Molecular systematics reveals the origins of subsociality in tortoise beetles (Coleoptera, Chrysomelidae, Cassidinae): Evolution of subsociality in Cassidinae

Abstract

Subsocial behaviour is known to occur in at least 19 insect orders and 17 families of Coleoptera. Within the leaf beetle family, Chrysomelidae, extended maternal care is reported in only 2 of 15 subfamilies: Cassidinae and Chrysomelinae. Although the emergence of subsociality in insects has received much attention, extensive analyses on the evolution of this behaviour based on phylogenetic approaches are missing. Subsociality is recorded in 33 species of tortoise beetles belonging to the tribes Mesomphaliini and Eugenysini. A molecular phylogenetic reconstruction of these tribes and the remaining five Neotropical tribes of cassidine tortoise beetles was used to investigate the evolution of maternal care and to elucidate the phylogenetic relationships among Neotropical cassidine tribes. A phylogeny was constructed using 90 species and three loci from both mitochondrial and nuclear genes (COI, CAD and 28S). Bayesian inference and maximum likelihood analyses based on a concatenated dataset recovered two independent origins, with no evidence of reversal to solitary behaviour. One origin comprises three Mesomphaliini genera tightly associated with Convolvulaceae, and the other consists of the genus Eugenysa Chevrolat (Eugenysini), a small clade embedded within a group feeding exclusively on Asteraceae. A previous hypothesis suggesting dual origins on different host plants was confirmed, whereas other hypotheses based on a phylogenetic reconstruction of Cassidinae could not be sustained. Our analysis also revealed that the tribe Mesomphaliini is a monophyletic taxon if Eugenysini is included, and for this reason, we re‐establish synonymy of both tribes. We also provide nine new records of subsociality for tortoise beetles species.

Full text

Systematic Entomology (2020), DOI: 10.1111/syen.12434
Molecular systematics reveals the origins
of subsociality in tortoise beetles (Coleoptera,
Chrysomelidae, Cassidinae)
MICHELE LEOCÁDIO1,2, MARIANNA V. P. SIMÕES3,
LUKAS SEKERKA4, CARLOS G. SCHRAGO1,5,
JOSÉ R. M. MERMUDES1,2andDONALD M. WINDSOR6
1Programa de Pós-graduação em Biodiversidade e Biologia Evolutiva, Instituto de Biologia, Universidade Federal do Rio de Janeiro,
Rio de Janeiro, Brazil, 2Laboratório de Entomologia, Departamento de Zoologia, Instituto de Biologia, Universidade Federal do Rio
de Janeiro, Rio de Janeiro, Brazil, 3Center of Natural History, University of Hamburg, Hamburg, Germany, 4Department of
Entomology, National Museum, Horní Poˇ
cernice, Czech Republic, 5Laboratório de Biologia Evolutiva Teórica e Aplicada,
Departamento de Genética, Instituto de Biologia, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil and 6Smithsonian
Tropical Research Institute, Balboa, Republica de Panamá
Abstract. Subsocial behaviour is known to occur in at least 19 insect orders and
17 families of Coleoptera. Within the leaf beetle family, Chrysomelidae, extended
maternal care is reported in only 2 of 15 subfamilies: Cassidinae and Chrysomelinae.
Although the emergence of subsociality in insects has received much attention, extensive
analyses on the evolution of this behaviour based on phylogenetic approaches are
missing. Subsociality is recorded in 33 species of tortoise beetles belonging to the
tribes Mesomphaliini and Eugenysini. A molecular phylogenetic reconstruction of these
tribes and the remaining ve Neotropical tribes of cassidine tortoise beetles was used to
investigate the evolution of maternal care and to elucidate the phylogenetic relationships
among Neotropical cassidine tribes. A phylogeny was constructed using 90 species and
three loci from both mitochondrial and nuclear genes (COI, CAD and 28S). Bayesian
inference and maximum likelihood analyses based on a concatenated dataset recovered
two independent origins, with no evidence of reversal to solitary behaviour. One origin
comprises three Mesomphaliini genera tightly associated with Convolvulaceae, and the
other consists of the genus Eugenysa Chevrolat (Eugenysini), a small clade embedded
within a group feeding exclusively on Asteraceae. A previous hypothesis suggesting
dual origins on different host plants was conrmed, whereas other hypotheses based on
a phylogenetic reconstruction of Cassidinae could not be sustained. Our analysis also
revealed that the tribe Mesomphaliini is a monophyletic taxon if Eugenysini is included,
and for this reason, we re-establish synonymy of both tribes. We also provide nine new
records of subsociality for tortoise beetles species.
Introduction
Subsocial behaviour represents the most complex stage of
parental care, in which one or both parents care for offspring
Correspondence: Michele Leocádio, Laboratório de Entomologia,
Departamento de Zoologia, Instituto de Biologia, Universidade Federal
do Rio de Janeiro, A1-107, Bloco A, Av. Carlos Chagas Filho, 373,
Cidade Universitária, Ilha do Fundão, Rio de Janeiro, RJ, PO BOX
68044, Brazil. E-mail: michelegleocadio@gmail.com
after hatching (Michener, 1969; Wilson, 1971). Care often takes
the form of guiding towards nutrients, shielding against preda-
tors and parasites or directly provisioning food through most
or all stages of development, and it ceases once offspring dis-
perse and mate (Clutton-Brock, 1991; Royle et al., 2012; Yip
& Rayor, 2014). The behaviour is thought to expose parents
to increased mortality and interfere with individual mainte-
nance and reproductive activities for an extended period of
time (Alonso-Alvarez & Velando, 2012). As such, parental care
© 2020 The Royal Entomological Society 1

2M. Leocádio et al.
represents an evolutionary trade-off between immediate repro-
ductive output and delayed gains through future offspring sur-
vival (Rankin et al., 2007; Elgar, 2015).
Subsociality is widespread among insects, occurring in at least
19 orders (including both hemi- and holometabolous) and 164
families (Machado & Trumbo, 2018). Within Coleoptera sub-
sociality is a rare behaviour recorded in only 17 of 189 families
(Machado & Trumbo, 2018). Subsociality in the leaf beetle fam-
ily, Chrysomelidae, occurs in two phylogenetically distinct sub-
families, the Cassidinae Gyllenhal, 1813, which includes both
tortoise beetles and leaf-mining beetles, and the Chrysomelinae
Gyllenhal, 1802, known commonly as broad-shouldered beetles
(Gómez-Zurita et al., 2008; Chaboo et al., 2014). However, the
absence of a well-supported phylogeny for each subfamily has
hindered appreciation of the number of origins and the ecolog-
ical factors that may have originally selected for subsociality
within each subfamily.
Cassidinae, the second largest subfamily of leaf beetles,
comprises close to 6400 species, distributed worldwide
(Staines, 2015; Borowiec & ´
Swi
¸
etoja´
nska, 2020). Within
the subfamily, maternal care (MC) is known in 33 species of
tortoise beetles distributed in only 2 of 35 tribes, Eugenysini
Hincks, 1952 and Mesomphaliini Hope, 1840, both conned
to the New World, with most of the diversity in the Neotrop-
ical region (Chaboo et al., 2014; Flinte et al., 2015; Macedo
et al., 2015). Females care for and defend all immature stages
(eggs, larvae and pupae) over a period of four or more weeks,
during which the tending mother largely forgoes feeding and
only rarely drinks water (Windsor, 1987; Chaboo et al., 2014).
The tribe Mesomphaliini contains 545 species in 25 gen-
era (Borowiec & ´
Swi
¸
etoja´
nska, 2020), with MC reported
from three genera (n∘of spp. with MC recorded/n∘of valid
spp.), Acromis Chevrolat (3/3), Paraselenis Spaeth (10/29)
and Omaspides Chevrolat (15/40) (Windsor, 1987; Chaboo
et al., 2014; Flinte et al., 2015; Macedo et al., 2015). Within
the 35 species classied in three genera of Eugenysini, MC
is reported in Agenysa Spaeth (1/9) and Eugenysa Chevrolat
(4/18) (Chaboo et al., 2014; Macedo et al., 2015) (Table S1).
Distinct origins of MC within Cassidinae were initially
hypothesized by Windsor & Choe (1994) after observing that
subsocial Eugenysini and Mesomphaliini taxa were exclusively
associated with different host plant families, Asteraceae and
Convolvulaceae, respectively. In addition, the two tribes differed
conspicuously in length and degree of spermathecal duct coiling
and in the construction of egg masses. The Eugenysini attach
eggs end-on to the substrate, forming a disc-like collar com-
pletely encircling the stem, whereas Mesomphaliini taxa con-
sistently attach eggs as a semi-pendant mass, usually attached
to the midrib of the natal leaf (Fig. 1). Later, based on a phy-
logenetic reconstruction of the subfamily using morphological
characters, Chaboo et al. (2014) proposed alternative evolution-
ary hypotheses for the origins of MC: (i) a single origin giving
rise to parental care species feeding on both Convolvulaceae
and Asteraceae (with subsequent loss in Echoma Chevrolat, a
nonmaternal care genus strictly associated with Mikania (Aster-
aceae)) or (ii) dual origins, one leading to an Omaspides clade
and a second giving rise to Acromis,Paraselenis and Eugenysini.
The phylogenetic relationships of the tribes Mesomphaliini
and Eugenysini are likely more complex than either of the
above studies indicated and are complicated by the paucity
of sampled taxa. Hope (1840) proposed the name Mesomphal-
idae to accommodate 11 genera (only 6 presently classied
in this tribe) but provided no diagnostic characters for the
group. Chapuis (1875) established tribal classication of Cas-
sidinae and characterized Mesomphaliini (as Mésomphaliites)
based on the head partially visible from above, the prosternum
expanded towards mouthparts and tarsal claws appendiculate.
Subsequently, Spaeth (1942) proposed tribe Calaspideitae for
three genera of large cassidines previously included in Mesom-
phaliini. However, the name does not meet the requirements
of the Article 13.1. ICZN (1999), and moreover, Calaspidea
Hope, 1840 proved to be a younger synonym of Eugenysa
Chevrolat, 1836. Finally, Hincks (1952) revised the tribes of
Cassidinae s. str. and proposed the name Eugenysini for Calaspi-
deitae; the only diagnostic character given in the key to dif-
ferentiate Mesomphaliini (incorrectly changed to Stolaini) from
Eugenysini was the expanded tarsomere V covering tarsal claws,
which are thus not visible from above.
Borowiec (1995) suggested a new tribal classication for
Cassidinae s. str. based on a cladistic analysis and stated that
Eugenysini should be synonymized with Mesomphaliini. Later,
Hsiao & Windsor (1999), using a small molecular dataset (12S
mtDNA), recovered Mesomphaliini (=Stolaini) polyphyletic.
Eugenysini, which had a single species sampled, was recovered
within Mesomphaliini. Chaboo (2007), utilizing morphological
characters of adults and immature stages, did not recover most of
the advanced cassidine tribes due to a polytomy; Mesomphaliini
were nonmonophyletic, with Eugenysini monophyletic and
derived from within one of the Mesomphaliini clades.
Despite interest in revealing the evolutionary relationships
among all Neotropical Cassidinae, and the Mesomphaliini and
Eugenysini in particular, the drivers and pathways of MC in this
group remain unresolved. Below, we combine ancestral charac-
ter state reconstruction and molecular systematics to reconstruct
the phylogenetic history of the New World Cassidinae s. str.
aiming to (i) elucidate the phylogenetic relationships between
Mesomphaliini and Eugenysini, (ii) test the monophyly of MC
groups and their relationships with non-MC species, (iii) inves-
tigate the homology and evolution of MC and (iv) incorporate
new records of MC in Cassidinae based on recent observations.
Material and methods
Taxon sampling
We selected taxa representing all seven tribes of tortoise
beetles (=Cassidinae sensu Borowiec, 1999) occurring in the
Neotropics and feeding on dicotyledonous plants (Table S2).
We selected 65 species representing 15 of 25 genera of Mesom-
phaliini and 6 species representing two of the three genera of
Eugenysini. All genera presently known to contain subsocial
species were included: Acromis (3 spp.), Paraselenis (7 spp.) and
Omaspides (7 spp.) and within Eugenysini, Eugenysa (4 spp.)
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

Evolution of subsociality in Cassidinae 3
Fig 1. Different egg arrangements of subsocial tortoise beetles. A, Disc-like egg mass encircling the stem constructed by subsocial Eugenysini
(Eugenysa coscaroni); B, Pendant egg mass constructed by Acromis sparsa (Mesomphaliini). [Colour gure can be viewed at wileyonlinelibrary.com].
and Agenysa (2 spp.). As the evolutionary relationships among
tribes of Cassidinae are partially disputed (Borowiec, 1995;
Hsiao & Windsor, 1999; Chaboo, 2007; López-Pérez
et al., 2017; Simões et al., 2018), we included in our anal-
ysis 19 species representing the ve closest tribes: Cassidini,
Dorynotini, Goniocheniini, Ischyrosonychini and Omocerini.
Voucher specimens are deposited at the Earl S. Tupper Cen-
ter in the Smithsonian Tropical Research Institute, Panama
(DW vouchers); at the Prof. José Alfredo Pinheiro Dutra col-
lection (DZRJ), Laboratório de Entomologia, Departamento de
Zoologia, Universidade Federal do Rio de Janeiro, Brazil (ML
vouchers); and at the Entomology Division of the University of
Kansas, USA (MS vouchers).
DNA extraction and gene sequencing
Genomic DNA was extracted from leg and thoracic mus-
cles using the Qiagen DNeasy Blood and Tissue kit (Valencia,
California) from specimens preserved in 80– 95% ethanol and
maintained at −20∘C. Five separate gene regions were ampli-
ed: one from the mitochondrial genome, cytochrome oxidase
I (COI, 471 bp), and four from the nuclear genome, two frag-
ments of 28S rRNA (D1-D3, 997bp; D2, 504 bp) and two frag-
ments of carbamoylphosphate synthetase (CAD, 558 bp; 616 bp)
(Table S3). The shorter 28S fragment (D2 region) was used
when the larger fragment was not available. Additional frag-
ments of 28S and COI were obtained from the GenBank repos-
itory (Table S2).
Polymerase chain reactions (PCRs) were performed with
different cycling conditions for each gene fragment (Table S4).
For both CAD fragments, we performed a nested PCR, which
consists of amplifying a longer fragment (∼1600 bp) and then
using its PCR product to amplify the target fragments (Wild &
Maddison, 2008).
The entire PCR product was used in the electrophoresis gel.
Bands were cut from this product based on the expected size
of the fragment. This product was then puried by Gelase
(Epicentre Biotechnologies) digestion. Puried PCR products
were then prepared for sequencing procedures with the BigDye
Terminator 3.1 cycle sequencing kit (Applied Biosystems).
Sequencing reactions were cleaned using Sephadex G-50
columns on Millipore Multiscreen 96-well ltration plates or
via BigDye Xterminator kit (Applied Biosystems). The clean
sequencing reaction was run through an ABI 31306 sequencer
at the Naos Marine and Molecular Laboratories of Smithsonian
Tropical Research Institute.
The resulting sequences were trimmed to exclude primers
and regions of uncertainty based on the chromatograms.
Protein-coding fragments were inspected in Sequencher 5.1
(Gene Codes Corporation) to ensure the absence of stop codons.
The consensus fragments were checked for contamination in
Basic Local Alignment Search Tool (BLAST).
Alignment
Multiple sequence alignment of protein-coding genes was
performed by ClustalW (Thompson et al., 1994) implemented
on MEGA7 (Kumar et al., 2016) with default parameters. Due
to the existence of an overlapping region (∼118bp), both CAD
fragments were aligned and treated as one fragment (1065 bp).
The 28S rRNA fragments were aligned on the MAFFT online
version (Katoh et al., 2017), using the Q-INS-i algorithm, which
considers the secondary structure of RNA (Katoh & Toh, 2008).
Data partitioning and model selection
Fragments were concatenated with SequenceMatrix (Vaidya
et al., 2011). 28S rRNA fragments were regarded as a single
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

4M. Leocádio et al.
partition. The protein-coding genes were partitioned apriori,
with codon positions 1 and 2 coded as a single partition and
another partition for codon positions 3. The concatenated data
were then used to estimate the best evolutionary models using
PartitionFinder2 (Lanfear et al., 2016). A search using the set
of models available in MrBayes was run for data undergoing
Bayesian analysis. The ‘all’ set of models was implemented
when searching best schemes for maximum likelihood (ML). In
both searches, we used the ‘greedy’ algorithm under the cor-
rected Akaike information criterion (Hurvich & Tsai, 1993).
The best models were then selected for each partition and analy-
sis: GTR +I+G was selected for all the partitions under the BI
analysis, with the exception of COI positions 3 (GTR +G); for
the ML analysis, GTR +I+G was selected to positions 1 and 2
of COI and CAD, TVMEF +I+G for 28S, K81UF +GforCOI
positions 3 and TVM +I+G for CAD positions 3 (Table S5).
Phylogenetic analyses
Bayesian inference (BI) was carried out with MrBayes 3.2.6
(Ronquist et al., 2012) on the Cipres Science Gateway server
(Miller et al., 2010). Two independent and simultaneous runs
were conducted with eight Markov chain Monte Carlo (MCMC)
chains for 40 million generations, with trees sampled every 1000
generations. The rst 25% of the MCMC samples were dis-
carded as burn-in. After calculating the posterior probabilities
(PP), the trees were summarized to generate a 50% majority-rule
consensus tree. Convergence of both runs was evaluated through
the average standard deviation of split frequencies. The effective
sample size (ESS) values were calculated in Tracer 1.6 (http://
beast.bio.ed.ac.uk/Tracer), and only ESS >200 runs were fur-
ther analysed.
ML topological estimation was conducted with IQ-TREE
1.6 (Nguyen et al., 2015). We performed 1000 ultrafast boot-
strap replicates (UFBoot) (Hoang et al., 2017) and 1000
Shimodaira-Hasegawa approximate likelihood ratio test
replicates (SH-aLRT) (Anisimova et al., 2011) to evaluate
nodal support in 10 independent runs. The run with the best
log-likelihood score was selected.
Consensus trees generated in both analyses were visualized
and rooted with Dorynotini in FigTree 1.4.2 (http://tree.bio.ed.ac
.uk/software/gtree/). The tribe Dorynotini was chosen because
it is arguably the most distantly related to Mesomphaliini
among the tribes sampled (Shin, 2015; López-Pérez et al., 2017;
Simões et al., 2018). A Bayesian posterior probability ≥0.95,
SH-aLRT ≥80 and UFBoot ≥95 were recognized as indicating
strong support for a given node (Erixon et al., 2003; Minh
et al., 2013, 2017).
Ancestral character state reconstruction
The evolution of MC was assessed under an ancestral charac-
ter state reconstruction with Mesquite 3.04 (Maddison & Mad-
dison, 2015). A matrix was constructed including all taxa and
the character ‘maternal care’ was coded either as ‘present’ or
‘absent’. All Mesomphaliini genera (Acromis,Paraselenis and
Omaspides) containing MC species were coded as ‘present’ for
MC based on data from personal observations by the authors
and published accounts (Table S1). Eugenysini species lack-
ing unambiguous observations on reproductive behaviour were
coded as ‘missing data’.
As past reports have suggested that host plant choice may well
be associated with the selection of subsocial traits (Windsor &
Choe, 1994), we also reconstructed the host plant preference in
our tree to elucidate whether such a pattern exists. Information
regarding host plants was based on the Cassidinae of the
world catalogue (Borowiec & ´
Swi
¸
etoja´
nska, 2020), personal
observations and identications by D. M. Windsor and L.
Sekerka. Both analyses were reconstructed under a parsimony
approach due to low data complexity as indicated by both binary
characters that were not widespread in the tree.
Results
New records of subsociality in Mesomphaliini
Genus Paraselenis.Adult females of Paraselenis
(Spaethiechoma) normalis (Germar) were observed guard-
ing eggs (Fig. 2A, C) by ML and JRM on multiple occasions
in Cachoeiras de Macacu (22∘24′58.8′′ S, 42∘44′20.3′′ W,
185 m a.s.l.), Rio de Janeiro, Brazil. Males were found on the
same plant but not close to females (Fig. 2B). Both larvae and
adults were feeding on leaves of Distimake dissectus var. eden-
tatus (Meisn.) Petrongari & Sim.-Bianch (Convolvulaceae),
which is the rst host plant record for the species. We observed a
eulophid wasp land on the female, possibly trying to oviposit in
the beetle’s egg mass (Fig. 2C). Eulophidae are the principal egg
parasitoids for Neotropical Cassidinae (Cuignet et al., 2007).
Adult females of Paraselenis (Spaethiechoma) puncticollis
(Spaeth) were observed by D. M. Windsor and L. Sekerka
on multiple occasions guarding eggs, larvae and pupae in
Potrerillos del Guenda (17∘40′S, 63∘27′W, 370 m a.s.l.), Santa
Cruz Department, Bolivia (Fig. 2D– F).
Subsociality was also observed in Paraselenis
(Spaethiechoma) scapulosa (Boheman) in two distinct locali-
ties: Los Jubas near Santa Cruz de la Sierra, Bolivia by D. M.
Windsor and in Campo Grande, Mato Grosso do Sul, Brazil by
Renan Oliver (Fig. 2G– H).
We have also noted three photographs on the iNaturalist plat-
form of adult females of Paraselenis marginipennis (Spaeth)
guarding eggs (https://www.inaturalist.org/observations/
2729457; https://www.inaturalist.org/observations/2749526;
https://www.inaturalist.org/observations/2685441). All records
were made in Peru.
Genus Omaspides.Adult females of the following species
of Omaspides were observed guarding eggs: Omaspides (s.
str.) collecta Spaeth (Fig. 3A) in Via Nuevo Paraiso, between
Cabañas Yankuam and Kusunts (Miazi), Canton Nangaritza,
Provincia de Zamora Chinchipe, Ecuador (4∘16′9.347′′ S
78∘38′35.304′′ W, 1034 ma.s.l.) by Dominik Hofer; Omaspides
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

Evolution of subsociality in Cassidinae 5
Fig 2. New records of subsociality for Paraselenis.A–C,Paraselenis (Spaethiechoma) normalis, female guarding eggs (A), male (B) and eulophid
parasitoid wasps indicated by red circles (C); D– F, Paraselenis (Spaethiechoma) puncticollis females guarding eggs (D), larvae (E) and pupae (F);
G–H,Paraselenis (Spaethiechoma) scapulosa, female guarding pupae (G), male (H). [Colour gure can be viewed at wileyonlinelibrary.com].
(s. str.) confusa Borowiec (Fig. 3B) in Las Cascadas, Napo
Province, Ecuador by D. M. Windsor and Río Hollín at
Narupa-Loreto Road, Napo Province, Ecuador (0∘43′04′′ S,
77∘38′19′′ W, 1068 m a.s.l.) by L. Sekerka; and Omaspides (s.
str.) trifasciata (Fabricius) (Fig. 3C) near Camp Patawa, French
Guiana by D. M. Windsor. An adult female of Omaspides (s. str.)
picaorensis Borowiec was recorded in a photograph guarding
eggs in Peru (O’Toole & Preston-Mafham, 1985). Finally,
Omaspides (s. str.) trichroa (Boheman) (Fig. 3D-E) were
observed in three different sites within the Santa Cruz Province
in Bolivia: Potrerillo del Guenda (17∘40.26′S, 63∘27.45′W,
370 m a.s.l.), San Sebastian near Concepción, (16∘21.58′S,
62∘00.00′W, 516 m a.s.l.), and Estación FCBC Alta Vista near
Concepción (16∘08.1′S, 61∘56.1′W, 425 m a.s.l.) by D. M.
Windsor and L. Sekerka.
Flinte et al. (2015) published on MC in O. trichroa from Rio
de Janeiro, Brazil. Based on the photographs in their paper, the
species was misidentied and in fact belongs to O. pallidipennis,
reported earlier to have MC (Ohaus, 1900). Omaspides trichroa
is restricted to the Andean foothills in Bolivia and Peru, and so
far, no reliable records exist from Brazil.
Genus Eugenysa.Subsociality in Eugenysa unicolor
(Borowiec & Dabrowska) is reported based on adult females
guarding eggs near Río Hollín in Ecuador (same locality data
as in O. confusa) on eld observations by L. Sekerka.
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

6M. Leocádio et al.
Fig 3. New records of subsociality for Omaspides. A– C, Females guarding eggs, Omaspides (s. str.) collecta. Photograph by Dominik Hofer (A),
Omaspides (s. str.) confusa (B) and Omaspides (s. str.) trifasciata (C); D and E, Omaspides (s. str.) trichroa females guarding eggs (D) and larvae (E).
[Colour gure can be viewed at wileyonlinelibrary.com].
Phylogenetic analyses
The concatenated data matrix contained 90 taxa and 2533
aligned base pairs. Both Mesomphaliini and Eugenysini were
individually recovered as nonmonophyletic groups, but when
considered a single clade, they comprised a highly supported
monophyletic group (PP =1, SH-aLRT =99, UFBoot =100)
(Fig. 4). All gene trees recovered broadly similar phylogenetic
patterns, with heterogeneous values of nodal support depending
on the optimality criterion (Figs. S1–S6). Moreover, BI and
ML topologies were nearly identical, with minor topological
incongruities (Fig. 5).
BI of the concatenated molecular dataset recovered
Mesomphaliini, Goniocheniini and Ischyrosonychini in a
strongly supported polytomy (PP =0.98). ML recovered
the clade comprising Goniocheniini +Orexita varians
(Guérin) (Ischyrosonychini) as sister taxon to Mesomphali-
ini with low support (SH-aLRT =0, UFBoot =42), but the
Goniocheniini +Ischyrosonychini clade was placed as sister
taxon to Mesomphaliini with high support (SH-aLRT =82,
UFBoot =96). Mesomphaliini were recovered as paraphyletic
and Eugenysini polyphyletic, with both Eugenysini genera
recovered within Meosmphaliini (see discussion under Tax o-
nomic classication).
Five clades were recovered within the ingroup: (i) clade dis-
tinct and highly supported, composed by Hilarocassis Spaeth,
Acromis,Paraselenis and Omaspides (PP =1, SH-aLRT =98,
UFBoot =99); (ii) a clade composed of Ogdoecosta Spaeth
and Terp si s Spaeth in ML (PP =0.83, SH-aLRT =99,
UFBoot =96); (iii) highly supported clade composed by Chely-
morpha Chevrolat (PP =1, SH-aLRT =100, UFBoot =100);
(iv) well-supported clade composed by Stolas Billberg,
Agenysa,Xenicomorpha Spaeth and Zatrephina Spaeth
(PP =0.97, SH-aLRT =99, UFBoot =79); and nally, (v)
a clade composed by Mesomphalia Hope, Cyrtonota Chevrolat,
Botanochara Dejean, Eugenysa,Echoma and Stolas (PP =0.97,
SH-aLRT =100, UFBoot =80) (Fig. 6).
Clade 1 is composed of the non-MC genus Hilarocassis,
recovered as a sister group to the Mesomphaliini MC genera,
which were recovered as a monophyletic clade—Acromis,
Paraselenis and Omaspides—with strong support (PP =1,
SH-aLRT =94, UFBoot =98). The genus Acromis emerged
as a sister to the clade comprising Paraselenis and Omaspides,
with high support (PP =0.99, SH-aLRT =87, UFBoot =93).
The genera Acromis (PP =1, SH-aLRT =100, UFBoot =100),
Omaspides (PP =1, SH-aLRT =98, UFBoot =100) and
Paraselenis (PP =1, SH-aLRT =100, UFBoot =95) were
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

Evolution of subsociality in Cassidinae 7
Fig 4. Maximum likelihood tree for 90 species of tortoise beetles resulting from a partitioned analysis of concatenated DNA sequence data from
three gene fragments (28S, COI and CAD). Support values intervals are shown in colours for every node. Subsocial lineages are shown in red branches.
Bayesian inference support values (posterior probabilities) are shown in congruent nodes (see Fig. 5 for conicting nodes). [Colour gure can be viewed
at wileyonlinelibrary.com].
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

8M. Leocádio et al.
Fig 5. Mirrored topologies recovered using maximum likelihood (left) and Bayesian inference (right). Conicting nodes between both analyses and
respective nodal support values are indicated. [Colour gure can be viewed at wileyonlinelibrary.com].
strongly supported. Clade 1 was recovered as a sister group to
the remaining Mesomphaliini.
Clade 2 includes Ogdoecosta biannularis (Boheman) as the
sister taxon of the remaining group comprised by Ogdoecosta
catenulata (Boheman) and Terpsis quadrivittata (Champion)
(PP =0.99, SH-aLRT =86, UFBoot =98). Stoiba sp. is
recovered as sister taxa to clade 2 in BI and as sister taxa
to the group comprising clades 3, 4 and 5 in ML. Clade 3
is composed of the genus Chelymorpha, and it was highly
supported (PP =1, SH-aLRT =100, UFBoot =100). Clade
4 recovers Stolas as paraphyletic, with Agenysa boliviana
Spaeth emerging as sister to Stolas ephippium (Lichtenstein)
(PP =0.97, SH-aLRT =100, UFBoot =100) and Xenicomor-
pha scapularis (Boheman) +the Zatrephina clade (PP =1,
SH-aLRT =91, UFBoot =99) sister to Stolas brachi-
ata (Fabricius) +Stolas aff. isthmica (Champion) (PP =1,
SH-aLRT =100, UFBoot =100).
Clade 5 is the largest group recovered within Mesomphali-
ini, with Mesomphalia as monophyletic and sister to the
remaining genera (PP =0.92, SH-aLRT =80, UFBoot =80),
which is composed of two groups: (i) strongly supported
clade comprising polyphyletic Cyrtonota and Botanochara
as monophyletic (PP =1, SH-aLRT =100, UFBoot =99) and
sister to (ii) the strongly recovered monophyletic clade of
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

Evolution of subsociality in Cassidinae 9
Fig 6. Genera recovered monophyletic in our phylogenetic reconstruction. Type species, when available, are indicated by stars. Species illustrated (from
top to bottom): Hilarocassis bordoni,Acromis sparsa,Paraselenis dichroa,Chelymorpha marginata,Zatrephina sexlunata,Botanochara subnervosa,
Eugenysa coscaroni and Echoma clypeata. [Colour gure can be viewed at wileyonlinelibrary.com].
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

10 M. Leocádio et al.
Fig 7. Mirrored topologies of ancestral character state reconstruction of subsociality (left) and host plant choice (right). Character states are indicated
as coloured branches and circles at terminal taxa. [Colour gure can be viewed at wileyonlinelibrary.com].
Eugenysa (PP =1, SH-aLRT =100, UFBoot =100) +clade
including Cyrtonota,Agenysa,Stolas and Echoma (PP =0.93,
SH-aLRT =79, UFBoot =83).
Ancestral character state reconstruction
Our reconstruction strongly supports the hypothesis that MC
has two separate origins within Cassidinae (Fig. 7). One ori-
gin occurs in the ancestor of the group comprised by Acromis,
Paraselenis and Omaspides (clade 1). All species observed
in these three genera exhibit MC. The second origin, within
clade 5, is represented by the genus Eugenysa where three
closely allied species, E. coscaroni Viana, E. columbiana (Bohe-
man) and E. unicolor, have been reported to exhibit MC. The
genus Agenysa contains a single questionable report of MC (A.
peruviana Spaeth; Chaboo et al., 2014). The two species in our
sample, A. boliviana and A. connectens Baly, were recovered
outside the subsocial clades, and their reproductive behaviour
remains unreported. No reversals from MC to solitary behaviour
were recorded.
Regarding host plant reconstruction, MC lineages are not asso-
ciated with particular host plant preferences. Ingroup taxa feed
exclusively on Convolvulaceae or Asteraceae species, and our
reconstruction recovers two well-dened groups, one for each
host plant preference (Fig. 7). The rst group feeds on Con-
volvulaceae and comprises clades 1, 2, 3 and 4 and two lineages
within clade 5, Botanochara and Cyrtonota (part). Asteraceae
feeders mostly comprised lineages within clade 5. As previously
suggested, the two MC lineages are associated with different
host plant preferences (Windsor & Choe, 1994). The subsocial
clade comprising Acromis,Omaspides and Paraselenis feeds
on Convolvulaceae, whereas the subsocial clade composed of
Eugenysa is recovered among other Asteraceae feeders.
Discussion
Taxonomic classication
Mesomphaliini were recovered as paraphyletic, whereas
Eugenysini were recovered as polyphyletic, contradicting cur-
rent classication as two distinct tribes. A monophyletic clade
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

Evolution of subsociality in Cassidinae 11
Fig 8. Mesomphaliini and Eugenysini tarsomeres morphology. A, Eugenysa sp. (Eugenysini); B, Agenysa parellina (Boheman) (Eugenysini); C,
Echoma clypeata (Mesomphaliini); D, Stolas antiqua (Sahlberg) (Mesomphaliini). [Colour gure can be viewed at wileyonlinelibrary.com].
comprising both taxa, however, was recovered with high sup-
port. Eugenysini were originally created to contain three genera,
Eugenysa,Agenysa and Miocalaspis Weise, which were previ-
ously allocated in Mesomphaliini. Since Hincks’ (1952) review,
Eugenysini are distinguished from Mesomphaliini by the “claw
segment strongly expanded towards apex and obscuring base or
whole of claws”. Viana (1968) revised Eugenysini, stating that
Eugenysa has strongly expanded tarsomeres (Fig. 8A), whereas
in Agenysa and Miocalaspis, the expansion is not as large
(Fig. 8B). Later, Borowiec (1995) claimed the tarsal character
proposed by Hincks (1952) was not exclusive to Eugenysini as
some Stolas species also have explanate last tarsomere. There-
fore, Borowiec synonymized Eugenysini with Mesomphaliini
but later treated both as separate tribes (Borowiec, 1999).
We studied tarsal morphology among Mesomphaliini and
Eugenysini and concluded that the morphology of tarsomere
V in Eugenysini is rather unique among tortoise beetles as
the widened part is formed by expanded ‘plate’; thus, the
tarsomere has anterior and lateral laminae and looks dorsally
at. The lateral laminae are very obvious in Eugenysa and
much less in Miocalaspis and Agenysa, but the anterior lamina
is obvious in all, and as a result, the anterior margin of the
tarsomere V is straight, and claws are not visible in dorsal view.
Based on a comparative study of ca. 85% of Mesomphaliini
species representing all genera, the width of the tarsomere V
is variable; however, the structure is different. Most species
have the tarsomere V rather slim, but, in particular, the large
Stolas species feeding on Asteraceae have it distinctly wider.
However, in all cases, the anterior margin is convex and more
or less projecting forwards only in the middle; thus, the claws
are clearly visible in dorsal view; the lateral laminae are not
present, and thus, the tarsomere is regularly convex. In addition,
‘Eugenysini’ seem to have tarsal claws directing more inwards
than true Mesomphaliini, again more obvious in Eugenysa
than in the other two genera. However, tarsal characters likely
evolved independently and are adaptive according to species’
environment. For example, multiple tarsal states are known to
occur within the genus Charidotella Weise (Cassidini) (Sek-
erka & Borowiec, 2015) or even a single species such as in
Erepsocassis rubella (Boheman) (Riley, 1982). Accordingly,
we assume that the extremely widened tarsomere V and more
inwards-directing tarsal claws in Eugenysa might be adaptive
characters associated with MC.
Thus far, only one phylogenetic reconstruction of Cassidinae
tribes includes more than one species of Eugenysini in the anal-
ysis. This reconstruction (Chaboo, 2007), based on morphologi-
cal characters from adults and immatures, recovered Eugenysini
monophyletic and, diverging from Mesomphaliini (part), sup-
ported by two synapomorphies: (i) the apical expansion of tar-
somere V – as proposed by Hincks (1952) – and (ii) a ‘very
long, tightly coiled ejaculatory duct’. Although the tarsomere
V expansion has always been treated as a binary character
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

12 M. Leocádio et al.
Tabl e 1. Relationships among Mesomphaliini, Eugenysini and other Cassidinae tribes recovered in previous works and the present study.
Work Data Tribes included Mesomphaliini position Eugenysini position
Cassidinae phylogeny
(Borowiec, 1995)
Morphology (adults)
Biology
Mesomphaliini
(=Eugenysini), Cassidini,
Dorynotini,
Ischyrosonychini,
Goniocheniini, Omocerini
Sister to Dorynotini +
Cassidini +
Ischyrosonychini
–
Cassidinae phylogeny
(Hsiao &
Windsor, 1999)
Molecular (12S) Mesomphaliini, Eugenysini,
Dorynotini, Cassidini,
Ischyrosonychini,
Goniocheniini, Omocerini
Nonmonophyletic,
closely related to
Ischyrosonychini,
Goniocheniini and
Cassidini (part)
Recovered within
Mesomphaliini (one
species sampled)
Cassidinae phylogeny
(Chaboo, 2007)
Morphology (adults and
immature stages)
Biology
Mesomphaliini, Eugenysini,
Dorynotini, Cassidini,
Ischyrosonychini,
Omocerini
Nonmonophyletic,
Cassidinae s. str.
recovered in a
polytomy
Diverging from
Mesomphaliini (part)
Ischyrosonychini
phylogeny (Shin, 2015)
Morphology (adults) Mesomphaliini, Eugenysini,
Dorynotini, Cassidini,
Ischyrosonychini,
Omocerini
Sister to Goniocheniini Recovered within
Mesomphaliini (one
species sampled)
Cassidini phylogeny
(López-Pérez
et al., 2017)
Morphology (adults and
larvae)
Mesomphaliini, Dorynotini,
Cassidini,
Ischyrosonychini
Sister to Ischyrosonychini
+Cassidini
–
Dorynotini phylogeny
(Simões et al., 2018)
Morphology (adults)
Molecular (COI, CAD, 28S)
Mesomphaliini, Dorynotini,
Cassidini,
Ischyrosonychini
Sister to Ischyrosonychini –
This paper Molecular (COI, CAD, 28S) Mesomphaliini, Eugenysini,
Dorynotini,
Ischyrosonychini,
Goniocheniini, Omocerini
Sister to Goniocheniini +
Ischyrosonychini (part)
Recovered within
Mesomphaliini
(expanded vs not expanded) and used to differentiate both tribes,
it seems more appropriate to treat it as a multiple state char-
acter. Intermediary states are observed in both Mesomphaliini
and Eugenysini, with tarsomere V showing different degrees
of expansion across these units (Fig. 8). Eugenysa presents the
most expanded state, entirely covering tarsal claws, and uniquely
expanded tarsomeres II and III (Fig. 8D), which are synapomor-
phic for the genus.
Sekerka (2016) stated that phylogenetic analysis of Cassidinae
based on adult morphology is dubious as most taxa are sustained
by a combination of characters that are, in part, homoplastic.
This view has been supported in recent phylogenies, which
show that classic morphological characters formerly regarded
as synapomorphic have emerged multiple times (López-Pérez
et al., 2017; Simões et al., 2018). Thus, based on the recovered
results and the lack of morphological synapomorphy to sustain
the current classication, we re-establish the synonymy of
Eugenysini with the tribe Mesomphaliini.
Mesomphaliini and their position within Cassidinae
Some studies have attempted to elucidate phylogenetic rela-
tionships among Cassidinae tribes (Table 1); however, little con-
sensus has emerged. Borowiec (1995) proposed a phylogeny for
Cassidinae using tribes as terminal taxa based on 19 characters
(17 morphological and 2 ecological). He recovered Mesom-
phaliini (including Eugenysini as synonym) as sister to Doryno-
tini +Cassidini +Ischyrosonychini (three clades placed in a
polytomy). As the author used tribes as terminal taxa, he was not
able to test their respective monophyly. Hsiao & Windsor (1999)
were the rst to use molecular data when attempting to eluci-
date tribal relationships in Cassidinae. Based on a phylogeny
reconstructed using a single mitochondrial gene fragment, 12S
mtDNA, a non-monophyletic Mesomphaliini was placed with
Cassidini (part), Ischyrosonychini and Goniocheniini recovered
within Mesomphaliini. Only one species of Eugenysini was
sampled, and it was recovered within Mesomphaliini. Cha-
boo (2007) proposed a phylogeny for Cassidinae based on
210 characters (adults and immatures morphology and biol-
ogy) for 98 species. This paper contains numerous question-
able identications of taxa, errors in character coding etc.
(´
Swi
¸
etoja´
nska, 2009; Sekerka, 2017) and resulted in most Cas-
sidinae s. str. grouped within a large polytomy, confound-
ing relationships among tribes. Mesomphaliini were recovered
within the polytomy comprising at least four distinct clades.
Shin (2015) tested the monophyly of Ischyrosonychini with a
cladistic analysis based on 155 morphological characters (adults
only). Mesomphaliini +Eugenysini were recovered as mono-
phyletic and sister to Goniocheniini. López-Pérez et al. (2017)
conducted a phylogenetic analysis to test the monophyly of Cas-
sidini using 96 morphological characters (95 from adults and 1
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

Evolution of subsociality in Cassidinae 13
from immature stages) and 93 species. The two Stolas species
representing Mesomphaliini were recovered as sister taxa to
the clade Ischyrosonychini +Cassidini +Aspidimorphini. Both
papers used a very similar character matrix to Chaboo (2007),
excluding characters that were not relevant to each sampling. A
total-evidence phylogenetic reconstruction of Dorynotini sam-
pled four species of Mesomphaliini recovered sisters to Ischy-
rosonychini (Simões et al., 2018).
Our study recovered Goniocheniini +Ischyrosonychini (part)
as sister taxa to Mesomphaliini with low support; however,
Goniocheniini +Ischyrosonychini was strongly supported as
sister to Mesomphaliini. All studies to date support a close
relationship among Mesomphaliini, Goniocheniini and Ischy-
rosonychini, but no consensus emerged regarding which tribe is
more closely related to Mesomphaliini. However, the majority
of previous studies did not sample the three tribes concomi-
tantly, complicating an adequate comparison among studies.
Goniocheniini is remarkably neglected, having been sampled in
only one published study (Hsiao & Windsor, 1999). Although
the present study did not reach a consensus, evidence points
towards Goniocheniini and Ischyrosonychini as closely related
to Mesomphaliini. Studies with greater sampling are still needed
to elucidate the positioning of Cassidinae tribes.
The origins of subsociality in Cassidinae
Although subsociality in insects and arachnids has
received much attention (Choe & Crespi, 1997; Costa, 2006;
Trumbo, 2012), there is still little information regarding the
evolution of this behaviour. The multiple independent origins of
subsociality remain poorly explored, with some evidence sup-
porting its evolutionary stability once evolved (Lin et al., 2004),
whereas others record multiple origins subject to reversals
(Crespi et al., 1998; Jordal et al., 2011; Yip & Rayor, 2014;
Ruch et al., 2015).
Within Cassidinae, two separate origins of subsociality are
clearly indicated – the rst in the ancestor of the Mesomphaliini
group comprised by Acromis,Paraselenis and Omaspides within
clade 1 and the second in the genus Eugenysa within clade
5. The possibility of an additional origin of MC behaviour
in Mesomphaliini cannot be entirely ruled out. The species
Agenysa peruviana is the sole species of the genus in which
parental care has been recorded (Chaboo et al., 2014). However,
this record requires further verication as few biological or
photographic details accompanied the report. The present paper
includes two Agenysa species, whose reproductive behaviours
have not been reported. Both these Agenysa species were
recovered in distant portions of the Mesomphaliini tree, neither
associated with subsocial clades.
The review of subsociality in Cassidinae by Chaboo
et al. (2014) is a morphology-based phylogenetic reconstruc-
tion testing the monophyly of the subfamily (Chaboo, 2007).
Two competing hypotheses were proposed to account for MC in
Cassidinae: (i) a single origin, with subsequent loss in Echoma,
or (ii) two independent origins, one in Omaspides and another
in the group comprised by Acromis +Paraselenis +Eugenysini.
The authors regarded the second hypothesis as equivocal due to
the polytomy present in their phylogeny.
Our results corroborate neither hypotheses as Echoma is
recovered distant from MC species, and Acromis,Paraselenis
and Omaspides (clade 1) are recovered as a strongly supported
monophyletic group. The added occurrence of MC in the
Eugenysa clade, distant from the clade 1, argues in favour of
two independent origins of MC in Mesomphaliini.
Windsor & Choe (1994) proposed that the most likely scenario
for the origin of subsociality in tortoise beetles would be
independent origins associated with clades occupying different
host plant families, the Asteraceae and Convolvulaceae. The
present study supports this hypothesis as clade 5 containing
Eugenysa spp. is known only from vining Asteraceae (Mikania
spp.), whereas species in the large clade 1 containing Acromis,
Paraselenis and Omaspides are known only to feed on vining
Convolvulaceae (Fig. 6). Although host plant choice denitely
plays an important role in tortoise beetle life history, it may
not be the only factor predisposing the origins of parental care.
Subsocial and nonsubsocial tortoise beetles species are known
to share the same host plant and to coexist on a single individual
and are therefore presumed to suffer from roughly the same
selection pressures (Chaboo et al., 2014).
MC in Cassidinae is often portrayed as a component part
of a defence arsenal developed in the face of a transition in
feeding habits – from leaf-mining larvae to exophagous lar-
vae (Vencl et al., 2011); however, the details of this transition
remain largely speculative. MC would be the most recent and
complex defensive behaviour acquired in tortoise beetles, pre-
ceded by exuvio-faecal shield construction and later by gre-
garious behaviour (Vencl & Srygley, 2013). Furthermore, larval
gregarious behaviour also occurs in nonsubsocial Mesomphali-
ini (Vencl et al., 2011; Simões & Monné, 2014; Leocadio &
Mermudes, 2019). Aggregated oviposition and larval gregarious
behaviour are most likely preconditions to the emergence of sub-
sociality as larval guarding requires offspring to be aggregated.
Larval gregarious behaviour is also commonly reported within
tribes Cassidini, Goniocheniini and Ischyrosonychini (Wind-
sor pers. obs.; Fiebrig, 1910; Flinte et al., 2008; Vencl & Sry-
gley, 2013), clearly predating the evolution of MC according
to our phylogeny. Our tree lacks resolution and strong sup-
port regarding the relationship between Ischyrosonychini and
Goniocheniini, indicating a larger set of taxa will be required
to determine with more precision the origin (or origins) of
larval gregarious behaviour. The construction of exuvio-faecal
shields likely predates our phylogeny as it is found in several
exophagous monocotyledonous feeders, such as those in the
tribes Hemisphaerotini and Spilophorini (Sekerka et al., 2014;
Albertoni & Leocadio, 2018). The shield is a complex struc-
ture differing considerably among tortoise beetles species (e.g.
with or without a faecal matrix, whether a liquid blob, a solid
shield or a delicate fan). Whether and how exuvio-faecal shields
function in chemical and structural defence of larval stages is
still in debate (Olmstead & Denno, 1993; Nogueira-de-Sá &
Trigo, 2005; Bottcher et al., 2009).
Within Chrysomelidae, multiple independent origins of sub-
sociality have been recovered in previous studies (Windsor
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

14 M. Leocádio et al.
et al., 2013; Chaboo et al., 2014). MC in Chrysomelinae, as in
Cassidinae, is provided solely by the female, which oviposit
on the host plants (Chaboo et al., 2014). However, maternal
guarding in Chrysomelinae lasts only through the nal larval
instar, ceasing as larvae descend into the leaf litter and soil
environment beneath the host plant to molt into pupa (Wind-
sor et al., 2013). Maternal guarding in Cassidinae, in contrast,
lasts through all larval instars, pupation and until adult offspring
begin to feed, mate or wander apart (Windsor, 1987; Windsor &
Choe, 1994). Moreover, subsociality in Chrysomelinae has been
reported in evolutionary distinct clades from Central and South
America, Europe, Asia and Australia (Reid et al., 2009; Wind-
sor et al., 2013), whereas subsocial Cassidinae are known only
from the Neotropics (Chaboo et al., 2014).
MC in Mesomphaliini presents an evolutionary pattern similar
to that reported in other subsocial insects (Lin et al., 2004; Jordal
et al., 2011; Windsor et al., 2013; Tsai et al., 2015), with multi-
ple origins and no reversals. It is striking that the two origins
we substantiate occur within a single Neotropical tribe, whereas
the subfamily is cosmopolitan and contains over 6000 species
(Borowiec & ´
Swi
¸
etoja´
nska, 2014). The pattern recovered thus
far suggests that MC behaviour has originated multiple times
and offers no evidence for subsequent loss as has been previ-
ously suggested.
The temporal framework for the two origins of subsocial-
ity we address remains to be investigated. Next steps regard-
ing the study of the subsociality evolution in tortoise beetles
should place Mesomphaliini within a time framework with a
dated phylogeny. Currently, the oldest available Cassidinae fos-
sil species represent hispoid lineages and are from the middle
Eocene (Chaboo & Engel, 2009), whereas the subfamily is cer-
tainly older. There is no reliable fossil record of true tortoise
beetles, Cassidinae s. str., and the known fossils are far related
to Mesomphaliini. Furthermore, secondary dating would not
provide a reasonable timing estimate as published dated phy-
logenies only include Cassidinae taxa of distantly related tribes
(Zhang et al., 2018; McKenna et al., 2019), which can hardly be
applied in our phylogeny. The discovery of new tortoise beetles
fossils and the estimate of larger dated phylogenies will allow
further analysis on this topic.
Conclusions
Although monophyly of Cassidinae is well established, some
historical relationships within the subfamily lack consensus or
are poorly supported. This paper represents the rst attempt
to test the monophyly of the tortoise beetle tribe Mesomphali-
ini. Our results indicate a monophyletic Mesomphaliini, after
synonymy with Eugenysini. We recovered several genera (e.g.
Acromis,Paraselenis,Omaspides) as monophyletic with strong
support. However, some of the most diverse genera in Mesom-
phaliini (e.g. Stolas and Cyrtonota) were not recovered as mono-
phyletic taxa, pointing to areas in need of further taxonomic
study. Furthermore, our results substantiate that subsociality
originated at least twice in Cassidinae, with no evidence of
reversals. This result corroborates past studies showing multiple
origins with a few or zero losses, indicating strong xation of
MC. Although the emergence of subsociality in Cassidinae is
unlikely to be causally linked to a single factor, association with
widespread, fast-growing herbaceous vines may have been a sig-
nicant factor selecting for increased protection of vulnerable
immature stages. With a well-supported phylogeny in hand, tar-
geted studies can be conducted to unravel what is behind the ori-
gin of such a complex and interesting behaviour. We encourage
future studies of tortoise beetles to ally molecular, morphology
(from both adults and immatures) and natural history to improve
the resolution and nodal support throughout the group.
Supporting Information
Additional supporting information may be found online in
the Supporting Information section at the end of the article.
Figure S1. Bayesian inference tree (28S).
Figure S2. Maximum likelihood tree (28S). Support val-
ues =SH-aLRT/UFBoot.
Figure S3. Bayesian inference tree (CAD).
Figure S4. Maximum likelihood tree (CAD). Support val-
ues =SH-aLRT/UFBoot.
Figure S5. Bayesian inference tree (COI).
Figure S6. Maximum likelihood tree (COI). Support val-
ues =SH-aLRT/UFBoot.
Tabl e S1. List of subsocial species recorded in Cassidinae.
Tabl e S2. Species names, localities, voucher numbers and
GenBank accession codes of the specimens used in this
study.
Tabl e S3. List of primers used to amplify the gene fragments
used in this study.
Tabl e S4. Polymerase chain reaction (PCR) cycling condi-
tions used to amplify the selected gene segments.
Tabl e S5. Best models selected with PartitionFinder2 for
each partition.
Acknowledgements
We thank Fernando Frieiro-Costa, the staff of the Reserva
Ecológica do Guapiaçu and especially Nicholas Locke for
eldwork assistance; Renan Oliver and Dominik Hofer for
new subsociality records; Mabelle Chong (STRI) for sequence
preparation and Márcio Felix (CEIOC-FIOCRUZ) for access
to specimens and the use of an automontage stereomicroscope;
Fernanda Petrongari for Convolvulaceae identication; and
Lic. Julieta Ledezma (MHNNKM) for export permits. We
thank the two anonymous reviewers for helpful comments
on the manuscript. ML received nancial support from
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

Evolution of subsociality in Cassidinae 15
CAPES – Coordenação de Aperfeiçoamento de Pessoal
de Nível Superior (fellowship no. 88887.162691/2018-00
CAPES/PROTAX-II). The Smithsonian Tropical Research
Institute provided a short-term fellowship to M. Leocádio and
long-term support of sequencing expenses to D. M. Windsor. L.
Sekerka received nancial support from the Ministry of Culture
of the Czech Republic (DKRVO 2019– 2023/5.I.b, National
Museum, 00023272). J. R. M. Mermudes received nancial
support by Conselho Nacional de Desenvolvimento Cientíco e
Tecnológico (PROTAX CNPQ/CAPES 440479/2015-0, CNPq
06105/2016-0). The authors declare no conict of interest.
Data availability statement
The data that supports the ndings of this study are available in
the supplementary material of this article
References
Albertoni, F.F. & Leocadio, M. (2018) The bromeliad leaf-scraper
tortoise beetle Spaethiella intricata (Boheman, 1850) from Brazil
(Coleoptera, Chrysomelidae, Cassidinae), description of immatures
and biology. Zootaxa,4531, 395– 418.
Alonso-Alvarez, C. & Velando, A. (2012) Benets and costs of parental
care. The Evolution of Parental Care (ed. by N.J. Royle, P.T. Smiseth
and M. Kölliker), pp. 1– 17. Oxford University Press, Oxford.
Anisimova, M., Gil, M., Dufayard, J.F., Dessimoz, C. & Gascuel,
O. (2011) Survey of branch support methods demonstrates accu-
racy, power, and robustness of fast likelihood-based approximation
schemes. Systematic Biology,60, 685– 699.
Borowiec, L. (1995) Tribal classication of the cassidoid Hispinae
(Coleoptera: Chrysomelidae). Biology, Phylogeny, and Classication
of Coleoptera: Papers Celebrating the 80th Birthday of Roy A.
Crowson (ed. by J. Pakaluk, S.A. ´
Slipi´
nski and R.A. Crowson), pp.
541– 558. Muzeum i Instytut Zoologii PAN, Warszawa.
Borowiec, L. (1999) A World Catalogue of the Cassidinae (Coleoptera:
Chrysomelidae). Biologica Silesiae, Wrocław.
Borowiec, L. & ´
Swi
¸
etoja´
nska, J. (2014) Cassidinae Gyllenhal, 1813.
Handbook of Zoology, Arthropoda, Insecta, Coleoptera, Beetles,Vol.
3(ed. by R.A.B. Leschen and R.G. Beutel), pp. 198–217. Walter de
Gruyter, Berlin & New York.
Borowiec, L. & ´
Swi
¸
etoja´
nska, J. (2020) World catalog of Cassidi-
nae. Wrocław, Poland. URL http://www.cassidae.uni.wroc.pl/katalog
%20internetowy/index.htm [accessed on 5 January 2019].
Bottcher, A., Zolin, J.P., Nogueira-de-Sá, F. & Trigo, J.R. (2009) Faecal
shield chemical defence is not important in larvae of the tortoise
beetle Chelymorpha reimoseri (Chrysomelidae: Cassidinae: Stolaini).
Chemoecology,19, 63– 66.
Chaboo, C.S. (2007) Biology and phylogeny of the Cassidinae Gyllenhal
sensu lato (tortoise and leaf-mining beetles) (Coleoptera: Chrysomel-
idae). Bulletin of the American Museum of Natural History,305,
1– 250.
Chaboo, C.S. & Engel, M.S. (2009) Eocene tortoise beetles from the
green river formation in Colorado, USA (Coleoptera: Chrysomelidae:
Cassidinae). Systematic Entomology,34, 202– 209.
Chaboo, C.S., Frieiro-Costa, F.A., Gómez-Zurita, J. & Westerduijn,
R. (2014) Origins and diversication of subsociality in leaf beetles
(Coleoptera: Chrysomelidae: Cassidinae: Chrysomelinae). Journal of
Natural History,48, 2325– 2367.
Chapuis, F. (1875) Histoire naturelle des insectes. Genera des coléop-
tères ou exposé méthodique et critique de tous les genres proposés
jusqu’ici dans cet ordre d’insectes. Tome onzième. Famille des phy-
tophages. Librairie Encyclopédique de Roret, Paris.
Choe, J.C. & Crespi, B.J. (1997) The Evolution of Social Behaviour in
Insects and Arachnids. Cambridge University Press, Cambridge.
Clutton-Brock, T.H. (1991) The Evolution of Parental Care. Princeton
University Press, Princeton.
Costa, J.T. (2006) The Other Insect Societies. Harvard University Press,
Cambridge.
Crespi, B.J., Carmean, D.A., Mound, L.A., Worobey, M. & Morris, D.
(1998) Phylogenetics of social behavior in Australian gall-forming
thrips: evidence from mitochondrial DNA sequence, adult morphol-
ogy and behavior, and gall morphology. Molecular Phylogenetics and
Evolution,9, 163– 180.
Cuignet, M., Hance, T. & Windsor, D.M. (2007) Phylogenetic rela-
tionships of egg parasitoids (Hymenoptera: Eulophidae) and corre-
lated life history characteristics of their Neotropical Cassidinae hosts
(Coleoptera, Chrysomelidae). Molecular Phylogenetics and Evolu-
tion,42, 573– 584.
Elgar, M.A. (2015) Integrating insights across diverse taxa: challenges
for understanding social evolution. Frontiers in Ecology and Evolu-
tion,3, 124.
Erixon, P., Svennblad, B., Britton, T. & Oxelman, B. (2003) Reliability
of Bayesian posterior probabilities and bootstrap frequencies in
phylogenetics. Systematic Biology,52, 665– 673.
Fiebrig, K. (1910) Cassiden und Cryptocephaliden Paraguays.
Ihre Entwicklungsstadien und Schutzvorrichtungen. Zoologische
Jahrbücher,12, 161– 264.
Flinte, V., Macedo, M.V. & Monteiro, R.F. (2008) Tortoise beetles
(Chrysomelidae: Cassidinae) of a tropical rain forest in Rio de
Janeiro, Brazil. Research on Chrysomelidae,Vol.1(ed. by P. Jolivet,
J. Santiago-Blay and M. Schmitt), pp. 194– 209 +pl. 17–18. Brill,
Leiden, Boston.
Flinte, V., Hentz, E., Morgado, B.M. et al. (2015) Biology and phenol-
ogy of three leaf beetle species (Chrysomelidae) in a montane forest
in Southeast Brazil. Research on Chrysomelidae,Vol.5,547 (ed. by
P. Jolivet, J. Santiago-Blay and M. Schmitt), pp. 119–132. ZooKeys,
Pensoft, Soa.
Gómez-Zurita, J., Hunt, T. & Vogler, A.P. (2008) Multilocus ribosomal
RNA phylogeny of 24 the leaf beetles (Chrysomelidae). Cladistics,
24, 34– 50.
Hincks, W.D. (1952) The genera of the Cassidinae (Coleoptera:
Chrysomelidae). Transactions of the Royal Entomological Society of
London,103, 327– 358.
Hoang, D.T., Chernomor, O., von Haeseler, A., Minh, B.Q. & Vinh, L.S.
(2017) UFBoot2: improving the ultrafast bootstrap approximation.
Molecular Biology and Evolution,35, 518– 522.
Hope, F.W. (1840) The coleopterist’s manual,part the third,containing
various families,genera,and species,of beetles,recorded by Linneus
and Fabricius. Also,descriptions of newly discovered and unpub-
lished insects. J. C. Bridgewater and Bowdery & Kerby, London.
Hsiao, T.H. & Windsor, D.M. (1999) Historical and biological rela-
tionships among Hispinae inferred from 12S mtDNA sequence data.
Advances in Chrysomelidae Biology,Vol.1(ed. by M.L. Cox), pp.
39– 50. Backhuys, Leiden.
Hurvich, C.M. & Tsai, C.L. (1993) A corrected Akaike information
criterion for vector autoregressive model selection. Journal of Time
Series Analysis,4, 271– 279.
International Commission of Zoological Nomenclature [ICZN] (1999)
International code of zoological nomenclature [the Code]. Fourth
edition. The International Trust for Zoological Nomenclature, c/o
Natural History Museum, London. i–xxix, +306 pp. URL http://
www.iczn.org/iczn/index.jsp.
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

16 M. Leocádio et al.
Jordal, B.H., Sequeira, A.S. & Cognato, A.I. (2011) The age and
phylogeny of wood boring weevils and the origin of subsociality.
Molecular Phylogenetics and Evolution,59, 708– 724.
Katoh, K. & Toh, H. (2008) Recent developments in the MAFFT
multiple sequence alignment program. Briengs in Bioinformatics,9,
286– 298.
Katoh, K., Rozewicki, J. & Yamada, K.D. (2017) MAFFT online
service: multiple sequence alignment, interactive sequence choice and
visualization. Briengs in Bioinformatics,20, 1160– 1166.
Kumar, S., Stecher, G. & Tamura, K. (2016) MEGA7: molecular evo-
lutionary genetics analysis version 7.0 for bigger datasets. Molecular
Biology and Evolution,33, 1870– 1874.
Lanfear, R., Frandsen, P.B., Wright, A.M., Senfeld, T. & Calcott,
B. (2016) PartitionFinder 2: new methods for selecting partitioned
models of evolution for molecular and morphological phylogenetic
analyses. Molecular Biology and Evolution,34, 772– 773.
Leocadio, M. & Mermudes, J.R.M. (2019) Description of immatures of
Stolas aenea (Olivier, 1790) and Stolas nudicollis (Boheman, 1850)
(Coleoptera: Chrysomelidae: Cassidinae: Mesomphaliini). Zootaxa,
4545, 61– 76.
Lin, C.P., Danforth, B.N. & Wood, T.K. (2004) Molecular phylogenetics
and evolution of maternal care in membracine treehoppers. Systematic
Biology,53, 400– 421.
López-Pérez, S., Zaragoza-Caballero, S., Ochoterena, H. & Morrone,
J.J. (2017) A phylogenetic study of the worldwide tribe Cassidini
Gyllenhal, 1813 (Coleoptera: Chrysomelidae: Cassidinae) based on
morphological data. Systematic Entomology,1813, 1– 15.
Macedo, M.V., Flinte, V., Abejanella, A. & Chaboo, C.S. (2015)
Three new reports of subsocial tortoise beetles from South America
(Chrysomelidae: Cassidinae). Annals of the Entomological Society of
America,108, 1088– 1092.
Machado, G. & Trumbo, S.T. (2018) Parental care. Insect Behav-
ior: From Mechanisms to Ecological and Evolutionary Conse-
quences (ed. by A. Córdoba-Aguilar, D. González-Tokman and
I. González-Santoyo), pp. 203– 218. Oxford University Press, Oxford.
Maddison, W.P. & Maddison, D. R. (2015) Mesquite: a modular system
for evolutionary analysis. Version 3.04. URL http://mesquiteproject
.org.
McKenna, D.D., Shin, S., Ahrens, D. et al. (2019) The evolution
and genomic basis of beetle diversity. Proceedings of the National
Academy of Sciences,116, 24729– 24737.
Michener, C.D. (1969) Comparative social behavior of bees. Annual
Review of Entomology,14, 299– 342.
Miller, M.A., Pfeiffer, W. & Schwartz, T. (2010) Creating the CIPRES
science gateway for inference of large phylogenetic trees. Gateway
Computing Environments Workshop, pp. 1–8. New Orleans.
Minh, B.Q., Nguyen, M.A.T. & Von Haeseler, A. (2013) Ultrafast
approximation for phylogenetic bootstrap. Molecular Biology and
Evolution,30, 1188– 1195.
Minh, B.Q., Trinopolous, J., Schrempf, D. & Schmidt, H.A. (2017)
IQ-TREE version 1.6.0: Tutorials and Manual Phylogenomic Soft-
ware by Maximum Likelihood. URL www.iqtree.org/doc/iqtree-doc
.pdf.
Nguyen, L.T., Schmidt, H.A., Von Haeseler, A. & Minh, B.Q. (2015)
IQ-TREE: a fast and effective stochastic algorithm for estimating
maximum-likelihood phylogenies. Molecular Biology and Evolution,
32, 268– 274.
Nogueira-de-Sá, F. & Trigo, J.R. (2005) Faecal shield of the tortoise
beetle Plagiometriona aff. avescens (Chrysomelidae: Cassidinae) as
chemically mediated defence against predators. Journal of Tropical
Ecology,21, 189– 194.
O’Toole, C. & Preston-Mafham, K. (1985) Insects in Camera. A
Photographic Essay on Behaviour. Oxford University Press, Oxford.
Ohaus, F. (1900) Bericht über eine entomologische Reise nach Central-
brasilien. Stettiner Entomologische Zeitung,61, 193– 274.
Olmstead, K.L. & Denno, R.F. (1993) Effectiveness of tortoise bee-
tle larval shields against different predator species. Ecology,74,
1394– 1405.
Rankin, D.J., Bargum, K. & Kokko, H. (2007) The tragedy of the
commons in evolutionary biology. Trends in Ecology & Evolution,
22, 643– 651.
Reid, C.A.M., Beatson, M. & Hasenpusch, J. (2009) The morphology
and biology of Pterodunga mirabile Daccordi, an unusual subsocial
Chrysomeline (Coleoptera: Chrysomelidae). Journal of Natural His-
tory,43, 373– 398.
Riley, E.G. (1982) Erepsocassis Spaeth, 1936: a valid genus
(Coleoptera: Chrysomelidae: Cassidinae). Journal of the Kansas
Entomological Society,55, 651– 657.
Ronquist, F., Teslenko, M., Van Der Mark, P. et al. (2012) Mrbayes 3.2:
efcient Bayesian phylogenetic inference and model choice across a
large model space. Systematic Biology,61, 539– 542.
Royle, N.J., Smiseth, P.T. & Kölliker, M. (2012) The Evolution of
Parental Care. Oxford University Press, Oxford.
Ruch, J., Riehl, T., May-Collado, L.J. & Agnarsson, I. (2015) Multiple
origins of subsociality in crab spiders (Thomisidae). Molecular
Phylogenetics and Evolution,82, 330– 340.
Sekerka, L. (2016) Taxonomic and nomenclatural changesin Cassidinae
(Coleoptera: Chrysomelidae). Acta Entomologica Musei Nationalis
Prague,56, 275– 344.
Sekerka, L. (2017) Taxonomy and Ecology of Neotropical Cassidi-
nae (Coleoptera: Chrysomelidae). PhD Thesis, University of South
Bohemia, ˇ
Ceské Budˇ
ejovice.
Sekerka, L. & Borowiec, L. (2015) Subgenera of Charidotella Weise
with description of a new subgenus and species from Brazil
(Coleoptera, Chrysomelidae, Cassidinae, Cassidini). ZooKeys,506,
61– 64.
Sekerka, L., Windsor, D. & Dury, G. (2014) Cladispa Baly: revision,
biology and reassignment of the genus to the tribe Spilophorini
(Coleoptera: Chrysomelidae: Cassidinae). Systematic Entomology,
39, 518– 530.
Shin, C. (2015) Systematics and morphology of Ischyrosonychini,
with an investigation of color change in the Geiger tortoise beetle
(Coleoptera: Chrysomelidae). PhD Thesis, University of Kansas,
Lawrence.
Simões, M.V.P. & Monné, M.L. (2014) Description of immatures of
Mesomphalia gibbosa (Fabricius, 1781) and Mesomphalia turrita
(Illiger, 1801) (Coleoptera: Chrysomelidae: Cassidinae: Mesomphali-
ini). Zootaxa,3861, 466– 478.
Simões, M.V.P., Baca, S.M., Toussaint, E.F.A., Windsor, D.M. & Short,
A.E.Z. (2018) Solving a thorny situation: DNA and morphology illu-
minate the evolution of the leaf beetle tribe Dorynotini (Coleoptera,
Chrysomelidae, Cassidinae). Zoological Journal of the Linnean Soci-
ety,185, 1123– 1136.
Spaeth, F. (1942) Cassidinae (Col. Chrysom.). Beitrage zur Fauna
Perus:nach der Ausbeute der Hamburger Sudperu- Expedition 1936,
anderer Sammlungen, wie auch auf Grund von Literaturangaben,Vol.
2(ed. by E. Titschak), pp. 11– 43. Fischer, Jena.
Staines, C. L. (2015) Catalog of the Hispines of the World. Smithsonian
National Museum of Natural History. URL http://entomology.si.edu/
Collections_Coleoptera-Hispines.html [accessed 10 January 2020].
´
Swi
¸
etoja´
nska, J. (2009) The Immatures of Tortoise Beetles with Biblio-
graphic Catalogue of all Taxa (Coleoptera: Chrysomelidae: Cassidi-
nae),Vol.16. Biologica Silesiae, Wroclaw.
Thompson, J.D., Higgins, D.G. & Gibson, T.J. (1994) CLUSTALW:
improving the sensitivity of progressive weighting, position-specic
gap penalties and weight matrix choice. Nucleic Acids Research,22,
4673– 4680.
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434

Evolution of subsociality in Cassidinae 17
Trumbo, S.T. (2012) Patterns of parental care in invertebrates. The
Evolution of Parental Care (ed. by N.J. Royle, P.T. Smiseth and
M. Kölliker), pp. 81– 100. Oxford University Press, Oxford.
Tsai, J.F., Kudo, S.I. & Yoshizawa, K. (2015) Maternal care in Acan-
thosomatinae (Insecta: Heteroptera: Acanthosomatidae) – correlated
evolution with morphological change. BMC Evolutionary Biology,
15, 258.
Vaidya, G., Lohman, D.J. & Meier, R. (2011) SequenceMatrix: con-
catenation software for the fast assembly of multi-gene datasets with
character set and codon information. Cladistics,27, 171–180.
Vencl, F.V. & Srygley, R.B. (2013) Enemy targeting, trade-offs, and the
evolutionary assembly of a tortoise beetle defense arsenal. Evolution-
ary Ecology,27, 237– 252.
Vencl, F.V., Trillo, P.A. & Geeta, R. (2011) Functional interactions
among tortoise beetle larval defenses reveal trait suites and escalation.
Behavioral Ecology and Sociobiology,65, 227– 239.
Viana, M. (1968) Revision sistematica de las “Eugenysini” Spaeth
con nuevas especies y catalogo bibliograco de la tribus (Coleopt.,
Chrysomelidae, Cassidinae). Revista del museo argentino de ciencias
naturales “Bernardino Rivadavia”, 3, 1–106.
Wild, A.L. & Maddison, D.R. (2008) Evaluating nuclear protein-coding
genes for phylogenetic utility in beetles. Molecular Phylogenetics and
Evolution,48, 877– 891.
Wilson, E.O. (1971) The Insect Societies. Harvard University Press,
Cambridge.
Windsor, D.M. (1987) Natural history of a subsocial tortoise beetle,
Acromis sparsa Boheman (Chrysomelidae, Cassidinae) in Panama.
Psyche,94, 127– 150.
Windsor, D.M. & Choe, J.C. (1994) Origins of parental care in
Chrysomelid beetles. Novel Aspects of the Biology of Chrysomelidae
(ed. by P. Jolivet, M.L. Cox and E. Petitpierre), pp. 111– 117. Kluwer
Academic Publishers, Dordrecht.
Windsor, D.M., Dury, G.J., Frieiro-Costa, F.A., Lanckowsky, S. & Pas-
teels, J.M. (2013) Subsocial Neotropical Doryphorini (Chrysomeli-
dae, Chrysomelinae): new observations on behavior, host plants and
systematics. Research on Chrysomelidae,Vol.4,332 (ed. by P. Jolivet,
J. Santiago-Blay and M. Schmitt), pp. 71– 93. ZooKeys, Pensoft,
Soa.
Yip, E.C. & Rayor, L.S. (2014) Maternal care and subsocial behaviour
in spiders. Biological Reviews,89, 427– 449.
Zhang, S.Q., Che, L.H., Li, Y., Liang, D., Pang, H., ´
Slipi´
nski, A. &
Zhang, P. (2018) Evolutionary history of Coleoptera revealed by
extensive sampling of genes and species. Nature Communications,9,
205.
Accepted 24 April 2020
© 2020 The Royal Entomological Society, Systematic Entomology, doi: 10.1111/syen.12434