An exquisitely preserved tiny bark-gnawing beetle (Coleoptera: Trogossitidae) from mid-Cretaceous Burmese amber and the phylogeny of Trogossitidae

2.1 Systematic paleontology

Order Coleoptera Linnaeus, 1758

Suborder Polyphaga Emery, 1886

Superfamily Cleroidea Latreille, 1802

Family Trogossitidae Latreille, 1802

Microtrogossita Li & Cai gen. nov.

Microtrogossita qizhihaoi Li & Cai sp. nov.

http://zoobank.org/urn:lsid:zoobank.org:act:D7A3BC9B-AD3B-4A39-A800-E6C55479BF89

2.2 Morphological phylogenetic analysis

The morphological character matrix was modified based on Kolibáč (2006, 2008), who conducted a phylogenetic analysis of 44 genera of the former “Trogossitidae.” Recent molecular studies have demonstrated that the traditional “Trogossitidae” is actually polyphyletic (Gimmel et al., 2019; Kolibáč et al., 2021; McKenna et al., 2019; Zhang et al., 2018). Thus, this morphological matrix cannot appropriately resolve the higher-level relationship between traditional trogossitid taxa. Because of above-mentioned progress in a study of Cleroidea, the goal of our analysis is merely to assess a placement of our fossil within the modern Trogossitidae.

Parsimony analyses using implied weighting (K = 12) yielded a single most parsimonious tree (Figure 5). The overall support values of key nodes of the tree were quite low, suggesting morphological information alone is not sufficient to solve the phylogeny of Trogossitidae. The traditional “Peltinae” was resolved as a basal paraphyletic grade, and the traditional “Lophocaterinae” appeared to be a monophyletic clade. The majority of modern Trogossitidae (not including Calityini) formed a monophyletic clade. The newly discovered fossil, M. qizhihaoi, was nested within the extant tribe Trogossitini (Trogossitidae), and sister to all other trogossitids except Alindria, though with low statistical support.

2.3 Molecular phylogenetic analysis

We focused on quantifying uncertainty in intertribal relationships in Trogossitidae (Trogossitinae sensu Kolibáč, 2013) by testing the robustness of Gimmel et al.'s (2019) recovered topology to different analytical methods. Specifically, we tested the effect of using a site-heterogeneous model to analyze the four-gene dataset. While more computationally intensive than the site-homogeneous models applied to the study of cleroid phylogeny in the past, site-heterogeneous models account for the unequal rate of evolution in sequences and yield phylogenies more robust to phylogenetic artifacts such as long branch attraction (Cai et al., 2020; Lartillot et al., 2007; Pisani et al., 2015).

We recovered a strongly supported clade (Bayesian Posterior Probabilities [BPP] = 0.91) consisting of Egoliini, Calityini, Gymnochilini, and Trogossitini (Figure 6). The Egolia + Calitys clade was well supported as sister to Gymnochilini and Trogossitini. In line with Gimmel et al. (2019), we recovered Gymnochilini nested within Trogossitini. The phylogenetic position of the fifth tribe traditionally associated with Trogossitinae, Larinotini, was poorly supported in the present analysis (BPP = 0.59) and recovered as sister to the genus Peltis (Peltidae).

3.1 Systematic position of Microtrogossita and comparison with extant relatives

A classification of the new genus within Trogossitidae is chiefly based on combination of (a) general habitus (elongate body), (b) pronotum with lateral carina, (c) asymmetrical antennal club, (d) all pairs of legs with ultimate tarsomeres as long as previous tarsomeres together, (e) projected bisetose empodium situated between simple tarsal claws, and (f) procoxal cavities externally closed (Kolibáč, 2013; Lawrence et al., 2014).

Microtrogossita gen. nov. further shares some morphological similarities with the tribe Trogossitini in (a) elongate and moderately convex body, (b) the absence of scales or tufts of setae on dorsal surface, (c) elytra with regular rows of punctures ordered in striae, (d) the absence of frontoclypeal suture and structure of anterior margin of head with reduced clypeal area, and (e) pronotum without projected anterior corners. This affinity between Microtrogossita and Trogossitini was also supported by our cladistic analysis (Figure 5).

On the other hand, Microtrogossita differs from all known trogossitid genera in (a) weakly asymmetrical club (in contrast to the distinctly asymmetrical club: figure 9C, D in Kolibáč et al., 2021; figure 14G in Kolibáč, 2013), (b) relatively wide separation of meso- and procoxal cavities (in contrast to the narrower separation: figure 1A in Kolibáč, 2013), (c) extremely coarse sculpture of dorsal and even ventral surface. Also, the absence of spines along outer margin of tibiae is rare within Trogossitidae. Considering its unique character combination in Trogossitidae, a more basal position for Microtrogossita cannot be confidently ruled out.

Currently, there are 13 extant genera in the tribe Trogossitini (Kolibáč, 2013). Microtrogossita shares shape of head (especially structure of anterior margin of clypeus; Figure 3c) with the most of trogossitids excepting Nemozoma, Corticotomus, and partly also Euschaefferia. The latter genera possess frons with pair of distinctly projecting horns which is believed to assist movement in narrow galleries inside wood (Hinson & Buss, 2016; Kolibáč, 2014a). The antennae of Microtrogossita only have 10 antennomeres (Figures 3e and 4d), which are rare in Trogossitini. The majority of Trogossitini (except for some species in Nemozoma) all possess 11-segmented antennae. Pronotum without distinctly extended anterior corners differs Microtrogossita from majority of trogossitids but the feature is shared with several genera including Airora, Parallelodera, Nemozoma, and Corticotomus. However, widened mid-part with narrowed anterior and posterior parts is exceptional (on contrary to cordate pronotum of, e.g., Tenebroides or some Temnoscheila). Coarsely facetted eyes of Microtrogossita may refer to hidden way of life, for example, under bark, in wood-borers galleries or even in litter as documented in various Cleroidea including Afrocyrona (Kolibáč, 2014b), Parapeltis, and Colydiopeltis (Ślipiński, 1992). In contrary, rapid flyers and surface dwellers among Trogossitidae (e.g., Anacypta, Xenoglena, Elestora, Parallelodera) and Lophocateridae (Ancyrona) often hunt for other insects in full sunlight on tree bark, branches, or leaves—such species mostly have eyes finely facetted (Kolibáč, 2013). Deep punctures in both dorsal and ventral surfaces as well as distinct striae in elytra also differ Microtrogossita from the extant trogossitids.

3.2 Notes on the intertribal relationships in Trogossitidae

Traditionally, five tribes were included in the subfamily Trogossitinae (or Trogossitidae in the modern sense): Calityini, Egoliini, Larinotini, Gymnochilini, and Trogossitini (e.g., Kolibáč, 2013). This arrangement was supported by the morphological phylogenetic analyses of Kolibáč (2006, 2008). However, in the molecular phylogeny of Gimmel et al. (2019), Calityini, Egoliini, and Larinotini did not form a clade together with Gymnochilini and Trogossitini, but rather occupied a position closer to Peltidae. Though Calityini indeed display a mixture of both peltid and trogossitid characters, the larvae of Egoliini clearly show an affinity to trogossitids (Kolibáč, 2013). In our re-analysis of Gimmel et al.'s (2019) data with the site-heterogeneous model CAT-GTR+G4, Calityini and Egoliini clustered with Gymnochilini and Trogossitini (Figure 6). Thus, the placement of Calityini and Egoliini remains yet to be convincingly resolved. Both Gymnochilini and Trogossitini were recovered as non-monophyletic in both the original analysis by Gimmel et al. (2019) and our re-analysis, indicating that their taxonomic status requires further attention.

3.3 Mesolophocateres Yu, Leschen & Ślipiński as a junior synonym of Burmacateres Kolibáč & Peris

Kolibáč and Peris (2021) reported the first Lophocateridae species from the mid-Cretaceous Burmese amber, Burmacateres longicoxa Kolibáč & Peris, 2020 (Figure 7, Figures S1 and S2). Burmacateres is characterized by 11-segmented antennae (Figure 7a), extremely enlarged procoxae (Figure 7b), and dorsal surface with tufts of hairs (Figure 7c). A few weeks later, Yu et al. (2021) described another two lophocaterids from Burmese amber, Mesolophocateres pengweii Yu, Leschen & Ślipiński, 2020 and Parayixianteres parvus Yu, Leschen & Ślipiński, 2020, in the same journal. However, the description and photographs of M. pengweii match well with those of B. longicoxa. Actually, we are unable to detect any significant differences between the two species. Thus, we herein suggest that Mesolophocateres syn. nov. is a junior synonym of Burmacateres, and M. pengweii syn. nov. is a junior synonym of B. longicoxa.

4.1 Materials

The Burmese amber specimens studied herein originated from amber mines near Noije Bum (26°20′N, 96°36′E), Hukawng Valley, Kachin State, northern Myanmar. The specimens are deposited in the Nanjing Institute of Geology and Palaeontology (NIGP), Chinese Academy of Sciences, Nanjing, China. The amber pieces were trimmed with a small table saw, ground with emery papers of different grit sizes, and finally polished with polishing powder.

4.2 Fossil imaging

Photographs under incident light were taken with a Zeiss Discovery V20 stereo microscope. Widefield fluorescence images were captured with a Zeiss Axio Imager 2 light microscope combined with a fluorescence imaging system. Confocal images were obtained with a Zeiss LSM710 confocal laser scanning microscope, using the 488-nm Argon laser excitation line (Fu et al., 2021). Images under incident light and widefield fluorescence were stacked in Helicon Focus 7.0.2 or Zerene Stacker 1.04. Confocal images were stacked with color coding for depth in ZEN 2.3 (Blue Edition). The holotype of M. qizhihaoi was also imaged using high-resolution X-ray microtomography (Zeiss Xradia 520 Versa) at the micro-CT laboratory of NIGP. Due to the comparatively small size of the fossil specimen, a CCD-based 4× objective was used, providing isotropic voxel sizes of 2.4165 μm with the help of geometric magnification. During the scanning, the acceleration voltage for the X-ray source was 50 kV. To improve the signal-to-noise ratio, 3001 projections over 360° were collected, and the exposure time for each projection was 1.5 s. The tomographic data were analyzed using VGStudio MAX 3.0. Images were further processed in Adobe Photoshop CC to enhance contrast.

The original confocal and micro-CT data are available in Zenodo repository (https://doi.org/10.5281/zenodo.3994902).

4.3 Measurements and description

The measurements were generally based on the micro-CT reconstruction. The measurements were taken as follows: body length as distance from mandibular apex to elytral apex when body fully stretched; body width as maximum width of body (at the middle part of elytra); head length as distance from mandibular apex to anterior margin of pronotum (from dorsal view); head width as maximum width of head across eyes; pronotal length as maximum length of pronotum; pronotal width as maximum width of pronotum; and elytral length as distance from anterior margin to apex.

The morphological terminology follows Lawrence and Ślipiński (2013) and Kolibáč (2005, 2013).

4.4 Morphological phylogenetic analysis

To evaluate the systematic placement of the new species, a morphological phylogenetic analysis was performed. The data matrix (Data S1 and S2) was mainly derived from a previously published dataset (Kolibáč, 2008). The coding of character 36 (shape of pronotum) in Kolibáč (2008) was problematic. The states “transverse” and “elongate” refer to the ratio of length and width, while the state “cordate” means the pronotum is narrowed toward base. A species may match the criteria of both “transverse” (or “elongate”) and “cordate” at the same time. Thus, this character was split into two characters in our analysis. The character 32 (shape of antennal club) in Kolibáč (2008) was also split into two characters. In addition, a new informative character, the shape of pronotal carina, was added. All characters were treated as non-additive. Rentonium and Rentonellum were selected as the out-group taxa, based on previous molecular results (Gimmel et al., 2019; McKenna et al., 2019). Parsimony analysis was performed under implied weights using the program TNT 1.5 (Goloboff & Catalano, 2016; Goloboff et al., 2008). Parsimony analysis has been shown to achieve the highest accuracy under a moderate weighting scheme (e.g., when concavity constants K are between 5 and 20) (Goloboff et al., 2018; Smith, 2019). Therefore, the concavity constant was set to 12 here, as suggested by Goloboff et al. (2018). Standard bootstrap analysis was implemented by 1000 pseudoreplicates, where the support values were shown as frequency differences (Goloboff et al., 2003). The tree was drawn with the online tool iTOL 5.7 (Letunic & Bork, 2021) and graphically edited with Adobe Illustrator CC 2017.

4.5 Molecular phylogenetic analysis

To test the relationships among the tribes of Trogossitidae, we used the most comprehensive dataset available for the group published to date, including all the currently recognized cleroid families, compiled by Gimmel et al. (2019). Four genes were included, namely the nuclear genes 18S and 28S, and the mitochondrial CYTB and COX1. All sequences were obtained from GenBank using the Batch Entrez tool. The accession numbers were the same as provided by Gimmel et al. (2019). Mitochondrial genes were aligned in MEGA X 10.0.5 with the “MUSCLE” tool (Kumar et al., 2018), while 18S and 28S were aligned in the MAFFT 7.313 plug-in in PhyloSuite 1.2.1 (Zhang et al., 2020) using the Q-INS-I algorithm (Katoh et al., 2002; Katoh & Toh, 2008). The CYTB sequence for Chaetosoma scaritides could not be unambiguously aligned and was excluded. The site-heterogeneous mixture model CAT-GTR+G4 was run in PhyloBayes mpi 1.7 (Lartillot et al., 2009). Two independent Markov chain Monte Carlo (MCMC) chains were run until convergence (maxdiff <0.3). Convergence was assessed by using the bpcomp program to generate output of the largest (maxdiff) and mean (meandiff) discrepancies observed across all bipartitions.

All data needed to run the molecular analysis and the output files are available in Zenodo repository (https://doi.org/10.5281/zenodo.4737037).

4.6 Nomenclatural acts

This published work and the nomenclatural acts it contains have been registered in ZooBank. The LSID for this publication is http://zoobank.org/urn:lsid:zoobank.org:pub:3694913C-FA40-4520-88F8-3C42E441EDC6.