Mitochondrial DNA barcoding of mosquito species (Diptera: Culicidae) in Thailand

Abstract

The correct identification of mosquito species is important for effective mosquito vector control. However, the standard morphological identification of mosquito species based on the available keys is not easy with specimens in the field due to missing or damaged morphological features during mosquito collections, often leading to the misidentification of morphologically indistinguishable. To resolve this problem, we collected mosquito species across Thailand to gather genetic information, and evaluated the DNA barcoding efficacy for mosquito species identification in Thailand. A total of 310 mosquito samples, representing 73 mosquito species, were amplified using mitochondrial cytochrome c oxidase subunit I (COI) primers. The average maximum intraspecific genetic variation of the 73 mosquito species was 1% ranged from 0-5.7%. While, average minimum interspecific genetic variation (the distance to the nearest neighbour) of the 73 mosquito species was 7% ranged from 0.3-12.9%. The identification of success rates based on the "Best Match," "Best Close Match," and "All Species Barcodes" methods were 97.7%, 91.6%, and 81%, respectively. Phylogenetic analyses of Anopheles COI sequences demonstrated a clear separation between almost all species (except for those between An. baimaii and An. dirus), with high bootstrap support values (97%-99%). Furthermore, phylogenetic analyses revealed potential sibling species of An. annularis, An. tessellatus, and An. subpictus in Thailand. Our results indicated that DNA barcoding is an effective molecular approach for the accurate identification of mosquitoes in Thailand.

Fig 1

Map of the study sites (A) and mosquito-collection methods used, including adult mosquito trapping (B) and larvae dipper (C). Mosquito samples were collected from 22 provinces from six geographic regions of Thailand, comprising the Northern region (green), including (1) Mae Hong Son, (2) Chiang Mai, and (3) Nan; the Western region (pink) including (4) Tak, (5) Kanchanaburi, (6) Ratchaburi, and (7) Phetchaburi; the Central region (gray), including (8) Nakhon Pathom and (9) Samut Songkhram; the Eastern region (red), including (10) Chachoengsao, (11) Chanthaburi, and (12) Trat; the Northeastern region (yellow), including (13) Chaiyaphum, (14) Nakhon Ratchasima, (15) Surin, and (16) Ubon Ratchathani; and the Southern region (blue), including (17) Phang Nga, (18) Surat Thani, (19) Nakhon Si Thammarat, (20) Krabi, and (21) Narathiwat. Free map provided by USGS National Map Viewer (public domain): http://viewer.nationalmap.gov/viewer/.

Fig 2. Scatter plots based on barcode gap analysis of 73 mosquito species.

(A) Maximum intraspecific distances compared with minimum interspecific distances (distance to the nearest species), and (B) Mean intraspecific distances compared with minimum interspecific distances. Species dots above the 1:1 line show the presence of a “barcode gap,” whereas those on and below the 1:1 line show the absence of a “barcode gap”.

Fig 3. Results of specimen identification success based on the “Best Match,” “Best Close Match,” and “All Species Barcodes” methods.

The threshold value for “Best Close Match” and “All Species Barcodes” was set at 1%.

Fig 4. Maximum likelihood (ML) tree based on 112 cytochrome c oxidase subunit I (COI) sequences representing 30 mosquito species in the subfamily Anophelinae.

Bootstrap values (1000 replicates) are shown near each branch (numbers in red). Vertical bars indicate species delimited using the assemble species by automatic partitioning (ASAP) (green bars) and the BIN-refined single linkage analysis (RESL) algorithm (blue bars) methods. Toxorhynchites splendens (OL743111) was used as an outgroup to root the tree. The pink branches showed subgrouping in the same mosquito species.

Fig 5. Maximum likelihood (ML) tree based on 198 cytochrome c oxidase subunit I (COI) sequences representing 43 mosquito species in the subfamily Culicinae.

Bootstrap values (1000 replicates) are shown near each branch (numbers in red). Vertical bars indicate species delimited using the assemble species by automatic partitioning (ASAP) (green bars) and the BIN-refined single linkage analysis (RESL) algorithm (blue bars) methods. Phlebotomus papatasi (MN086383).

Fig 6. Maximum likelihood (ML) tree based on cytochrome c oxidase subunit I (COI) sequences of Anopheles annularis from this study and their sibling species from GenBank.

This tree was constructed using the Tamura 3-parameter substitution model with gamma distribution. Bootstrap values (1000 replicates) are shown near each branch (numbers in red). Anopheles pallidus (MH330200) was used as an outgroup to root the tree. The bold font is samples in this study.

Fig 7. Maximum likelihood (ML) tree based on cytochrome oxidase subunit I (COI) sequences of Anopheles tessellatus from this study and their sibling species for GenBank.

This tree was constructed using the Tamura 3-parameter substitution model with gamma distribution. Bootstrap values (1000 replicates) are shown near each branch (numbers in red). Anopheles sinensis (OL742920) was used as an outgroup to root the tree. The bold font is samples in this study.

Fig 8. Maximum likelihood (ML) tree based on cytochrome c oxidase subunit I (COI) sequences of Anopheles subpictus from this study and their sibling species for GenBank.

This tree was constructed using the Tamura 3-parameter substitution model with gamma distribution. Bootstrap values (1000 replicates) are shown near each branch (numbers in red). Anopheles minimus (KX668149) was used as an outgroup to root the tree. The bold font is samples in this study.