Phylogenomic relationships of bioluminescent elateroids define the 'lampyroid' clade with clicking Sinopyrophoridae as its earliest member

Abstract

Bioluminescence has been hypothesized as aposematic signalling, inter-sexual communication and a predatory strategy, but origins and relationships among bioluminescent beetles have been contentious. We reconstruct the phylogeny of the bioluminescent elateroid beetles (i.e. Elateridae, Lampyridae, Phengodidae and Rhagophthalmidae), analysing genomic data of Sinopyrophorus Bi & Li, and in light of our phylogenetic results, we erect Sinopyrophoridae Bi & Li, stat.n. as a clicking elaterid-like sister group of the soft-bodied bioluminescent elateroid beetles, that is, Lampyridae, Phengodidae and Rhagophthalmidae. We suggest a single origin of bio-luminescence for these four families, designated as the 'lampyroid clade', and examine the origins of bioluminescence in the terminal lineages of click beetles (Elateridae). The soft-bodied bioluminescent lineages originated from the fully sclerotized elateroids as a derived clade with clicking Sinopyrophorus and Elateridae as their serial sister groups. This relationship indicates that the bioluminescent soft-bodied elateroids are modified click beetles. We assume that bioluminescence was not present in the most recent common ancestor of Elateridae and the lampyroid clade and it evolved among this group with some delay, at the latest in the mid-Cretaceous period, presumably in eastern Laurasia. The delimitation and internal structure of the elaterid-lampyroid clade provides a phylogenetic framework for further studies on the genomic variation underlying the evolution of bioluminescence.

Full text

Systematic Entomology (2020), DOI: 10.1111/syen.12451
Phylogenomic relationships of bioluminescent
elateroids define the ‘lampyroid’ clade with clicking
Sinopyrophoridae as its earliest member
DOMINIK KUSY
1*,JIN-WU HE
2*, SETH M. BYBEE
3,
MICHAL MOTYKA1,WEN-XUAN BI
2, LARS PODSIADLOWSKI
4,
XUE-YAN LI
2and LADISLAV BOCAK1
1Laboratory of Biodiversity and Molecular Evolution, Palacky University, Olomouc, Czech Republic, 2State Key Laboratory of
Genetic Resources and Evolution, Kunming Institute of Zoology CAS, Kunming, China, 3Department of Biology, Monte L. Bean
Museum, Brigham Young University, Provo, UT, U.S.A. and 4Zoological Research Museum Alexander Koenig, Centre of Molecular
Biodiversity Research, Bonn, Germany
Abstract. Bioluminescence has been hypothesized as aposematic signalling, inter-
sexual communication and a predatory strategy, but origins and relationships among
bioluminescent beetles have been contentious. We reconstruct the phylogeny of the
bioluminescent elateroid beetles (i.e. Elateridae, Lampyridae, Phengodidae and
Rhagophthalmidae), analysing genomic data of Sinopyrophorus Bi & Li, and in light
of our phylogenetic results, we erect Sinopyrophoridae Bi & Li, stat.n. as a clicking
elaterid-like sister group of the soft-bodied bioluminescent elateroid beetles, that is,
Lampyridae, Phengodidae and Rhagophthalmidae. We suggest a single origin of bio-
luminescence for these four families, designated as the ‘lampyroid clade’, and examine
the origins of bioluminescence in the terminal lineages of click beetles (Elateridae).
The soft-bodied bioluminescent lineages originated from the fully sclerotized elateroids
as a derived clade with clicking Sinopyrophorus and Elateridae as their serial sister
groups. This relationship indicates that the bioluminescent soft-bodied elateroids are
modied click beetles. We assume that bioluminescence was not present in the most
recent common ancestor of Elateridae and the lampyroid clade and it evolved among
this group with some delay, at the latest in the mid-Cretaceous period, presumably
in eastern Laurasia. The delimitation and internal structure of the elaterid-lampyroid
clade provides a phylogenetic framework for further studies on the genomic variation
underlying the evolution of bioluminescence.
Introduction
Bioluminescence has been intensively studied by numerous
researchers (Costa, 1975; Branham & Wenzel, 2003; Nakatsu
et al., 2006; Fallon et al., 2018). The production of light occurs
Correspondence: Ladislav Bocak, Laboratory of Biodiversity and
Molecular Evolution, Palacky University, Olomouc, Czech Republic,
E-mail: ladislav.bocak@upol.cz; and Xueyan Li, State Key Laboratory
of Genetic Resources and Evolution, Kunming Institute of Zoology,
Chinese Academy of Sciences, Kunming, Yunnan, 650223, China,
E-mail: lixy@mail.kiz.ac.cn.
∗These authors contributed equally to the study.
sporadically in insects (Watkins et al., 2018; Martin et al., 2019);
nevertheless, beetle bioluminescence is well known to both
researchers and the general public. Most luminous beetles
belong to Elateroidea and we know of ∼2000 rey species
(Lampyridae, Fig. 1F– J), ∼300 glow-worms or railroad-worm
beetles (Phengodidae, Rhagophthalmidae), as well as >100 bio-
luminescent click beetle species (Elateridae; Fig. 1A–D), espe-
cially in the Neotropical region (Costa, 1975, 1984). With the
recent discovery of the rst Palearctic bioluminescent click-
ing beetle, the number of bioluminescent beetle lineages has
increased (He et al., 2019; Bi et al., 2019; Fig. 1D–E).
The phylogenetics of Elateroidea have been studied with
morphology since the late 1980s, where the bioluminescent
© 2020 The Royal Entomological Society 1

2D. Kusy et al.
Fig 1. Morphological diversity of the elaterid-lampyroid clade. Elateridae: (A) Denticollis sp.; (B) Agrypnus murinus (L.); (C) click beetle larva;
(D) Ampedus sp. The lampyroid clade. Sinopyrophoridae: (D, E) Sinopyrophorus schimmeli Bi & Li; Lampyridae: (F) Asymmetricata circumdata
(Motschulsky); (G) Lampyris noctiluca (L.), larva; (H) ditto, in copula; (I, J) Lamprohiza splendidula (L.), female. Photographs by M. Motyka (B, D,
G– J), L. Bocak (A,C) and Z.-W. Dong (F); (D, E) from Bi et al., 2019. [Colour gure can be viewed at wileyonlinelibrary.com].
soft-bodied ‘cantharoids’ and fully sclerotized click beetles were
merged in a single superfamily, Elateroidea (Lawrence, 1988;
Branham & Wenzel, 2003; Lawrence et al., 2011). Later
studies using DNA data supported showed that the soft-bodied
elateroids were polyphyletic (Bocakova et al., 2007). Some
relationships among these groups remained inconclusive
when topologies were inferred from a few widely used
molecular markers and the controversies persisted even
when protein-coding nuclear fragments were employed
(McKenna et al., 2015), or when taxon sampling was sub-
stantially increased (Kundrata et al., 2014; Bocak et al., 2016;
Linard et al., 2018) (Fig S1A–C). A clade of bioluminescent
elateroids, that is, Elateridae, Lampyridae, Phengodidae and
Rhagophthalmidae, was rst identied by the analyses of 13
mitochondrial genes (Timmermans et al., 2010) and conrmed
by further analyses of these mitogenomes but with more exten-
sive taxon sampling (Amaral et al., 2016; Bocak et al., 2016).
Nevertheless, the analyses were relatively limited by the volume
of data and resulted in ambiguous support for critical clades and
contradictory results (Kundrata et al., 2014; Linard et al., 2018).
Recent progress has been made possible by high throughput
sequencing, transcriptomes and whole-genome analyses. Zhang
et al. (2018a) used 99 nuclear markers and Kusy et al. (2018a)
analysed ∼4000 orthologous genes. Both studies provided
evidence for a clade that included ve elateroid families with
at least some bioluminescent taxa. However, these two studies
either used a low number of orthologs or restricted taxon sam-
pling. A phylogenomic analysis for all of Coleoptera provided
a timeframe for the evolution of the order based on multi-
ple calibration points, conrming the elateroid backbone but
contained only six elateroid terminals (McKenna et al., 2019)
(Fig S1D,E).
Recently, the subfamily Sinopyrophorinae Bi & Li was pro-
posed in Elateridae for the bioluminescent Sinopyrophorus
schimmeli Bi & Li. Bi et al. (2019) included homologous
fragments of S. schimmeli in a phylogenetic reconstruction
from earlier published click beetle rRNA and mtDNA genes
(Bocakova et al., 2007; Timmermans et al., 2010, 2016; Kun-
drata et al., 2014; Amaral et al., 2016) and used glow-worms
as outgroups. They recovered Sinopyrophorus with the elaterid
subfamilies Hemiopinae and Oestodinae, but with limited statis-
tical support for many relationships among elaterid subfamilies.
Herein, several thousand orthologs of S. schimmeli are used
to investigate the evolution of the clade of bioluminescent
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12451

Phylogenomics of luminescent elateroids 3
elateroid families. Large genomic datasets are proving empiri-
cally necessary if the relationships among old lineages can be
recovered with some level of condence (Misof et al., 2014;
Kusy et al., 2018a; McKenna et al., 2019). Based on these
results, we discuss (i) the origins of bioluminescence in
Elateroidea, (ii) loss of the clicking mechanism, and (iii)
loss of a fully sclerotized body in most bioluminescent
elateroids. This phylogeny will enable further evaluation of
the similarity of the luciferase genes and track their evolution
(Fallon et al., 2018).
Methods
Genomic data
The genomic DNA of a single male adult of S. schimmeli from
Yunnan Province (Collecting data: Husa village, 1770 m a.s.l.,
Longchuan County, 15– 25 Jun 2017, coll. Wenxuan Bi) was
shotgun-sequenced on the Illumina HiSeq4000 by Novogene
Co., Ltd. (Tianjing, China) for 150 bp paired-end reads and ∼30
Gbp of total data (Bi et al., 2019). Raw paired-end reads were
ltered using fastp v.0.20.0 (Chen et al., 2018) and low-quality
reads were removed. The draft genome was assembled using
SPAdes v.3.13.1 (Bankevich et al., 2012), with k-mer sizes of 21,
33, 55, 77 and 99. Contigs were used to train ‘Augustus’ (Stanke
& Waack, 2003) for species-specic gene models with BUSCO
v.3 (Waterhouse et al., 2018). Predicted models were used for ab
initio gene predictions and protein-coding sequences were used
for analyses. Basic statistics of genome assembly were evaluated
with QUAST v.5 (Mikheenko et al., 2018). We performed k-mer
counts on the ltered data in Jellysh 2.2.10 using 17 and 21-mer
sizes (Marcais & Kingsford, 2011). Based on the distribution
of k-mer occurrences, we estimated the genome size using
GenomeScope (Vurture et al., 2017).
We compiled the phylogenomic dataset using Sinopyrophorus
and 42 publicly available transcriptomes or genomes (Poelchau
et al., 2014; Sanders & Hall, 2015; Amaral et al., 2017,
2019; Wang et al., 2017; Fallon et al., 2018; Ye et al., 2018;
Kusy et al., 2018b, 2019; McKenna et al., 2019) (Table S1).
The single-copy ortholog set was collated by searching the
OrthoDB v.9.1 database (Zdobnov et al., 2016) (Tables S2,
S3). We carried out Orthograph v.0.6.3 searches (Petersen
et al., 2017) on assembled transcriptomes and protein-coding
gene sets. Terminal stop codons were removed and internal
stop codons at the translational and nucleotide levels were
masked using the Perl script summarize_orthograph_results.pl
(Petersen et al., 2017). The amino acid sequences were aligned
using Mafft v.7.407 with the L-INS-i algorithm (Katoh &
Standley, 2013). Resulting alignments from each ortholog
group were checked for the presence of outliers using the
script checker_complete.1.3.1.2.pl and earlier reported meth-
ods (Misof et al., 2014; Peters et al., 2017). Outlier sequences
were removed from alignments. Then, the non-elateriform taxa
were removed. The multiple sequence alignments of nucleotides
were generated using Pal2Nal (Suyama et al., 2006) and Alis-
core v.2.2 (Misof & Misof, 2009; Kück et al., 2010) was used
to identify ambiguous and randomly similar aligned sections.
Aliscore was invoked with a custom –r 1027 option, with a
scoring approach for gap-lled amino acid sites (option -e).
After that, we used Alinuc.pl (Misof et al., 2014) to create a
list of corresponding codons to be removed from nucleotide
alignments. Identied random or ambiguous similarities were
masked using ALICUT v.2.3 (Kück et al., 2010). In each gene
alignment, the short randomly aligned fragments were replaced
with gaps and sequences with ≥80% missing data were removed
using Python scripts (Zhang et al., 2020). We used MARE
v.0.1.2-rc (Misof et al., 2013) to calculate the information con-
tent of each gene partition. Partitions with zero information con-
tent were removed.
For the remaining 4199 genes, we used AMAS
(Borowiec, 2016) to calculate statistics (alignment length,
GC content, number of missing taxa, number of parsimony
informative sites) and individual gene alignments were retained
for coalescent analyses. The concatenated datasets were gen-
erated using FasConCat-G v.1.4 (Kück & Longo, 2014).
From the 4199 genes, we generated the datasets A-4199-AA
and B-4199-NT (designation: name-number of taxa-amino
acid or nucleotide level, Table S4). To reduce the effect
of saturation (Breinholt & Kawahara, 2013), we created
dataset C-4199-NT12, with third-codon positions excluded. To
increase the data decisiveness, we constructed additional super-
matrices using only partitions with all 43 species present:
datasets D-968-AA and E-968-NT. The degree of missing
data and overall completeness scores across all datasets was
inspected using AliStat v.1.7 (https://github.com/thomaskf/
AliStat).
Compositional heterogeneity tests
To explore the effect of compositional heterogeneity, we
inspected the dataset A-4199-AA with BaCoCa v.1.105 (Kück
& Struck, 2014). We considered compositional heterogeneity
among species in a given partition to be high when the overall
RCFV value was ≥0.1 (Fernandez et al., 2016; Vasilikopoulos
et al., 2019). Heterogeneous partitions were excluded from
the dataset A-4199-AA to generate dataset F-2195-AA. We
used Maximum Symmetry Test (Naser-Khdour et al., 2019)
to exclude the deviating genes from the dataset B-4199-NT
(P-value cutoff <0.05) and the dataset G-958-NT contains only
partitions that passed the test. The software SymTest v.2.0.49
(https://github.com/ottmi/symtest) was used to calculate the
deviation from stationarity, reversibility and homogeneity
(Jermiin et al., 2008) (SRH). Heatmaps were generated
for all datasets to visualize the pairwise deviations from
SRH conditions. To eliminate synonymous signal (Kawa-
hara et al., 2011), we used Degen v.1.4 (Zwick et al., 2012)
(http://www.phylotools.com/ptdegendocumentation.htm). All
sites with synonymous substitutions were replaced by the
corresponding ambiguity codes. The synonymous signal was
removed from dataset B-4199-NT and H-4199-DEGEN-NT
was generated.
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12451

4D. Kusy et al.
Phylogenetic analyses
Phylogenetic reconstructions were performed using maxi-
mum likelihood (ML) criterion with IQ-TREE v.2.0-rc2 (Minh
et al., 2020). Model selection for each gene was performed
with ModelFinder (Chernomor et al., 2016; Kalyaanamoorthy
et al., 2017) using the -MFP option. For amino acid superma-
trices, the substitution models LG, DCMUT, JTT, JTTDCMUT,
DAYHOFF, WAG, and free rate models LG4X and LG4M were
tested and all combinations of rate heterogeneity among sites
were allowed (options: -mrate E,I,G,I+G,R -gmedian -merit
AICc). We used the edge-linked partitioned model for tree
reconstructions (-spp option) allowing each gene to have its
own rate. The optimized partition schemes and best-tting
model were inferred for the datasets A-4199-AA, B-4199-NT,
D-968-AA, and E-968-NT using -m MFP+MERGE −merit
AICc -gmedian options and considering the same substitu-
tion models as above. The fast-relaxed clustering algorithm
was used (Lanfear et al., 2017). The top 10% of partitions
schemes –rclusterf 10 and maximum 10 000 partitions pairs
–rcluster-max 10 000 were considered for all datasets except for
D-968-AA and E-968-NT, where –rclusterf 10 –rcluster-max
5000 were used. Ultrafast bootstrap (Hoang et al., 2018) and
SH-like approximate likelihood ratio test (SH-aLRT) were cal-
culated using options -bb 3000 and -alrt 10 000.
To account for variation among gene trees owing to incom-
plete lineage sorting and to account for potential gene tree
heterogeneity and discordance (Degnan & Rosenberg, 2006;
Edwards, 2009; Mirarab et al., 2016), the datasets A-4199-AA,
B-4199-NT and C-4199-NT12 were analysed using the
coalescent-based species-tree method. For every single-gene
partition, we calculated an ML gene tree, with 1000 ultra-
fast bootstrap replicates (-bb option) and using the same
substitution models as earlier. For coalescent species tree
estimation, the Accurate Species Tree Algorithm was used
[ASTRAL-III v.5.6.3 (Zhang et al., 2018b)]. ASTRAL accu-
racy is reduced when poorly resolved gene trees are included
(Barrow et al., 2018). Therefore, we calculated average ultrafast
bootstrap and branch length for every gene tree using an R script
(https://github.com/marekborowiec/good_genes/tree_props.R)
and reduced gene trees datasets were created. To account for
very poorly resolved branches on gene trees, branches with
ultra-fast bootstrap ≤10 were collapsed using Newick utilities
v.1.6 (Junier & Zdobnov, 2010) in every ASTRAL analysis.
Local posterior probabilities (Erfan & Mirarab, 2016) and
quartet frequencies of the internal branches in every species
tree were calculated using the parameter ‘-t =2’. Furthermore,
we used DiscoVista v.1.0 (Erfan et al., 2018) to visualise
gene-tree quartet frequencies of three topologies around
focal internal branches of the inferred ASTRAL nucleotide
species tree in the datasets A-4199-AA, B-4199-NT, and
C-4199-NT12. Here, the following regularly recovered mono-
phyletic groups were considered: nine elateriformian outgroups,
Throscidae, Lycidae, Cantharidae, Elateridae (except Elateri-
nae), Elaterinae, Sinopyrophorus, Lampyridae, Phengodidae,
Rhagophthalmidae.
Analyses of alternative relationships
We used Four-cluster Likelihood mapping (Strimmer &
von Haeseler, 1997; Misof et al., 2013) (FcLM) analysis to
investigate alternative topologies in the datasets A-4199-AA,
B-4199-NT and C-4199-NT12. The analysis determines if
incongruent or confounding signals are present, which may be
obscured in a multispecies phylogenetic tree. The tree-likeness
graph for the three possible quartet topologies shows the support
for each topology. Additionally, we tested positions of Elaterinae
as (i) a sister to clade of Sinopyrophorus, Lampyridae, Phen-
godidae, Rhagophthalmidae (paraphyletic Elateridae) and (ii)
a sister to other Elateridae (monophyletic Elateridae) by eval-
uating which gene partitions of the datasets A-4199-AA and
B-4199-NT favour the alternatives. We calculated log-likelihood
scores and differences of each pL score for each gene partition
using option -wpl in IQ-TREE (Minh et al., 2020).
66-gene dataset
We assembled a dataset from earlier published data (Kusy
et al., 2018a,b, 2019; Zhang et al., 2018a) and newly produced
homologs (Table S5). Sequences of 84 elateroids and outgroups
were used as an input to Orthograph, 66 beetle orthologs were
extracted, and aligned at an amino acid level using mafft v.7.394
with the L-INS-i algorithm (Katoh & Standley, 2013). Multi-
ple sequence alignments of nucleotides were generated using
Pal2Nal (Suyama et al., 2006) to produce datasets J-66-NT and
K-66-AA. We then manually checked all alignments for the
presence of outliers and alignment errors. The matrices for both
amino acid and corresponding nucleotide alignments were gen-
erated using FasConCat-G v.1 (Kück & Longo, 2014). The soft-
ware SymTest v.2.0.49 (https://github.com/ottmi/symtest) was
used to calculate the overall SRH deviation. The fully parti-
tioned reconstructions were performed using the ML criterion
with IQ-TREE. Full methods and the complete list of analyses
are provided in File S1.
Results
Newly generated shotgun DNA reads of S. schimmeli produced
a genome assembly with ∼190x read coverage (Fig S34A).
Despite fragmentation, the assembly contained a sufcient
amount of information for ortholog extraction and downstream
phylogenomic analyses (Table S3). The genome completeness
and assembly statistics are summarized in Figs S33A, B, S34A.
The genome size of S. schimmeli was estimated to 171.8 and
190.7 million base pairs (Mbp) (Fig S34B, C) using k-mer sizes
17 and 21, respectively.
Denition and relationships of lineages
Two sets of topologies were produced which differ in the
density of sampling and the number of orthologs. The rst
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12451

Phylogenomics of luminescent elateroids 5
Fig 2. Phylogenetic relationships. The topology recovered by the ASTRAL coalescent phylogenetic method applied to the full set of single-gene trees
inferred at nucleotide level from the dataset B-4199-NT; only the elaterid-lampyroid clade shown, see the full trees in Figs S17, S18. [Colour gure can
be viewed at wileyonlinelibrary.com].
set was based on 4199 orthologs and 43 taxa, nine of them
as outgroups; the topologies were inferred under the ML cri-
terion and the coalescent method by the analyses of amino
acids and nucleotides (Fig. 2). The second set was based on
the 66-orthologs, 83 elateroids and the topologies were inferred
by ML analyses (Figs 3A, B, S2–S22). All analyses recovered
a clade containing exclusively the families with at least some
taxa that produce light, that is, Elateridae (∼2% of taxa biolumi-
nescent, in three subfamilies), Sinopyrophoridae (Sinopyropho-
rus, 1 sp., bioluminescent in an adult stage, larvae unknown),
Phengodidae, Rhagophthalmidae and Lampyridae (all biolumi-
nescent at least in the larval stage, except Phengodidae: Cydis-
tinae for which larvae and females are unknown and males are
non-luminescent). Hereafter, these widely accepted families and
Sinopyrophorus are designated as the ‘elaterid-lampyroid clade’
(Figs 2–4).
Sinopyrophorus was found outside of Elateridae and as a sister
to Lampyridae, Phengodidae and Rhagophthalmidae. The topol-
ogy was stable regardless of the dataset and inference method
employed (Figs 2– 3, S2 – S22). Its position was also supported
by the FcLM analysis (Figs 5A, B, S25, Table S6) and alter-
native topologies were rejected by an approximately unbiased
test (AU test, Table S7). All phylogenomic analyses indicate
that phenotypically elaterid-like Sinopyrophorus (originally Ela-
teridae: Sinopyrophorinae, Fig. 6) and the here redened Ela-
teridae do not share a common exclusive ancestor and form a
serial paraphylum towards the branch of Lampyridae, Rhagoph-
thalmidae and Phengodidae. The clade of Sinopyrophorus and
the soft-bodied bioluminescent families is here designated as the
‘lampyroid clade’.
In contrast to the relatively well-supported lampyroid clade,
we found confounding signal for some relationships among ela-
terid lineages (Figs 2, 3, 5B, D). Most analyses of genomic
datasets suggested the paraphyly of Elateridae with Elaterinae
rooted as a sister to the lampyroid clade and other elaterids
as a sister to both of them (Figs S2–33). Such a result was
inferred from all ML analyses and the coalescent analysis of the
amino acid data, but not by the coalescent method applied to
the nucleotide dataset (Fig. 2). The distribution of the signal for
alternative topologies prefers the monophyly of Elateridae. The
FcLM analysis of the nucleotide dataset B-4199-NT returned
58.7% for the monophyly of Elateridae versus 39.6% for their
paraphyly (Figs 5B, S25). Similarly, the DiscoVista relative fre-
quency analysis of the B-4199-NT dataset (node 9; Fig. 5D)
preferred the monophyly of Elateridae, but paraphyly is sup-
ported by the datasets A-4199-AA and C-4199-NT12 (Figs S23,
24). The 66-gene amino acid dataset suggests Elateridae without
Sinopyrophorus as a serial paraphylum of three elaterid groups
to the lampyroid clade (Fig. 3A), but the same dataset produced
a monophyletic Elateridae at the nucleotide level (Fig. 3B).
With regard to the positions of the clicking and non-clicking
forms in the elaterid-lampyroid clade, our analyses always
recovered the clicking forms as the earliest splits. Biolu-
minescent taxa are always recovered in a derived position
within the elaterid-lampyroid clade, regardless of monophyly
or paraphyly of click beetles inferred from various analyses
(Figs 2–3), that is, the most recent common ancestor of the
elaterid-lampyroid clade was clicking and non-bioluminescent.
All topologies and additional information on data are shown
in Figs S2–S34.
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12451

6D. Kusy et al.
Fig 3. (A) The topology recovered by the maximum likelihood analysis of the 66-gene dataset J-66-AA at the amino acid level. (B) The topology
recovered by the maximum likelihood analysis of the 66-gene dataset J-66-NT at the nucleotide level. The dark blue branches are signicantly supported
(both SH-aLRT >80% and UFboot >95%). [Colour gure can be viewed at wileyonlinelibrary.com].
Taxonomy
Sinopyrophoridae Bi & Li, 2019, new status
Sinopyrophorinae Bi & Li, 2019 in Bi et al., 2019: 83.
Type genus: Sinopyrophorus Bi & Li, 2019 in Bi et al.,
2019: 89.
=Sinopyrophorus He et al., 2019: 565, unavailable name
due to the absence of a description in the study where the
name was proposed (International Committee for Zoological
Nomenclature, 1999).
Based on the recovered relationships, Sinopyrophorus (earlier
Elateridae: Sinopyrophorinae) cannot be a member of Elateri-
dae. Despite its morphological similarity and a shared clicking
mechanism, its phylogenetic position requires it to be classi-
ed as a separate taxon of the same rank as true click beetles,
that is, family. The proposed rank fulls the requirement of the
reciprocal monophyly of all taxa and, simultaneously, keeps tra-
ditionally recognized families valid.
Discussion
Phylogenomic relationships
The uncertain homology of characters in phenotypically dis-
parate elateroids, ambiguities in length variable alignments and
low information content in the Sanger data have produced con-
tradicting phylogenies (Bocakova et al., 2007; Sagegami-Oba
et al., 2007; Lawrence et al., 2011; Kundrata et al., 2014;
McKenna et al., 2015). Here, we present analyses based on more
than 4000 orthologs that support three principal relationships
that have not been well supported in earlier studies:
1 The monophyly of the elaterid-lampyroid clade (Figs 2 – 4;
Timmermans et al., 2010; Amaral et al., 2016; Bocak
et al., 2016; Kusy et al., 2018b; McKenna et al., 2019) is
preferred over alternative hypotheses (Bocakova et al., 2007;
Sagegami-Oba et al., 2007; Kundrata et al., 2014; McKenna
et al., 2015; Zhang et al., 2018a).
2 We prefer the monophyly of redened click beetles, that
is, including Elaterinae and Lissominae without Sinopy-
rophorus, as suggested by the B-4199-NT dataset and the
coalescent method, the J-66-NT dataset and ML methods,
and as additionally supported by the FcLM analyses, Disco-
Vista analyses and AU test (Figs 2, 3B, 5B, D, S20, S22B,
S23–S26; Table S7) over a paraphyletic Elateridae obtained
from all ML analyses of the genomic datasets, from the
coalescent method analyses of C-4199-NT12/A-4199-AA
datasets, and the ML analysis of the I-66-AA dataset
(Figs 3A, S2–S19, S21, S22A). The paraphyletic or poly-
phyletic Elateridae were recovered in earlier estimates where
datasets representing all Coleoptera were analysed (McKenna
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12451

Phylogenomics of luminescent elateroids 7
Fig 4. The summary cladogram of Elateroidea with bars showing the distribution of estimated origin of Elateroidea, the earliest split within the
bioluminescent clade, and the splits within the [Lampyridae (Rhagophthalmidae, Phengodidae)] clade. The origin of Sinopyrophoridae is supposed
after the origin of the bioluminescent clade and before the earliest split between Lampyridae and Rhagophthalmidae +Phengodidae. [Colour gure can
be viewed at wileyonlinelibrary.com].
et al., 2015, 2019; Zhang et al., 2018a). In this case, only
a slight majority of genes recover the monophyly of click
beetles over their paraphyly (Figs 5B, S26). Generally, the
amino acid and nucleotide rst +second codon position
sequences are highly conservative when the set of orthologs
for whole Coleoptera is assembled and possibly, due to
incomplete lineage sorting, the analyses of these datasets
do not support the monophyly of click beetles is analysed
using the maximum likelihood approach (Figs 3A, S2– 16).
The ASTRAL coalescent phylogenetic method applied to the
full set of gene trees inferred the monophyly of Elateridae
from the dataset B-4199-NT at the nucleotide level (Fig. 2).
Additionally, the FcLM and DiscoVista relative frequency
analyses of the dataset B-4199-NT support, albeit weakly,
the monophyly of Elateridae (Figs 5B, D, S24, 25). Similarly,
the 66-gene dataset is based on highly conservative genes and
Elateridae are split into three serial groups of the lampyroid
clade if amino acids are analysed (Figs 3A, S17), but the
family is monophyletic when the tree is inferred with the
maximum likelihood analysis at the nucleotide level (Figs 3B,
S18). The earlier 99-gene (among them the here employed
66 orthologs) analysis of the whole Coleoptera by Zhang
et al. (2018a) suggested the polyphyly of click beetles when
Lissominae was the sister to net-winged beetles and the rest
of Elateridae formed two serial sister groups to the lampyroid
families. We are aware that further analyses of a much larger
dataset will be needed to test the monophyly of Elateridae
but, based on the current results, we prefer to accept the
monophyly of Elateridae. Although the genome-scale data
offer a powerful tool for phylogenetics, the relationships
of elaterid subfamilies remain poorly supported and need
rigorous testing with a dense sampling of taxa and a higher
number of orthologs specically designed for this question.
3 Sinopyrophoridae is decisively placed as a sister to Lampyri-
dae, Phengodidae and Rhagophthalmidae (Figs 2, 3A,
S2–S25), or as a sister to Lampyridae (Fig. 3B) rather than
a member of the click beetles (Bi et al., 2019). Therefore,
we propose to accept Sinopyrophorus as a member of
the ‘lampyroid clade’, that is, the ancient lineage closely
related to reies and glow-worms (Figs 2, 3, S2– S26).
The elaterid-like morphology of Sinopyrophorus supports
the idea that the characteristic well-sclerotized body form
and clicking escape mechanism were abandoned multiple
times during the evolution of Elateroidea. Here, a single
shift to incomplete sclerotization in the common ancestor of
soft-bodied reies and glow-worms was recovered by all
analyses (Figs 2, 3). Such a process was earlier inferred for
drilids and omalisids now included in Elateridae (Kundrata
et al., 2014; Kusy et al., 2018a). The clicking mechanism is
unique, relatively complex and we can assume that it evolved
once if the probabilities of the origin and the loss of the
clicking mechanism are substantially different (Trueman
et al., 2004).
Formal classication
The elevation of Sinopyrophoridae as a separate family is our
preferred, but not a single possible solution of the formal clas-
sication. Our decision is conservative in the sense that it keeps
the family rank for all earlier designated reciprocally mono-
phyletic lineages, that is, click beetles, reies, glow-worms and
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12451

8D. Kusy et al.
Fig 5. Four cluster Likelihood Mapping tests of the selected phylogenetic hypothesis applied at the nucleotide level of the dataset B-4199-NT. (A) The
test of the position of Sinopyrophorus. (B) The test of the monophyly of Elateridae. (C) Evaluated topology. (D) DiscoVista relative frequency analyses
of the dataset B-4199-NT. [Colour gure can be viewed at wileyonlinelibrary.com].
railroad-worm beetles. There are several alternative solutions.
We could merge Sinopyrophoridae, Phengodidae, Rhagoph-
thalmidae and Lampyridae under a single family, that is, the here
recovered lampyroid clade will be called Lampyridae Latreille,
and contain ve subfamilies. Lampyridae in a new wide sense
would contain several morphologically distinct lineages with
differing natural history and the whole internal classication
of reies would be down-ranked. The formal requirement for
the monophyly of all named taxa would also be fullled by the
redenition of Elateridae sensu lato that would contain eight
traditional families Elateridae, Drilidae, Omalisidae, Plastoceri-
dae, Lampyridae, Phengodidae, Rhagophthalmidae and Sinopy-
rophorinae. (Kusy et al., 2018a,b; Bi et al., 2019). Considering
also the logical extreme, we could accept whole Elateroidea
as a single family (sans the artemotopodid clade) as recently
mentioned by Muona & Taräväinen (2020). The clade would
be dened by the clicking mechanism that was secondarily lost
in half of its members. Nevertheless, for convenience, we pre-
fer to keep the family status for all traditional, widely accepted
families whenever possible. Click beetles and reies are known
to non-specialists and they are intuitively distinguished also by
the general public, for example, naturalist involved in the Fire-
y Citizen Science Project by the Natural History Museum of
Utah among others. Therefore, the family rank is appropriate for
their separation and the expansion of reies to a heterogeneous
assemblage of biologically and morphologically disparate forms
would not be practical.
Origins of the bioluminescence
Our analyses indicate a unique origin of bioluminescence
in the common ancestor of the clade Sinopyrophoridae,
Rhagophthalmidae, Phengodidae and Lampyridae clade and
further independent origin of bioluminescence within click
beetles as dened here, that is, without Sinopyrophorus.The
relatively distant position of bioluminescent taxa in the formal
classication was conrmed by the earliest molecular phyloge-
nies (Bocakova et al., 2007; Sagegami-Oba et al., 2007) and has
been inferred due to the genetic structure around the luciferase
genes from the genomes of Photinus,Aquatica and Ignelater
(Fallon et al., 2018). The position of photic organs is variable
in adult click beetles: prothoracic spots are reported in Balgus,
Campyloxenus, and Pyrophorini: Nyctophyxina, the prothoracic
and abdominal organs in Pyrophorini: Hapsodrilina and most
Pyrophorina, and only the abdominal photic organs in the
pyrophorine genus Hifo (Costa, 1975, 1984). Unlike these, only
abdominal photic organs are shared by all adult bioluminescent
forms in the lampyroid clade (Branham & Wenzel, 2003; Bi
et al., 2019).
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12451

Phylogenomics of luminescent elateroids 9
Fig 6. The summary of Elateroidea relationships with the distribution of clicking, soft-bodied, and bioluminescent forms. Photographs J. Klvá ˇ
cek and
authors. [Colour gure can be viewed at wileyonlinelibrary.com].
About 2300 species of the lampyroid clade inherited bio-
luminescence from their most recent common ancestor and
we must suppose that if this ancestor was a fully sclerotized
elaterid-like beetle, the origin of bioluminescence preceded
the shift to being soft-bodied (Fig. 6). Among bioluminescent
groups, only lampyrids are species-rich today (∼2000 spp.)
and the second largest bioluminescent clade are glow-worms
(>200 spp.). Both are common in the Neotropical region along
with bioluminescent elaterids (>100 spp.) (Costa, 1975, 1984).
The effectiveness of bioluminescence as an aposematic signal
depends on the number of taxa and individuals sharing the sig-
nal. As ashing can serve as an isolating mechanism (Branham
& Wenzel, 2003), the intensive interactions among biolumines-
cent elateroids might have played a role in the high diversity of
extant reies, glow-worms and luminous click beetles in the
Neotropical region (Ellis & Oakley, 2016).
Since South America is a major epicentre for bioluminescent
species, the region has been hypothesized as the ancestral region
for the diversication of reies (Amaral et al., 2016). The sister
group position of the elaterid-like East Asian Sinopyrophori-
dae (Figs 2, 3), the predominantly Laurasian distribution of two
of the deepest subfamilies of reies, Ototretinae and Lucioli-
nae (Kundrata et al., 2014; Martin et al., 2017, 2019; Zhang
et al., 2018a) (Fig. 3), Southeast Asian Rhagophthalmidae, and
the presence of Phengodidae in Burmese amber (unpublished
data) suggest as an alternative hypothesis that the early diver-
sication of the bioluminescent lineages took place in eastern
Laurasia and was followed by subsequent intensive radiations
of reies in the Neotropical region.
Bioluminescence is scattered among some terminal lineages
of click beetles, each of them being fully sclerotized and with
a clicking mechanism. Further investigation of the origins of
bioluminescent elaterids will need denser sampling of all biolu-
minescent genera, including as many species as possible. Nev-
ertheless, we support the earlier proposed hypothesis that ela-
terid photic organs evolved several times (Bocakova et al., 2007;
Sagegami-Oba et al., 2007; Fallon et al., 2018). The known
examples of elaterid bioluminescence are placed in three dif-
ferent subfamilies (Costa, 1975, 1984), and a non-luminescent
most recent common ancestor and sister taxa have already
been recovered for Balgus (Thylacosterninae) and Pyrophorini
(Agrypninae) (Bocakova et al., 2007; Kundrata et al., 2014).
How old is elateroid bioluminescence?
We do not attempt a formal dating analysis as our data
partly overlap with the datasets used earlier for the analyses
of all beetles with multiple fossil calibration points (McKenna
et al., 2015, 2019; Bocak et al., 2016; Toussaint et al., 2017;
Kusy et al., 2018a; Zhang et al., 2018a). Most calibration points
would not be available if we attempt an analysis restricted
only to elateroids. Previous studies and the elateroid fossil
records already serve as a guide to estimating of the origins
of both bioluminescence and Sinopyrophoridae. The recov-
ered topologies place Sinopyrophorus between the rst split
of the elaterid-lampyroid clade and the deepest split between
soft-bodied bioluminescent reies, railroad and glow-worms
(Figs 2, 3). Most published analyses dated the earliest split
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12451

10 D. Kusy et al.
within the elaterid-lampyroid clade to the lower Cretaceous peri-
ods, ∼135 Ma (Toussaint et al., 2017; Zhang et al., 2018a;
McKenna et al., 2019) (Ma; Fig. 4). More shallow estimations
of 115–125 Ma (McKenna et al., 2015; Fallon et al., 2018), as
well as a deeper one at 165 Ma (Kusy et al., 2018a,b) are also
hypothesized. The estimations for the earliest split between the
reies and glow-worm clade are similarly inconclusive, at ∼98
Ma (Zhang et al., 2018a), ∼122 Ma (Martin et al., 2019), and
∼140 Ma (Kusy et al., 2018a,b), but the presence of a lucioline
rey in Burman amber supports older dates (Kazantsev, 2015).
If we consider median estimations, the elaterid-lampyroid clade
would have originated in the lower Cretaceous and the subse-
quent split of the Lampyridae, Rhagophthalmidae and Phengo-
didae clade in the mid-Cretaceous. As a result, the ancestor of
Sinopyrophoridae had to have split from their closest relative
∼120 Ma (Fig. 4). That period must also be considered as a
moment when bioluminescence evolved in Elateroidea for the
rst time.
Dating analyses using different taxon sampling, analysed
markers, applied models and calibrations often come to dif-
ferent age estimates (Fig. 4). Some studies recovered the early
origins of the bioluminescent elateroids (Bocak et al., 2016;
Kusy et al., 2018a) and we suggest these should be inspected as
a possibility. The shallower estimates (McKenna et al., 2015,
2019; Toussaint et al., 2017; Fallon et al., 2018; Zhang et al.,
2018a) set the origin of the elaterid-lampyroid clade of which
click beetles is a principal branch to 115–140 Ma in contrast
with the upper Jurassic 152–158 Ma old Karatau deposits
contain a rich click beetle fauna (Doludenko et al., 1990).
Additionally, Elaterophanes (Whalley, 1985) was assigned to
Elateridae, and even if it is not a true click beetle but an ances-
tral non-artematopodid elateroid lineage, its age is older than
some estimates of the origin of Elateroidea (Fig. 4). Similarly,
false click beetles, Eucnemidae, were reported from the lower
Jurassic Xiwan deposits (Lin, 1986) and multiple species are
known also from the upper Jurassic– lower Cretaceous period
(Chang et al., 2011). These fossils also support quite an early
origin of the Elateroidea. Consequently, an earlier origin of
bioluminescence is also possible.
Conclusion
The phylogenetic studies dealing with Elateroidea produce
conicting results (Bocakova et al., 2007; Sagegami-Oba
et al., 2007; Lawrence et al., 2011) and similarly, the placement
of Sinopyrophorus as a terminal clade within Elateridae is
ambiguous (Bi et al., 2019; He et al., 2019). Phylogenomic data
presented here provide strong evidence that Sinopyrophoridae,
stat.n., is the sister group of reies, railroad and glow-worm
beetles, despite its morphological similarity to the extant click
beetles (Fig. 6). Our nding suggests that lampyrids and their
closest relatives are in fact modied click beetles. The denition
of the Sinopyrophoridae+Phengodidae+Rhagophthalmidae+
Lampyridae clade supports an early origin of bioluminescence
in their common clicking ancestor, likely ∼120 Ma, in the lower
Cretaceous, the delayed shift to being soft-bodied and further
radiation and signal diversication in the lineages which already
used some form of luminescence.
Author contributions
DK analysed genomic data, carried out sequence alignments
and phylogenetic analyses, XYL and JWH produced genomic
data, MM and JWH participated in analyses, all co-authors
contributed to the draft of the manuscript; WXB collected the
specimen, DK, LB, LP, and SB conceived and designed the
study, XYL and LB coordinated the study, LB, DK, SB, XYL
and JWH drafted the manuscript. All authors commented the
drafts and gave nal approval for publication.
Supporting Information
Additional supporting information may be found online in
the Supporting Information section at the end of the article.
Figure S1. Classication of Elateroidea: an overview of
hypothesized phylogenies.
Figure S1. An overview of Elateroidea topologies recovered
by earlier studies.
Figure S2– S22. Topologies recovered by individual analy-
ses.
Figure S23– S24. DiscoVista relative frequency analyses.
Figure S25. Four cluster likelihood mapping (FcLM) analy-
ses.
Figure S26. Calculated log-likelihood difference analyses.
Figure S27– S32. AliStat and SymTest analyses.
Figure S33. BUSCO3 and QUAST assessment for draft
genome assembly.
Figure S34. K-mer coverage depth of assembled Spades
scaffolds.
Table S1. The list of taxa, accession numbers.
Table S2. Overview of gene sets used for ortholog assess-
ment.
Table S3. Descriptive statistics and results of the orthology
assignment.
Table S4. Detailed information and statistics of each gener-
ated dataset.
Table S5. The list of taxa included in the 66-gene datasets.
Table S6. Results of four-cluster likelihood mapping and
approximately unbiased test.
Table S7. Result of approximately unbiased (AU) test for the
dataset A-4199-AA.
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12451

Phylogenomics of luminescent elateroids 11
Acknowledgements
We are obliged for the generous advice and pipeline sharing
to all colleagues from the Museum of A. Koenig in Bonn
and R. Bilkova is cordially acknowledged for her assistance.
J. Klváˇ
cek. and Z.-W. Dong generously provided their pho-
tographs. We specially thank Wen Wang and other members in
his laboratory for their support. We thank to the editor and two
anonymous reviewers for invaluable comments on the earlier
version of the manuscript. The study was funded by the GACR
and IGA projects (18-14942S; LB, DK, and MM; PrF-2020;
DK), the projects of NNSF China (31472035; X-YL) and CAS
(‘Light of West China’; X-YL), the NSF project (DEB-1655981;
SMB). The authors declare no conicts of interest.
Data availability statement
The data that support the ndings of this study are openly
available in GenBank, accession number PRJNA64573 and the
datasets are openly available in the Mendeley Data repository at
https://www.doi.org/10.17632/g4xkckycph.1.
References
Amaral, D.T., Mitani, Y., Ohmiya, Y. & Viviani, V.R. (2016) Organiza-
tion and comparative analysis of the mitochondrial genomes of biolu-
minescent Elateroidea (Coleoptera: Polyphaga). Gene,586, 254– 262.
Amaral, D.T., Silva, J.R. & Viviani, V.R. (2017) Transcriptional com-
parison of the photogenic and non-photogenic tissues of Phrixothrix
hirtus (Coleoptera: Phengodidae) and non-luminescent Chauliog-
nathus avipes (Coleoptera: Cantharidae) give insights on the origin
of lanterns in railroad worms. Gene Reports,7, 78– 86.
Amaral, D.T., Silva, J.R. & Viviani, V.R. (2019) RNA-Seq analysis
of the bioluminescent and non-bioluminescent species of Elateridae
(Coleoptera): comparison to others photogenic and non-photogenic
tissues of Elateroidea species. Comparative Biochemistry and Physi-
ology Part D: Genomics and Proteomics,29, 154 – 165.
Bankevich, A., Nurk, S., Antipov, D. et al. (2012) SPAdes: a new
genome assembly algorithm and its applications to single-cell
sequencing. Journal of Computational Biology,19, 455– 477.
Barrow, L.N., Lemmon, A.R. & Lemmon, E.M. (2018) Targeted
sampling and target capture: assessing phylogeographic concordance
with genome-wide data. Systematic Biology,67, 979– 996.
Bi, W.X., He, J.W., Chen, C.C., Kundrata, R. & Li, X.Y. (2019) Sinopy-
rophorinae, a new subfamily of Elateridae (Coleoptera, Elateroidea)
with the rst record of a luminous click beetle in Asia and evidence
for multiple origins of bioluminescence in Elateridae. ZooKeys,864,
79– 97.
Bocak, L., Kundrata, R., Andújar-Fernández, C. & Vogler, A.P. (2016)
The discovery of Iberobaeniidae (Coleoptera: Elateroidea): a new
family of beetles from Spain, with immatures detected by environ-
mental DNA sequencing. Proceedings of the Royal Society B: Bio-
logical Sciences,283, 20152350.
Bocakova, M., Bocak, L., Hunt, T., Teräväinen, M. & Vogler, A.P. (2007)
Molecular phylogenetics of Elateriformia (Coleoptera): evolution of
bioluminescence and neoteny. Cladistics,23, 477– 496.
Borowiec, M.L. (2016) AMAS: a fast tool for alignment manipulation
and computing of summary statistics. PeerJ,4, e1660.
Branham, M.A. & Wenzel, J.W. (2003) The evolution of photic
behaviour and the evolution of sexual communication in reies
(Coleoptera: Lampyridae). Cladistics,84, 565– 586.
Breinholt, J.W. & Kawahara, A.Y. (2013) Phylotranscriptomics: sat-
urated third codon positions radically inuence the estimation of
trees based on next-gen data. Genome Biology and Evolution,5,
2082– 2092.
Chang, H., Kirejtshuk, A.G. & Ren, D. (2011) On taxonomy and
distribution of fossil Cerophytidae (Coleoptera: Elateriformia) with
description of a new Mesozoic species of Necromera Martynov, 1926.
Annales Societe Entomologique de France,47, 33–44.
Chen, S., Zhou, Y., Chen, Y. & Gu, J. (2018) Fastp: an ultra-fast
all-in-one FASTQ preprocessor. Bioinformatics,34, i884–i890.
Chernomor, O., von Haeseler, A. & Minh, B.Q. (2016) Terrace aware
data structure for phylogenomic inference from supermatrices. Sys-
tematic Biology,65, 997– 1008.
Costa, C. (1975) Systematics and evolution of the tribes Pyrophorini
and Heligmini, with description of Campyloxeninae, new subfamily
(Coleoptera, Elateridae). Arquivos de Zoologia,26, 49– 190.
Costa, C. (1984) Note on the bioluminescence of Balgus schnusei
(Heller, 1914) (Trixagidae, Coleoptera). Revista Brasileira de Ento-
mologia,28, 397– 398.
Degnan, J.H. & Rosenberg, N.A. (2006) Discordance of species trees
with their most likely gene trees. PLoS Genetics,2, e68.
Doludenko, M.P., Ponomarenko, A.G. & Sakulina, G.V. (1990) La
géologie du gisement unique de la faune et de la ore du jurassique
supérieur d’Aulié (Karatau, Kazakhstan du Sud). Académie des
Sciences de l’URSS, Institut Géologique, Moscow.
Edwards, S.V. (2009) Is a new and general theory of molecular
systematics emerging? Evolution,63, 1– 19.
Ellis, E.A. & Oakley, T.H. (2016) High rates of species accumulation in
animals with bioluminescent courtship displays. Current Biology,26,
1–6.
Erfan, S. & Mirarab, S. (2016) Fast coalescent-based computation of
local branch support from quartet frequencies. Molecular Biology and
Evolution,33, 1654– 1668.
Erfan, S., James, J.B., Whiteld, B. & Mirarab, S. (2018) DiscoVista:
interpretable visualizations of gene tree discordance. Molecular
Phylogenetics and Evolution,122, 110– 115.
Fallon, T.R., Lower, S.E., Chang, C.-H. et al. (2018) Firey genomes
illuminate parallel origins of bioluminescence in beetles. eLife,7,
e36495.
Fernandez, R., Edgecombe, G.D. & Giribet, G. (2016) Exploring phy-
logenetic relationships within Myriapoda and the effects of matrix
composition and occupancy on phylogenomic reconstruction. System-
atic Biology,65, 871– 889.
He, J.W., Bi, W., Dong, Z. et al. (2019) The mitochondrial genome of the
rst luminous click-beetle (Coleoptera: Elateridae) recorded in Asia.
Mitochondrial DNA Part B,4, 565 –567.
Hoang, D.T., Chernomor, O., von Haeseler, A., Minh, B.Q. & Vinh, L.S.
(2018) UFBoot2: improving the ultrafast bootstrap approximation.
Molecular Biology and Evolution,35, 518– 522.
International Committee for Zoological Nomenclature (1999) Interna-
tional Code of Zoological Nomenclature. International Trust for Zoo-
logical Nomenclature, London.
Jermiin, L.S., Jayaswal, V., Ababneh, F. & Robinson, J. (2008) Phylo-
genetic model evaluation. Bioinformatics, Data, Sequence Analysis
and Evolution,Vol.I(ed. by J. Keith), pp. 331–363. Humana Press,
Tot owa , N ew J e rse y.
Junier, T. & Zdobnov, E.M. (2010) The Newick utilities:
high-throughput phylogenetic tree processing in the UNIX shell.
Bioinformatics,26, 1669– 1670.
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12451

12 D. Kusy et al.
Kalyaanamoorthy, S., Minh, B.Q., Wong, T.K.F., von Haeseler, A. &
Jermiin, L.S. (2017) ModelFinder: fast model selection for accurate
phylogenetic estimates. Nature Methods,14, 587– 589.
Katoh, K. & Standley, D.M. (2013) MAFFT multiple sequence align-
ment software version 7: improvements in performance and usability.
Molecular Biology and Evolution,30, 772– 780.
Kawahara, A.Y., Ohshima, I., Kawakita, A. et al. (2011) Increased
gene sampling strengthens support for higher-level groups within
leaf-mining moths and relatives (Lepidoptera: Gracillariidae). BMC
Evolutionary Biology,11, 182.
Kazantsev, S.V. (2015) Protoluciola albertalleni gen.n., sp.n., a new
Luciolinae rey (Insecta: Coleoptera: Lampyridae) from Burmite
amber. Russian Entomological Journal,24, 281– 283.
Kück, P. & Longo, G.C. (2014) FASconCAT-G: extensive functions for
multiple sequence alignment preparations concerning phylogenetic
studies. Frontiers in Zoology,11, 81.
Kück, P. & Struck, T.H. (2014) BaCoCa – a heuristic software tool for
the parallel assessment of sequence biases in hundreds of gene and
taxon partitions. Molecular Phylogenetics and Evolution,70, 94– 98.
Kück, P., Meusemann, K., Dambach, J., Thormann, B., von
Reumont, B.M., Wägele, J.W. & Misof, B. (2010) Parametric
and non-parametric masking of randomness in sequence alignments
can be improved and leads to better resolved trees. Frontiers in
Zoology,7, 10.
Kundrata, R., Bocakova, M. & Bocak, L. (2014) The comprehensive
phylogeny of the superfamily Elateroidea (Coleoptera: Elateriformia).
Molecular Phylogenetics and Evolution,76, 162– 171.
Kusy, D., Motyka, M., Andújar, C. et al. (2018a) Genome sequencing
of Rhinorhipus Lawrence exposes an early branch of the Coleoptera.
Frontiers in Zoology,15, 21.
Kusy, D., Motyka, M., Bocek, M., Vogler, A.P. & Bocak, L. (2018b)
Genome sequences identify three families of Coleoptera as mor-
phologically derived click beetles (Elateridae). Scientic Reports,8,
17084.
Kusy, D., Motyka, M., Bocek, M., Masek, M. & Bocak, L. (2019)
Phylogenomic analysis resolves the relationships among net-winged
beetles (Coleoptera: Lycidae) and reveals the parallel evolution of
morphological traits. Systematic Entomology,44, 911– 925.
Lanfear, R., Frandsen, P.B., Wright, A.M., Senfeld, T. & Calcott,
B. (2017) PartitionFinder 2: new methods for selecting partitioned
models of evolution for molecular and morphological phylogenetic
analyses. Molecular Biology and Evolution,34, 772– 773.
Lawrence, J.F. (1988) Rhinorhipidae, a new beetle family from Aus-
tralia, with comments on the phylogeny of the Elateriformia. Inverte-
brate Taxonomy,2, 1 –53.
Lawrence, J.F., Slipinski, A., Seago, A.E. et al. (2011) Phylogeny of the
Coleoptera based on morphological characters of adults and larvae.
Annales Zoologici,61, 1– 217.
Lin, Q.B. (1986) Early Mesozoic fossil insects from South China.
Palaeontologia Sinica,170, 1–112.
Linard, B., Crampton-Platt, A., Moriniere, J. et al. (2018) The contribu-
tion of mitochondrial metagenomics to large-scale data mining and
phylogenetic analysis of Coleoptera. Molecular Phylogenetics and
Evolution,128, 1– 11.
Marcais, G. & Kingsford, C. (2011) A fast, lock-free approach for
efcient parallel counting of occurrences of k-mers. Bioinformatics,
27, 764– 770.
Martin, G.J., Branham, M., Whiting, M.F. & Bybee, S.M. (2017)
Total evidence phylogeny and the evolution of adult bioluminescence
in reies (Coleoptera: Lampyridae). Molecular Phylogenetics and
Evolution,107, 564– 575.
Martin, G.J., Stanger-Hall, K.F., Branham, M.A. et al. (2019)
Higher-level phylogeny and reclassication of Lampyridae
(Coleoptera: Elateroidea). Insect Systematics and Diversity,3,
11.
McKenna, D.D., Wild, A.L., Kanda, K. et al. (2015) The beetle tree
of life reveals that Coleoptera survived end-Permian mass extinction
to diversify during the Cretaceous terrestrial revolution. Systematic
Entomology,40, 835– 880.
McKenna, D.D., Shin, S., Ahrens, D. et al. (2019) The evolu-
tion and genomic basis of beetle diversity. Proceedings of the
National Academy of Sciences of the United States of America,116,
24729– 24737.
Mikheenko, A., Prjibelski, A., Saveliev, V., Antipov, D. & Gurevich,
A. (2018) Versatile genome assembly evaluation with QUAST-LG.
Bioinformatics,34, i142– i150.
Minh, B.Q., Schmidt, H.A., Chernomor, O., Schrempf, D., Woodhams,
M.D., von Haeseler, A. & Lanfear, R. (2020) IQ-TREE 2: new models
and efcient methods for phylogenetic inference in the genomic era.
Molecular Biology and Evolution,37, 1530– 1534.
Mirarab, S., Bayzid, M.S. & Warnow, T. (2016) Evaluating summary
methods for multilocus species tree estimation in the presence of
incomplete lineage sorting. Systematic Biology,65, 366– 380.
Misof, B. & Misof, K.A. (2009) A Monte Carlo approach successfully
identies randomness in multiple sequence alignments: a more
objective means of data exclusion. Systematic Biology,58,21–34.
Misof, B., Meyer, B., von Reumont, B.M., Kück, P., Misof, K. & Meuse-
mann, K. (2013) Selecting informative subsets of sparse supermatri-
ces increases the chance to nd correct trees. BMC Bioinformatics,
14, 348.
Misof, B., Liu, S., Meusemann, K. et al. (2014) Phylogenomics resolves
the timing and pattern of insect evolution. Science,346, 763– 767.
Muona, J. & Taräväinen, M. (2020) A re-evaluation of the Eucnemidae
larval characters. Papeis Avulsos de Zoologia,60. https://doi.org/10
.11606/1807-0205/2020.60.special- issue.28
Nakatsu, T., Ichiyama, S., Hiratake, J., Saldanha, A., Kobashi, N.,
Sakata, K. & Kato, H. (2006) Structural basis for the spectral
difference in luciferase bioluminescence. Nature,440, 372– 376.
Naser-Khdour, S., Minh, B.Q., Zhang, W., Stone, S.A. & Lanfear, R.
(2019) The prevalence and impact of model violations in phylogenetic
analysis. Genome Biology and Evolution,11, 3341– 3352.
Peters, R.S., Krogmann, L., Mayer, C. et al. (2017) Evolutionary history
of the Hymenoptera. Current Biology,27, 1013– 1018.
Petersen, M., Meusemann, K., Donath, A. et al. (2017) Orthograph: a
versatile tool for mapping coding nucleotide sequences to clusters of
orthologous genes. BMC Bioinformatics,18, 111.
Poelchau, M., Childers, C., Moore, G. et al. (2014) The i5k
Workspace@NAL enabling genomic data access, visualization
and curation of arthropod genomes. Nucleic Acids Research,43,
D714– D719.
Sagegami-Oba, R., Takahashi, N. & Oba, Y. (2007) The evolutionary
process of bioluminescence and aposematism in cantharoid beetles
(Coleoptera: Elateroidea) inferred by the analysis of 18S ribosomal
DNA. Gene,400, 104– 113.
Sanders, S.E. & Hall, D.W. (2015) Variation in opsin genes correlates
with signalling ecology in north American reies. Molecular Ecol-
ogy,24, 4679–4696.
Stanke, M. & Waack, S. (2003) Gene prediction with a hidden Markov
model and a new intron submodel. Bioinformatics,19, 215– 225.
Strimmer, K. & von Haeseler, A. (1997) Likelihood-mapping: a simple
method to visualize phylogenetic content of a sequence alignment.
Proceedings of the National Academy of Sciences of the United States
of America,94, 6815– 6819.
Suyama, M., Torrents, D. & Bork, P. (2006) PAL2NAL: robust con-
version of protein sequence alignments into the corresponding codon
alignments. Nucleic Acids Research,34, W609– W612.
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12451

Phylogenomics of luminescent elateroids 13
Timmermans, M.J.T.N., Dodsworth, S., Culverwell, C.L. et al. (2010)
Why barcode? High-throughput multiplex sequencing of mitochon-
drial genomes for molecular systematics. Nucleic Acids Research,38,
e197.
Timmermans, M.J.T.N., Barton, C., Haran, J. et al. (2016) Family-level
sampling of mitochondrial genomes in Coleoptera: compositional
heterogeneity and phylogenetics. Genome Biology and Evolution,8,
161– 175.
Toussaint, E., Seidel, M., Arriaga-Varera, E. et al. (2017) The peril of
dating beetles. Systematic Entomology,42, 1– 10.
Trueman, J.W.H., Pfeil, B.E., Kelchners, S.A. & Yeates, D.K. (2004)
Did stick insects really regain their wings? Systematic Entomology,
29, 138– 139.
Vasilikopoulos, A., Balke, M., Beutel, R.G. et al. (2019) Phylogenomics
of the superfamily Dytiscoidea (Coleoptera: Adephaga) with an
evaluation of phylogenetic conict and systematic error. Molecular
Phylogenetics and Evolution,135, 270– 285.
Vurture, G.W., Sedlazeck, F.J., Nattestad, M., Underwood, C.J.,
Fang, H., Gurtowski, J. & Schatz, M.C. (2017) GenomeScope: fast
reference-free genome proling from short reads. Bioinformatics,33,
2202– 2204.
Wang, K., Hong, W., Jiao, H. & Zhao, H. (2017) Transcriptome
sequencing and phylogenetic analysis of four species of luminescent
beetles. Scientic Reports,7, 1814.
Waterhouse, R.M., Seppey, M., Simão, F.A. et al. (2018) BUSCO appli-
cations from quality assessments to gene prediction and phyloge-
nomics. Molecular Biology and Evolution,35, 543– 548.
Watkins, O.C., Sharpe, M.L., Perry, N.B. & Krause, K.L. (2018) New
Zealand glowworm (Arachnocampa luminosa) bioluminescence is
produced by a rey-like luciferase but an entirely new luciferin.
Scientic Reports,8, 3278.
Whalley, P.E.S. (1985) The systematics and palaeogeography of the
lower Jurassic insects of Dorset, England. Bulletin of the British
Museum of Natural History,39, 107– 189.
Ye, B., Zhang, Y., Shu, J., Wu, H. & Wang, H. (2018) RNA-sequencing
analysis of fungi-induced transcripts from the bamboo wireworm
Melanotus cribricollis (Coleoptera: Elateridae) larvae. PLoS One,13,
e019118.
Zdobnov, E.M., Tegenfeldt, F., Kuznetsov, D. et al. (2016)
OrthoDBv9.1: cataloging evolutionary and functional annota-
tions for animal, fungal, plant, archaeal, bacterial and viral orthologs.
Nucleic Acids Research,45, D744– D749.
Zhang, S.Q., Che, L.H., Li, Y., Dan Liang, Pang, H., ´
Slipi´
nski, A. &
Zhang, P. (2018a) Evolutionary history of Coleoptera revealed by
extensive sampling of genes and species. Nature Communications,9,
205.
Zhang, C., Rabiee, M., Sayyari, E. & Mirarab, S. (2018b) ASTRAL-III:
polynomial time species tree reconstruction from partially resolved
gene trees. BMC Bioinformatics,19(S6), 153.
Zhang, J., Lindsey, A.R.I., Peters, R.S. et al. (2020) Conicting signal
in transcriptomic markers leads to a poorly resolved backbone
phylogeny of chalcidoid wasps. Systematic Entomology. https://doi
.org/10.1111/syen.12427.
Zwick, A., Regier, J.C. & Zwickl, D.J. (2012) Resolving discrepancy
between nucleotides and amino acids in deep-level arthropod phy-
logenomics: differentiating serine codons in 21-amino-acid models.
PLoS One,7, e47450.
Accepted 23 July 2020
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12451