3.1 Diglyphus sabulosus Erdӧs, 1951(Newly Recorded Species from China)
Diglyphus sabulosus Erdös, 1951:197.
Diglyphus sabulosus Thuróczy, 1992: 167.
Specimens examined. 15♀♀3♂♂, CHINA, Xinjiang Uygur Autonomous Region, Akto County Oyatake Glacier Park, 38°53′54″N, 75°12′5″E, 2764.51 m, 17. Ⅶ. 2022. Hongying Hu et al., sweeping.
Diagnosis. Female. Body length 1.30–1.41 mm (n = 15). Body (Fig. 1F) with a dark green metallic sheen. Antenna scape white, pedicel and funicle light brown (Fig. 1C); proximal 1/2 of femur brown, tibia pale yellow, entire tarsus yellowish-brown; wing veins pale yellow. F1 1.18× as long as F2. Malar sulcus prominent, POL 2.56× that of OOL (Fig. 1A). Thorax (Fig. 1D) with reticulated engraving pattern. Foring veins slightly thickened (Fig. 1E), with postmarginal vein subequal in length to stigmal vein which possesses a stalk; speculum minute. Gaster (Fig. 1F) approximately oblong-ovate, ovipositor extends slightly from the end of the gaster.
Male. Veins of both forewings and hindwings are distinctly thickened (Fig. 1H); all other characters silimar with those of the female.
Hosts. Agromyzidae (Yefremova et al.2011; UCD 2023).
Distribution. China (Xinjiang); Czech Republic, Hungary, Romania, Slovakia, Sweden, Turkey (UCD 2023).
Comments. This species is closely similar to Diglyphus poppoea, with the main difference being the color of the hind tibia. In Diglyphus poppoea, the distal two-thirds of the hind tibia are dark green (Jafarlu et al. 2023), whereas in this species, they are entirely yellow. Additionally, the male wings of this species have distinctly thickened, yellow veins, a feature that differs from those of other congeneric species.
3.2 Mitogenome Organization
The complete mitogenome of Diglyphus sabulosus is a circular molecule of 15,690 bp, encoding 13 PCGs, 22 transfer RNA genes (tRNAs), and 2 ribosomal RNA genes (rRNAs). The organization of 37 genes and one non-coding region, together with the GC content and GC skew along the mitogenome, is presented in Fig. 2. A total of 10 PCGs, 15 tRNA genes, and 2 rRNA genes are located on the major coding strand (H-strand); 3 PCGs (cytb, nad2, and nad6) and 7 tRNA genes (trnS2, trnT, trnE, trnK, trnN, trnW, and trnQ) are situated on the minor coding strand (L-strand).
There is one long non-coding region in the Diglyphus sabulosus mitogenome. The control region is located between trnI and trnM. Basic information on all the genes and non-coding region is listed in Table 2, including strand positions, lengths, tRNA anticodons, the start and stop codons of the PCGs, and intergenic spacer lengths. The mitochondrial adjacent genes of Diglyphus sabulosus exhibit variable-length intergenic spacers and gene overlap. In this study, 19 intergenic gaps and 15 gene overlaps were identified in the mitogenome of Diglyphus sabulosus. Among these gaps, the longest (99 bp) was detected between nad2 and trnI, and the shortest (2 bp) was present between two gene pairs: trnT-trnP and trnC-trnN. For gene overlaps, the longest spanned 21 bp and was located between rrnL and trnL1.
Table 2
Positions and basic features of genes and non-coding region of the Diglyphus sabulosus mitogenome
| Gene | From | To | Length/bp | Intergenic nucleotide/bp | Anticodon | Start codon | Stop codon | Coding strand |
| trnA | 1 | 64 | 64 | | TGC | | | H |
| rrnL | 80 | 1394 | 1315 | 15 | | | | H |
| trnL1 | 1374 | 1441 | 68 | -21 | TAG | | | H |
| nad1 | 1457 | 2389 | 933 | 15 | | ATT | TAA | H |
| trnS2 | 2388 | 2454 | 67 | -2 | TGA | | | L |
| cytb | 2453 | 3592 | 1140 | -2 | | ATG | TAA | L |
| nad6 | 3592 | 4113 | 522 | -1 | | ATT | TAA | L |
| trnP | 4170 | 4236 | 67 | 56 | TGG | | | H |
| trnT | 4239 | 4306 | 68 | 2 | TGT | | | L |
| nad4L | 4324 | 4593 | 270 | 17 | | ATG | TAA | H |
| nad4 | 4608 | 5939 | 1332 | 14 | | ATG | TAA | H |
| trnH | 5923 | 5989 | 67 | -17 | GTG | | | H |
| nad5 | 5993 | 7690 | 1698 | 3 | | ATT | TAA | H |
| trnF | 7671 | 7734 | 64 | -20 | GAA | | | H |
| trnE | 7734 | 7798 | 65 | -1 | TTC | | | L |
| cox1 | 7802 | 9337 | 1536 | 3 | | ATG | TAA | H |
| trnL2 | 9333 | 9399 | 67 | -5 | TAA | | | H |
| cox2 | 9425 | 10102 | 678 | 25 | | ATT | TAA | H |
| trnK | 10112 | 10180 | 69 | 9 | TTT | | | L |
| trnD | 10188 | 10255 | 68 | 7 | GTC | | | H |
| atp8 | 10280 | 10441 | 162 | 24 | | ATT | TAA | H |
| atp6 | 10435 | 11109 | 675 | -7 | | ATG | TAA | H |
| cox3 | 11109 | 11897 | 789 | -1 | | ATG | TAA | H |
| trnG | 11908 | 11973 | 66 | 10 | TCC | | | H |
| nad3 | 11971 | 12324 | 354 | -3 | | ATA | TAA | H |
| trnR | 12328 | 12386 | 59 | 3 | TCG | | | H |
| trnC | 12402 | 12467 | 66 | 15 | GCA | | | H |
| trnN | 12470 | 12536 | 67 | 2 | GTT | | | L |
| trnY | 12536 | 12602 | 67 | -1 | GTA | | | H |
| trnS1 | 12602 | 12663 | 62 | -1 | TCT | | | H |
| trnW | 12663 | 12732 | 70 | -1 | TCA | | | L |
| nad2 | 12731 | 13723 | 993 | -2 | | ATT | TAA | L |
| trnI | 13823 | 13890 | 68 | 99 | GAT | | | H |
| Control region | 13891 | 14692 | 802 | 0 | | | | |
| trnM | 14693 | 14757 | 65 | 0 | CAT | | | H |
| trnV | 14758 | 14833 | 76 | 0 | TAC | | | H |
| rrnS | 14841 | 15600 | 760 | 7 | | | | H |
| trnQ | 15608 | 15676 | 69 | 7 | TTG | | | L |
3.3 Nucleotide Composition
The whole mitogenome of Diglyphus sabulosus consists of 37.5% A, 43.5% T, 7.9% C, and 11.1% G. It shows an obvious bias towards A + T (81.0%), with a negative AT skew of -0.074 and a positive GC skew of 0.171 (Table 3), similar to what is observed in other Eulophid species (Tang et al. 2021; Tian et al. 2021). The AT content of the four datasets comprising the complete mitogenome, protein-coding genes, tRNA genes, and rRNA genes of Diglyphus sabulosus were 81.0%, 79.8%, 86.6%, and 86.0%, respectively. The AT content of 13 protein-coding genes ranged from 73.7% to 86.0%, with cox1 exhibiting the lowest value at 73.7% and nad2 the highest at 86.0%. Analysis of AT skew in each protein-coding gene revealed that, with the exception of the atp8 gene, 12 remaining protein-coding genes exhibited T bias. In contrast, in GC skew analysis, 4 protein-coding genes (atp8, cytb, nad2, and nad6) showed C bias, while the remaining 9 protein-coding genes exhibited G bias.
Table 3
Nucleotide composition and skewness of the Diglyphus sabulosus mitogenome.
| Regions | Size(bp) | T | C | A | G | AT(%) | GC(%) | AT skewness | GC skewness |
| Full genome | 15690 | 43.5 | 7.9 | 37.5 | 11.1 | 81.0 | 19.0 | -0.074 | 0.171 |
| PCGs | 11082 | 45.7 | 9.4 | 34.1 | 10.8 | 79.8 | 20.2 | -0.146 | 0.073 |
| tRNAs | 1469 | 42.0 | 5.2 | 44.6 | 8.2 | 86.6 | 13.4 | 0.030 | 0.218 |
| rRNAs | 2075 | 43.6 | 4.8 | 42.4 | 9.2 | 86.0 | 14.0 | -0.013 | 0.313 |
| Control region | 802 | 39.0 | 12.0 | 31.3 | 17.7 | 70.3 | 29.7 | -0.110 | 0.190 |
| atp6 | 675 | 44.6 | 10.5 | 34.1 | 10.8 | 78.7 | 21.3 | -0.134 | 0.014 |
| atp8 | 162 | 37.7 | 8.6 | 47.5 | 6.2 | 85.2 | 14.8 | 0.116 | -0.167 |
| cox1 | 1536 | 43.0 | 12.0 | 30.7 | 14.3 | 73.7 | 26.3 | -0.166 | 0.089 |
| cox2 | 678 | 42.2 | 9.7 | 34.4 | 13.7 | 76.6 | 23.4 | -0.102 | 0.170 |
| cox3 | 789 | 46.4 | 9.9 | 30.7 | 13.1 | 77.1 | 23.0 | -0.204 | 0.138 |
| cytb | 1140 | 42.1 | 13.1 | 33.1 | 11.8 | 75.2 | 24.9 | -0.120 | -0.053 |
| nad1 | 933 | 46.0 | 8.0 | 33.7 | 12.3 | 79.7 | 20.3 | -0.155 | 0.211 |
| nad2 | 993 | 47.3 | 9.3 | 38.7 | 4.7 | 86.0 | 14.0 | -0.101 | -0.324 |
| nad3 | 354 | 47.5 | 7.3 | 35.6 | 9.6 | 83.1 | 16.9 | -0.143 | 0.133 |
| nad4 | 1332 | 47.7 | 7.4 | 33 | 11.8 | 80.7 | 19.2 | -0.182 | 0.227 |
| nad4L | 270 | 50.7 | 4.4 | 35.2 | 9.6 | 85.9 | 14.0 | -0.181 | 0.368 |
| nad5 | 1698 | 48.8 | 7.2 | 34.4 | 9.6 | 83.2 | 16.8 | -0.173 | 0.144 |
| nad6 | 522 | 46.4 | 9.6 | 39.1 | 5.0 | 85.5 | 14.6 | -0.085 | -0.316 |
3.4 Protein-Coding Genes
The total tandem length of 13 protein-coding genes in Diglyphus sabulosus is 11,082 bp, accounting for 70.63% of the entire mitogenome and encoding 3,681 amino acids. The amino acids were ranked by content in descending order as follows: Leu > Ile > Phe > Ser > Met > Asn > Val > Tyr > Gly > Lys > Pro = Thr > Ala > Trp > Glu > His > Gln > Asp > Arg > Cys, with Pro and Thr having identical percentages. Among these, Leu exhibited the highest relative content at 13.69%, whereas Cys showed the lowest, accounting for only 0.08%. RSCU values were calculated to measure the codon usage bias (Fig. 3). The codons with an RSCU value > 1.0 were defined as abundant codons. Relative codon usage analysis was performed on Diglyphus sabulosus, revealing 64 distinct codons in its coding sequences. A total of 29 codons showed RSCU > 1.0, of which 14 terminated in U and 15 in A (Table 4). Among these preferred codons, UUA had the highest usage frequency (n = 393, RSCU = 4.68). Notably, all amino acids exhibited distinct codon usage biases, with preferences specifically for codons terminating in A or U (A/U at the third position). As illustrated in Fig. 3, notable examples include: Leu preferring the codon UUA, Ile preferring AUU, Phe preferring UUU, and Met preferring AUA, among others. All PCGs start with typical ATN initiation codons, including one ATA (nad3), six ATTs (nad1, nad6, nad5, cox2, atp8, and nad2), and six ATGs (cytb, nad4L, nad4, cox1, atp6, and cox3). All PCGs terminate with conventional stop TAA codon (Table 2).
Table 4
Relative synonymous codon usage (RSCU) in the mitochondrial genome of Diglyphus sabulosus
| Codon | Count | RSCU | Codon | Count | RSCU | Codon | Count | RSCU | Codon | Count | RSCU |
| UUU(F) | 388 | 1.87 | UCU(S) | 110 | 2.29 | UAU(Y) | 161 | 1.85 | UGU(C) | 29 | 2.00 |
| UUC(F) | 26 | 0.13 | UCC(S) | 8 | 0.17 | UAC(Y) | 13 | 0.15 | UGC(C) | 0 | 0.00 |
| UUA(L) | 393 | 4.68 | UCA(S) | 128 | 2.66 | UAA(*) | 13 | 2.00 | UGA(W) | 76 | 1.85 |
| UUG(L) | 22 | 0.26 | UCG(S) | 3 | 0.06 | UAG(*) | 0 | 0.00 | UGG(W) | 6 | 0.15 |
| CUU(L) | 57 | 0.68 | CCU(P) | 65 | 2.39 | CAU(H) | 53 | 1.86 | CGU(R) | 17 | 1.48 |
| CUC(L) | 6 | 0.07 | CCC(P) | 4 | 0.15 | CAC(H) | 4 | 0.14 | CGC(R) | 0 | 0.00 |
| CUA(L) | 24 | 0.29 | CCA(P) | 36 | 1.32 | CAA(Q) | 44 | 1.57 | CGA(R) | 24 | 2.09 |
| CUG(L) | 2 | 0.02 | CCG(P) | 4 | 0.15 | CAG(Q) | 12 | 0.43 | CGG(R) | 5 | 0.43 |
| AUU(I) | 421 | 1.90 | ACU(T) | 59 | 2.17 | AAU(N) | 201 | 1.83 | AGU(S) | 21 | 0.44 |
| AUC(I) | 22 | 0.10 | ACC(T) | 8 | 0.29 | AAC(N) | 19 | 0.17 | AGC(S) | 1 | 0.02 |
| AUA(M) | 322 | 1.86 | ACA(T) | 41 | 1.50 | AAA(K) | 129 | 1.80 | AGA(S) | 101 | 2.10 |
| AUG(M) | 24 | 0.14 | ACG(T) | 1 | 0.04 | AAG(K) | 14 | 0.20 | AGG(S) | 13 | 0.27 |
| GUU(V) | 77 | 1.74 | GCU(A) | 47 | 2.16 | GAU(D) | 46 | 1.74 | GGU(G) | 53 | 1.25 |
| GUC(V) | 5 | 0.11 | GCC(A) | 0 | 0.00 | GAC(D) | 7 | 0.26 | GGC(G) | 4 | 0.09 |
| GUA(V) | 84 | 1.90 | GCA(A) | 38 | 1.75 | GAA(E) | 61 | 1.58 | GGA(G) | 74 | 1.74 |
| GUG(V) | 11 | 0.25 | GCG(A) | 2 | 0.09 | GAG(E) | 16 | 0.42 | GGG(G) | 39 | 0.92 |
3.5 Transfer RNAs and Ribosomal RNAs
Diglyphus sabulosus has 22 tRNAs, ranging from 59 to 76 bp in length. Fifteen of these tRNAs are located on the major coding strand (H-strand), and seven on the minor coding strand (L-strand). These tRNAs exhibit an AT bias (86.6%), similar to the overall mitogenomic composition. The AT and GC skew values are 0.030 and 0.218, respectively (Table 3). The predicted secondary structure of all tRNAs in the Diglyphus sabulosus mitogenome is presented in Fig. 4. With the exception of trnR and trnS1, which lack both the dihydrouracil arm (DHU arm) and dihydrouracil ring (DHU ring), all the tRNAs are folded into typical clover-leaf secondary structures, consisting of four domains and a variable loop. trnS1 lacks the DHU arm, which is almost ubiquitous in insect mitogenomes (Sun et al. 2010; Jühling et al. 2012; Cameron 2014). In the 22 tRNAs of the Diglyphus sabulosus mitogenome, a total of 11 G-U base mismatches were detected: one mismatch in each of trnD, trnL1, trnL2, and trnP; two mismatches in each of trnF and trnY; and three mismatches in trnA (Fig. 4). The mitogenome of Diglyphus sabulosus also contains 2 rRNAs. rrnL is 1,315 bp long and positioned between trnA and trnL1, while rrnS is 760 bp long and located between trnV and trnQ (Table 2).
3.6 Phylogenetic Relationships
In the current study, we analyzed phylogenetic relationships using complete mitogenome sequences. The primary objective was to gain insights into interrelationships within the Eulophidae clade, focusing on Diglyphus sabulosus. The phylogenetic relationships of Eulophidae were subsequently inferred via the ML and BI methods, based on the complete mitogenomes from Diglyphus sabulosus and other selected species. The two trees exhibited nearly congruent topologies. Therefore, we present the consensus tree, along with both Bayesian posterior probabilities and maximum likelihood bootstrap values, in Fig. 5. The phylogenetic results revealed that all species of the family Eulophidae form a single well-supported clade, confirming that Eulophidae is a monophyletic group with strong statistical support. Within Eulophidae, the two tribes Eulophinae and Tetrastichinae were also recovered as monophyletic. Diglyphus and Necremnus form a tightly clustered subclade, with the four Diglyphus species coalescing into a single clade. Within this clade, Diglyphus sabulosus and Diglyphus poppoea are sister groups, representing the closest relatives (posterior probability of 1; bootstrap value of 100%). Rao et al. (2025) reconstructed a ML phylogenetic tree using 13 protein-coding genes derived from four species within the family Eulophidae. The phylogenetic results revealed that all species belonging to the family Eulophidae form a single well-supported clade, confirming that Eulophidae is a monophyletic group with strong statistical support. The phylogenetic relationships within the family Eulophidae inferred in this study are generally consistent with previous findings [54–55] and congruent with conclusions from traditional taxonomy.