Reliability of wing morphometrics for species identification of human-biting black flies (Diptera: Simuliidae) in Thailand

Abstract

Background: Fast and reliable species identification of black flies is essential for research proposes and effective vector control. Besides traditional identification based on morphology, which is usually supplemented with molecular methods, geometric morphometrics (GM) has emerged as a promising tool for identification. Despite its potential, no specific GM techniques have been established for the identification of black fly species.

Methods: Adult female black flies collected using human bait, as well as those reared from pupae, were used in this study. Here, landmark-based GM analysis of wings was assessed for the first time to identify human-biting black fly species in Thailand, comparing this approach with the standard morphological identification method and DNA barcoding based on the mitochondrial cytochrome c oxidase subunit I (COI) gene. To explore genetic relationships between species, maximum likelihood (ML) and neighbor-joining (NJ) phylogenetic trees were built. Additionally, three different methods of species delimitation, i.e., assemble species by automatic partitioning (ASAP), generalized mixed yule coalescent (GMYC), and single Poisson tree processes (PTP), were utilized to identify the morphologically defined species. The effectiveness of a COI barcode in identifying black fly species was further examined through the best match (BM) and best close match (BCM) methods.

Results: Seven black fly species, namely Simulium tenebrosum Takaoka, Srisuka & Saeung, 2018 (complex), S. doipuiense Takaoka & Choochote, 2005 (complex), S. nigrogilvum Summers, 1911, S. nodosum Puri, 1933, S. asakoae Takaoka & Davies, 1995, S. chamlongi Takaoka & Suzuki, 1984, and S. umphangense Takaoka, Srisuka & Saeung, 2017 were morphologically identified. Compared with the standard method, the GM analysis based on wing shape showed high success in separating species, achieving an overall accuracy rate of 88.54%. On the other hand, DNA barcoding surpassed wing GM for species identification with a correct identification rate of 98.57%. Species delimitation analyses confirmed the validity of most nominal species, with an exception for S. tenebrosum complex and S. doipuiense complex, being delimited as a single species. Moreover, the analyses unveiled hidden diversity within S. asakoae, indicating the possible existence of up to four putative species.

Conclusions: This study highlights the potential of wing GM as a promising and reliable complementary tool for species identification of human-biting black flies in Thailand.

Keywords: Simulium; DNA barcodes; Hematophagous insect; Medical entomology; Morphometric analysis; Species delimitation.

Fig. 1

A representative image of black fly wing showing ten landmarks used in geometric morphometric analysis

Fig. 2

Scatter plot showing the allometric relationship between the wing shape (the first principal component, PC1) and wing size (centroid size) of seven black fly species. The red line indicates the linear regression prediction with 95% confidence intervals (shaded areas), while the sapphire dots represent individual samples

Fig. 3

Violin plot overlaid with a box plot showing the distribution of wing centroid sizes (CS) in millimeters (mm) of seven black fly species

Fig. 4

Shape differences in wing venation based on anatomical landmarks of seven black fly species. (A) Scatter plot showing residual coordinates of ten landmarks aligned by Procrustes analysis and (B) a wireframe graph showing the superposition of the overall mean shape

Fig. 5

Factor map based on discriminant analysis (DA) showing the shape divergence of seven black fly species. Each polygon represents a different species, with dots indicating individual specimens and a sun cross marking the mean values for each species

Fig. 6

UPGMA dendrogram based on the Mahalanobis distances between average group shapes showing the phenetic relationship of wing shape among seven black fly species. The scale bar represents the Mahalanobis distance

Fig. 7

Frequency distribution of intraspecific and interspecific K2P genetic distances based on the COI gene of seven human-biting black fly species

Fig. 8

Maximum likelihood phylogenetic tree based on 585 bp COI gene of seven human-biting black fly species and their related species. Bootstrap support values (ML/NJ) greater than 50% are indicated near the branches. Some distinct clades were collapsed for clearer presentation, and the number of sequences falling within those clades is indicated in square brackets. Sequences obtained in this study are highlighted in bold type. All sequences used for constructing the tree are detailed in Table S2

Fig. 9

Summary of the three species delimitation analyses (ASAP, GMYC, and PTP) based on the COI haplotypes from seven different nominal species of black flies. The maximum clade credibility tree built from BEAST is colored according to morphospecies. The vertical bars at the tips of the tree correspond to the result of each species delimitation method and morphological identification, respectively. Detailed information of each haplotype is provided in Table S2