Phylogenetic position and morphological polymorphism of the chafer, Clinterocera nigra (Coleoptera: Scarabaeidae: Cetoniinae) from Taiwan.

Three mitochondrial genomes of the cetoniine beetle, Clinterocera nigra (Kano, 1931) were assembled via next-generation sequencing. The newly sequenced mitogenomes all have 37 genes, showing standard gene order and annotation as the other insects. To examine their phylogenetic positions and relationships between their elytral color (red-spot and melanistic forms) and sequence variation, a total of 118 public mitogenomes of Scarabaeidae were used to infer a maximum-likelihood (ML) tree. Our results show that the melanistic form is grouped within red-spot ones, revealing a population level variation on the elytra color. Our work also provides the first mitogenomic reference of myrmecophilous chafers.

The myrmecophilous genus, Clinterocera Motschulsky, 1858 (Coleoptera: Scarabaeidae: Cetoniinae: Cremastocheilini), is distributed in Southeast Asia, consisting of 29 species (Xu et al. 2018). Their diagnostic features have been distinctly characterized by reduced tarsomeres, triangularly enlarged antennal scape and cup-shaped prementum which might be adapted to symbiosis with ants (Qiu and Xu 2016). Kano (1926) first reported Callinomes davidis Fairmaire, 1878 from Taiwan, and it was lately transferred to Clinterocera by Medvedev (1964). The species can be easily recognized by well-developed red spots on elytra. However, there was a subspecies, Cl. nigra, named by Kano (1931) with the elytral red spots largely reduced (Figure 1). This taxonomic arrangement was followed and has never been challenged (i.e. Sakai and Nagai 1998; Smetana 2006; Bezděk 2016) until Xu et al. (2018) restricted the distribution of Cl. davidis in mainland China and raised Cl. d. nigra to specific rank. We herein conducted a mitogenomic comparison with three Clinterocera specimens in different color form from Taiwan.

Figure 1.

The ML phylogeny of the Scarabaeidae based on 118 mitogenomes using IQ-TREE. The dorsal view of red-spot and melanistic forms of Clinterocera nigra is shown. Nodal supports of SH-aLRT/ML: SH-like approximate likelihood ratio test (Guindon et al. 2010)/ML ultrafast bootstrapping (Minh et al. 2013).

For evaluating mitogenomic variation, three specimens were collected in this work: two red elytral spots specimens of Cl. nigra were collected from Taipei (coordinate: N: 24.9928, E: 121.5484; DNA code: 20CL07001) and Xitou, Nantou (N: 23.6738, E: 120.7969; DNA code: 20CL07002); the melanistic specimen was collected at its type locality (Neimoupu, Nantou; N: 23.6898, E: 120.8505; DNA code: 20CL07003). Genomic DNAs were extracted using Gentra Puregene Tissue Kit (Gentra Systems, Minneapolis, MN) and all the extracts were fragmented into 200–300 bp for next-generation sequencing via Illumina Novaseq 6000 platform (San Diego, CA).

Each NGS dataset was trimmed out low-quality region (

Cl. nigra, a total of 118 mitogenomes, comprising three our samples and 115 public scarabaeids, were used to infer their phylogenetic relationships. Sequences were separately aligned based on gene region using MUSCLE (Edgar 2004), implied in MEGA-X (Kumar et al. 2018). All the aligned genes were concatenated (15,521 bp in length) into a sequence matrix, and manually formatted the matrix as phylip or nexus formats for downstream analysis. The Hybosorus sp. (Hybosoridae) was set to outgroup and 16 partitions (13 protein-coding genes, rrnL, rrnS, and tRNAs) with GTR + G model were set. The maximum-likelihood (ML) phylogeny was reconstructed using IQ-TREE (Nguyen et al. 2015), and the nodal supports were evaluated by 1000 replicates of bootstrapping and the SH-like approximate likelihood ratio test (SH-aLRT) (Guindon et al. 2010).

The phylogenetic relationships show that three Cl. nigra individuals are strongly grouped in a clade, sister to Dicronocephalus under the tribe Cremastocheilini (Figure 1). The result is consistent with the previous study (Šípek et al. 2016). Meanwhile, the melanistic Cl. nigra (20CL07003) are grouped within the other two red-spot samples (Figure 1), even genetic distance between melanistic and red-spot forms (K2P genetic distance: 0.3% and 0.9%) is smaller than the distances between two red-spot samples (K2P genetic distance: 1.2%), supporting the taxonomic treatment by Xu et al. (2018).