The complete mitogenomes of fruit flies Zeugodacus (Javadacus) calumniatus, Z. (Javadacus) heinrichi and Z. (Sinodacus) hochii have similar gene order and contain 37 genes and a non-coding region. They share an identical start codon for the respective protein-coding genes (PCGs), an identical TAA stop codon for 11 PCGs, TAG for cob, and an incomplete T stop codon for nad5. The cloverleaf structure of most of the tRNAs is similar in the three Zeugodacus species. Phylogenetic analyses reveal Z. (Parasinodacus) cilifer to be external to two main clades: (A) monophyletic subgenus Zeugodacus; and (B) subgenera Javadacus and Sinodacus. The present results indicate that the taxonomic status of some taxa needs clarification. Z. calumniatus is genetically very similar to Z. tau and is not congruent with its current placement in the munda complex. Z. mukiae NC_067083 is genetically very similar to Z. scutellaris, but differs significantly from Z. mukiae MG683384 of the arisanicus (arisanica) complex. On the other hand, Z. proprediaphorus is genetically distinct from and not a synonym of Z. diaphorus. Z. caudatus sensu stricto from Indonesia forms a sister lineage with Z. diversus, instead of with the Malaysian and Chinese taxa of Z. caudatus sensu lato. A notable incongruence is the sister lineage of Z. (Sinodacus) hochii and Z. (Javadacus) heinrichi among other taxa of subgenus Javadacus. A more extensive taxon sampling, particularly the subgenus Sinodacus (and other subgenera), is needed to clarify/resolve their subgenus status.

Fruit fly, mitogenomics, phylogeny, systematics, Zeugodacus subgenera

The fruit fly genus Zeugodacus Hendel, 1927 (considered previously, and still by some researchers, as a subgenus of genus Bactrocera Macquart, 1835) consists of 13 subgenera with some 200 species worldwide (Hancock and Drew 2018a, 2018b). The larvae of many Zeugodacus species have cucurbits as host plants. Some 17 species have been listed as pest of cucurbits: 11 species of fruit pest and six species of flower pest (Doorenweerd et al. 2018). Zeugodacus cucurbitae (Coquillett, 1899), a fruit pest, is the economically most important species. Based on a global checklist, 195 species are found in the Asia-Pacific and one species in Africa (Doorenweerd et al. 2018). New species are, however, continuously being discovered (Yong et al. 2015a; Kunprom and Pramual 2019; Prabhakar et al. 2019; Leblanc et al. 2019; Drew and Romig 2022).

To date, the subgenus names of genus Zeugodacus have not been applied consistently, for example, Z. cucurbitae has been treated as a member of subgenus Javadacus (Hancock and Drew, 2018b; Leblanc, 2022; Starkie et al., 2022) and subgenus Zeugodacus (San Jose et al. 2018; Zhang et al. 2023), Z. tau as subgenus Javadacus (Hancock and Drew 2018b; Starkie et al. 2022) and subgenus Zeugodacus (San Jose et al. 2018; Zhang et al. 2023), and Z. triangularis as subgenus Sinodacus (Starkie et al. 2022; Zhang et al. 2023) and subgenus Zeugodacus (Hancock and Drew 2018b).

Based on molecular phylogenetic analysis, some of the subgenera (as applied by the researchers) within the genus Zeugodacus are recovered as polyphyletic or paraphyletic (San Jose et al. 2018; Starkie et al. 2022; Zhang et al. 2023). However, the assignments of some Zeugodacus species to subgenera are subjected to emendation (Hancock and Drew 2018a, 2018b). Hancock and Drew (2018b) opined that analysing small groups of subgenera separately would enable fine-tuning of the subgeneric limits established so far.

Mitochondrial genomes (mitogenomes) of insects have been extensively studied and applied particularly to studies regarding phylogeny and evolution (Cameron 2014). They have been shown to be suitable for resolving higher-level phylogeny of Paraneopteran insects (Li et al. 2015). Compared to partial sequences of single or multiple mitochondrial and nuclear genes (Yong et al. 2015a; San Jose et al. 2018; Prabhakar et al. 2019; Starkie et al. 2022), there are relatively few studies of the genus Zeugodacus based on complete mitogenomes. To date, the mitogenomes of some 14 species of the genus Zeugodacus (not including the unnamed cryptic species such as in Z. caudatus species complex) are available in the NCBI GenBank. Of these, four species are fruit pests, five species are flower pests, and five species are non-pest. Furthermore, fewer Zeugodacus mitogenomes have been reported compared to genus Bactrocera (Yong et al. 2021; Zhang et al. 2023).

In view of the lack of mitogenomic studies in the genus Zeugodacus and the unresolved systematic status of some taxa, we sequenced and annotated the complete mitogenomes of Z. (Javadacus) calumniatus (Hardy 1970), Z. (Javadacus) heinrichi (Hering, 1941) and Z. (Sinodacus) hochii (Zia, 1936) to determine their genomic features, and phylogenetic relationships with other congeners. Z. calumniatus and Z. heinrichi are non pest, while Z. hochii is a Cucurbitaceae fruit pest (Doorenweerd et al. 2018). At the time of this study (sequencing performed in September 2018), there were no reports on the mitogenomes of these three species. The present study is still the first report on the mitogenomes of Z. calumniatus and Z. heinrichi. These whole mitogenomes will serve as a useful dataset for studying the genetics, systematics and phylogeny of the Zeugodacus genus and subgenera in particular, and tephritid fruit flies in general.

Like other tephritid fruit flies, as well as other insects, the mitogenomes of Z. calumniatus, Z. heinrichi and Z. hochii have the three main clusters of characteristic tRNA genes (Fig. 1): (1) I-Q-M (isoleucine, glutamate and methionine); (2) W-C-Y (tryptophan, cysteine and tyrosine); and (3) A-R-N-S1-E-F (alanine, arginine, asparagine, serine S1, glutamate and phenylalanine) (Cameron 2014). They also have the atypical cloverleaf structure of serine S1 (trnS1), which is common in all Metazoa (Jühling et al. 2012).

The A-T rich control region of the three Zeugodacus mitogenomes possesses both similar and dissimilar features, such as a long poly-A stretch, a long poly-T stretch, tandem repeats and palindromes. Due to its high variability, lack of purifying selection and higher substitution rate, this non-coding control region has been explored for its phylogenetic utility. For example, it has been reported to be of possible phylogenetic utility in some groups of Hemiptera (Li and Liang 2018), a powerful marker for phylogenetic inference in echinoids (Bronstein et al. 2018), and successful for differentiating the BPH (Brown Plant Hopper, Nilaparvata lugens) populations (Anand et al. 2022).

The cox1 gene, with very low Ka/Ks ratio (0.006 to 0.017) in the three Zeugodacus mitogenomes of the present study, representing fewer changes in amino acids, supports its use as a molecular marker for species differentiation and DNA barcoding (Doorenweerd et al. 2020; Lopez-Vaamonde et al. 2021). Genes with very low Ka/Ks ratio, such as cox1, atp6 and cox3 (Fig. 3), reflect the purifying selection that acts on most protein-coding genes. This suggests that any mutations that reduce their function would be quickly eliminated from the population due to their deleterious effects on fitness. In this study, the atp8 gene has a comparatively higher Ka/Ks ratio (Fig. 3). There are similar results in other insect groups, such as the true bugs in which atp8 shows sign of positive selection (Gonçalves et al. 2022).

In the present study, Z. mukiae NC_067083 is genetically very similar to Z. (Zeugodacus) scutellaris, with p = 0.4–0.9% based on 15 mt-genes (Table S7), and p = 0.3–0.4% based on partial cox1 sequences (Table S8), indicating that it may be a misidentification, as it differs from the taxon Z. mukiae MG683384 with p = 12.6% based on partial cox1 sequences (Table S8; Fig. 6). A similar incorrect taxonomic identification has also been inferred for Z. calumniatus, with maximum intraspecific genetic distance of p = 8.8% based on partial cox1 sequences (Kunprom and Pramual 2019).

Previous work has shown that the ‘canonical’ Z. mukiae is a member of the arisanicus (arisanica) complex and not the scutellaris complex (Hancock and Drew 2018b; San Jose et al. 2018). Additionally, it was shown that Z. mukiae MG683384 forms a lineage with Z. trilineatus and Z. arisanicus (San Jose et al. 2018). However, based on partial cox1 sequences of selected taxa (this study), Z. arisanicus is sister to subgenus Zeugodacus and does not form a sister lineage with Z. mukiae and Z. trilineatus (Fig. 6). Z. arisanicus was previously assigned to the subgenus Parazeugodacus (San Jose et al. 2018).

Zeugodacus calumniatus is genetically very similar to Z. tau with p = 0.8% based on 15 mt-genes (Table S7; Fig. 5); the intraspecific genetic distance of Z. tau is p = 0.2–0.7%. Based on partial cox1 sequences, the genetic distance between Z. calumniatus and Z. tau is p = 0.6–1.0% (Table S8; Fig. 6); the intraspecific genetic distance of Z. calumniatus is p = 0.4%. In an earlier study based on partial cox1 sequences from bp 50–700, the intraspecific uncorrected genetic distance of the Z. tau taxa from China, Bangladesh, India (Meghalaya, north of Bangladesh) and Malaysia ranges from p = 0 to p = 0.72% (Yong et al. 2017). In the finding of Kunprom and Pramual (2019), the closest genetic distance between Z. calumniatus from Indonesia and Z. tau is p = 0.2% based on partial cox1 sequences. Z. calumniatus is placed in the munda complex by Hancock and Drew (2018b); it is grouped with the lineage (Z. cucurbitae – Z. tau) based on molecular phylogeny (San Jose et al. 2018).

The taxonomic status of Z. calumniatus needs clarification as it is morphologically very similar to Z. tau (Drew and Romig 2013, 2016); it “is similar in most respects to tau (Walker) and is differentiated by the presence of the prominent brown mark extending over the m crossvein” (Hardy 1974). Nonetheless, in the present study, Z. calumniatus is distinctly separated from the Z. tau taxa in the calumniatus-tau sister lineage of subgenus Javadacus (Fig. 5). There are similar examples of closely related tephritid fruit flies with small genetic distance, for example, Bactrocera carambolae and B. dorsalis (currently accepted as good species) with p = 1.2% based on 15 mt-genes (Yong et al. 2016a).

In the current taxonomic treatment, Zeugodacus proprediaphorus (previously Bactrocera proprediaphora Wang et al., 2008) is synonymised with Zeugodacus diaphorus (previously Bactrocera diaphora) (Drew and Romig 2013). The present phylogenetic analysis reveals that Z. proprediaphorus is genetically distinct from Z. dia­phorus, with a genetic distance of p = 2.9–3.0% based on 15 mt-genes; the intraspecific genetic distance of Z. diaphorus is p = 0.1% (Table S7). This is concordant with the phylogenetic analysis by Wang et al. (2020) which indicates the two taxa to be closely related. An integrative study based on multiple individuals and comprehensive sampling is needed to elucidate the species status of Z. proprediaphorus.

It is noteworthy that Z. diversus forms a sister lineage with Z. caudatus Indonesia in a subclade containing Z. caudatus Malaysia and Z. caudatus China (Fig. 5). Earlier molecular phylogeny has shown the Malaysian and Chinese taxa of Z. caudatus to be genetically very different from and hence not conspecific with Z. caudatus sensu stricto from Indonesia (Yong et al. 2015a, 2016b). Further taxonomic work is needed to formally erect the Malaysian population as a new species.

A notable incongruence in the present study is the grouping of Z. (Sinodacus) hochii with Z. (Javadacus) heinrichi among other taxa of subgenus Javadacus (Fig. 5). An earlier study on molecular phylogeny also indicates the grouping of Z. (Sinodacus) hochii with Z. (Javadacus) heinrichi in the same clade (San Jose et al. 2018). This raises the question whether Z. heinrichi is a member of subgenus Sinodacus, or Z. hochii a member of subgenus Javadacus. Assuming that the subgenus Sinodacus forms a lineage in the same clade as the subgenus Javadacus, the subgenus status of Z. (Javadacus) depressus (the external taxon to the other taxa in Clade B) also needs clarification as it has been assigned to subgenus Paradacus Perkins, 1938 in some studies (see Jeong et al. 2017). The recent study of Zhang et al. (2023) on the mitogenomes of tephritid fruit flies has recovered paraphyletic/polyphyletic subgenera within the genus Zeugodacus. Our present results add to these inconsistencies. A more extensive taxon sampling, particularly the subgenus Sinodacus (and other subgenera), is needed to clarify/resolve their subgenus status. Independent sources of information from across the genome (e.g. independent nuclear genes) are also important to confirm/establish their taxonomic relationships.

In summary, we have successfully sequenced and annotated the whole mitochondrial genomes of Z. (Javadacus) calumniatus, Z. (Javadacus) heinrichi and Z. (Sinodacus) hochii. The genome features are similar in the three species. Phylogenetic analysis based on the mt-genes reveals two major clades of the Zeugodacus taxa: (A) monophyletic subgenus Zeugodacus, and (B) subgenera Javadacus and Sinodacus; Z. (Parasinodacus) cilifer is external to the two main clades. It reveals the incongruence of Z. (Sinodacus) hochii forming a sister lineage with Z. (Javadacus) heinrichi. It also indicates the need to clarify the taxonomic status of Z. mukiae NC_067083 and Z. calumniatus. On the other hand, the results indicate the possible valid species status of Z. proprediaphorus (genetically distinct from and likely not a synonym of Z. dia­phorus). A broad taxon sampling of subgenus Sinodacus and other subgenera will help to clarify their taxonomic status and phylogeny.

We thank our respective institutions for their support of our research on tephritid fruit flies. H-S Yong is supported by MoHE-HIR Grant (H-50001-00-A000025) and Universiti Malaya (H-5620009). Our thanks also go to Dr. Leonardo Gonçalves and another reviewer for their valuable suggestions in improving the manuscript.