Early evolution of Anamorphidae (Coleoptera: Coccinelloidea): the oldest known anamorphid beetles from Upper Cretaceous amber of northern Myanmar and the first report of potential glandular pores in the family
Abstract
In order to place newly discovered fossil taxa (Palaeosymbius gen. nov. with P. groehni and P. mesozoicus spp. nov.) from the mid-Cretaceous amber from northern Myanmar, we investigated the relations of extant and extinct lineages of the coccinellid group of Coccinelloidea with emphasis on the family Anamorphidae. We assembled a taxonomic sampling of 34 taxa, including 15 genera and 19 species of Anamorphidae, the most comprehensive sampling of Anamorphidae at the generic level in a phylogenetic analysis. A morphological dataset of 47 characters was built as well as a molecular alignment of 7140 bp including fragments of eight genes (12S, 16S, 18S, 28S, COI, COII, H3 and CAD). Five anamorphid and one endomychid species were sequenced for the first time and added to the dataset. We performed parsimony-based analysis of the morphological dataset and Bayesian inference analysis of the combined matrix (morphological plus molecular data). Our results confirm that Palaeosymbius belongs to Anamorphidae and represents the oldest known member of this family so far. Among Anamorphidae, Symbiotes (with extant and known Eocene species) was recovered as the most probable closest relative of Palaeosymbius. Our morphological studies additionally revealed the presence of probable glandular openings in the anterolateral corners of the pronotal margins in Asymbius sp. and Anamorphus sp., representing the first report of secretory openings in the family Anamorphidae. Similar openings are found in other cucujiform beetles such as Cryptophagidae and Boganiidae with possible defensive purposes.
Introduction
The last decade has seen an increase in the discovery of fossil organisms that represent the oldest known members for many lineages in Coleoptera (e.g. Yamamoto et al., 2018; Liu et al., 2019; Li et al., 2022a, c; Telnov et al., 2023). This has given a new impulse to evolutionary studies into the temporal dimension of the evolution of the most diverse group of animals in the world (Toussaint et al., 2017; McKenna et al., 2019; Cai et al., 2022). The vast majority of these recent discoveries have been made utilizing specimens embedded in amber from northern Myanmar, from the Cenomanian age of the Cretaceous period, and dated to ca. 99 Myr (Shi et al., 2012). Among the vast fauna recovered from this deposit we can find taxa that show different levels of resemblance to known recent organisms that range from relative morphological stasis (Cai et al., 2019; Yamamoto, 2021; Li et al., 2024) to species that are hard to associate with known clades and are assumed to represent extinct branches of the beetle tree of life (e.g. Tihelka et al., 2022; Kundrata et al., 2024). An inherent challenge for the recovery of the correct phylogenetic position of these fossil organisms is the reconstruction of their relationships on the basis of a fragmentary set of morphological features. Because molecular characters cannot be used for insect fossil taxa in phylogenetic analyses, the analysis of available morphological data under a parsimony criterion seems to be the most reasonable approach to reconstructing the position of these extinct species (Goloboff et al., 2018; Smith, 2019). On the other hand, in groups with comparatively robust phylogenetic hypotheses obtained based on rich molecular sampling in terms of terminals and genes, combined analysis (morphology plus molecules) has been conducted in order to understand the relations of extant and extinct taxa in various insect lineages including beetles (e.g. Gustafson et al., 2017; Weirauch et al., 2019; Zhou et al., 2020; Żyła et al., 2021).
The handsome fungus beetles, Endomychidae sensu lato (Shockley et al., 2009a; Tomaszewska, 2010; Robertson et al., 2015; Tomaszewska et al., 2023) are a case of a beetle group where fossil species, from Cenozoic and Mesozoic, are being discovered and described at an unprecedented pace (e.g. Kirejtschuk and Nel, 2009; Alekseev and Tomaszewska, 2018; Tomaszewska et al., 2018; Li et al., 2022b, 2023; Tomaszewska et al., 2022; Arriaga-Varela et al., 2023b). However, the interpretation of the relations of these extinct taxa with modern representatives is very challenging, mostly owing to discrepancies between the results of phylogenetic analyses based on morphological vs. molecular datasets (see Tomaszewska et al., 2023, 2023).
The concept of the classification of all handsome fungus beetles, in just one family, Endomychidae, containing 12 subfamilies (Tomaszewska, 2000, 2005), and as sister group to Coccinellidae and Coccinellidae + Corylophidae, respectively, was changed with the recognition of different families. The results of the molecular-based study by Robertson et al. (2015) confirmed the endomychid subfamilies Mycetaeinae, Eupsilobiinae and Anamorphinae as independent lineages within Coccinelloidea, with Anamorphidae recovered as sister to Corylophidae (minute hooded beetles). The latter had already been suggested in earlier molecular phylogenetic analyses that included anamorphines (e.g. Hunt et al., 2007; Robertson et al., 2008, 2013; Bocak et al., 2014) and was confirmed by the recent study of Arriaga-Varela et al. (2023c).
Anamorphidae are a group of small beetles of predominantly subhemisphaerical shape which contains about 175 described species in 35 genera (Shockley et al., 2009a; Tomaszewska et al., 2023). Although very little is known about the natural history of most members of this family, it is suspected that most species feed predominantly on fungal spores (Shockley et al., 2009b; Arriaga-Varela et al., 2023a). They can either be found on exposed surfaces, such as fungi-infested logs, or collected in more cryptic environments like forest leaf litter. The most noteworthy known feature of their natural history has evolved in species mostly found in exposed surfaces; the larvae of at least three genera protect themselves by avoiding detection by predators using debris particles attached to specially modified dorsal setae as camouflage (Leschen and Carlton, 1993; Arriaga-Varela et al., 2023a). Owing to the small size and rather cryptic habits of the members of this family, there is a large gap in our knowledge of different aspects of their life histories. For example, species of various lineages in Coccinelloidea are known to have developed different strategies to protect themselves from predation. However, besides debris-cloaking and potential aposematism, no other defensive mechanism has been observed in Anamorphidae. For this reason it is crucial to consider every source of information of the present and past Anamorphidae fauna in search of clues about the natural history of organisms in this family.
Our ongoing studies of specimens embedded in Upper Cretaceous amber of northern Myanmar have suggested the existence of members of Anamorphidae in that period. These fossil specimens show overall similarities to the extant genera Symbiotes and Asymbius. Such a discovery would contribute to the understanding of the origin of the main evolutionary lineages in Coccinelloidea.
In order to assess the phylogenetic position of these fossils, described here as Palaeosymbius groehni and P. mesozoicus gen. et spp. nov., a comprehensive study of representatives of Anamorphidae and allied groups was performed incorporating morphological and molecular evidence. Moreover, a detailed study of the morphology of Anamorphidae necessary to better reflect their evolutionary affinities has revealed morphological features in extant taxa that so far had eluded our knowledge. These include the first report in Anamorphidae of external pores putatively associated with glands in two genera from the Neotropical and Oriental regions. Additionally, we explore the presence and nature of the shallower pits or pores seen in the integument of most anamorphids, in both extant and fossil taxa.
Material and methods
Preparation, examination, imaging and deposition of fossil taxa
- Leibniz-Institut zur Analyse des Biodiversitätswandels, Hamburg, Germany (formerly the Geolog.-Palaeontolog. Institut of University Hamburg (GPIH), later CeNak—Centrum of Natural History, coll. Carsten Gröhn); and
- Museum of Amber Inclusions at the University of Gdańsk, Poland (MAIG).
The types of the new species are deposited in these institutions and their acquisition and handling were done in accordance with all ethical standards for studying fossils from Myanmar (Shi et al., 2021). The age of amber coming from northern Myanmar is generally considered to be of Late Cretaceous age (earliest Cenomanian, ca. 99 Ma) based on U–Pb zircon dating (Shi et al., 2012). However, in order to confirm the origin of our amber pieces, Fourier transform infrared spectra (FT-IR) were obtained for the amber samples containing the specimens of Palaeosymbius species described here. Analyses were performed at the laboratory Amber Experts in Gdańsk, Poland with a Nicolet 380 FT-IR spectrometer with an attenuated total reflectance accessory.
The amber-embedded specimens were initially studied using an Olympus SZX16 stereomicroscope. Photographs were taken using an Olympus BX43F microscope. The measurements criteria follow Tomaszewska et al. (2018, 2022): total length from apical margin of clypeus to apex of elytra; pronotal length from the middle of anterior margin to margin of basal foramen; pronotal width at the widest part; elytral length across the sutural line including scutellum; and elytral width across both elytra at the widest part. New taxa described here are registered in Zoobank (https://zoobank.org/) under the URN for the publication: urn:lsid:zoobank.org:pub:1F72964B-05E6-482F-BA90-EF7A0903BD14.
In order to have a better visualization of different details of the morphology X-ray micro-computed tomography observations were conducted for the amber piece containing the specimen described here as P. groehni gen. et sp. nov. Initial exploration of the amber piece and embedded specimen was performed using a Skyscan1172 Bruker system, voltage 80 kVp, with an aluminium filter (Al 0.5 mm) and an exposure time of 0.5 s per projection at the Palaeontology Department, National Museum, Prague. Later it was scanned at higher resolution at the National Laboratory for X-ray Micro-Computed Tomography of the Australian National University in Canberra (https://ctlab.anu.edu.au/capabilities/micro-ct.php) using a high-resolution micro-computed tomography system (ANU4). Scanning parameters were as follows: isotropic voxel size, 4.03 μm; aluminium filter 0.25 mm; beam current, 32 μA; acceleration voltage, 60 kV; exposure time, 11 s; and number of projections, 4768. Videos with 3D reconstructions of the specimens were uploaded and are available at the Zenodo depository (https://zenodo.org/) under the doi 10.5281/zenodo.10320209
Morphological dataset preparation and analysis
The aim of the morphology-based cladistic analysis was to find the closest relatives of the newly described extinct genus Palaeosymbius gen. nov. and its species, and to recognize, and thus testing our predictions, their proper family and/or subfamily placement within handsome fungus beetles and their relatives.
This dataset is focused on the coccinelloid group of families of Coccinelloidea (sensu Robertson et al., 2015). Members of the families collectively known as handsome fungus beetles (Endomychidae sensu Tomaszewska, 2000) were sampled as an in-group, Endomychidae, Eupsilobiidae, Mycetaeidae and a broad representation of Anamorphidae, including the type species for genera when specimens were available for study. Representatives of Coccinellidae and Corylophidae were also included. Hobartius eucalypti (Blackburn) (Hobartiidae) was used as a more distant outgroup. The taxon sampling and characters used for the analysis were mainly based on Tomaszewska et al. (2023) (with references to Tomaszewska, 2000 and 2005), and modifications were made in an attempt to provide resolution between Palaeosymbius gen. nov. and possibly related taxa.
Our morphological data matrix included 34 taxa (29 ingroup taxa and five outgroups) (see Appendix S1 for sampled taxa in the analyses) scored for 47 multistate characters. Unknown character states were treated as missing and coded as “?”. The morphological characters are listed in Appendix S2 and the character matrix can be found in Appendix S3.
The maximum parsimony method was used for phylogenetic analysis of our dataset. The analysis was conducted in TNT (ver. 1.5, see https://www.lillo.org.ar/phylogeny/tnt/; Goloboff and Catalano, 2016) in order to find the most parsimonious trees (MPTs). We used the Traditional Search (TS) option under the following parameters: memory set to hold 1 000 000 trees, tree bisection–reconnection branch-swapping algorithm with 1000 replications saving 100 trees per replicate; zero-length branches collapsed after the search. We also used the New Technology (NT) option and in turn, the Driven Search with Sectorial Search, Ratchet, Drift and Tree fusing options activated with standard settings. All characters were treated as unordered and analyses were first performed under equal weights. Subsequently, the Implied Weighting option was used to reduce the effects of homoplasy. Other settings were unchanged. The analysis was repeated with values of the concavity constant K = 3, 5, 10 and 15. Bremer support on the preferred tree was calculated using the TNT Bremer function, using suboptimal trees up to 20 steps longer. Character mapping was done in Winclada (ver. 1.00.08, see www.diversityoflife.org/winclada; Nixon, 2002) with unambiguous optimization.
Combined data preparation and analysis
In order to integrate different sources of biological information aiming to reconstruct a more robust hypothesis on the position of the new fossil taxa among Coccinelloidea and Anamorphidae, we constructed a combined dataset of morphological plus molecular characters. The taxon sampling overlaps as much as possible with that of the morphological dataset. Members of families Coccinellidae, Eupsilobiidae, Mycetaeidae, Endomychidae, Corylophidae and Anamorphidae are represented, and the sampling of Anamorphidae consists of 19 species, in 15 genera. Our sampling is the most comprehensive to date in terms of extant genera included for the family, in terms of both molecular or morphological datasets. In order to root the tree, H. eucalypti (Hobartiidae) was chosen as distantly related outgroup.
The molecular data for the majority of the species was taken from the dataset of Robertson et al. (2015) or the GenBank database (www.ncbi.nlm.nih.gov/genbank/). Nevertheless, five new species were sequenced ex novo. These species are: Exysma laevigata Gorham (Mexico), Coryphus sp. (Papua New Guinea), Anagaricophilus sp. (Madagascar), Mychothenus sp. (Republic of South Africa) and Endomychus coccineus (Linnaeus) (Romania). For the full list of species included in the analysis see Appendix S1. The molecular part of the dataset includes the eight genetic markers used by Robertson et al. (2015): nuclear 18S rDNA (18S), 28S rDNA (28S), histone subunit 3 (H3), carbamoyl-phosphate synthetase 2, aspartate transcarbamoylase and dihydroorotase (CAD), mitochondrial 12S rDNA (12S), 16S rDNA (16S), cytochrome c oxidase subunit I (COI) and cytochrome c oxidase subunit II (COII). For these newly added species, amplification of seven of these markers was attempted. Mitochondrial 16S was excluded owing to previous problems with the amplification using the primers employed by Robertson et al. (2015). Amplification was performed in protocols of 37 cycles. The CAD marker was amplified in a semi-nested protocol. For the first reaction we used the pair of primers CD338F + CD688R, and for the second reaction we used CD338F + CD668R. For the whole list of primers and annealing temperatures used in the amplification of all fragments see Appendix S4. The trace files of the new sequences obtained were uploaded into GENEIOUS (Kearse et al., 2012) for inspection, assembly and editing. All newly generated sequences were submitted to GenBank under accession numbers PP501119-PP501122, PP501444-PP501454, PP505961-PP505966, PP511208-PP511213, PP515079-PP515080, PP528884-PP528887. The list of vouchers, associated data and GenBank accession numbers of the sequences can be seen in Appendix S1.
The combined (molecular + morphological) dataset was composed of 34 species including three fossil species (the two new species of Palaeosymbius and Symbiotes borussiaeorientalis Alekseev & Tomaszewska known from the Eocene Baltic amber). Molecular sequences were included for all extant species, except for Asymbius sp., for which DNA-grade specimens were not available. For Symbiotes latus Redtenbacher only one marker, the barcoding region of COI gene, was available from Genbank. Since the sampling of the morphological dataset and the available molecular dataset did not overlap at specific level in all cases, we constructed three interspecific molecular-morphological chimeras at the generic level for a merging of these data in the best possible way and to minimize missing data on any terminal (see Appendix S1). Sequences were aligned in GENEIOUS software using the ClustalW (Thompson et al., 2003) algorithm with default settings. The alignment of DNA sequences was trivial for COI, COII, H3 and CAD. Alignments for rDNA sequences (12S, 16S, 18S and 28S) were inspected by eye and corrected manually or by subsequent iterations of realignments of hypervariable zones using ClustalW or MUSCLE (Edgar, 2004) algorithms in GENEIOUS. Concatenation of sequences was performed manually in the GENEIOUS software. Final concatenated combined alignment was composed as follows: (bp/character range in the final alignment is provided in parentheses)—12S, 358 bp (1–358); 16S, 522 bp (359–881); 18S, 1917 bp (882–2794); 28S, 886 bp (2795–3680); COI, 1239 bp (3681–4919); COII, 627 bp (4920–5546); H3, 327 bp (5547–5873); and CAD, 915 bp (5874–6788); morphological characters, 47 (6789–6835). The final concatenated combined dataset can be found in Appendix S5. This dataset was used to infer the phylogenetic relationships using Bayesian inference (BI). Protein-coding genes were partitioned by codon position and ribosomal genes were analysed unpartitioned. The BI analyses were performed using MRBAYES v.3.2.6 (Ronquist et al., 2012) implemented on the CIPRES Science Gateway v.3.3 (Miller et al., 2010). Seven relaxed topological constraints were added based on the phylogenetic relations recovered in the analysis of molecular sequences by Robertson et al. (2015). The constraints are: (i) monophyly of the ingroup (coccinellid group) with respect of the outgroup (H. eucalypti); (ii) monophyly of Coccinellidae; (iii) monophyly of (Coccinellidae + Eupsilobiidae); (iv) monophyly of (Mycetaeidae + (Coccinellidae + Eupsilobiidae)); (v) monophyly of (Endomychidae + (Mycetaeidae + (Coccinellidae + Eupsilobiidae))); (vi) monophyly of (Anamorphidae + Corylophidae); and (vii) monophyly of Corylophidae. Four simultaneous independent Markov chain Monte Carlo runs with six chains each and 40 million generations were used for the analysis, with a tree sampled every 10 000 generations to calculate posterior probabilities (PP). The convergences of the runs and their parameters were assessed in TRACER v.1.7 (Rambaut et al., 2018). A conservative 50% burn-in fraction was excluded from the data prior to the construction of the final maximum-credibility tree. The tree resulting from the phylogenetic analysis of the combined data under BI is presented in Fig. 2.
Study of cuticular cavities and pores
During the preparation of the morphological matrix special attention was given to the cuticular pores or pits found in the pronotum, elytra and meso- and metaventrites of Anamorphidae species. In order to have a better idea of the nature of the pits observed in the fossil taxa, including their structure and potential use, we focused in particular on examining the integument of extant representatives of Symbiotes and Asymbius, since these genera are presumably the closest relatives of our fossil genus based on preliminary morphological comparisons. Apart from the pits commonly present on meso- and metaventrites of many Anamorphidae and Endomychidae members, we have additionally found the cavities or pores distributed mainly near the base of the pronotum or near the lateral margins of pronotum and elytra, and discovered a series of smaller punctiform perforations located in the anterolateral corners of the thickened pronotal borders in species of Asymbius and Anamorphus. This led us to explore in detail their structure and compare it with that of the pores found in other anamorphids.
Specimens of Anamorphidae species where the series of perforations on the pronotal borders or other pits in the pronotum, elytra or ventrites were observed (Symbiotes gibberosus, Asymbius sp., Anamorphus sp.) were treated with 10% KOH at 100°C for 20 min, then allowed to cool down for 2 h and afterwards rinsed with distilled water. This procedure was performed even in the case of a specimen that was previously treated with proteinase K for the DNA isolation (Anamorphus sp.). The structure of the pores was observed and photographed using an Olympus BX43F microscope. The structure of the perforated pronotal borders in specimens of Asymbius and Anamorphus was examined and micrographs were taken using a Hitachi S-3400N scanning electron microscope under low vacuum conditions in the Electron Microscopy Laboratory at the MIZ PAS.
Results
Parsimony analysis of morphological dataset
Different search strategies used in our phylogenetic analysis of morphological characters resulted in sets of at least partly similar resolutions. The maximum parsimony analysis (MP) under equal weights under TS (MP TS) resulted in 24 MPTs with a length (L) of 114 steps, a consistency index (CI) of 48 and a retention index (RI) of 74, and under NT search (MP NT) resulted in three MPTs with L = 117 steps, CI = 47 and RI = 73. For both search strategies a strict consensus tree was calculated—from the 24 MPTs from the TS (L = 134 steps; CI = 41; RI = 66) (Appendix S6) and from three MPTs from the NT (L = 133 steps; CI = 41; RI = 66) (Appendix S7). Both these consensus trees are nearly identical with small differences within Anamorphidae clade, which in both cases is displayed mostly as an unresolved polytomy.
The MP analysis under implied weighting (IW) with the concavity constant K set as K = 3, 5, 10 and 15 resulted in nine trees in the TS option (for each K parameter) with L = 114 steps, CI = 48 and RI = 74, and two trees in the NT search option for K = 3, with L = 114 steps, CI = 48 and RI = 74 (other K parameters resulted in three or four trees). The consensus trees calculated from trees of both search options obtained under IW (from the nine MPTs from the TS (L = 118 steps; CI = 46; RI = 73) (Appendix S8) and from two MPTs from the NT (L = 116 steps; CI = 47; RI = 74) (Fig. 1)) are also nearly identical and display most relationships within Anamorphidae clade as well resolved.
All calculated consensus trees (Fig. 1, Appendices S6–S8) display all recognized families/ subfamilies used in the analysis as monophyletic groups, although with relationships between many of them unresolved. The sister relationships of Eupsilobiidae + Coccinellidae, Corylophidae + Anamorphidae, the monophyly of Endomychidae (sensu stricto) and the newly described fossil taxa being a part of Anamorphidae clade, are recovered in all consensus trees (from TS and NT, under equal weights and with IW option analyses). The major clades Mycetaeidae, (Eupsilobiidae + Coccinellidae), Endomychidae (sensu stricto) and (Corylophidae + Anamorphidae) are recovered as an unresolved tetratomy.
Among the obtained resolutions, we selected as our preferred tree the best-resolved topology—the consensus tree of two MPTs from the MP NT search under IW with K = 3 (Fig. 1), which mostly agrees with Robertson et al. (2015) and at least partly agrees with the results of Tomaszewska (2005) and Tomaszewska et al. (2023).
It appears that rich taxon sampling of Anamorphidae used in our analysis (we included members of 14 of 35 extant genera and the newly described fossil species of Palaeosymbius) and some morphological characters not used in previous analyses (e.g. hind wings details), had an impact for recovering Anamorphidae outside Endomychidae and as sister to Corylophidae, in agreement with molecular based phylogeny by Robertson et al. (2015), and contrary to previous morphology only based analyses by Tomaszewska (2000, 2005) and Tomaszewska et al. (2023).
Bayesian analysis of the combined dataset
The combined analysis resulted in a tree that shows Anamorphidae forming a strongly supported (PP = 1.00) clade-including Palaeosymbius gen. nov. (Fig. 2). The topology recovered for Anamorphidae revealed two main clades, one composed of Symbiotes + Palaeosymbius and the second including the rest of the genera. The main difference from the results of the morphological analysis is the position of Asymbius, which in the parsimony analysis is recovered as a sister taxon to Symbiotes but in the combined analysis it is shown as sister to the rest of taxa, although with a low support value (PP = 0.64). The rest of the genera are arranged in a clade with low support value (PP = 0.42). This group itself is divided in two clades. One with a relatively high support (PP = 0.94) is composed mainly of Neotropical genera: Bystus, Micropsephodes, Anamorphus, Catapotia and an Australian species of Geoendomychus. The other clade, with a support of 0.72 (PP) is composed mainly of representatives from the Old World: Papuella and Coryphus from Papua New Guinea, Bystodes from the Oriental region, and Anagaricophilus and Mychothenus from the Afrotropics. Meanwhile, the Neotropical Exysma and Nearctic Clemmus minor (Crotch) are also nested in this clade. Support values among these taxa range from 1 to 0.56 (PP). It is worth noting that the Neotropical Catapotia was recovered in the morphological analysis as the sister group to all remaining Anamorphidae. This placement is based on the following character states in Catapotia: pronotum with basal and/or lateral sulci/modifications absent (present in remaining Anamorphidae (19:1)): mesoventral and metaventral postcoxal pits absent (pits present at least on meso-, or metaventrite in other anamorphids (28:1 and 30:1)). Interestingly all of these character states in Catapotia are shared with Coccinellidae, as well as the glabrous body (pubescent in most other anamorphids), although this character was not used in the analysis.
Phylogenetic position and taxonomy of Palaeosymbius
The BI and MP analyses resulted in the monophyly of Palaeosymbius, and provide unequivocal support for a placement of this genus within the family Anamorphidae. The BI combined analysis recovered the relationships within Anamorphidae as ((Symbiotes + Palaeosymbius) + (Asymbius + remaining genera)) with high support (PP = 1.00); meanwhile morphological analysis recovered these relationships as ((Catapotia + (Palaeosymbius + (Asymbius + Symbiotes))) + (Exysma + (Clemmus + remaining genera included in analysis))).
In the combined analysis, a group formed by Symbiotes + Palaeosymbius is moderately supported (PP = 0.84), while in morphological analysis the monophyly of Asymbius + Symbiotes is well supported by an uncontroverted synapomorphy, elytra with parasutural striae strongly bent outwards before the base of each elytron (47:2), and the monophyly of Palaeosymbius + (Asymbius + Symbiotes) is also supported by an uncontroverted synapomorphy, elytron with a pit antero-laterally (46:1).
The placement of Asymbius remains unclear owing to the lack of molecular information. In this paper we tentatively consider it as close to Symbiotes according to the results of the morphological analysis.
Palaeosymbius is moderately supported as a monophyletic group (PP = 0.62), and morphologically it is defined by one homoplastic character, a pronotum with a pair of pits located postero-laterally, close to the lateral margin (21:1), shared with extant species of Symbiotes.
The following taxonomical decision is based on the results of conducted phylogenetic analyses.
Taxonomy
Order COLEOPTERA Linnaeus, 1758.
Suborder POLYPHAGA Emery, 1886.
Superfamily COCCINELLOIDEA Latreille, 1807.
Family ANAMORPHIDAE Strohecker, 1953.
Palaeosymbius Arriaga-Varela, Szawaryn and Tomaszewska, gen. nov.
Zoobank LSID: urn:lsid:zoobank.org:act:5BBB51E6-3B3C-4679-8C9D-5D5AB664FC65.
Etymology: The new genus-group name Palaeosymbius is composed of “palaeo” (meaning old, ancient) and “-symbius”, part of the names Asymbius and Symbiotes, the extant, habitually most similar anamorphide genera. The gender is masculine.
Composition: The new genus is composed of two species.
Type species: Palaeosymbius groehni Arriaga-Varela, Szawaryn and Tomaszewska, sp. nov.
Diagnosis
Palaeosymbius resembles most Symbiotes and Asymbius, sharing with both genera overall similarities of habitus and having a pair of pits on the elytra, placed antero-laterally. However, the new genus can be distinguished from these genera by the elytra having very weakly marked parasutural striae (parasutural striae are distinct, complete and strongly bent outwards before the base of each elytron in Asymbius and Symbiotes). Moreover, Palaeosymbius can be separated from Symbiotes by the absence of lateral, submarginal groove on pronotum that extends from the anterior margin to one-third of the basal length in Symbiotes. From Asymbius, it can be distinguished by a pronotum with a pair of pits placed postero-laterally, between sulci and lateral margins (pits absent in Asymbius); pronotal lateral margins simple antero-laterally (thickened anteriorly and provided with row of pores in Asymbius); fore tarsi with four tarsomeres (three tarsomeres in Asymbius); and pretarsal claws simple (modified, provided with a basal tooth in Asymbius).
Description
Body length 1.5–1.8 mm; width 0.90–1.06 mm (at the widest point in about the anterior quarter of elytral length). Body oval, 1.6–1.7 times as long as wide (Figs 3a–c and 4a–c); dorsal surfaces covered by short decumbent setae.
Head weakly transverse, retracted into prothorax to the hind margin of the eyes (Figs 3a and 5c). Clypeus transverse and flat. Eyes large, rounded and prominent. Antennal insertions visible from above. Antenna (Figs 3d and 4d,e) with 10 or 11 antennomeres, moderately long, extending shortly beyond posterior margin of pronotum, 0.35–0.40 times as long as the body; scape stout, distinctly longer than pedicel; pedicel moderately stout, antennomeres 3–8 getting progressively shorter; antennomeres 9–11 (or 9–10) forming a narrow, loose, somewhat flattened club; antennomeres 9–10 about the same length, 10 slightly wider; antennomere 11 (if present) almost twice as long as 9, ovoid, tapering apically on lateral view (Fig. 3d). Maxillary palpi four-segmented (Fig. 5d); palpomere 1 smallest, palpomeres 2 and 3 of about same length, terminal palpomere about 1.5 times as long as preceding one, apically acuminate with a finely rounded tip. Labium with mentum transverse, sub-trapezoidal, about twice as wide as long; palpi apparently three-segmented, palpomere 1 shortest/smallest; palpomere 2 transverse, widening apically; terminal palpomere weakly widening towards the apex, about as long as wide, nearly 2 times longer than palpomere 2, widely truncate at apex.
Prothorax about twice as wide as long (Figs 3e and 4f); widest at about mid-length. Lateral margins with narrow, slightly thickened margin throughout; anterior angles rounded, not produced; posterior angles right-angled to weakly obtuse; disc more distinctly convex medially; basolateral impressions on disc in form of subtriangular or narrow longitudinal sulci near lateral margins; basal sulcus absent; posterior margin nearly straight. Prosternum with intercoxal process separating coxae from 0.6 to 1.3 times the coxal width, extending beyond coxae by a fifth to a half of the coxal length, apex rounded to almost truncate (Figs 3c,f and 4c).
Mesothorax with scutellar shield small, triangular, transverse. Mesoventrite transverse, slightly convex medially, with two large oval pits antero-laterally to intercoxal process (Figs 3c,h and 4c); mesoventral intercoxal process somewhat pentagonal in shape, slightly to distinctly wider that mesocoxal diameter, with sides between coxae parallel or weakly narrowing backwards. Mesocoxal cavities widely closed laterally (Figs 3c,h and 4h). Elytra short-oval, weakly longer than wide, 2.2–2.7 times as long and 1.1–1.2 times as wide as pronotum (Figs 3a and 4a); relatively weakly convex; lateral margins from narrowly to comparatively broadly flattened (Fig. 3g); with a small pit situated near the anterolateral corners of each elytron (Figs 4a and 5e), placed sub-laterally on the sides of convex area of elytral disc; sutural striae weak, only distinguishable in the anterior fifth; epipleura broad, narrowing backwards and vanishing before the apex of elytra (Figs 3c and 4c).
Metaventrite transverse, about twice as wide as long, between 1.7 and 2.0 times as long as mesoventrite (Figs 3c,h, 4c and 5f); intercoxal process straight between mesocoxae; anterior margin with one or two rounded pits posterior to each coxa. Metacoxae transverse, broadly separated. Hind wings not visible.
Legs moderately long (Figs 3c and 4c). Femur weakly clavate, sparsely pubescent. Tibiae weakly curved and widening towards apices, without modifications; tibial spurs absent. Tarsal formula 4–4–4. Tarsi long (Fig. 3i), tarsomeres 1–3 subequal in length, ranging from slightly longer than wide to 2.5 times as long as wide; tarsomeres 2 and 3 slightly lobed ventrally; tarsomere 4 slightly longer than tarsomeres 2–3 combined.
Abdomen with six ventrites (Figs 3c and 4c). Ventrite 1 about as long as the following three ventrites combined, postcoxal lines absent; ventrites 2–4 subequal in length; terminal ventrite rounded apically.
Palaeosymbius groehni Arriaga-Varela, Szawaryn and Tomaszewska, sp. nov.
Zoobank LSID: urn:lsid:zoobank.org:act:FEEA5B90-2628-4535-8FA1-1BCA6B1DDBF5.
Holotype: Sex unknown, GPIH no. 5074 (CCGG no. 11939). In an amber piece triangular-based prismatic shape of 12 mm along the side and 3 mm in height. Syninclusions: vegetal detritus and trichomes. FT-IR spectrum, Amber Experts Laboratory no. 230614008857.
Repository: Carsten Gröhn collection of the Geologisch-Palaeontologischen Institut of the University of Hamburg, Germany (GPIH).
Etymology: This new species is named after Carsten Gröhn (Glinde, Germany), collector of the type specimen, who made it available for our study.
Locality and horizon: Myanmar, Kachin State, Hukawng Valley, Myanmar amber; unnamed horizon; Cretaceous, Upper Albian to Lower Cenomanian.
Diagnosis
Palaeosymbius groeheni can be differentiated from its congener by the following characters: pronotum with lateral sulci subtriangular (Fig. 3e); prosternal process very long, rounded apically (Fig. 3c,f); mesoventral process not wider than mesocoxal diameter, narrowing backwards (Fig. 3c,h); and metaventrite with two pairs of large pits (Figs 3c and 5b,f). It also has antenna with 11 antennomeres and a three-segmented club (Fig. 3d,c), which potentially differs from P. mesozoicus sp. nov.
Description
Length 1.5 mm, width 0.9 mm. Body elongate-oval, 1.6 times as long as wide (Figs 3a,b and 5a,b), pale yellowish-brown (probably a teneral specimen) with dorsal surfaces covered by short decumbent yellowish setae.
Antenna with 11 antennomeres (Fig. 3c,d), 0.4 times as long as the body; club formed by three antennomeres. Antennomere 3 about 1.5 times as long as wide, following antennomeres getting progressively shorter towards antennomere 8, which is as long as wide; antennomere 9 about 1.1 times as long as wide; antennomere 10 about as long as wide anteriorly, terminal antennomere sub-ovoid, about twice longer than wide.
Pronotum 2 times as wide as long, widest near mid-length (Fig. 3e). Lateral margins not distinctly thickened or raised. Disc raised, accentuating towards anterior half. Base with subtriangular impressions as lateral sulci. Prosternum with intercoxal process long and narrow (Fig. 3c,f), slightly convex near mid-length and about 0.6 times as wide as procoxal diameter, extending beyond coxae by about a half of coxal length, apex rounded.
Mesoventrite with intercoxal process somewhat pentagonal in shape about 0.8 times as wide at posterior margin as mesocoxal diameter (Fig. 3h); process between coxae narrowing backwards. A pair of large pits present antero-laterally to process (Fig. 3c). Elytra about 1.0 mm long, 1.15 times as long as wide, 2.7 times as long and 1.2 times as wide as pronotum (Figs 3a and 5a). Widest at anterior fifth then continuously converging towards rounded apex. Humeri weakly raised. Lateral margins widely explanate, continuously narrowing posteriorly (Fig. 3g); width of explanate section near base of elytron nearly a fifth of the total width of elytron at base.
Metaventrite with two pairs of large pits at anterior margin below mesocoxae (Figs 3c and 5f); lateral-most pit larger than the internal one.
Legs with metatibia about one-third the length of elytra. Meso- and metatarsi with tarsomeres 2 and 3 relatively long (Fig. 3i), about 2.5 times as long as wide.
Palaeosymbius mesozoicus Arriaga-Varela, Szawaryn and Tomaszewska, sp. nov.
Figure 4
Zoobank LSID: urn:lsid:zoobank.org:act:28B418AF-516C-43D2-9034-DEBE17421709.
Holotype: Sex unknown, MAIG 7077. In 10 × 8 × 2 mm amber piece of rectangular prismatic shape. Syninclusions: Hemiptera: Enicocephalomorpha, vegetal detritus and trichomes. FT-IR spectrum, Amber Experts Laboratory no. 230614008858.
Repository: Museum of Amber Inclusions, University of Gdańsk, Poland (MAIG).
Etymology: This new species is named after the Mesozoic geological era (252–66 million years ago) from which this species comes.
Locality and horizon: Myanmar, Kachin State, Hukawng Valley, Myanmar amber; unnamed horizon; Cretaceous, Upper Albian to Lower Cenomanian.
Diagnosis
Palaeosymbius mesozoicus sp. nov. differs from P. groehni primarily in the pronotum with lateral sulci long and linear (Fig. 4f); prosternal process subtruncate at apex (Fig. 4c); mesoventral process distinctly wider than mesocoxal diameter, parallel-sided (Fig. 4c); and metaventrite with one small pit posterior to each mesocoxa (Fig. 4c). It also potentially differs from P. groehni in having antenna with 10 antennomeres and a two-antennomere club (Fig. 4d,e) (see also Remarks for that species).
Description
Length 1.8 mm, width 1.06 mm. Body elongate-oval, 1.7 times as long as wide (Fig. 4a,b), reddish-brown with dorsal surfaces covered by short decumbent dark reddish-brown setae.
Antenna with 10 antennomeres (possibly missing terminal antennomere in both antennae), 0.35 times as long as body (Fig. 4d,e); club formed by two antennomeres. Antennomere 3 about three times as long as wide, following antennomeres getting progressively shorter towards antennomere 8, which is about 2.2 times longer than wide. Antennomeres 9 and 10 about twice as long as wide.
Pronotum about 1.7 times as wide as long, widest near mid-length (Fig. 4f). Lateral margins slightly thickened and raised. Disc raised, gradually accentuating towards its anterior half. Base with lateral longitudinal sulci near sides, positioned subparallel to lateral margins and reaching shortly after mid-length of pronotum. Prosternum with intercoxal process comparatively short and wide (Fig. 4c), almost flat in mid-length, at mid-length about 1.3 times wider than procoxa, extending beyond coxae by about one-fifth of the coxal length, apex truncate.
Mesoventrite with intercoxal process about 1.2 times as wide at the posterior margin as the mesocoxal diameter (Fig. 4c); between coxae subparallel; carinae delimitating process weakly narrowing anteriorly in the anterior half. Elytra about 1.0 mm long, 1.17 times as long as wide, 2.2 times as long and 1.1 times as wide as the pronotum (Fig. 4a). Widest at the anterior fifth then continuously converging towards the rounded apex. Humeri not distinctly raised (Fig. 4g). Lateral margins narrowly explanate, slightly wider near elytral pore, then continuously narrowing posteriorly to near apex.
Metaventrite with one small pit at anterior margin posterior to each mesocoxa near its lateral margin (Fig. 4c).
Legs with metatibia about one-quarter of the elytral length. Meso- and metatarsi with tarsomeres 2 and 3 relatively short, about 1.5 times as long as wide.
Remarks
The type specimen of this species shows both antennae with 10 antennomeres. It is not completely clear if this is the natural condition of this species or a result of either a teratological condition or the mechanical loss of the apical antennomeres. Although the symmetrical loss of the apical antennomeres in both antennae seems unprovable, the truncate antennomere 10 and the overall similarity of both species of Palaeosymbius lead us to think that the possibility of 11-segmented antennae in P. mesozoicus cannot be rejected.
Fourier transformed infrared spectra
Our results of the FT-IR (attenuated total reflectance) spectrum analyses (Appendix S9) show absorption peaks that have been considered characteristic of Burmite from the Kachin region in northern Myanmar on the basis of the of the location and relative intensity at the wavenumber regions of 1226, 1155, 1031 and 974 cm−1 (see Giłka et al., 2020, 2022; Arriaga-Varela et al., 2023b), supporting the origin of the specimens.
Cuticular pores in Anamorphidae
Our observations on a selected set of anamorphids with the purpose of scoring morphological characters in the matrix has revealed the presence of previously unreported cuticular pores on the anterolateral corners of the pronotum in members of two genera: Anamorphus and Asymbius (Fig. 6a,c). These pores or openings are situated in a thickened portion of the pronotal lateral margin that is slightly exposed towards the dorsal side. Along this thickened part of the margin a series of narrow pores was found. The approximate number of pores on each side is seven in Anamorphus and five in Asymbius, although these can sometimes be confluent. In order to understand the structure of these pores the cleared pronota were observed under the compound light microscope. We found associated internal modifications that suggest glandular structures of the pores. In Asymbius these appear as four or five large pyriform saccular structures connected to the pores by short terminal ducts (Fig. 6d,e). In this genus these potentially glandular structures occupy a large part of the prothorax, extending to about a third of the width of each lateral side and more than a half of its length. In Anamorphus these modifications are smaller, appearing as a series of six or seven ducts acutely pointed internally and apparently merged towards the pronotal margin (Fig. 6b).
These pores were found in an unidentified, probably undescribed species of Asymbius from Laos but we failed to find similar modifications in other species of this genus in a preliminary survey. On the other hand, all species of Anamorphus seem to have similar pores. A preliminary survey for these kinds of modifications among members of Anamorphidae revealed the presence of similar pores and associated internal gland-like structures in Rhymbomicrus hemisphaericus (Champion) from Mexico. These are very similar to those found in Anamorphus spp. The discovery of these structures represents the first report of potentially exocrine glands for defensive purposes located in the prothorax in members of Anamorphidae. An exhaustive review of the morphology in Anamorphidae is necessary in order to understand the distribution of these glands and associated pores among Anamorphidae.
The structure of other cuticular pores or deep concavities present on dorsal surfaces of the body in Anamorphidae was explored in other studied specimens. These modifications include the pits located at the base of the pronotum, associated with lateral sulci, present in species of many genera of Anamorphidae including Palaeosymbius (and also Endomychidae), the pair of pits located postero-laterally, close to lateral margin of the pronotum, present in Symbiotes and Palaeosymbius, and the pit located near the anterolateral corner of each elytron present in Symbiotes, Asymbius (Fig. 6f–h), and in Palaeosymbius. These concavities turned out to be of a different nature as compared with those putatively glandular ones. These are single, relatively wide concavities without clear evidence of internal opening and in which the cuticle remains of the same thickness. No associated internal modification in the prothorax or elytra was observed in the form of a potential glandular organ. At least in Symbiotes, the pronotal pits are associated with dense patch of short setae (Fig. 6g,h). Spores of possible fungal origin attached to this patches of setae were observed, even after being treated with proteinase or KOH. Certainly, the detailed study of the pits in Palaeosymbius was not possible but their placement, compatible with extant taxa and overall similarities, allows us to hypothesize that these are homologous structures.
Discussion
Phylogenetic analyses and the position of Palaeosymbius
Our study represents the most comprehensive taxonomic study of relationships within Anamorphidae at the genus level. We included 15 genera of the family in both morphological and combined datasets. The closest representation was used by Robertson et al. (2015), where 10 anamorphid genera were included. Our MP analyses resulted in an unresolved backbone for the families of the coccinelloid group. The exceptions are the sister-group relationship between both Eupsilobiidae and Coccinellidae, and between Corylophidae and Anamorphidae. The latter sister relationship is supported by three uncontroverted synapomorphies: mesocoxal cavities widely closed outwardly (26:1), a hind wing with deep medial embayment (44:1) and a hind wing with an apical field at least 0.6 times as long as the total length of the wing (45:1). This relationship is in concordance with the results of the molecular phylogenetic analyses of Hunt et al. (2007) and Robertson et al. (2015). It is noteworthy that the previous MP analyses of morphological datasets failed to recover this relationship (Tomaszewska, 2000, 2005; Tomaszewska et al., 2023). Finding the most congruent evolutionary hypothesis as well as the shared synapomorphies between the lineages in Coccinelloidea, where parallelisms are abundant, is a challenging task that requires reciprocal illumination between the efforts to analyse and interpret both morphological and molecular evidence (Gustafson et al., 2021). The present contribution represents a step forward in the understanding of the morphological evolution in these lineages.
The combined analysis was mainly aimed at reaching a better resolution on the internal relations of Anamorphidae and thus obtaining a clearer insight regarding the position of Palaeosymbius. For this analysis, Anamorphidae was taken as the ingroup and relaxed topological constraints were imposed on the other coccinelloid families constituting the outgroup in order to avoid artefacts in the relationships of family level evolutionary lineages owing to the limited sampling represented in the dataset. The positions of the two Palaeosymbius species were, however, unconstrained and free to find their positions among any of the clades.
Nevertheless, the internal resolution within the family and potential synapomorphies for the main clades remain not fully and convincingly resolved. This uncertainty is mainly a result of the discrepancies between the results of the parsimony analysis of morphological characters and the Bayesian analysis of the combined dataset. The main discrepancies between both results are found in the position of genera Catapotia and Asymbius. The MP analysis recovered Catapotia as sister group to the rest of Anamorphidae genera instead of related to a group of mainly Neotropical taxa as in the combined analysis. On the other hand, Asymbius, which is a part of the clade including Symbiotes and Palaeosymbius in the morphology-only analysis, in the combined dataset analysis is recovered as sister group of all remaining Anamorphidae genera. However, the combined analysis shows very low posterior probability values for most of the basal nodes in the family. This includes the node supporting Asymbius + the rest of the anamorphids except for Symbiotes + Palaeosymbius (PP = 0.54) and the node containing the rest of Anamorphidae except for Symbiotes + Palaeosymbius (PP = 0.47). This low support could be explained at least in part by taxa such as Asymbius missing the molecular section of the alignment. In order to have a better notion of the effect of the missing data of Asymbius for the support values of basal-most nodes in Anamorphidae we performed an exploratory test repeating the combined analysis with the same parameters but with Asymbius removed. Our results, (Appendix S10) show considerable higher support values for the three basalmost nodes containing the anamorphids excluding Symbiotes + Palaeosymbius (PP = 1.0). This suggests the possibility that the position of Asymbius in the combined analysis is an artefact caused by the lack of molecular data in the alignment.
The position of Palaeosymbius gen. nov. in the clade including Symbiotes is, however, a constant among our results, and sister relationships of the clade Palaeosymbius + (Asymbius + Symbiotes) in our MP analysis are supported by one uncontroverted synapomorphy, elytron with a pit located antero-laterally (46:1). This undoubtedly supports the membership of the new genus within Anamorphidae with a probable close relationship to Symbiotes and possibly Asymbius as well. The combined analysis resulted in the tree showing the Anamorphidae with an inclusion of Palaeosymbius as strongly supported (PP = 1.00).
Our discoveries strongly support our predictions that the new genus Palaeosymbius represents the earliest known representative of the family Anamorphidae, and this confirms that the origin of the crown-group Coccinelloidea goes back to at least the Jurassic, in agreement with the dating results of McKenna et al. (2019) and Cai et al. (2022). Recent discoveries of representatives of the coccinelloid group of Coccinelloidea from the Cretaceous amber of Myanmar suggest that most of the main evolutionary lineages of this group were already well stablished 100 million years ago (see Li et al., 2022b; Tomaszewska et al., 2018, 2022).
Previously known fossil taxa of Anamorphidae date back to the Eocene (Alekseev and Tomaszewska, 2018) or the Miocene (Shockley, 2006). However, different aspects of the morphology of most of these taxa were not studied, owing mainly to the preservation of the specimens and the visibility of morphological details. For this reason any analysis or discussion of the affinities of those taxa within Anamorphidae is not possible at this point. A notable exception of a known fossil anamorphid that can be confidently assigned to a particular lineage within Anamorphidae is Symbiotes borussiaeorientalis Alekseev and Tomaszewska, 2018 from Baltic amber (Eocene). This species shows a close resemblance to the extant members of the genus. The suggested affinity between Cretaceous Palaeosymbius and extant Symbiotes, and their morphological similarity, suggest an early evolution of this body plan within the family and relative morphological stasis among their members. Morphological stasis of different lineages of Coleoptera has been suggested in diverse clades (e.g. Cai et al., 2019; Yamamoto, 2021).
Nature of the pores and its relation to defensive strategies
Given the shape of the pores in the anterolateral corners of the pronotum and associated internal organs of Asymbius sp. and Anamorphus spp., we consider it is possible that these have an exocrine secretory function. Cuticular openings in immature and adult Coleoptera are usually assumed to be of a defensive nature as exocrine defensive glands secreting haemolymph (Bünnige and Hilker, 2005) or for sexual communication purposes (Pierre et al., 1996). Given the observed shape of these structures, their placement on the body and presence in both sexes, it is fair to assume at this point that the nature of these secretory units is defensive.
The position and shape of these structures found in Anamorphidae species resemble those found among members of Boganiidae (Crowson, 1990; Escalona et al., 2015) and Cryptophagidae: Cryptophagini (Leschen, 1996). Crowson (1990) assumed that the presence of prothoracic glandular pores in the same position for both sexes in Boganiidae would make it more likely for them to be defensive than pheromonal. Among the coccinellid groups of Coccinelloidea (sensu Robertson et al., 2015) “reflex bleeding” (secretion of haemolymph) is known among some groups, particularly in Endomychidae and Coccinellidae (Shockley et al., 2009b; Arriaga-Varela et al., 2023a). In those cases the haemolymph is secreted from the tibio-femoral joint and it is produced when the adult or immature beetle gets disturbed. However, the behavioural and biochemical implications of this defensive strategy have only been seriously studied in Coccinellidae (see Knapp et al., 2018). Another coccinelloid family known to have pore openings is Discolomatidae (Cline and Ślipiński, 2010; Szawaryn and Kupryjanowicz, 2019). Most members of this group show tubular openings along the external margin of the pronotum and elytra. These putatively glandular pores have not been studied in depth but they have been assumed to be linked to the myrmecophilous lifestyle of members of this family “in a positive or negative sense” (John, 1959). Nevertheless, behavioural, histological and biochemical studies need to be performed in order to fully understand the nature of the cuticular openings and associated internal modifications in Anamorphidae.
The nature of other, more common, cuticular concavities/pores in Anamorphidae and related groups like Endomychidae still needs to be comprehensively studied. These pores or pits are commonly found in the pronotum, meso- and metaventrites and sometimes in the abdomen. They seem to have a different structure as compared with the putatively glandular ones. Therefore it is unlikely that they have secretory functions. The discovery of fungal spores persistently attached to the setose patches associated with these cavities would suggest that the dispersion of fungal propagules is a direction in which future studies could be focused. However, although the possibility of a mycangial purpose for these pores cannot be discarded (Grebennikov and Leschen, 2010), no clear mechanisms for their use as means of transportation and inoculation of fungal spores is evident as of now.
Conclusions
In this study we have shown, through the phylogenetic analysis of morphological and molecular evidence, that the two species of Palaeosymbius gen. nov. described here belong in Anamorphidae and are closely related to extant Symbiotes and very likely to Asymbius. They represent the oldest member of Anamorphidae known to this date (ca. 99 Myr). Although preliminary, our phylogenetic analyses of Anamorphidae is the most taxonomically comprehensive analysis at the generic level and shows patterns that should be insightful for future studies on the evolution of this family. Palaeosymbius exhibits cavities or pores in the pronotum and elytra, which share characteristics with those seen in Symbiotes and Asymbius. The general structure of these cavities or pores is, however, similar to that of those found in the base of the pronotum, thoracic ventrites or abdomen of many other members of Anamorphidae or Endomychidae. On the other hand, we have discovered that various species of at least three genera in Anamorphidae, Asymbius, Anamorphus and Rhymbomicrus, have a series of small punctiform pores in the anterolateral corner of the pronotum and associated internal structures that resemble an exocrine glandular complex. This represents the first report of glandular pores in the family. Although it is possible that the purpose of these glands is the secretion of substances for defence, their real function is yet to be investigated.
Acknowledgements
We are grateful to Carsten Gröhn (Glinde, Germany), and the curators of the depository institutions for letting us examine the amber pieces containing the described specimens. This research was funded by the National Science Centre of Poland (Narodowe Centrum Nauki; grant no. UMO-2020/36/C/NZ8/00584 to EAV). Additionally, this work was possible owing to the financial support from grants by the National Science Centre of Poland (Narodowe Centrum Nauki; grant no. UMO-2021/43/D/NZ8/02781 to KS) and the Ministry of Culture of the Czech Republic (DKRVO 2024–2028/2.III.a, National Museum, 00023272 to JB). We thank Magdalena Kowalewska-Groszkowska (MIZ) for the help with the scanning electron microscopy illustrations, and Ajay Limaye (Australian National University, Canberra) for rendering the micro-computed tomography reconstruction of the specimen. Dagmara Żyła (Leibniz Institute for the Analysis of Biodiversity Change, Zoological Museum Hamburg), Alexey Solodovnikov (Natural History Museum of Denmark, University of Copenhagen) and Joseph V. McHugh (Collection of Arthropods, Georgia Museum of Natural History, University of Georgia, Athens, USA) are thanked for the loan of specimens used for the study of the morphology of Anamorphidae.
Conflict of interest
None declared.
Open Research
Data availability statement
The authors confirm that the data supporting the findings of this study are available within the article, its supplementary materials and in GenBank Nucleotide Database at https://www.ncbi.nlm.nih.gov/genbank/,accessible under the accession numbers: PP501119-PP501122, PP501444-PP501454, PP505961-PP505966, PP511208-PP511213, PP515079-PP515080, PP528884-PP528887.
