Revalidation of morphological characteristics and multiplex PCR for the identification of three congener invasive Liriomyza species (Diptera: Agromyzidae) in China

Insects

The three species of Liriomyza spp. were collected from areas where leafminers occur in China. In this study, 263 individuals of three species were selected for further data analysis (Table S1). These were collected at the larval stage, tagged with relevant information and transported to the laboratory for pupation and emergence as adults. After preliminary morphological identification, adults were labeled, immersed in 70% ethanol and stored at −20 °C. After dissecting and photographing the samples, the remaining tissues were stored in 100% ethanol for DNA extraction and molecular analysis.

Morphological identification

Samples were examined with a stereomicroscope (Zeiss Stemi 2000c) and photographed with a wide depth of field (Zeiss Smartzoom 5). Male genitalia and wings were dissected, and slides were prepared and photographed with the Axio imager A2 (Zeiss, Germany).

Differences in the ratios of ultimate section lengths of vein CuA₁ among different Liriomyza species were determined by one-way analysis of variance (ANOVA), followed by Tukey’s multiple comparisons. All statistical analyses were performed using SPSS v. 16.0 (SPSS, Chicago, IL, USA), and statistical significance was determined when P <  0.05.

Molecular identification and primer selection for multiplex PCR

Genomic DNA of Liriomyza species was extracted using the AxyPrep™ Multisource Genomic DNA Kit (Axygen, USA). A partial sequence of the mitochondrial cytochrome oxidase I (COI) gene was amplified with common primers F, 5′-CAACATTTATTTTGATTTTTTGG-3′ and R, 5′- TCCAATGCACTAATCTGCCATATTA-3′ (Simon et al., 1994; Yang, Cao & Du, 2010) using protocols described by Chen et al. (2019), to molecular cross-checking and verification all of Liriomyza species in this study using sequencing, accession number can be found in Table S1.

For multiplex PCR, full-length COI genes of three Liriomyza species were downloaded from NCBI (https://www.ncbi.nlm.nih.gov/) and aligned using Clustal X. To develop a rapid identification method, three species-specific primers and a common reverse primer were mixed to amplify DNA from different Liriomyza species. The PCR conditions were as follows: denaturation at 94 °C for 3 min; 35 cycles at 94 °C for 1 min, 58 °C for 1 min and 72 °C for 1 min; followed by extension at 72 °C for 10 min. PCR was conducted in a 25 µL reaction volume containing 2 µL (100 ng) of DNA template, 1 µL (10 µM) of each primer, 12.5 µL of 2 × Taq Master mix (Vazyme Biotech Co., Ltd) and 6.5 µL ddH₂O. PCR products were separated in 1.0% agarose gels, and primers that amplified only one specific band for each species are shown in Table 1.

Morphological identification

The distiphallus, which is part of the male genitalia, is a very small, fragile structure enclosed by membranes located at the terminus of the aedeagus. For L. trifolii, the morphological characteristics of the distiphallus include one distal bulb with marked constriction between lower and upper halves in dorsoventral view; the bulb is lightly sclerotized with a long basal stem (Fig. 1A). For L. sativae, the distiphallus is characterized by one distal bulb with a slight constriction between upper and lower halves in the dorsoventral view; the bulb is more intensely sclerotized with a shorter basal stem (Fig. 1B). For L. huidobrensis, the distiphallus contains two distal bulbs; these meet at rims that extend in an anteroventral orientation (Fig. 1C).

With respect to vertical setae, L. trifolii exhibits inner and outer vertical setae on a yellow background; whereas vertical setae are present on a black background for L. huidobrensis. In L. sativae, outer and inner vertical setae are presented on black and yellow backgrounds, respectively (Spencer, 1973). In this study, only 86.1% (192/223) of L. trifolii had yellow inner and outer vertical setae; 9.9% (22/223) had yellow inner vertical setae and undetermined color for outer setae, and 4.0% (9/223) had yellow inner and black outer vertical setae (Table 2; Figs. 2A–2C). For L. sativae, 17.6% (6/34) had black inner and outer vertical setae, 58.8% (20/34) had yellow inner and black outer vertical setae, and 23.5% (8/34) had outer black setae with an undetermined color for inner vertical setae (Table 2; Figs. 2D–2F). For L. huidobrensis, 100% (6/6) exhibited black inner and outer vertical setae (Table 2; Figs. 2G–2I). These results show that characteristics of vertical setae are not reliable for identifying Liriomyza species.

Wing pattern ratios were calculated as the length of the ultimate section of vein CuA₁ divided by the penultimate section (‘a’ and ‘b’, see Figs. 3A–3C). In this study, ‘a’ was 2.70 ±  0.31 times the length of ‘b’ in L. trifolii, and ‘a’ was 2.72 ±  0.37 times the length of ‘b’ in L. sativae. For L. huidobrensis, ‘a’ was 2.20 ± 0.24 times the length of ‘b’ (F_2,237 = 7.345, P <  0.05) (Fig. 4). Although the ratio of L. huidobrensis was significantly different from the other two species (P <  0.05), there was no significant difference between L. trifolii and L. sativae ( P = 0.907). Many L. trifolii individuals exhibited truncated or missing dm-cu cross veins. Furthermore, we noted inconsistency between left and right forewing patterns within individual samples (Fig. 3A, with dashed lines).

In L. trifolii, the 2nd–5th visible tergites were generally divided by a yellow medial furrow in male adults; furthermore, there was a yellow patch at the 5th black visible tergite that can distinguish L. trifolii from other Liriomyza species (Figs. 5A–5C). In L. sativae and L. huidobrensis, only the second visible tergite is divided by a yellow medial furrow and no yellow patch is evident on the 5th tergite (Figs. 5D–5I).

Molecular detection of Liriomyza spp.

Candidate primers for species-specific detection of Liriomyza were based on the alignment of 262 (L. sativae), 612 (L. trifolii), and 959 (L. huidobrensis) COI sequences. We designed one reverse primer, 1181 R, that was common to all three Liriomyza species. The position of forward primers was selected to produce < 1,000 bp amplicons when paired with the reverse primer with at least 300 bp nucleotides between species. In addition, sites were selected where the number of differential nucleotides was >2 bp to increase the specificity of the primers (Fig. 6).

The three Liriomyza species could be differentiated by specific PCR products in 1.0% agarose gels, and the resulting PCR products were 569, 919, and 222 bp for L. trifolii, L. sativae and L. huidobrensis, respectively (Fig. 7A). The validity of multiplex PCR for identification was further confirmed by using the system with different developmental stages; the approach worked equally well for larvae, pupae and adults of the three Liriomyza species (Fig. 7B). Populations from different geographical regions were also obtained to evaluate the reliability of species-specific primers. The results obtained by multiplex PCR (Fig. S1) and subsequent sequence analysis of COI (Fig. S2) showed that geography did not impact the reliability of primers.