerinae, Dryophthorinae and
Platypodinae retain the hypothesized “ancestral” male genital
structure (pedotectal aedeagus) characteristic of other weevil
families (Supplementary Material online), they are each also
morphologically derived groups. The remaining true weevils—
the higher Curculionidae—possess a more derived male gen-
ital structure (pedal aedeagus) and are robustly established
here as a monophyletic group (congruent with Thompson’s
[1992] more restricted concept of Curculionidae) and thus
provide an important point of congruence between conclu-
sions based on morphological and molecular data. Among
other things, these results highlight the potential value of phy-
logenomic data for advancing the classification of diverse
groups such as weevils, in which morphology-based classifica-
tions have historically been unstable. For example, our results
suggest that a change in the classification of the subfamily
Brachycerinae is warranted (Supplementary Material online),
and we found that the small aquatic subfamily Bagoinae con-
stitutes an isolated lineage forming the sister group of the
higher curculionids (fig. 1,Supplementary Material online),
raising questions about the botanical and ecological associa-
tions of the stem lineage of this more inclusive clade. Although
our results concur with those of Gillett et al. (2014) and sup-
port the conclusion that the former tribe Bagoini be re-
elevated to subfamily rank (previously its placement in
Curculionidae was uncertain; Oberprieler, Caldara et al.
2014), others have found Bagous to be more closely related
to taxa with the pedotectal type of male genitalia (Oberprieler,
Anderson et al. 2014;Gunter et al. 2016), so our results also
suggest a more complex interpretation of evolutionary
changes in the genital structure of weevils (Supplementary
Material online).
The pinhole borer and palm weevils (Platypodinae and
Dryophthorinae, respectively) form a monophyletic clade as
sister group of a subset of Brachycerinae (the erirhinines),
with moderate to low support (fig. 1,supplementary figs.
S1, S2 and S4–S6,Supplementary Material online). This result
is unsurprising given the close relationship between these two
groups found by McKenna et al. (2009,2015), Haran et al.
(2013),andGillett et al. (2014) and also the support for this
relationship from larval characters (Marvaldi 1997). However,
it is also surprising given their very divergent anatomy and the
morphological similarities between Platypodinae and
Scolytinae (bark beetles). The systematic position and rank
of Platypodinae has long been the subject of considerable
debate. Sharing many morphological features with the simi-
larly wood-boring/tunneling bark beetles (Kuschel et al. 2000;
Marvaldi et al. 2002;Jordal et al. 2011;Hulcr et al. 2014), both
groups have been considered either as distinct families (e.g.,
Morimoto and Kojima 2006)orascloselyordistantlyrelated
groups within the true weevils, and it has even been suggested
that pinhole borers are nested inside the bark beetles
(rejected family and subfamily status: Kuschel et al. 2000).
Interestingly, in our NT Bayesian analysis, Austroplatypus
(Platypodinae) formed the sister group of a monophyletic
Dryophthorinae and Notoplatypus (Platypodinae) the sister
group of a subset of Scolytinae (supplementary fig. S3,
Supplementary Material online, 0.62 PP). Patterns of codon
usage and data from 4-fold degenerate sites (FDS) for third-
codon positions may help explain this incongruent result.
Mean FDS GC content was 38.9% (61.1% AT), indicating
that our data are generally AT-rich (supplementary table
S6,Supplementary Material online). Notably, the FDS GC
content of Austroplatypus reflects substantial AT bias (82%
AT), whereas that of Notoplatypus is much less AT-rich (59%
AT; slightly less than the study mean) (supplementary fig. S16,
Supplementary Material online). Thus, differences in muta-
tional bias (resulting in bias of synonymous codon usage) in
Platypodinae may contribute to the observed incongruence.
Our results support a deep phylogenetic split dividing the
higher weevils into two main clades (fig. 1). These clades,
called the “CCCMS clade” and “CEGH clade” (Gunter et al.
2016), have been variously found and supported in other
studies (Marvaldi et al. 2002,inpart;McKenna et al. 2009;
Haran et al. 2013;Gillett et al. 2014;Gunter et al. 2016;
Supplementary Material online). The CCCMS clade, compris-
ing seven current curculionid subfamilies and constituting
one of the most diverse plant-feeding groups of beetles
(>34,000 described species), was monophyletic in all analyses
but maximally supported only in AA-based analyses (fig. 1;
supplementary figs. S1–S6,Supplementary Material online).
In the CCCMS clade, the hugely diverse subfamilies
Curculioninae and Molytinae are clearly not monophyletic,
and the subfamily Conoderinae was polyphyletic in all anal-
yses (fig. 1). Scolytinae, while also polyphyletic in all our anal-
yses (itself an interesting though not entirely unexpected
result given some previous analyses, e.g., Kuschel et al. 2000;
Gillett et al. 2014), were more closely related to other groups
in the CCCMS clade than to Platypodinae. The CEGH clade,
comprising over 13,000 species of the traditionally defined
“broad-nosed” weevils (e.g., Kuschel 1995;Marvaldi 1997),
was maximally supported in all analyses (fig. 1;supplementary
table S2,Supplementary Material online), but neither of the
two large subfamilies in this clade (Entiminae and
Cyclominae) was strictly monophyletic (fig. 1,supplementary
figs. S1–S6,Supplementary Material online; but see below).
This is notable because Entiminae are relatively well charac-
terized morphologically, and support for their monophyly has
been found in some previous analyses of molecular (Haran
et al. 2013;Gillett et al. 2014)andmorphologicaldata
(Marvaldi et al. 2014). Cyclominae are maximally supported
as a monophyletic group if the tribe Gonipterini is included
(fig. 1;supplementary figs. S1–S6, see also Supplementary
Material online), though analyses of AA data also placed
the entimine genus Naupactus in this group (but also in sev-
eral other positions depending on analysis and data; supple-
mentary figs. S1–S15,Supplementary Material online). The
nested and maximally supported position of Gonipterini in
the CEGH clade in all our analyses confirms its relationship to
the “broad-nosed” weevils, although further studies are
Phylogenomic Data Yield Insights into Weevil Evolution .doi:10.1093/molbev/msx324 MBE
5
Downloaded from https://academic.oup.com/mbe/advance-article-abstract/doi/10.1093/molbev/msx324/4765916
by University of Memphis - Library, sciaridae1@gmail.com
on 05 February 2018
required to clarify its exact position in this clade (Haran et al.
2013;Gillett et al. 2014) and to resolve the higher-level clas-
sification of the “broad-nosed” weevils more generally. This is
an interesting result both because of the lack of morpholog-
ical support for the monophyly of Cyclominae (Oberprieler
2010) and because of the ecological implications of our
results. The Australo-Pacific tribe Gonipterini, likely mono-
phyletic based on morphological characters (Oberprieler,
Caldara et al. 2014) and with ectophytic larvae (feeding ex-
ternally on plants) as far as known, has been variously allied
with taxa in both the CEGH and CCCMS clades (Kuschel
1995;Marvaldi 1997;Marvaldi et al. 2002;Hundsdo¨rfer et al.
2009;McKenna et al. 2009). With the placement of Hyperinae
(also with ectophytic larvae, see Supplementary Material on-
line) in the CEGH clade in all our analyses, and nearly uni-
formly as the sister group of the rest of the CEGH clade
(supplementary figs. S1–S15,Supplementary Material online),
our results firmly refute the hypothesis that this group may
belong in the CCCMS clade (e.g., Oberprieler et al. 2007,in
Curculioninae). These results also build on and reinforce with
nuclear phylogenomic data the hypothesis presented by
Haran et al. (2013) that the major phylogenetic split within
the higher weevils may also correspond to an important eco-
logical division within this clade, wherein the CEGH clade
mainly comprises groups with larvae that live (and
hence feed) external to the plant (ectophytic) whereas
the majority of taxa in the CCCMS clade seemingly retain
the putative ancestral larval habit of living (and hence also
feeding) inside plant tissues (endophytic) (Haran et al. 2013;
Oberprieler 2014c,2014d;Oberprieler, Anderson et al. 2014;
Oberprieler, Caldara et al. 2014). Exceptions to this pattern
(e.g., CCCMS clade: Cionus is ectophytic) combined with fu-
ture more broadly sampled studies of particularly the CCCMS
clade may allow for tests of diversification in relation to the
larval feeding habit within weevils. For example, by robustly
supporting the placement of Hyperinae and Gonipterini in
different positions in the CEGH clade (fig. 1), our data indicate
that changes in larval ecology may be associated with taxo-
nomic diversification. Both of these tribes differ from the
majority of “broad-nosed” weevils with ectophytic but gen-
erally subterranean larvae (Oberprieler, Anderson et al. 2014)
in having subaerial larvae. These larvae are instead active on
exposed plant surfaces, and several of their taxa, like those of
other ectophytic groups, have developed morphological and
behavioral traits likely connected with their subaerial or ecto-
phytic lifestyles (e.g., Costa et al. 2004;Oberprieler, Caldara
et al. 2014;Skuhrovec and Bogusch 2016).
Some relationships deep in our trees (e.g., involving the
placement of Belidae) were not strongly supported or con-
sistently found in analyses of NT data (supplementary figs. S1,
S3, and S5;supplementary table S2,Supplementary Material
online). In contrast, we consistently found strong support for
all the deep splits in trees reconstructed from AA data
(all >97% MLBS, 1.0 PP; supplementary figs. S2, S4, and S6
and, table S2,Supplementary Material online). Several of the
nodes that were strongly supported in our AA-based analyses
(fig. 1) but poorly supported in our NT-based ones, such as
the higher Curculionidae, the placement of the subfamily
Bagoinae and the CCCMS clade, have also been found in
other phylogenetic studies (Marvaldi 1997;Marvaldi et al.
2002;McKenna et al. 2009;Haran et al. 2013;Gillett et al.
2014), consistent with the observation that deep splits are
often more accurately reconstructed from analyses of AA
data (Rota-Stabelli et al. 2013;Cox et al. 2014). The observed
incongruence and differences in nodal support between our
NT- and AA-based analyses may relate to substitution model-
ing problems among large numbers of genes, codon usage
bias and/or mutational bias (e.g., Inagaki et al. 2004;Inagaki
and Roger, 2006;Rota-Stabelli et al. 2013;Cox et al. 2014)or
even patterns of missing data (e.g., Xi et al. 2016), all of which
may have contributed to the instability in the placement of
Platypodinae and a few other taxa. Holder et al. (2008) pro-
posed that the application of the CAT substitution model in
Bayesian analyses (in PhyloBayes) should offer a solution to
the AA–NT incongruence problem. As already noted, higher-
level (backbone) relationships were largely congruent and
strongly supported by both these data sets in our analyses.
This was true also for trees derived from analyses of AA
and NT data using the CAT–GTR models in PhyloBayes
v4.1 (Lartillot et al. 2009), but these trees showed some
incongruence in lower-level relationships, for example, in
Curculionidae (supplementary figs. S12 and S13,
Supplementary Material online). These trees also placed the
early-diverging family Belidae in a different phylogenetic po-
sition than in otherwise identical AA and NT analyses
employing the GTR model. The position of Belidae in both
of the CAT–GTR trees is the same as in the tree based on the
“dayhoff6”-recoded AA data (where the 20-AA data set was
recoded to represent only six Dayhoff categories; supplemen-
tary material and fig. S11, Supplementary Material online).
Trees generated from dayhoff6-recoded analyses of the AA
data (supplementary fig. S11,Supplementary Material online)
differed in lower-level relationships from trees found by both
the GTR and the CAT–GTR-based NT analyses (supplemen-
tary figs. S1, S3, S5, and S13,Supplementary Material online).
Analyzing our data sets under different models therefore did
not completely resolve the observed incongruence between
the NT and AA phylogenies. Based on these CAT–GTR and
dayhoff6-recoded analyses, some of the groupings present
only in the trees derived from GTR-based ML analyses of
NT data consist of clades characterized by similar composi-
tional issues (see also supplementary figs. S16 and S17,
Supplementary Material online).Wesuggest,then,thatthe
GTR-based ML NT trees are probably biased and affected by
systematic errors related to the uneven distribution of base
composition across the tree.
In NT-based partitioned ML analyses, the placement of
Belidae lacks strong nodal support (<65% MLBS, supplemen-
tary fig. S5,Supplementary Material online). In results of the
NT-based analyses using the GTR model, Belidae are recov-
ered either as the sister group of Attelabidae (supplementary
fig. S5,Supplementary Material online,63%MLBS)orofthe
clade (Attelabidae (Caridae (Curculionidae þBrentidae)))
(supplementary fig. S3,Supplementary Material online, 0.5
PP). But in all AA-based analyses using the GTR model,
Belidae are included in a strongly supported clade along
Shin et al. .doi:10.1093/molbev/msx324 MBE
6
Downloaded from https://academic.oup.com/mbe/advance-article-abstract/doi/10.1093/molbev/msx324/4765916
by University of Memphis - Library, sciaridae1@gmail.com
on 05 February 2018
with Nemonychidae: Rhinorhynchinae and Anthribidae
(supplementary fig. S2, 97% MLBS; supplementary fig.
S6,98%MLBS;supplementary fig. S4, 1.0 PP)—a relation-
ship not found by previous studies, and Attelabidae are
the sister group of the clade (Caridae
(Brentidae þCurculionidae)), with strong support (sup-
plementary figs. S2, S6,both99%MLBS;supplementary
fig. S4, 1.0 PP; Supplementary Material online). To explore
this incongruence further, codon-usage bias in the NT
data was also investigated using skew (a measure of de-
viation from expected base frequencies) for arginine, leu-
cine and serine (following Rota-Stabelli et al. 2013)
(supplementary fig. S17,Supplementary Material online).
Skew values range from þ1to1 (e.g., for the arginine
codon sets CGN/AGR: 1 ¼only CGN, –1 ¼only AGR,
0¼CGN and AGR used equally). Looking specifically at
arginine, skew values ranged from –0.37 to 0.45 and were
low (AGR-rich) in clades containing taxa whose phyloge-
netic placement was different in NT versus AA analyses
and/or poorly supported, for example, Attelabidae,
Belidae and Nemonychidae: Rhinorhynchinae (supple-
mentary fig. S17,Supplementary Material online).
Interestingly, in NT-based analyses these taxa tended to
group together with other taxa showing similar (low)
skew values for arginine. Although the CGN/AGR arginine
skew values for both species of Belidae in our analysis
range from –0.03 to -0.07 (mean: –0.05), these represent
only a slight bias compared with other taxa in the analysis
and the values for Nemonychidae and Anthribidae, while
similarly biased, are actually much higher (e.g.,
Anthribidae mean: –0.29; Rhinorhynchinae mean: –0.21;
supplementary fig. S17,Supplementary Material online).
Moreover, the mean of arginine skew values for
Attelabidae is 0.13, indicating a bias instead toward
codons in the CGN set. This is interesting because in all
trees except the C3-only tree and the partitioned NT tree
(in which Belidae were the sister-group of Attelabidae),
Belidae were the sister-group of the clade
Anthribidae þRhinorhynchinae with varying MLBS sup-
port (particularly high in AA analyses), suggesting a pos-
sible but unclear effect of codon usage bias on the NT-
based analyses. Synonymous codon-usage bias has been
shown to produce incongruence between NT- and AA-
based phylogenies (Inagaki et al. 2004;Rota-Stabelli et al.
2013;Cox et al. 2014), and the alternative placements of
the pinhole-borer weevils Austroplatypus and
Notoplatypus may be the clearest example of incongru-
ence explained by codon-usage bias (and related factors)
in our study. These two genera, while exhibiting similar
bias for serine codon sets, demonstrate bias in opposite
directions for both arginine and leucine codon sets (sup-
plementary fig. S17,Supplementary Material online),
which may both explain their alternate placements (and
low support for Platypodinae monophyly) among our
analyses (supplementary figs. S1–S15 and table S2,
Supplementary Material online) and reinforce our con-
clusion that mutational bias may have impacted the
placement of these genera in our NT-based trees.
Reconstructing the Temporal Framework of Weevil
Evolution
The radiation of weevils in association with plants is—on
account of the large number of species involved—one of
the most notable plant–insect coevolutionary stories (e.g.,
Oberprieler et al. 2007,McKenna et al. 2009). We leveraged
our robust AHE phylogeny to investigate the temporal origins
of major weevil groups via divergence time analyses imple-
mented in RelTime (Tamura et al. 2012,2013;Kumar et al.
2016). Our divergence time analyses support an early Triassic
origin of the clade Phytophaga (weevils plus their sister group,
the superfamily Chrysomeloidea [leaf and longhorned bee-
tles]; Haddad and McKenna 2016)andalateTriassic(Norian
208.5–228 Ma) to early Jurassic (Toarcian 174.1–182.7 Ma)
origin for the divergence of early weevil lineages (fig. 1,sup-
plementary figs. S18–S25,Supplementary Material online).
This latter result implies an age of origin for the weevils
40–50 Ma older than the oldest known weevil fossils
(Nemonychidae; Arnoldi 1977;supplementary table S4,
Supplementary Material online). Consistent with McKenna
et al. (2009) and Gunter et al. (2016), our results further sug-
gest that the diversification of weevils into families (except
Brentidae and Curculionidae) occurred on gymnosperms in
the Mesozoic and that the early branching events occurred
well before the origin of angiosperms. Age estimates for family
and other higher-level nodes in our study are generally higher
than in other studies of weevils (McKenna et el. 2009;Misof
el al. 2014;Gunter et al. 2016) and higher than estimates in
higher-level studies of Coleoptera (beetles) that include wee-
vils (Hunt et al. 2007;McKenna and Farrell 2009;McKenna
et al. 2015), with the exception of the Coleoptera-wide study
of Toussaint et al. (2017), which found similar estimated ages
for the deep splits in the clade Phytophaga. Unlike previous
divergence time studies that suggested an age of origin and
diversification largely contemporaneous with the evolution of
flowering plants, the older age estimates from our RelTime
analysis (fig. 1) effectively exclude the possibility that the ear-
liest ancestral weevils fed on angiosperms but instead suggest
that these beetles exclusively fed on and diversified in associ-
ation with nonangiosperm (likely gymnosperm) hosts, lend-
ing further support to the hypothesized ancestral association
of weevils with conifers (Oberprieler et al. 2007).
Our new timetree indicates that most of the extant family/
subfamily-level lineages/splits occurred after the origin of the
flowering plants (fig. 1). The fungus-feeding habits of most
Anthribidae (most are associated with fungi growing in/on
angiosperm wood) may have evolved from feeding on the
decaying cones of conifers or cycads (Oberprieler 1999;
Oberprieler et al. 2007), and a similar trajectory may have
occurred in the Attelabidae (also largely associated with fungi
on angiosperms but with a few ancient lineages on conifers;
Oberprieler et al. 2007). The angiosperm-associated sister
groups Brentidae and Curculionidae diverged during the
Late Jurassic (Tithonian, 145.0–152.1 Ma) to early–middle
Cretaceous (Aptian, 112.0–125.0 Ma). This overlaps with
the estimated timing of first appearance of angiosperms
(fossils 132–141 Ma [Brenner 1996], molecular dating
Phylogenomic Data Yield Insights into Weevil Evolution .doi:10.1093/molbev/msx324 MBE
7
Downloaded from https://academic.oup.com/mbe/advance-article-abstract/doi/10.1093/molbev/msx324/4765916
by University of Memphis - Library, sciaridae1@gmail.com
on 05 February 2018
140–180 Ma [Bell et al. 2005]) and monocots (stem-group
origin 140–150 Ma [Chaw et al. 2004]; crown-group origin
112 Ma, from fossil pollen [Friis et al. 2004], 130 Ma molecular
dating [Bremer 2000]) and includes the interval over which
angiosperms rose to ecological dominance (70–120 Ma;
Lidgard and Crane 1988,1990).
Our new timetree also suggests that the K–Pg extinction
event played a larger role in the diversification of the two
major clades of higher weevils than previously thought. The
estimated dates for the diversification of the CCCMS
(>34,000 species) and CEGH (>13,000 species) clades in
our timetree are very close to the K–Pg boundary and not
clearly pre-K–Pg, as reported in previous studies (McKenna
et al. 2009;Gunter et al. 2016)—though the estimated age of
origin for these groups is similar to that of previous studies
(85–90 Ma; fig. 1). Therefore, diversification of these two
clades may actually be more closely correlated with post-
K–Pg (early-to-mid-Cenozoic) evolutionary processes and
was perhaps initiated by the K–Pg event, but then facilitated
by early-to-middle Paleogene warming and associated plant
diversification (especially of core eudicots) (McKenna et al.
2009). The extraordinarily ecologically diverse CCCMS and
CEGH clades also contain most weevil species and are among
the most species-rich clades of plant-feeding animals on Earth.
We explored the effects of using alternative fossils as age
constraints in our timetree analysis (supplementary figs. S18–
S25 and supplementary table S4,Supplementary Material on-
line). Among other alternatives (see supplementary material
and supplementary tables S4 and S5, Supplementary Material
online), we explored two alternate minimum ages for each of
the CEGH and CCCMS clades representing significantly dif-
ferent interpretations for the minimum ages of the relevant
taxa (Entiminae for CEGH constraint, Scolytinae for CCCMS
constraint, Eocene vs. Cretaceous minimum age alternatives
for each clade). Ages of the early-divergent nodes are not
significantly affected by the minimum age calibration placed
ontheCCCMSandCEGHclades(fig. 1,supplementary table
S4,Supplementary Material online). For example, the age of
the divergence between Brentidae and Curculionidae remains
near the Jurassic–Cretaceous boundary, which is approxi-
mately 40 Ma older than any known fossil assignable to this
clade (Gunter et al. 2016). When a less conservative (older)
fossil calibration (the controversial fossil Microborus inertus
(Scolytinae) [Cognato and Grimaldi, 2009]) is applied as a
minimum age for the CCCMS clade (supplementary figs.
S20, S21, S24, and S25,Supplementary Material online), it
causes the crown age of the higher Curculionidae to be
very near that of the crowns of the CEGH and CCCMS de-
scendent nodes (supplementary figs. S21 and S24,
Supplementary Material online). This is quite unlike the
results from our other analyses, which use a more conserva-
tive (younger) Eocene Baltic amber fossil to constrain the age
of the CCCMS clade (supplementary figs. S18, S19, S22, and
S23,Supplementary Material online). Therefore, although it is
possible that these internodes are very short, it may also be
worth further investigating whether the Microborus fossil
truly belongs to the 98þ-My-old Burmese amber (see
Gunter et al. 2016). Another major and consistent difference
between the results of using a younger or older calibration age
for Scolytinae is a dramatic compression in the time intervals
(internodes) between the Brentidae–Curculionidae clade and
the CCCMS–CEGH clade (especially different in supplemen-
tary figs. S21 and S25,Supplementary Material online).
Dorotheus guidensis, a fossilized elytron from the late
Cretaceous of southern Chile (Kuschel 1959), is considered
to be the oldest fossil evidence of Entiminae and was there-
fore used to constrain the age of the CEGH clade (supple-
mentary figs. S18–S21,Supplementary Material online).
However, due to uncertainty concerning its identification
(characters of the elytra observed in Dorotheus are possibly
plesiomorphic, occurring also in the CCCMS clade), we sep-
arately applied a minimum age for the CEGH clade using
more reliably identified entimines from Baltic amber
(Polydrusus spp.: Yunakov and Kirejtshuk 2011). However,
we found no significant difference in the results for the age
of the CEGH clade in analyses using either the older
(Dorotheus)oryounger(Polydrusus)constraint(supplemen-
tary figs. S18–S25,Supplementary Material online).
Conclusions
This work establishes a uniquely robust phylogenetic and
temporal framework for evaluating patterns of ecological
and taxonomic diversification in weevils, something that
has not yet been possible due, in part, to the lack of a robust
phylogenetic framework (e.g., Oberprieler et al. 2007). Further,
it provides molecular support for several groups morpholog-
ically defined in Curculionoidea, indicates strong support for
other relationships and suggests the need for future changes
to the classification of weevils. Most notably, it solidifies the
phylogenetic relationships and classificatory position of a
number of taxa of controversial placement, such as
Ithycerinae, Microcerinae, Platypodinae and Bagoinae.
Further, our results demonstrate the utility of AHE for recon-
structing the phylogeny of weevils—especially regions of the
tree that have long been recalcitrant to resolution—and sug-
gest several ways in which our AHE probe set (the first for
beetles) might be improved, and taxon sampling expanded,
to help further reconstruct the phylogeny and evolution of
weevils. The probes we use in this article were designed to be
informative for analyses spanning all beetles and will likely
also be useful for analyses of insect groups closely related to
beetles, such as lacewings, dobsonflies, snakeflies, scorpionflies
and twisted-wing parasites. Our AHE probes proved highly
effective at robustly resolving many critical nodes within wee-
vils, including nodes spanning the backbone and terminal
nodes of the tree. Our timetree suggests that some critical
early events in the coevolution of weevils and plants may
have occurred 40–50 My earlier than previously thought
and that major transitions in the evolutionary history and
diversification of both groups are closely correlated. Our di-
vergence time analyses suggest that changes in the floristically
dominant plant group (gymnosperms to angiosperms) were
tracked by the evolution of new weevil taxa radiating first in
association with angiosperms and again later, after the K–Pg
mass extinction event. The latter may have promoted the
Shin et al. .doi:10.1093/molbev/msx324 MBE
8
Downloaded from https://academic.oup.com/mbe/advance-article-abstract/doi/10.1093/molbev/msx324/4765916
by University of Memphis - Library, sciaridae1@gmail.com
on 05 February 2018
radiation of the CEGH and CCCMS clades, the two most
speciose groups of higher weevils (fig. 1). With a robust back-
bone phylogeny in place, including a clear understanding of
the composition of the family Curculionidae, we are now well
positioned to further sample/resolve relationships within this
extraordinarily diverse family and further explore the tempo-
ral dimension of evolution at the beetle–plant interface.
Materials and Methods
Probe Design and Locus Selection
For probe design, we chose 26 species (here called
“models”) whose genomes and/or transcriptomes had
been sequenced (supplementary table S1,Supplementary
Material online, also see Haddad et al. 2018). These com-
prised six species of Curculionoidea: Arrhenodes minutus
(Transcriptome, 1KITE unpublished data), Dendroctonus
ponderosae (Transcriptome, PRJNA178770), Ips typogra-
phus (Transcriptome, 1KITE unpublished data), Pissodes
strobi (Transcriptome, PRJNA186387), Rhynchophorus fer-
rugineus (Transcriptome, PRJNA79205), Sitophilus oryzae
(Transcriptome, PRJNA79583, Pauchet et al. 2010), four of
Cerambycidae: Anoplophora glabripennis (Genome,
PRJNA274806, McKenna et al. 2016), Phymatodes amoenus
(Genome, Mitchell R, unpublished data), Rhamnusium bicolor
(Transcriptome, 1KITE unpublished data), Xylotrechus colonus
(Genome, Mitchell R, unpublished data), six of Chrysomelidae:
Callosobruchus maculatus (Genome, PRJNA308906), Diabrotica
undecimpunctata (Genome, Robertson H, unpublished data),
Gastrophysa viridula (Transcriptome, PRJNA79577, Pauchet
et al. 2010), Leptinotarsa decemlineata (Transcriptome,
PRJNA79581, Pauchet et al. 2010), Donacia marginata
(Transcriptome, 1KITE unpublished data), Chrysomela tremu-
lae (Transcriptome, PRJNA79423, Pauchet et al. 2010), one of
Cryptophagidae: Atomaria fuscata (Transcriptome, 1KITE
unpublished data), one of Tenebrionidae: Tribolium casta-
neum (TCAST) (Genome, PRJNA12540, Richards et al. 2008),
one of Carabidae: Calosoma scrutator (Genome, McKenna D,
unpublished data), one of Coccinellidae: Harmonia axyridis
(Genome, McKenna D, unpublished data), one of
Hydroscaphidae: Hydroscapha redfordi (Genome, McKenna
D, unpublished data), one of Cupedidae: Priacma serrata
(Genome, McKenna D, unpublished data), one of Byturidae:
Byturus ochraceus (Transcriptome, 1KITE unpublished data),
one of Cleridae: Thanasimus formicarius (Transcriptome,
1KITE unpublished data), one of Strepsiptera: Mengenillidae:
Mengenilla moldryzki (Genome, PRJNA181027), and one of
Megaloptera: Corydalidae: Chauliodes pectinicornis (Genome,
McKenna D, unpublished data).
Anchored hybrid enrichment probes were developed tar-
geting 941 orthologous nuclear loci of known utility for phy-
logenetic analysis (e.g., Niehuis et al. 2012;Misof et al. 2014;
also see Haddad et al. 2018). These loci were located in con-
served “anchor regions” of the genomes and/or transcrip-
tomes of the model species, were flanked by less conserved
regions and were selected based on their presence as 1:1
orthologs across the model species used in this study. Loci
were therefore included only if they were single-copy
orthologs. The probes were designed to target loci suitable
for use across all insects and to ensure utility across all
Neuropteroidea (Coleoptera, Neuropterida, Strepsiptera).
The 941 target loci were selected from a pool of 1,200
candidate loci that were identified by seeking the intersection
of a genome-based data set [4,485 1:1 orthologs from
Holometabola; Niehuis et al. (2012)]withatranscriptome-
based data set [1,478 1:1 orthologs from Misof et al. (2014)].
Those resulting candidate loci were then sought in the afore-
mentioned genomes and transcriptomes from the 26 model
species, to confirm their presence and assess their phyloge-
netic utility using criteria enumerated by Lemmon et al.
(2012: 728–729). Accordingly, we filtered the candidate loci
for unique (single-copy), widely dispersed genomic loci lacking
any indels or repetitive elements and with a 240-bp con-
served center of the probe region but variable sites in at least
one of the flanks. A core set of 236 loci with variation and
phylogenetic utility across Arthropoda, but primarily in
Insecta, was chosen for inclusion in the probe set, along with
705 loci chosen because they were phylogenetically informa-
tive across Neuropteroidea (Coleoptera, Strepsiptera and
Neuropterida; McKenna and Farrell 2010;Beutel and
McKenna 2016;McKenna 2016). Probes were tiled approxi-
mately every 50 bp for each of the 26 model species (2.4
coverage per species), starting at the beginning of the align-
ment. Alignments for the 941 target loci containing the 26
model species were used to identify enrichment probes.
Final alignments and probe sequences are available from
Dryad (accession numbers doi: 10.5061/dryad.v0b7v and
doi:10.5061/dryad.42d3b).
Taxon Sampling and DNA Extraction
Our taxon sample included 63 weevil species representing all
extant families of Curculionoidea and all subfamilies of
Curculionidae following the classification of the Handbook
of Zoology,Coleoptera Vol. III (Leschen and Beutel 2014)(sup-
plementary table S3, see details in Supplementary Material
online; see details in supplementary material). Genomic DNA
was extracted from one to six legs, thoracic muscle or the
whole body of specimens preserved in 80–100% ethanol,
depending on their size and state of preservation. Total ge-
nomic DNA (100–500 ng was ultimately used) was
extracted from air-dried specimens using the OmniPrep kit
(G Biosciences, St. Louis, MO) and was treated with RNase A.
DNA amount and concentration was estimated using a Qubit
fluorometer, and DNA quality (degree of fragmentation/deg-
radation and contamination with RNA) was assessed via gel
electrophoresis. Voucher specimens and specimen parts are
preserved in 100% ethanol in the McKenna Laboratory at the
University of Memphis, at the Harvard University Museum of
Comparative Zoology or at the Commonwealth Scientific
and Industrial Research Organization.
Library Preparation, Enrichment, and Sequencing
Genomic DNAs were sent to the Center for Anchored
Phylogenomics (www.anchoredphylogeny.com) at Florida
State University (FSU) for library preparation, enrichment
Phylogenomic Data Yield Insights into Weevil Evolution .doi:10.1093/molbev/msx324 MBE
9
Downloaded from https://academic.oup.com/mbe/advance-article-abstract/doi/10.1093/molbev/msx324/4765916
by University of Memphis - Library, sciaridae1@gmail.com
on 05 February 2018
and high-throughput sequencing. Two rounds of library prep-
aration and sequencing were performed. Data were collected
following the general methods of Lemmon et al. (2012).Each
genomic DNA sample (ranging from 80 to 1,000 ng) was
sonicated to a fragment size of 150–350 bp using an E220
Focused-ultrasonicator (Covaris Inc., www.covarisinc.com,
Woburn, MA). Library preparation and indexing was per-
formed on a Beckman-Coulter Biomek FXp liquid-handling
robot following a protocol modified from Meyer and Kircher
(2010). One modification to this protocol was a size-selection
step after blunt-end repair using SPRIselect beads (Beckman-
Coulter Inc., www.beckman.com, Brea, CA; 0.9ratio of bead
to sample volume). Indexed samples were then pooled at
equal quantities (12–16 samples per pool), and enrichment
was performed on each multisample pool using an Agilent
Custom SureSelect kit (Agilent Technologies Inc., www.agi-
lent.com, Santa Clara, CA) containing probes designed to
anneal to the targeted loci. After enrichment, each set of
six (for the first round) or three (for the second round) en-
richment reactions was pooled in equal quantities for PE150
sequencing at the Translational Science Laboratory in the
College of Medicine at FSU on three lanes of an Illumina
HiSeq 2000 sequencer (Illumina Corp., www.illumina.com,
San Diego, CA), shared with samples from other projects
(77 Gb raw DNA sequence data were used for this study).
Read Processing and Assembly
Quality control and assembly for sequenced raw reads was
performed for each species using the AHE bioinformatics
pipeline described in Prum et al. (2015). Prior to assembly,
overlapping reads were identified and merged following the
methods of Rokyta et al. (2012). Most (50–75%) of the se-
quenced library fragments had an insert size of 150–300 bp.
Therefore, the majority of the paired reads were overlapping
such that they could be merged before assembly. For each
read, we computed the probability of obtaining the observed
number of matches by chance for each degree of overlap and
selected the degree of overlap that produced the lowest prob-
ability. A P-value of less than 10
10
was used to determine
when to merge reads. Mismatched reads were reconciled us-
ing base-specific quality scores after merging, which were
combined to form the new quality scores for the merged
read (see Rokyta et al. [2012] for details). We kept separate
the reads that failed to meet the probability threshold but still
used them in the final assembly. The merging process gener-
ated fastq files; one containing merged reads and two con-
taining the unmerged reads. Merged reads were assembled
(using the script Assembler.class) by mapping reads to the
proberegionsforeachlocus(seetaxausedasreferencesbe-
low) and by extending the flanking regions using a de novo
approach. Orthology-matched NT probe regions from the
following genomes and transcriptomes were used as referen-
cesforlocusassembly:A. minutus (Brentidae, transcriptome,
768 loci), P. strobi (Curculionidae, partial transcriptome, 220
loci), R. ferrugineus (Curculionidae, transcriptome, 789 loci),
S. oryzae (Curculionidae, transcriptome, 746 loci), D. ponder-
osae (Curculionidae, transcriptome, 780 loci) (Andersson
et al. 2013), and T. castaneum (Tenebrionidae, genome,
941 loci) (Richards et al. 2008). The TCAST reference data
covered 100% of the targeted loci, so we used it as the main
reference gene set. Assembly parameters and scripts were
obtained from Prum et al. (2015). After the reads were as-
sembled, we checked for possible cross-contamination using
(for each taxon) an all-versus-all Blast search (following
Camacho et al. 2009).
AHE Data Harvesting
Although the assembled AHE fasta files were mapped in a
previous step, we also employed the protein-based orthology
search pipeline Orthograph for strict orthology assessment
(Petersen et al. 2017). Orthograph removes possible paralo-
gous genes using HMM (hidden Markov model)-based
orthology searches of protein-translated sequences. For this
pipeline, the official gene sets (OGSs) from three insect taxa
selected from among the holometabolous insects included in
OrthoDB 7 (Waterhouse et al. 2013;Kriventseva et al. 2015)
were used as a reference for orthology prediction. These in-
cludedonebeetle(TCAST:Richards et al. 2008;theonly
beetle genome available at the time in OrthoDB) and two
other holometabolous insects: Danaus plexippus (DPLEX;
Lepidoptera: Danaidae, Zhan et al. 2011)andNasonia vitri-
pennis (NVITR;Hymenoptera:Pteromalidae;Werren et al.
2010). We used the database tool OrthoDB 7 (Waterhouse
et al. 2013;Kriventseva et al. 2015) to generate a table of
clusters of orthologous genes (COGs) for the three selected
OGSs. In this step, the TCAST 941 AHE reference locus set
was remapped by BlastX (E<1e–6) against the reference
OGS for TCAST (OGS 3.0; Richards et al. 2008). This exercise
recovered 663 genes (each comprised sequences from one or
more of the targeted loci). AHE reference assemblies can
generate spurious duplicates of single-copy orthologs if the
flanking regions of different targeted loci (here, different exons
from the same gene) overlap between loci. This occurs when
the assembled flanking regions extend beyond short introns
flanking a target exon, resulting in the recovery of neighboring
exons that may also be targets. Based on the BlastX results, we
ultimately settled on a total of 522 COGs that matched with
all single-copy COGs of the aforementioned three OGSs in
the Orthograph pipeline (we excluded 141 of the 663 total
genes represented by our 941 target loci because they had
multiple copies in at least one of the three OGSs).
The COG tables and OGS sequences selected from the
previous step were loaded into Orthograph as the reference
database for subsequent strict protein-based orthology
searches. All 63 fasta files from the assembly steps (one for
each terminal taxon) were used as Orthograph input. In the
first step of the Orthograph pipeline, all DNA sequences were
translated into the six possible reading frames (finding a cor-
rect AA translation point for each locus), and then the result-
ing library of AA sequences was searched using profile HMMs
that were trained by the three OGSs we selected from
OrthoDB 7. For each taxon-based result from the
Orthograph pipeline, the orthology of our targeted (522)
genes was assessed using a reciprocal Blast search and the
results were stored in both AA and NT format, following
Petersen et al. (2017). The resulting fasta-formatted NT files
Shin et al. .doi:10.1093/molbev/msx324 MBE
10
Downloaded from https://academic.oup.com/mbe/advance-article-abstract/doi/10.1093/molbev/msx324/4765916
by University of Memphis - Library, sciaridae1@gmail.com
on 05 February 2018
for each species were screened for vector contamination us-
ing UniVec (Cochrane and Galperin 2010). We used an all-
versus-all Blast to further assess our data set for apparent
cross-contamination.
Phylogenomic Pipeline for the AHE Analysis
Orthology search pipelines generate fasta files using OrthoDB
7 identifiers (e.g., EOG7B3CRG.nt.fas) for every species and
gene. Thus, a custom bioinformatics pipeline was required to
process the files. Specifically, reference taxa had to be re-
moved and headers on each fasta file had to be modified
for later phylogenetic analysis. The AA and NT fasta files
from the Orthograph pipeline (these lack introns) included
information from the results in their headers, in addition to
three OGS sequences for each gene. The headers on each
sequenceweremodifiedtoHamstradformat(Orthograph
package by Petersen et al. [2017]). Then, the reference genes
from the OGSs were removed for each fasta file. As such, each
fasta file then had only one target sequence for each gene and a
taxon name that includes an OrthoDB 7 ID. After this trim-
ming step, and prior to alignment, we used a script
“summarize_orthograph_results.pl” from Orthograph by
Petersen et al. (2017) to combine the fasta files for each gene
basedontheOrthoDB7IDs.Next,eachAAsequencewas
aligned in MAFFT 7 with L-INS-i and all default options. After
AA sequences were aligned each NT sequence was aligned by
PAL2NAL using the corresponding codon alignments
(Suyama et al. 2006). Before concatenation, we changed the
Hamstrad header in all resulting files to a simple taxon code
using a custom script included in the pipeline. Finally, the final
aligned files were concatenated with AMAS 0.97 (Borowiec
2016). All scripts and other details of our pipeline are available
from Dryad (accession number doi:10.5061/dryad.42d3b).
Phylogenetic Analysis
Most analyses were run on the HPC (high-performance com-
puting) cluster at the University of Memphis. Model selection
and partitioning for both the AA and NT data sets was
performed using PartitionFinder 1.1.1 (Lanfear et al. 2012).
The AA and NT matrices were analyzed separately in
RAxML (Stamatakis 2014) (10 replicate ML searches; 1,000
rapid bootstrap replicates). Results from the bootstrap anal-
yses were mapped onto the resulting ML trees (AA and NT).
Trees based on nonpartitioned and partitioned AA and NT
data sets are provided (supplementary figs. S1–S2 and S5–S6,
respectively, Supplementary Material online). We used the
command line version of MEGA7 (Kumar et al. 2016)for
codon usage bias analyses (e.g., Inagaki et al. 2004;
Inagaki and Roger 2006;Rota-Stabelli et al. 2013;Cox
et al. 2014) based on the partitioned NT ML tree. We
also analyzed the AA and NT data using BI implemented
in MrBayes 3.2.5 (Ronquist et al. 2012)(supplementary
figs. S3-S4) and PhyloBayes-MPI v1.7a (Lartillot et al.
2013)(supplementary figs. S12-S13,Supplementary
Material online). Bayesian analyses were only conducted
on the nonpartitioned (concatenated) data set (see
Supplementary Material online). The coalescent species
tree analysis was performed on our AA and NT data sets
using ASTRAL 4.11.1 (Mirarab, Bayzid, et al. 2014)withthe
weighted statistical binning scripts from Mirarab, Reaz et al.
(2014)andBayzid et al. (2015) (Supplementary Material on-
line; results are presented in supplementary figs. S14-S15,
Supplementary Material online). Several papers have consid-
eredapossiblepreferenceforresultsofAA-overNT-based
phylogenetic analyses, especially for resolving deep divergen-
ces (e.g., Inagaki et al. 2004;Cox et al. 2014). Although the
major weevil lineages diverged much later than groups stud-
ied by those authors, our analyses of codon usage and base
composition suggested that similar compositional biases
may also be present in our weevil NT data set (see Results
and Discussion). Our phylogenetic analyses also strongly in-
dicated saturation at (at least) the third-codon positions. We
therefore used our AA-based results for summarizing our
results and conducting divergence time analyses.
Divergence Time Analyses
We calculated divergence times using the program RelTime
(Tamura et al. 2012,2013;Kumar et al. 2016). Local clocks
were used for each lineage, with no clock rates merged (see
Bond et al. 2014). The LG substitution model was employed
withagammadistributedmodelandfivediscretegamma
categories. The “Use all site” option was used for all analyses.
In our analyses, we used dated fossils to apply constraints to
up to ten nodes in the preferred phylogeny (Fig. 1; partitioned
ML, AA data set) and tested different combinations of min-
imum/maximum ages to gauge the effect of including/ex-
cluding some key fossils and min–max combinations (see
details in Supplementary Material online).
Supplementary Material
Supplementary data are available at Molecular Biology and
Evolution online.
Acknowledgments
We thank Matt Krull, Hannah Ralicki and Michelle Kortyna,
who assisted with lab work, and Robert Mitchell, Hugh
Robertson and Yannick Pauchet for providing unpublished
genomes or transcriptomes to assist with probe design. We
thank Alexander Donath, Lars Podsiadlowski, Xin Zhou, Karl
Kjer, James Kjer, Bernhard Misof, Ralph S. Peters, Jonas Beller,
Martin Kubiak, Eric Anton, Karen Meusemann, Adam
Siplinski, Hermes Escalona, Rolf Beutel, and the 1KITE
Coleoptera subproject for allowing access to unpublished
transcriptome assemblies. We also thank Udo Schmidt for
permission to use his weevil photos in figure 1.Wethank
the following colleagues for helpful discussions and/or for
specimens used in this study: Robert Anderson, Sam
Brown, Roberto Caldara, Steven Chown, Larry Kirkendall,
Richard Leschen, Harald Letsch, Geoff Monteith, Al
Newton, Charles O’Brien, Serban Proches, Petr Svacha, and
Margaret Thayer. We acknowledge funding to D.D.M. from
the University of Memphis FedEx Institute, the University of
Memphis College of Arts and Sciences, the U.S. National
Science Foundation (DEB1355169), and the U.S.
Department of Agriculture, Animal and Plant Health
Phylogenomic Data Yield Insights into Weevil Evolution .doi:10.1093/molbev/msx324 MBE
11
Downloaded from https://academic.oup.com/mbe/advance-article-abstract/doi/10.1093/molbev/msx324/4765916
by University of Memphis - Library, sciaridae1@gmail.com
on 05 February 2018
Inspection Service (cooperative agreement 16-8130-0547-
CA). A.E.M. acknowledges support from research grants by
the National Scientific and Technical Research Council
(CONICET, Argentina, PIP 0355) and the National Agency
of Promotion of Science (ANPCyT, Argentina, PICT 2573).
References
Alonso-Zarazaga MA. 2007. On terminology in Curculionoidea
(Coleoptera). Bol SEA. 40:210.
Anderson RS. 1993. Weevils and plants: phylogenetic versus ecological
mediation of evolution of host-plant associations in Curculioninae
(Coleoptera, Curculionidae). Mem Entomol Soc Can. 125(S165):
197–232.
Anderson RS. 1995. An evolutionary perspective on diversity in
Curculionoidea. Mem Entomol Soc Wash. 14:103–114.
Andersson MN, Grosse-Wilde E, Keeling CI, Bengtsson JM, Yuen MMS,
Li M, Hillbur Y, Bohlmann J, Hansson BS, Schlyter F. 2013. Antennal
transcriptome analysis of the chemosensory gene families in
the tree killing bark beetles, Ips typographus and Dendroctonus
ponderosae (Coleoptera: Curculionidae: Scolytinae). BMC
Genomics 14(1): 198.
Arnoldi LV. 1977. Eobelidae. In: Arnoldi LV, Zherikhin VV, Nikitrin LM,
Ponomarenko AG, editors. Mesozoic Coleoptera. Moscow (Russia):
Nauka Publishers. p. 144–176. [English translation, 1991, New Delhi
(India): Oxonian Press. p. 195–241].
Bayzid MS, Mirarab S, Boussau B, Warnow T. 2015. Weighted statistical
binning: enabling statistically consistent genome-scale phylogenetic
analyses. PLoS One 10(6): e0129183.
Bell C, Soltis DE, Soltis PS. 2005. The age of the angiosperms: a molecular
timescalewithoutaclock.Evolution 59(6): 1245–1258.
Beutel RG, McKenna DD. 2016. Coleoptera: systematic position,
basal branching pattern and early evolution. In: Leschen RAB,
Beutel RG, editors. Handbook of zoology. Vol. I: Arthropoda:
Insecta. Coleoptera, Beetles. Berlin (Germany): Walter de
Gruyter. p. 1–11.
Bocak L, Barton C, Crampton-Platt A, Chesters D, Ahrens D, Vogler AP.
2014. Building the Coleoptera tree-of-life for >8000 species: compo-
sition of public DNA data and fit with Linnaean classification. Syst
Entomol. 39(1): 97–110.
BondJE,GarrisonNL,HamiltonCA,GodwinRL,HedinM,AgnarssonI.
2014. Phylogenomics resolves a spider backbone phylogeny and
rejects a prevailing paradigm for orb web evolution. Curr Biol.
24(15): 1765–1771.
Borowiec ML. 2016. AMAS: a fast tool for alignment manipulation and
computing of summary statistics. PeerJ 4:e1660.
Breinholt JW, Earl C, Lemmon AR, Moriarty LE, Xiao L, Kawahara AY.
2018. Resolving relationships among the megadiverse butterflies and
moths with a novel pipeline for anchored phylogenomics. Syst Biol.
67(1): 78–93.
Bremer K. 2000. Early Cretaceous lineages of monocot flowering plants.
Proc Natl Acad Sci U S A. 97(9): 4707–4711.
Brenner G. 1996. Evidence for the earliest stage of angiosperm pollen
evolution: a paleoequatorial section from Israel. In: Taylor DW,
Hickey LJ, editors. Flowering plant origin, evolution, and phylogeny.
New York: Chapman & Hall. p. 91–115.
CamachoC,CoulourisG,AvagyanV,MaN,PapadopoulosJ,BealerK,
Madden TL. 2009. BLASTþ: architecture and applications. BMC
Bioinformatics 10:421.
Chaw SM, Chang CC, Chen HL, Li WH. 2004. Dating the monocot–dicot
divergence and the origin of core eudicots using whole chloroplast
genomes. J Mol Evol. 58(4): 424–441.
Cochrane GR, Galperin MY. 2010. The 2010 nucleic acids research data-
base issue and online database collection: a community of data
resources. Nucleic Acids Res. 38(Database issue): D1–D4.
Cognato AI, Grimaldi D. 2009. 100 million years of morphological con-
servation in bark beetles (Coleoptera: Curculionidae: Scolytinae). Syst
Entomol. 34(1): 93–100.
Costa JT, Fitzgerald TD, Pescador-Rubio A, Mays J, Janzen DH. 2004.
Social behavior of larvae of the neotropical processionary weevil
Phelypera distigma (Boheman) (Coleoptera: Curculionidae:
Hyperinae). Ethology 110(7): 515–530.
Cox CJ, Li B, Foster PG, Embley TM, Civan P. 2014. Conflicting phylog-
enies for early land plants are caused by composition biases among
synonymous substitutions. Syst Biol. 63(2): 272–279.
Crowson RA. 1985. The systematic position of Nemonyx Redtenbacher
(Coleoptera: Curculionoidea: Nemonychidae). Entomol Gener. 11(1/
2): 57–58.
Farrell BD. 1998. “Inordinate Fondness” explained: why are there so many
beetles? Science 281(5376): 555–559.
Friis EM, Pedersen KR, Crane PR. 2004. Araceae from the early
Cretaceous of Portugal: evidence on the emergence of monocoty-
ledons. Proc Natl Acad Sci U S A. 101(47): 16565–16570.
Gillett CPDT, Crampton-Platt A, Timmermans MJTN, Jordal BH,
Emerson BC, Vogler AP. 2014. Bulk de novo mitogenome assembly
from pooled total DNA elucidates the phylogeny of weevils
(Coleoptera: Curculionoidea). MolBiolEvol. 31(8): 2223–2237.
Gunter NL, Oberprieler RG, Cameron SL. 2016. Molecular phylogenetics
of Australian weevils (Coleoptera: Curculionoidea): exploring rela-
tionships in a hyperdiverse lineage through comparison of indepen-
dent analyses. Aust Entomol. 55(2): 217–233.
Haddad S, McKenna DD. 2016. Phylogeny and evolution of the super-
family Chrysomeloidea (Coleoptera: Cucujiformia). Syst Entomol.
41(4): 697–716.
Haddad S, Shin S, Lemmon AR, Lemmon EM, Svacha P, Farrell B,
Slipi
nski
A, Windsor D, McKenna DD. 2018. Anchored hybrid enrichment
provides new insights into the phylogeny and evolution of long-
horned beetles (Cerambycidae). Syst Entomol. 43(1): 68–89.
Haran J, Timmermans MJTN, Vogler AP. 2013. Mitogenome sequences
stabilize the phylogenetics of weevils (Curculionoidea) and establish
the monophyly of larval ectophagy. Mol Phylogenet Evol.67(1):
156–166.
Holder MT, Zwickl DJ, Dessimoz C. 2008. Evaluating the robustness of
phylogenetic methods to among-site variability in substitution pro-
cesses. Philo Trans R Soc B. 363(1512): 4013–4021.
Holloway BA. 1982. Anthribidae (Insecta: Coleoptera). Fauna of New
Zealand 3. Manaaki (New Zealand): Whenua Press.
Hulcr J, Atkinson TH, Cognato AI, Jordal BH, McKenna DD. 2014.
Morphology, taxonomy, and phylogenetics of bark beetles. In:
Vega F, Hofstetter R, editors. Bark beetles: biology and ecology of
native and invasive species. London: Academic Press. p. 41–81.
Hundsdo¨rfer AK, Rheinheimer J, Wink M. 2009. Towards the phylogeny
of the Curculionoidea (Coleoptera): reconstructions from mito-
chondrial and nuclear ribosomal DNA sequences. Zool Anz.
248(1): 9–31.
Hunt T, Bergsten J, Levkanicova Z, Papadopoulou A, John OS, Wild R,
Hammond PM, Ahrens D, Balke M, Caterino MS, et al. 2007. A
comprehensive phylogeny of beetles reveals the evolutionary origins
of a superradiation. Science 318(5858): 1913–1916.
Inagaki Y, Roger AJ. 2006. Phylogenetic estimation under codon models
can be biased by codon usage heterogeneity. Mol Phylogenet Evol.
40(2): 428–434.
Inagaki Y, Simpson AGB, Dacks JB, Roger AJ, Naylor G. 2004. Phylogenetic
artifacts can be caused by leucine, serine, and arginine codon usage
heterogeneity: dinoflagellate plastid origins as a case study. Syst Biol.
53(4): 582–593.
Jordal BH, Sequeira AS, Cognato AI. 2011. The age and phylogeny of
wood boring weevils and the origin of subsociality. Mol Phylogenet
Evol. 59(3): 708–724.
Kriventseva EV, Tegenfeldt F, Petty TJ, Waterhouse RM, Simao FA,
Pozdnyakov IA, Ioannidis P, Zdobnov EM. 2015. OrthoDB v8: update
of the hierarchical catalog of orthologs and the underlying free soft-
ware. Nucleic Acids Res. 43(Database issue): D250–D256.
Kumar S, Stecher G, Tamura K. 2016. MEGA7: Molecular Evolutionary
Genetics Analysis version 7.0 for bigger datasets. Mol Biol Evol. 33(7):
1870–1874.
Shin et al. .doi:10.1093/molbev/msx324 MBE
12
Downloaded from https://academic.oup.com/mbe/advance-article-abstract/doi/10.1093/molbev/msx324/4765916
by University of Memphis - Library, sciaridae1@gmail.com
on 05 February 2018
Kuschel G. 1959. Un curculi
onido del cret
aceo superior, primer insecto
f
osil de Chile. Invest Zool Chil. 5:49–54.
Kuschel G. 1971. Entomology of the Aucklands and other islands south
of New Zealand: Coleoptera: Curculionidae. Pac Insects Monogr.
27:225–259.
Kuschel G. 1983. Past and present of the relict family Nemonychidae.
GeoJournal 7:499–504.
Kuschel G. 1995. A phylogenetic classification of Curculionoidea to
families and subfamilies. Mem Entomol Soc Wash. 14:5–33.
Kuschel G. 2003. Nemonychidae, Belidae, Brentidae (Insecta: Coleoptera:
Curculionoidea). Fauna N Z. 45:1–100.
Kuschel G, Leschen RAB. 2011. Phylogeny and taxonomy of the
Rhinorhynchinae (Coleoptera: Nemonychidae). Invertebr Syst.
24(6): 573–615.
Kuschel G, Leschen RAB, Zimmerman EC. 2000. Platypodidae under
scrutiny. Invertebr Syst. 14(6): 771–805.
Lanfear R, Calcott B, Ho SY, Guindon S. 2012. Partitionfinder: combined
selection of partitioning schemes and substitution models for
phylogenetic analyses. Mol Biol Evol. 29(6): 1695–1701.
Lartillot N, Lepage T, Blanquart S. 2009. PhyloBayes 3: a Bayesian software
package for phylogenetic reconstruction and molecular dating.
Bioinformatics 25(17): 2286–2288.
Lartillot N, Rodrigue N, Stubbs D, Richer J. 2013. PhyloBayes MPI: phy-
logenetic reconstruction with infinite mixtures of profiles in a parallel
environment. Syst Biol. 62(4): 611–615.
Leache AD, Rannala B. 2011. The accuracy of species tree estimation
under simulation: a comparison of methods. Syst Biol. 60(2):
126–137.
Lemmon AR, Emme SA, Lemmon EM. 2012. Anchored hybrid enrich-
ment for massively high-throughput phylogenomics. Syst Biol. 61(5):
727–744.
Lemmon EM, Lemmon AR. 2013. High-throughput genomic data in
systematics and phylogenetics. Annu Rev Ecol Evol Syst. 44(1):
99–121.
Leschen RAB, Beutel RG, editors. 2014. Handbook of zoology, Coleoptera
Vol. III: Morphology and systematics (Phytophaga). Berlin
(Germany): Walter de Gruyter.
Lidgard S, Crane PR. 1988. Quantitative analyses of the early angiosperm
radiation. Nature 331(6154): 344–346.
Lidgard S, Crane PR. 1990. Angiosperm diversification and Cretaceous
floristic trends: a comparison of palynofloras and leaf macrofloras.
Paleobiology 16(01): 77–93.
Marvaldi AE. 1997. Higher level phylogeny of Curculionidae (Coleoptera:
Curculionoidea) based mainly on larval characters, with special ref-
erence to broad-nosed weevils. Cladistics 13(4): 285–312.
Marvaldi AE, Duckett CN, Kjer KM, Gillespie JJ. 2009. Structural align-
ment of 18S and 28S rDNA sequences provides insights into phy-
logeny of Phytophaga (Coleoptera: Curculionoidea and
Chrysomeloidea). Zool Scr. 38(1): 63–77.
Marvaldi AE, Lanteri AA, del R
ıo GM, Oberprieler RG. 2014. 3.7.5
Entiminae Schoenherr, 1823. In: Leschen RAB, Beutel RG, editors.
Handbook of zoology, Vol. IV: Arthropoda: Insecta. Part 38
Coleoptera,beetles,Vol.III:Morphologyandsystematics
(Phytophaga). Berlin (Germany): Walter de Gruyter. p. 503–522.
Marvaldi AE, Morrone JJ. 2000. Phylogenetic systematics of weevils
(Coleoptera: Curculionoidea): a reappraisal based on larval and adult
morphology. Insect Syst Evol. 31(1): 43–58.
Marvaldi AE, Sequeira AS, O’Brien CW, Farrell BD. 2002. Molecular and
morphological phylogenetics of weevils (Coleoptera,
Curculionoidea): do niche shifts accompany diversification? Syst
Biol. 51(5): 761–785.
McKenna DD. 2011. Temporal lags and overlap in the diversification of
weevils and flowering plants: recent advances and prospects for
additional resolution. Am Entomol. 57(1): 54–55.
McKenna DD. 2014. Molecular phylogenetics and evolution of
Coleoptera. In: Leschen RAB, Beutel RG, editors. Handbook of zool-
ogy, Vol. IV: Arthropoda: Insecta. Part 38 Coleoptera, beetles, Vol. III:
Morphology and systematics (Phytophaga). Berlin (Germany):
Walter de Gruyter. p. 1–10.
McKenna DD. 2016. Molecular systematics of Coleoptera. In: Beutel RG,
Leschen RAB, editors. Handbook of zoology, Vol. I Arthropoda:
Insecta. Coleoptera, beetles. Berlin (Germany): Walter de Gruyter.
p. 23–34.
McKenna DD, Farrell BD. 2009. Beetles (Coleoptera). In: Hedges SB,
Kumar S, editors. The timetree of life.NewYork:OxfordUniversity
Press. p. 278–289.
McKenna DD, Farrell BD. 2010. 9-genes reinforce the phylogeny of
Holometabola and yield alternate views on the phylogenetic place-
ment of Strepsiptera. PLoS One 5(7): e11887.
McKenna DD, Scully ED, Pauchet Y, Hoover K, Kirsch R, Geib SM,
Mitchell RF, Waterhouse RM, Ahn S-J, Arsala D, et al. 2016.
Genome of the Asian longhorned beetle (Anoplophora glabripennis),
a globally significant invasive species, reveals key functional and evo-
lutionary innovations at the beetle-plant interface. Genome Biol.
17(1): 227.
McKenna DD, Sequeira AS, Marvaldi AE, Farrell BD. 2009. Temporal lags
and overlap in the diversification of weevils and flowering plants.
Proc Nati Acad Sci U S A. 106(17): 7083–7088.
McKenna DD, Wild AL, Kanda K, Bellamy CL, Beutel RG, Caterino MS,
Farnum CW, Hawks DC, Ivie MA, Jameson ML, et al. 2015. The beetle
tree of life reveals that Coleoptera survived end-Permian mass ex-
tinction to diversify during the Cretaceous terrestrial revolution. Syst
Entomol. 40(4): 835–880.
Mello B, Tao Q, Tamura K, Kumar S. 2017. Fast and accurate estimates of
divergence times from big data. Mol Biol Evol. 34(1): 45–50.
Meyer M, Kircher M. 2010. Illumina sequencing library preparation for
highly multiplexed target capture and sequencing. Cold Spring Harb
Protoc. 2010(6): pdb.prot5448.
Mirarab S, Bayzid MS, Boussau B, Warnow T. 2014. Statistical binning
enables an accurate coalescent-based estimation of the avian tree.
Science 346(6215): 1250463.
Mirarab S, Reaz R, Bayzid MS, Zimmermann T, Swenson MS, Warnow T.
2014. ASTRAL: genome-scale coalescent-based species tree estima-
tion. Bioinformatics 30(17): I541–I548.
Misof B, Liu S, Meusemann K, Peters RS, Donath A, Mayer C, Frandsen
PB, Ware J, Flouri T, Beutel RG, et al. 2014. Phylogenomics resolves
the timing and pattern of insect evolution. Science 346(6210):
763–767.
Morimoto K. 1962. Comparative morphology and phylogeny of the
superfamily Curculionoidea of Japan. J Fac Agric Kyushu Univ.
11(4): 331–373.
Morimoto K, Kojima H. 2006. New taxa. In: Morimoto K, Kojima H,
Miyakawa S. editors. The insects of Japan Vol III. Curculionoidea:
general introduction and Curculionidae: Entiminae (part 1)
Phyllobiini, Polydrusini and Cyphicerini (Coleoptera). Fukuoka
(Japan):ToukaShoboCo.Ltd.p.1–406.
Niehuis O, Hartig G, Grath S, Pohl H, Lehmann J, Tafer H, Donath A,
Krauss V, Eisenhardt C, Hertel J, et al. 2012. Genomic and morpho-
logical evidence converge to resolve the enigma of Strepsiptera. Curr
Biol. 22(14): 1309–1313.
Oberprieler RG. 1999. Systematics and evolution of the cycad-associated
weevil genus Apinotropis Jordan (Coleoptera: Anthribidae). Afr
Entomol. 7:1–33.
Oberprieler RG. 2000. The larvae of the weevil tribe Eurhynchini and the
phylogeny of the Brentidae (Coleoptera: Curculionoidea). Invertebr
Syst. 14(6): 755–770.
Oberprieler RG. 2010. A reclassification of the weevil subfamily
Cyclominae (Coleoptera: Curculionidae). Zootaxa 2515:1–35.
Oberprieler RG. 2014a. 3.6 Brentidae Billberg, 1820. In: Leschen RAB,
Beutel RG, editors. Handbook of zoology, Vol. IV: Arthropoda:
Insecta.Part38Coleoptera,beetles,Vol.III:Morphologyandsystem-
atics (Phytophaga). Berlin (Germany): Walter de Gruyter. p.
363–364.
Oberprieler RG. 2014b. 3.6.1 Eurhynchinae Lacordaire, 1863, Ithycerinae
Schoenherr, 1823 and Microcerinae Lacordaire, 1863. In: Leschen
RAB, Beutel RG, editors. Handbook of zoology, Vol. IV: Arthropoda:
Insecta. Part 38 Coleoptera, beetles, Vol. III: Morphology and
Phylogenomic Data Yield Insights into Weevil Evolution .doi:10.1093/molbev/msx324 MBE
13
Downloaded from https://academic.oup.com/mbe/advance-article-abstract/doi/10.1093/molbev/msx324/4765916
by University of Memphis - Library, sciaridae1@gmail.com
on 05 February 2018
systematics (Phytophaga). Berlin (Germany): Walter de Gruyter. p.
364–384.
Oberprieler RG. 2014c. 3.7 Curculionidae Latreille, 1802. In: Leschen RAB,
Beutel RG, editors. Handbook of zoology, Vol. IV: Arthropoda:
Insecta. Part 38 Coleoptera, beetles, Vol. III: Morphology and system-
atics (Phytophaga). Berlin (Germany): Walter de Gruyter. p.
423–424.
Oberprieler RG. 2014d. 3.7.1 Brachycerinae Billberg, 1820. In: Leschen
RAB, Beutel RG, editors. Handbook of zoology, Vol. IV:
Arthropoda: Insecta. Part 38 Coleoptera, beetles, Vol. III:
Morphology and systematics (Phytophaga). Berlin (Germany):
Walter de Gruyter. p. 424–451.
Oberprieler RG, Anderson RS, Marvaldi AE. 2014. Curculionoidea
Latreille, 1802: Introduction, Phylogeny. In: Leschen RAB, Beutel
RG, editors. Handbook of zoology, Vol. IV: Arthropoda: Insecta.
Part 38 Coleoptera, beetles, Vol. III: Morphology and systematics
(Phytophaga). Berlin (Germany): Walter de Gruyter. p. 285–300.
Oberprieler RG, Caldara R, Skuhrovec J. 2014. Bagoini Thomson, 1859;
Gonipterini Lacordaire, 1863; Hyperini Marseul, 1863. In: Leschen
RAB, Beutel RG, editors. Handbook of zoology, Vol. IV:
Arthropoda: Insecta. Part 38 Coleoptera, beetles, Vol. III:
Morphology and systematics (Phytophaga). Berlin (Germany):
Walter de Gruyter. p. 452–476.
Oberprieler RG, Marvaldi AE, Anderson RS. 2007. Weevils, weevils, wee-
vils everywhere. Zootaxa 1668:491–520.
Oberprieler RG, Oberprieler SK. 2012. Talbragarus averyi gen. et sp. n.,
the first Jurassic weevil from the southern hemisphere (Coleoptera:
Curculionoidea: Nemonychidae). Zootaxa 3478:256–266.
Pauchet Y, Wilkinson P, Chauhan R, Ffrench-Constant RH. 2010.
Diversity of beetle genes encoding novel plant cell wall degrading
enzymes. PLoS One 5 (12): e15635.
Peters RS, Krogmann L, Mayer C, Donath A, Gunkel S, Meusemann K,
Kozlov A, Podsiadlowski L, Petersen M, Lanfear R, et al. 2017.
Evolutionary history of the Hymenoptera. Curr Biol. 27(7):
1013–1018.
Petersen M, Meusemann K, Donath A, Dowling D, Liu S, Peters RS,
Podsiadlowski L, Vasilikopoulos A, Zhou X, Misof B, et al. 2017.
Orthograph: a versatile tool for mapping coding nucleotide sequen-
ces to clusters of orthologous genes. BMC Bioinformatics 18:111.
Prum RO, Berv JS, Dornburg A, Field DJ, Townsend JP, Lemmon EM,
Lemmon AR. 2015. A comprehensive phylogeny of birds (Aves)
using targeted next-generation DNA sequencing. Nature
526(7574): 569–573.
Richards S, Gibbs RA, Weinstock GM, Brown SJ, Denell R, Beeman RW,
Gibbs R, Beeman RW, Brown SJ, Bucher G. 2008. The genome of the
model beetle and pest Tribolium castaneum.Nature 452(7190):
949–955.
Rokyta DR, Lemmon AR, Margres MJ, Aronow K. 2012. The venom-
gland transcriptome of the eastern diamondback rattlesnake
(Crotalus adamanteus). BMC Genomics 13:312.
Ronquist F, Teslenko M, van der Mark P, Ayres DL, Darling A, Hohna S,
Larget B, Liu L, Suchard MA, Huelsenbeck JP. 2012. MrBayes 3.2:
efficient Bayesian phylogenetic inference and model choice across
alargemodelspace.Syst Biol. 61(3): 539–542.
Rota-Stabelli O, Lartillot N, Philippe H, Pisani D. 2013. Serine codon-
usage bias in deep phylogenomics: pancrustacean relationships as
acasestudy.Syst Biol. 62(1): 121–133.
Skuhrovec J, Bogusch P. 2016. The morphology of the immature stages
of Metadonus vuillefroyanus (Capiomont, 1868) (Coleoptera,
Curculionidae, Hyperini) and notes on its biology. Zookeys
589:123–142.
Stamatakis A. 2014. RAxML version 8: a tool for phylogenetic analysis
and post-analysis of large phylogenies. Bioinformatics 30(9):
1312–1313.
Suyama M, Torrents D, Bork P. 2006. PAL2NAL: robust conversion of
protein sequence alignments into the corresponding codon align-
ments. Nucleic Acids Res. 34(Web Server): W609–W612.
Tamura K, Battistuzzi FU, Billing-Ross P, Murillo O, Filipski A, Kumar S.
2012. Estimating divergence times in large molecular phylogenies.
Proc Natl Acad Sci U S A. 109(47): 19333–19338.
Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. 2013. MEGA6:
Molecular Evolutionary Genetics Analysis version 6.0. MolBiolEvol.
30(12): 2725–2729.
Thompson RT. 1992. Observations on the morphology and classification
of weevils (Coleoptera, Curculionoidea) with a key to major groups. J
Nat Hist. 26(4): 835–891.
ToussaintE,SeidelM,Arriaga-VarelaE,H
ajek J, Kr
al D, Sekerka L, Short A,
Fik
a
cek M. 2017. The peril of dating beetles. Syst Entomol.42(1):
1–10.
Waterhouse RM, Tegenfeldt F, Li J, Zdobnov EM, Kriventseva EV. 2013.
OrthoDB: a hierarchical catalog of animal, fungal and bacterial
orthologs. Nucleic Acids Res. 41(Database issue): D358–D365.
Werren JH, Richards S, Desjardins CA, Niehuis O, Gadau J, Colbourne JK,
Werren JH, Richards S, Desjardins CA, Niehuis O, et al. 2010.
Functional and evolutionary insights from the genomes of three
parasitoid Nasonia species. Science 327(5963): 343–348.
Wink M, Mikes Z, Rheinheimer J. 1997. Phylogenetic relationships in
weevils (Coleoptera: Curculionoidea) inferred from nucleotide
sequences of mitochondrial 16S rDNA. Naturwissenschaften 84(7):
318–321.
Xi Z, Liu L, Davis CC. 2016. The impact of missing data on species tree
estimation. Mol Biol Evol. 33(3): 838–860.
Young AD, Lemmon AR, Skevington JH, Mengual X, Stahls G, Reemer M,
Jordaens K, Kelso S, Lemmon E, Moriarty Hauser M, et al. 2016.
Anchored enrichment dataset for true flies (order Diptera) reveals
insights into the phylogeny of flower flies (family Syrphidae). BMC
Evol Biol. 16(1): 143.
Yunakov NN, Kirejtshuk AG. 2011. New genus and species of broad-
nosed weevils from Baltic amber and notes on fossils of the
subfamily Entiminae (Coleoptera, Curculionidae). ZooKeys
160:73–96.
Zhan S, Merlin C, Boore JL, Reppert SM. 2011. The monarch butterfly
genome yields insights into long-distance migration. Cell 147(5):
1171–1185.
Zimmerman EC. 1994. Australian weevils (Coleoptera: Curculionoidea).
Vol. I. Orthoceri, Anthribidae to Attelabidae. Melbourne (VIC):
CSIRO Publishing.
Shin et al. .doi:10.1093/molbev/msx324 MBE
14
Downloaded from https://academic.oup.com/mbe/advance-article-abstract/doi/10.1093/molbev/msx324/4765916
by University of Memphis - Library, sciaridae1@gmail.com
on 05 February 2018
