Introduction

Complete mitochondrial genomes (mitogenomes) have been widely used to infer phylogenetic relationships specifically for wide-ranging groups of organisms such as insects. Mitogenomes have been considerably used as molecular tools for phylogenetic investigations, and comparative and evolutionary relationship studies1,2,3. The insect mitogenomes are relatively small in size, show rapid evolution rates, have low level of recombination, and possess maternal inheritance4,5,6,7. Therefore, the utilization of mitogenome is expected to provide novel information concerning the classification of insects and assessments of their evolutionary features.

The insect mitochondrial genome is typically a double-stranded, circular molecule that is 14–19 kb in length and is composed of 13 protein-coding genes (PCGs): two ATPase genes (atp6 and atp8), three cytochrome c oxidase genes 1–3 (cox1-cox3), one cytochrome B (cob), seven NADH dehydrogenase genes (nad1-6 and nad4L), 22 transfer RNA (tRNA), two ribosomal RNA (rrnL and rrnS) genes and non-coding A + T- rich region2,5,8,9,10.

There are about 1,57,424 described moth species worldwide, from 45 superfamilies, and belonging to 139 families11. Analysis of the phylogenetic relationships of lepidopteran moths using mitogenomes has increased rapidly during the last decade12. However, the primary focus has been on 11 moth superfamilies namely Bombycoidea, Cossoidea, Gelechioidea, Geometroidea, Hepialoidea, Noctuoidea, Pyraloidea, Tineoidea, Tortricoidea, Yponomeutoidea, and Zygaenoidea.

Noctuoidea is the largest superfamily of the order Lepidoptera, comprising 42, 407 species11. The monophyly of Noctuoidea is supported by the existence of a gained apomorphic character, metathoracic tympanal organ13. Phylogenetic studies of Noctuoidea were primarily analyzed using molecular methods based on one or two genes and with limited taxon sampling14,15,16,17. The molecular phylogenetic relationship of Noctuoidea has been analyzed based on single mitochondrial (cox1) and seven nuclear genes (EF-1α, wingless, RpS5, IDH, MDH, GAPDH, and CAD) from 152 species18. Zahiri et al.18 have proposed a novel perception, separating the traditional group of quadrifid noctuids, and re-establishing Erebidae and Nolidae as families. This result contrasted meaningfully with previous investigations of both morphological and molecular studies. Nevertheless, this analysis failed to resolve the phylogenetic relationships between Erebidae subfamilies19. The phylogenetic relationships of the family Erebidae were analyzed using the mitogenomes20,21,22,23,24,25,26,27,28,29,30,31,32,33,34. Several mitogenome sequencing studies were carried out in the family Noctuidae for their phylogenetic utilization[35,36,37,38,39,40,41,42] and single mitogenomic analysis was done in the family Euteliidae12.

Hyblaeoidea is among the smallest superfamilies in the order Lepidoptera and consists of only one family Hyblaeidae comprising only two genera Hyblaea and Erythrochrus. The family Hyblaeidae contains only 20 species distributed all over the new and old-world tropics and subtropics. The genus Hyblaea is known as a serious forest pest43. Twort et al.44 have done the whole genome sequence and analyzed the phylogenetic relationships of Hyblaea puera and Hyblaea madagascariensis. They used the dataset of 162 taxa for analyses, which showed the stable placement of Hyblaea as sister to the Pyraloidea member group with strongly supported values.

At present, 84 complete mitogenomes of Noctuoidea from eight families have been deposited in GenBank12,20,21,24,25,26,28,30,31,36,37,39,41,45,46,47,48,49,50. In the present study, we sequenced the complete mitochondrial genomes of four species representing superfamilies Noctuoidea (Oraesia emarginata, Actinotia polyodon, Odontodes seranensis) and Hyblaeoidea (Hyblaea puera) for the first time. The mitogenomes of these species were annotated and the general characteristics of the mitogenome sequences were analyzed and compared. We analyzed phylogenetic relationships of mitogenomes from 90 lepidopteran taxa. In addition, the phylogenetic tree was reconstructed using the maximum likelihood method and Bayesian inference to evaluate the relationships among the lepidopteran moths.

Results and discussion

Genome organization and base composition

In this study, we sequenced and characterized the complete mitogenomes of three Noctuoidea species; Oraesia emarginata (GenBank Accession no. MW648382), Actinotia polyodon (GenBank Accession no. MW697903), Odontodes seranensis (GenBank Accession no. MW719565) and one Hyblaeoidea species Hyblaea puera (GenBank Accession no. MW885970). The sequences were deposited in GenBank (Table 1). The total lengths of the mitogenomes of O. emarginata, A. polyodon, O. seranensis, and H. puera were 16, 668 bp, 15, 347 bp, 15, 419 bp, and 15, 350 bp, respectively. The sizes of mitogenome sequenced so far in the superfamily Noctuoidea ranged from 15, 229 bp in Helicoverpa gelotopoeon to 16, 346 bp in Spodoptera frugiperda. The mitogenome sequence lengths of A. polyodon, O. seranensis, and H. puera fell within the range of mitogenome of other sequenced Noctuoid moths. However, the mitogenome of Oraesia emarginata was larger than that of Gynaephora jiuzhiensis. The organization of newly sequenced mitogenomes of four species are presented in Fig. 1.

Table 1 List of the complete mitogenome of the superfamily Noctuoidea reported so far.
Full size table
Figure 1
figure 1

Circular maps of the newly sequenced complete mitochondrial genomes of (A) Actinotia polyodon, (B) Odontodes seranensis, (C) Oraesia emarginata, (D) Hyblaea puera.

Full size image

All four species comprised a distinctive metazoan mitogenome composition of 13 protein-coding genes PCGs viz. ATPase subunits 6 and 8 (atp6 and atp8), cytochrome c oxidase subunits 1–3 (cox1, cox2 and cox3), NADH dehydrogenase subunits 1–6 (nad1, nad2, nad3, nad4, nad5 and nad6), subunit 4L nad (nad4l) and cytochrome B (cob), 22 transfer RNA (tRNA) genes, two ribosomal genes (rrnL and rrnS) and a control region (A + T-rich region). Four of the thirteen PCGs (nad1, nad4, nad4l, and nad5) and eight tRNAs (trnQ, trnC, trnY, trnF, trnH, trnP, trnL1 and trnV) and two rRNAs were encoded on the N-strand, whereas the other 23 genes (9 PCGs and 14 tRNAs) and the control-region (A + T-rich) were encoded on the J-strand (Table 2). All genes were organized in the same way without the rearrangement phenomenon.

Table 2 Details on gene organization of four newly determined Lepidoptera mitogenomes.
Full size table

The nucleotide compositions of the four moth mitogenomes had a high A + T bias: 79.72% in O. emarginata, 81.69% in A. polyodon, 81.09% in O. seranensis, and 81.21% in H. puera. Among the 88 Noctuoid species for which mtDNA data was available, the lowest A + T content was 77.83% in O. lunifer, while the highest A + T content was 81.69% in Gabala argentata. The mitogenome of A. polyodon was also highest among the known Noctuoid mitogenomes. All four mitogenomes showed a negative AT-skew on the majority strand and negative GC-skew as it occurred mostly among other Noctuoid mitogenomes. The AT and GC skew values on the majority strand of the four moths’ species are O. emarginata (−0.002 and −0.200), A. polyodon (−0.008 and −0.167), O. seranensis (−0.023 and −0.196), and H. puera (−0.000 and −0.178) (Supplementary Table 1). Similar patterns of nucleotide negative skew have also been found in the mitogenomes of other Noctuoid taxa36,45,46,49.

Protein-coding genes and codon usage

The total lengths of the 13 PCGs of O. emarginata, A. polyodon, O. seranensis and H. puera were 11, 182 bp, 11, 213 bp, 11, 208 bp, and 11, 195 bp accounting for 67.08%, 73.06%, 72.68%, and 72.93% of the mitogenomes respectively. The locations and orientations of the 13 PCGs within the four mitogenomes were identical to those of most Noctuoid species. The nucleotide PCGs translated into 3716–3725 amino acid-coding codons, excluding the stop codons. Similar to the PCGs nad5 and atp8 were observed to be the largest (1727–1746 bp) and smallest (162–165 bp) genes, respectively. The majority of PCGs stringently started with an ATN (ATG/ATT/ATA) start codon, except cox1 gene which started with CGA in O. emarginata, A. polyodon and H. puera and with TTG in O. seranensis (Table 1). The majority of PCGs terminated with a complete and canonical stop codon (TAA/TAG) except in O. emarginata where gene nad5 terminated with TTA. The genes cox2, nad4 (O. emarginata, A. polyodon), cox1, cox2, nad4 (O. seranensis), cox1, cox2, and nad5 (H. puera) were found to have a truncated termination codon (T) and it might be altered by post-transcriptional polyadenylation. The presence of an incomplete stop codon was also a common phenomenon in metazoan mitochondrial genes (Sheffield et al. 2010). The average A + T contents of the 13 PCGs within the four mitogenomes ranged from 78.85 to 80.36% (Supplementary Table 1).

Relative synonymous codon usage (RSCU) was calculated in the mitogenomes of four lepidopterans (Fig. 2 and Supplementary Table 2). The most frequently utilized codons were almost similar within the four Noctuoid species. UUU (Phe), UUA (Leu), AUU (Ile), AAU (Asn) and AAA (Lys) were the most consistently used codons (> 232) within the PCGs of the four mitogenomes; however, GUG (V), ACG (Thr), CCG (Pro), GCC (Ala), and UGC (Cys) were the smallest used codons (< 10). We found average relative synonymous codons of 3737 (A. polyodon), 3736 (O. seranensis), 3 727 (O. emarginata) and 3731 (H. puera), not including stop codons, that were predicted for codon usage of all the four mitogenomes.

Figure 2
figure 2

Relative Synonymous Codon Usage (RSCU) in Noctuoidea PCGs. The species name represents the superfamilies Noctuoidea & Hyblaeoidea. RSCU of Actinotia polyodon, RSCU of Odontodes seranensis, RSCU of Oraesia emarginata and RSCU of Hyblaea puera.

Full size image

Overlapping and intergenic spacer regions

The intergenic sequences and overlapping regions were analyzed. The small intergenic spacers (IGS) were seen ranging in size 1–63 bp, and totalling 174 bp in A. polyodon, 256 bp in O. seranensis, 527 bp in O. emarginata and 262 bp in H. puera (Table 2). The longest intergenic spacer was located between tRNAGln and nad2 with the length of 48 bp in H. puera, 56 bp in A. polyodon, 57 bp in O. seranensis, and 63 bp in O. emarginata.

In addition, the overlapping regions were also analyzed. The numbers of overlapping regions in four Noctuoid moths were mostly inconstant from 1 to 45 bp. However, the longest overlapping region was located between rrnL and rRNAVal, with a length of 45 bp in A. polyodon (Table 2).

Transfer RNA genes (tRNA)

The mitogenomes of H. puera, A. polyodon, O. seranensis, O. emarginata had 22 tRNA genes (Table 2). The total lengths of the 22 tRNA genes were 1479 bp (O. emarginata), 1472 bp (A. polyodon), 1476 bp (O. seranensis) and 1471 bp (H. puera); however individual tRNA genes typically ranged from 65 to 71 bp, among which, eight tRNAs were encoded on the N-strand and the remaining 14 were encoded on the J-strand. The putative secondary structures of tRNA genes recognized in these Noctuoid mitogenomes are given in Supplementary Figs. 1–4. All the predicted tRNAs revealed the typical putative secondary structure except for trnS1 (AGN) where dihydrouridine (DHU) arm lacked and formed a simple loop which has been found in many lepidopterans22,30,39,40. The lack of dihydrouridine (DHU) arm in trnS1 (AGN) was observed in all species, while TΨC arm disappeared only in trnE of O. emarginata. In addition, TΨC loop was seen lacking in trnY and trnF of O. emarginata, and trnI, trnT, trnF of H. puera. Several mismatching base pairs occurred in tRNA clover-leaf secondary structures in all four lepidopteran mitogenomes. A total of 16 mismatches of 9 U-G and 7 G-U wobble pairs were noticed in 13 tRNA genes of A. polyodon; 18 mismatches (1 U-U and 9 U-G) and 8 G-U wobble pairs were detected in 15 tRNA genes of O. seranensis; 20 mismatches (1 A-A, 3 U-U and 10 U-G) and 6 G-U wobble pairs were observed in 13 tRNA genes of H. puera; 18 mismatches (1 U-U and 10 U-G) and 7 G-U wobble pairs were observed in 14 tRNA genes of O. emarginata.

Ribosomal RNA genes

Two rRNA genes (rrnL and rrnS) are extremely conserved in Noctuoid mitogenomes, and each of the four mitogenomes contained these two rRNA genes. rrnL gene lengths were 1370 bp for A. polyodon, 1336 bp for O. seranensis, 1276 bp for H. puera and 1317 bp for O. emarginata; whereas rrnS were 777, 781, 762, and 738 bp (Supplementary Table 1). The rRNA genes of currently sequenced mitogenomes displayed a negative AT skew (−0.007 to −0.036) and GC skew (−0. 356 to −0. 483). rrnL gene was located between trnL1 and trnV, and rrnS was located between trnV and the control region (Table 2).

The A + T-rich region

The A + T-rich regions of O. emarginata, A. polyodon, O. seranensis, and H. puera were 287, 259, 343 and 439 bp in size respectively, all positioned between the rrnS and tRNAMet (Table 2). The A + T content of these regions was 93.38%, 94.72%, 95.35% and 97.25%, respectively (Supplementary Table 1). The A + T-rich regions exhibited negative AT and GC-skew values. The conserved structure that connected the motif “ATAGA + poly-T stretch” was located downstream of the rrnS gene in the A + T-rich region of H. puera, A. polyodon and O. seranensis, which was not observed in O. emarginata (Fig. 3). We found that the motif ‘ATAGA’, might be the origin of light-strand replication51, directly connecting to the poly-T stretch in A. polyodon instead of (A)n which connected to the poly-T structure in H. puera and O. seranensis. Multiple tandem repeats are naturally existing in the A + T-rich region of most lepidopteran insects. We detected the presence of tandem repeats in the mitochondrial A + T-rich region in A. polyodon, O. seranensis and H. puera, but not in O. emarginata. The A + T-rich region of H. puera consisted of three tandem repeats each of size 128 bp, 113 bp and 108 bp. Only one tandem repeat (103 bp) was found in the A + T-rich region of A. polyodon. In O. seranensis, the A + T-rich region consisted of two tandem repeats (57 bp and 47 bp). However, the tandem repeat was not observed in the A + T-rich region of O. emarginata; similarly, the tandem was also not present in Dysgonia stuposa32.

Figure 3
figure 3

Alignment of initiation site for A + T-rich region of 16 species completely sequenced lepidopteran mitogenomes. The boxed nucleotides indicate the conserved motif ATAGA and the shaded nucleotides indicate poly-T stretch. * newly sequenced mitogenomes presented in this study.

Full size image

Additionally, two dinucleotide microsatellites and three motifs were detected in H. puera, referred to as (TA)10, (TA)3 in repeat 5, motif (ATAGA)2, (ATTTA)16, and TAATAATAA. In A. polyodon, one dinucleotide microsatellite (TA)7 and three motifs (ATTTA)5, (ATAGA)1 and TAATAATAA were also observed. One dinucleotide (TA)3, one trinucleotide (TAATAATAA)2 microsatellites, and three motifs (ATAGA)2, (ATTTA)4 and (ATATTA)10 were found in O. seranensis. Similarly, one dinucleotide microsatellite (TA)10 and two motifs (ATTTA)4 and (ATATTA)3 were found in O. emarginata. Furthermore, the ‘ATCTAA’ block in H. puera upstream of the origin of light-strand replication was different from the ‘ATACAA’ block in A. polyodon (Fig. 4).

Figure 4
figure 4

Motifs and microsatellites found in the A + T-rich region of Odontodes seranensis, Actinotia polyodon, Hyblaea puera and Oraesia emarginata. These are indicated by specific colours and highlights. Motifs (ATAGA) are shown in dark blue high lights. Poly-T stretch are shown in darker gold accent highlights. Microsatellites (ATATTA) are shown in pink highlights. Microsatellites (ATTTA) are shown in green highlight. All tandem repeats are underlined. (ATACAA) block is shown in light blue highlight. Microsatellite (TA)10 and (TA)7 are shown in yellow colour.

Full size image

Phylogenetic relationships

We performed the phylogenetic study on the mitogenomes of 99 lepidopteran species representing five Noctuoid families (Erebidae, Euteliidae, Noctuidae, Nolidae, Notodontidae), one Hyblaeoid (Hyblaeidae), one Pyraloid (Crambidae), one Geometroid (Geometridae), three Bombycoid (Sphingidae, Saturniidae, Bombycidae), and one Lasiocampoid (Lasiocampidae) with two outgroup species (Papilio polytes and Trogonoptera brookiana) using the Maximum likelihood (ML) method and Bayesian inference (BI). The analyses were conducted on the dataset 13 PCGs + two rRNAs of the mitochondrial genomes which acquired similar tree topology (Figs. 5 and 6). We obtained the concatenated amino acid sequences to reconstruct the phylogenetic relationships (Figs. 5 and 6). The topology of the families based on mitogenomes in this study was consistent with the previous morphological and molecular studies11,12,32,52,53,54.

Figure 5
figure 5

Phylogenetic tree of superfamily Noctuoidea moths using IQ-TREE. The phylogeny was reconstructed using 13 PCGs and two rRNA of the 90 species with maximum likelihood (ML) method (1000 replications). The species Papilo polytes and Trogonoptera brookiana mitogenomes were used as outgroups.

Full size image
Figure 6
figure 6

Phylogenetic tree of superfamily Noctuoidea moths using MrBayes. The phylogeny was reconstructed using 13 PCGs and two rRNA of the 90 species with Bayesian Inference. Posterior probability values lower than 50 were not shown.

Full size image

The phylogenetic trees consisted of 6 clades corresponding to 12 major lepidopteran families (Figs. 5 and 6). The family Erebidae formed a major clade including 43 species with high bootstrap proportion and posterior probability (BP ≥ 100; PP: 1). This clade is further comprised of two subclades; subclade I with strongly supported values (BP ≥ 100; PP: 1) involving five subfamilies; Arctiinae Erebinae, Calpinae, Herminiiinae, and Aganainae. The type species Oraesia emarginata belonging to the tribe Calpini was closely related to the Catocala sp., Grammodes geometrica, and Parallelia stuposa with well-supported values (BP ≥ 100; PP: 1); subclade II comprised of 22 species belonging to the subfamilies Lymantriinae and Hypeninae with high nodal support (BP ≥ 96; PP: 1), out of which a single species Paragabara curvicornuta belonging to the subfamily Hypeninae was clustered separately in the same clade. The subfamily Arctiinae is closely related to the subfamily Erebinae rather than Lymantriinae. The Erebidae clade showed the relationship as; ((Aganainae + Calpinae + Herminiinae + Erebinae + (Arctiinae + (Hypeninae + Lymantriinae)))).

The newly sequenced species O. seranensis and Eutelia adulatricoides clustered into a single clade with high bootstrap proportion (BP ≥ 100; PP: 1). These species belonged to the family Euteliidae and strongly supported a monophyletic group. This branch consisted of two subfamilies, Stictopterinae and Euteliinae. Formerly the Euteliinae and Stictopterinae were treated as separate subfamilies of Noctuidae55,56,57. Later these two were placed as subfamilies of Erebidae53. Afterwards, the position of subfamily Euteliinae was raised to the family level and Stictopterinae was placed into the subfamily of Euteliidae based on molecular study18. The present observation was well supported by the molecular study of Zahiri et al.18. The recently reconstituted family Nolidae with species Gabala argentata, Sinna extrema and Risoba prominens clustered in a single clade with high bootstrap proportion (BP ≥ 100; PP: 1) and observed more closely related to the clade (Euteliidae + Noctuidae), instead of Erebidae as proposed by Zahiri et al.58.

The target species A. polyodon and thirty-one species belonging to the family Noctuidae were clustered into single branches with high nodal support values (BP ≥ 100; PP: 1). The species A. polyodon is clustered with Hadeninae clade with a high support value (BP ≥ 100; PP: 1). This phylogenetic analysis showed the main topology: ((Plusiinae + (Heliothinae + (Amphipyrinae + (Acronictinae + (Xyleninae + ((Hadeninae + Noctuinae)))))))).

Notodontiidae was strongly supported as a monophyletic group (BP ≥ 99; PP: 1). The clade consisted of two subfamilies Phalerinae (Phalera flavescens), and Thaumetopoeinae (Ochrogaster lunifer, Clostera anachoreta, Clostera anastomosis, and Thaumetopoea pityocampa).

In the past decade, a number of studies have explored the molecular phylogenetic relationships among the Noctuoidea species. Zahiri et al.18 proposed the following among these families: (Notodontidae + (Euteliidae + (Noctuidae + Erebidae + Nolidae))). In comparison with this, Yang et al.12 published different study in which the following assemblage was proposed: (Notodontidae + (Erebidae + Nolidae + Euteliidae + Noctuidae))). All analyses clearly supported the monophyletic relationships of the 16 subfamilies within Noctuoidea (Figs. 5 and 6). The reformulated family Noctuidae clustered with the newly erected family Euteliidae. Our findings indicated that the branch of Noctuidae and Euteliidae was sister to the newly constituted family Nolidae. The family Erebidae was sister to the clade of (Nolidae + (Euteliidae + Noctuidae)). Family Notodontidae members formed as a single clade consisting of subfamilies, Phalerinae and Thaumetopoeinae. Notodontidae was the sister group to the other Noctuoid families. Our analysis revealed a topology within Noctuoidea as follows: (Notodontidae + (Erebidae + (Nolidae + (Euteliidae + Noctuidae)))). The superfamily Noctuoidea relationships further confirmed that Noctuoidea was a monophyletic group, which was also supported by many previous mitogenome phylogenies12,32,33,34.

In the present analyses, a total of 9 species were included belonging to the superfamilies Bombycoidea, Lasiocampoidea, and Geometroidea. The phylogenetic tree analyses showed that Saturniidae (Actias selene and Antheraea pernyi), Sphingidae (Manduca sexta), Bombycidae (Bombyx mandarina and Bombyx mori), Lasiocampidae (Kunugia undans and Trabala vishnou guttata) and Geometridae (Biston panterinaria and Phthonandria atrilineata) formed a clade with high nodal support values (BP≥89; PP: 0.89), this is consistent with earlier molecular study59. The tree topologies indicate that the relationships are ((Geometridae + (Lasiocampidae + (Bombycidae + ((Sphingidae + Saturniidae)))))). The phylogenetic analyses also revealed the relationships in the superfamilies Bombycoidea, Geometroidea, Lasiocampoidea, and Noctuoidea with strongly supported values (BP≥100; PP: 1). This relationship is the resemblance to the novel Lepidoptera classification revised by van Nieukerken et al.11 and the superfamilies are designated as the Macroheterocera clade.

Four species belonging to the family Crambidae formed a separate clade (BP ≥ 99; PP: 1) which was placed sister to the family Hyblaeidae and both families belong to the clade Obtectomera11. The present analysis is analogous to the molecular analysis by Twort et al.44 which also showed that Hyblaea is sister to Pyraloidea. The newly sequenced species H. puera is separately clustered with moderate support in ML analysis and high support in BI analysis (BP ≥ 61; PP: 0.96). This species was earlier classified under the family Noctuidae (Hampson, 1894). During the same year, the family Hyblaeidae was placed under the superfamily Pyraloidea based on the morphology characters60. Afterward, it got its own superfamily rank Hyblaeoidea and was placed under the Obtectomera clade11. In the present study, the species H. puera deviated from the Noctuoidea member groups. This mitogenome study is well supported by morphological11 and molecular studies18. Unfortunately, the presence of only one mitogenome of Hyblaeidae restricted the discussion of its relationships; more species need to be added for a meaningful inference.

Conclusion

The complete mitochondrial genome sequences of O. emarginata, O. seranensis, A. polyodon and H. puera were successfully determined. The mitogenomes of these fourmoth species were all double-stranded single-circular molecules with similar gene arrangements (Fig. 1). The overall genomic characteristics (gene order, gene size, base composition, PCG codon usage, and tRNA cloverleaf structure) of the lepidopteran mitogenomes were typically constant with those of reported Lepidoptera mitogenomes. The longest intergenic spacer was present between trnQ and nad2; this was a unique feature in all sequenced species. Based on the phylogenetic analyses, the amino acid datasets supported the monophyly of Noctuoidea and its relationships (Notodontidae + (Erebidae + (Nolidae + (Euteliidae + Noctuidae)))). However, more mitochondrial genome samples need to be used to further resolve the relationships among the Noctuoidea.

Materials and methods

Sample collection and genomic DNA extraction

The samples of the four species, O. emarginata (11° 41′ 181″ N 76° 72′ 07″ E), A. polyodon, O. seranensis (10° 23′ 5367˝ N 77° 49′ 2933˝ E) and H. puera (10° 27′ 045˝ N 77° 53′ 3633˝ E) were collected from the Tamil Nadu part of Western Ghats. K. Sivasankaran identified all the species, which were preserved in absolute ethanol and stored at -80ºC until DNA isolation. The genomic DNA was extracted from thorax tissue of moths using Quick-DNA Tissue/Insect Microprep Kit (Cat No-D6016-HSN CODE-38220090, Zymo Research, USA) with the manufacturer’s protocol. The DNA samples and quality were checked using Nanodrop 1000 and confirmed with 1% agarose gel.

Mitogenome sequencing

The quality-check passed samples were subjected further for the library preparation. In brief, 100 ng of DNA was subjected to prepare indexed library using Truseq Nano library preparation kit (Illumina #20,015,964). Final libraries were quantified using Qubit 4.0 fluorometer (Thermofisher #Q33238) using DNA HS assay kit (Thermofisher #Q32851) following manufacturer’s protocol. To identify the insert size of the library, we queried it on Tapestation 4150 (Agilent) utilizing highly sensitive D1000 screen tapes (Agilent # 5067–5582) following manufacturers’ protocol. The next-generation sequencing was performed by Molsys Scientific Pvt. Ltd (Bangalore, India). Finally, NOVASEQ 6000 platform (Illumina, San Diego, California USA) was used to sequence 151 bp read lengths about 4 GB in size.

Sequence assembly and annotation

The raw sequences were assembled using the NOVOPLASTY Ver 4.2 (https://github.com/ndierckx/NOVOPlasty)61. The sequences’ annotations were executed using MITOS2 (http://mitos2.bioinf.uni-leipzig.de/index.py)62 using the genetic code for invertebrate mitogenomes. The sequences were also annotated and verified for accurate lengths of the 13 protein-coding genes using CHLOROBOX-GeSeq-Annotation of Organellar Genomes (https://chlorobox.mpimp-golm.mpg.de/geseq.html)63. The composition skewness was calculated using the formula: AT skew = [A − T]/[A + T]; similarly, GCskew = [G − C]/[G + C] (https://en.vectorbuilder.Com/tool/gc-content-calculator). The tRNA genes and their cloverleaf structures were predicted with MITOS2 software and analyzed by comparison with the nucleotide sequence of other lepidopteran tRNA sequences. Tandem repeats at the A + T-rich region were identified using the online Tandem Repeats Finder tool (http://tandem.bu.edu/trf/trf.html). Relative Synonymous Codon Usage (RSCU) of PCGs was determined using MEGA X64. The circular maps of the four complete mitogenomes were drawn using the OGDRAW-Draw Organelle Genome Maps (https://chlorobox.mpimp-golm.mpg.de/OGDraw.html)65.

Phylogenetic analyses

A total of 90 species (4 newly sequenced in this study, 86 available from GenBank) representing 7 families of Lepidoptera11 were used to reconstruct the phylogenetic relationships among them. The ingroup consisted of 43 species of Erebidae, 2 species of Euteliidae, 32 species of Noctuidae, 3 species of Nolidae, 5 species of Notodontidae, 1 species of Hyblaeidae, and 4 species of Crambidae. Species Papilio polytes and Trogonoptera brookiana mitogenomes were selected as outgroups (Table 1).

The amino acid sequences of 13 protein-coding genes and two rRNA genes were used in phylogenetic analysis. We used MAFFT to align and concatenate each of the 13 PCGs and rRNAs genes. Further the concatenated amino acid sequences from the 13 PCGs and rRNA genes were used for reconstructing the phylogenetic tree, which was performed using the Model-based Maximum Likelihood method using the IQ-TREE in PhyloSuite V1.2.2 program https://github.com/dongzhang0725/PhyloSuite66. The appropriate model General Reversible mitochondrial (mtREV) Gamma distributed with invariant sites (G + I) was used to infer the phylogenetic relationships based on 5000 bootstraps of ultrafast replicates.

The analysis of Bayesian inference (BI) was conducted for the dataset. The BI analysis was performed through the MrBayes 3.2.6 in PhyloSuite V1.2.266 using the GTR + I + R model. Invgamma (+ I + G proportion invariable, remaining gamma rate variations across sites were presented and performed. The convergence of Markov Chain Monte Carlo (MCMC), which was observed by the average standard deviation of split frequencies, reached below 0.01. Four chains (three hot and one cold) were run with a dataset for one million generations with the tree being sampled every 1000 generations with a burn-in of 2500. FigTree v1.4.4 (http://tree.bio.ed.ac.uk/software/figtree/) was practised to visualise the phylogenetic tree.