Anopheles maculatus s.l as potential malaria vectors with opportunistic feeding in Meghalaya, Northeast India

Abstract

A recent study conducted in West Khasi Hills (WKH) and West Jaintia Hills (WJH) districts of Meghalaya has documented a low occurrence of the primary malaria vectors, Anopheles baimaii and An. minimus, alongside a relatively higher abundance of other anophelines, particularly members of the Maculatus Group. Plasmodium infection in the An. maculatus sensu lato (s.l.) was last reported in 1941, with no subsequent implication of An. maculatus s.l. in malaria transmission from Meghalaya or Northeast (NE) India for over eight decades. Given this apparent gap between species abundance and its recognised vector status, the present study was undertaken to examine the species composition and vector potential of predominant species using molecular tools in the Khasi Hills and Garo Hills regions of Meghalaya. In this study, we employed molecular techniques for precise species identification and detection of parasites. Our findings confirm natural infections of Plasmodium falciparum in An. maculatus sensu stricto (s.s.) and An. pseudowillmori—two member species of the Maculatus Group in Meghalaya. Resting adult Anopheles mosquitoes were collected between 2019 and 2023 (excluding 2021 due to the COVID-19 pandemic) in the Khasi Hills and Garo Hills regions using a mouth aspirator and torchlight, both indoors and outdoors. Out of 215 specimens analysed, 177 (82.32%) were identified as An. maculatus s.s., 33 (15.34%) as An. pseudowillmori, and 5 (2.32%) as An. willmori. Notably, Plasmodium infection was detected in 49 (27.68%) of 177 An. maculatus s.s. specimens and 16 (48.48%) of 33 An. pseudowillmori specimens, reinforcing their potential vectorial role in malaria transmission. F_ST analysis revealed significant genetic differentiation between An. maculatus s.s. populations from the Khasi Hills Region (KHR) and Garo Hills Region (GHR), further supported by haplotype network analysis. Additionally, blood meal analysis of specimens collected from cattle sheds during evening resting hours indicated diverse host feeding patterns: 15% had fed on humans, 16% on cattle, and 17% on both. The presence of mixed blood meals suggests that An. maculatus s.s. exhibits opportunistic and flexible feeding behaviour. These results provide molecular evidence of malaria transmission by specific members of the Maculatus Group in Meghalaya, underscoring their epidemiological importance. The genetic divergence and feeding plasticity highlight the need for targeted vector control in the region.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-49468-1.

Introduction

The northeastern part of India comprises 8 states. Among these states, excluding Sikkim, malaria is a significant public health concern in the remaining seven states, which include Assam, Arunachal Pradesh, Meghalaya, Mizoram, Nagaland, Tripura, and Manipur¹. In recent years, the malaria problem has been primarily concentrated in three states, including Meghalaya (National Centre for Vector Borne Diseases Control, NCVBDC).

Meghalaya, one of the states in Northeast (NE) India, is situated on the Shillong plateau, a part of the Indian peninsula, and is primarily composed of three major hills: the Khasi, Garo, and Jaintia hills. Studies conducted in Meghalaya since 1998 demonstrated significant infections of both Plasmodium falciparum and P. vivax. P. falciparum has been historically predominant and mostly found in the Garo and West Khasi Hills (WKH), while overall less abundant P. vivax is more common in the Jaintia Hills and Ri-Bhoi district^2–8. Due to the combined effects of various factors, the epidemiology of malaria in Meghalaya is highly complex².

Earlier studies considered Anopheles minimus, An. fluviatilis and An. baimaii as primary vectors of malaria in NE India⁹, but later studies revealed An. fluviatilis to be a winter variant of An. minimus due to melanisation leading to misidentification as An. fluviatilis^10,11. However, some other species were also considered to play a secondary role in malaria transmission in NE India^12–14.

The recent study in Meghalaya on anopheline diversity, conducted between 2018 and 2021 in the West Khasi Hills (WKH) and West Jaintia Hills (WJH) regions, has reported the rarity of known vectors, An. baimaii and An. minimus and abundance of the Maculatus Group of mosquitoes⁶. However, there is no report of An. maculatus sensu lato (s.l.) as a vector from Meghalaya or the whole NE India in the last 80 years, with the last report of An. maculatus being implicated as a definite vector of malaria in the region dates back to a study conducted between 1930 and 1941^15,16. Mosquitoes belonging to the Maculatus Group are under genus Anopheles, subgenus Cellia and Neocellia Series. Currently, the Maculatus Group comprises 10 species, including An. dispar, An. greeni, An. pseudowillmori, An. willmori, An. dravidicus, An. maculatus, An. maculatus Javanese form, An. notanandai, An. rampae, and An. sawadongporni¹⁷. Based on the present findings, it can be assumed that in earlier studies carried out on the Maculatus Group, the exact species was unknown. In India, An. maculatus was earlier considered as a single species with a variety. However, Singh et al.. (2012)¹⁸ reported, for the first time, the presence of six members of this Complex, along with their detailed geographical distribution in the north-eastern part of India. Studies conducted in various places in Southeast Asia reported An. maculatus sensu stricto (s.s.), An. pseudowillmori, An. sawadwongporni, and An. willmori as an important vector of human malaria¹⁹. Recently, An. maculatus s.s. was implicated as a vector in the Alipurduar district of West Bengal, India, which is a corridor linking NE with the rest of India²⁰, and we have found it to be a potential vector in Tripura, a malaria-endemic state of NE India (ICMR-RMRC unpublished studies) as well. The level of vector competence and involvement in malaria transmission can vary among different populations and species of Anopheles mosquitoes.

Considering all these factors, it is important to explore the potential of different abundant species such as An. maculatus s.l., other than the known primary vectors, as a malaria vector in malaria-endemic zones of Meghalaya. Such studies are also important from the perspective of malaria elimination. In this study, we have investigated the archived samples collected by the State Entomologists of Meghalaya, focusing on suspected An. maculatus group species. As it is difficult to distinguish the sibling species morphologically and often, the samples are not intact, leading to further identification challenges^21,22, molecular methods were used to understand the species composition and their potential as malaria vectors in different villages under the Khasi Hills Region (KHR) and Garo Hills Region (GHR), the two malaria-endemic districts of Meghalaya.

Materials and methods

Study location

Adult Anopheles mosquitoes were collected between 2019 and 2023 (2021 excluded as no collection could be done due to COVID-19 pandemic situation) from two major geographical locations of Meghalaya State in NE India, namely KHR and GHR, as these two regions are the most malaria-endemic areas of Meghalaya⁸, with the former comprising of WKH, South West Khasi Hills (SWKH) and Eastern West Khasi Hills (EWKH) districts while West Garo Hills (WGH), South Garo Hills (SGH) and South West Garo Hills (SWGH) districts covering GHR. The villages Goreng A, Goreng B, Khalu B, Bokchung, Riangdim, Nongshilliang, Mawpyllun, Khonjoy, Mawkhap, Aradonga, and Riangchi were selected from KHR, and Salmanpara, Sanjengpara, Sanknigre, Darengre, Madragre, and Karukolgre were selected from the GHR. The selection of villages was based on high malaria incidence reported over the past five years.

Methodology for land use land cover (LULC) preparation

The LULC mapping was prepared using geometrically corrected Indian Remote Sensing satellite data, Cartosat-1 (2.5 m) and LISS-IV (5.8 m), in a GIS environment using ArcMap 10.8.1. Major LULC categories and subcategories were classified and delineated using visual interpretation techniques. These were updated using the latest data pertaining to the years 2019 to 2023 on the spatial layer that was initially prepared under NRSC/ISRO’s SIS-DP project (Space-based Information Support for Decentralised Planning at Panchayat level), which was a national initiative of preparing basic spatial layers useful in the planning process at the grassroots level. The mapping was done at a 1:10,000 scale. Field verifications were conducted by the project team to verify the accuracy of the interpreted data. Location of households gathered from GPS were also plotted on the map for analysis of the proximity of the households to the jhum fields/plantations/forests, etc.

Some of the major LULC categories classified and grouped under the present study are croplands, agricultural plantations, scrublands, forest lands, jhumlands, water bodies and rubber plantations. Crop lands (agricultural lands) are those lands which are used for the production of food crops. Crop areas appear in bright red to red in colour with varying shapes and sizes in a contiguous to non-contiguous pattern on False Colour Composite (FCC) images. Agricultural plantations are areas under agricultural tree crops planted, adopting agricultural/horticulture management techniques. Plantations appear in a dark-red to red tone of different sizes with regular shapes and patterns. Forest lands (or tree lands) are those areas with tree canopy cover of more than 10% in a broad ecological perspective, irrespective of whether they are within a defined forest boundary or not. Canopy covers more than 40% are categorised as dense forests while those which are less than 40% as open forests. Scrublands are deteriorated land with sparse vegetation, such as abandoned jhumlands. Rivers, streams, lakes and ponds are classified as water bodies. Shifting cultivation fields are those with current jhums, which are temporary croplands that involve alternation with fallow and forests.

Emphasis was given to the new area cleared for shifting cultivation or for new horticultural plantations, as well as rubber crops. Shifting cultivation fields could be identified visually based on their spectral signatures and by comparing the current year’s satellite data with that of the preceding years. The pre-burnt jhum plots have higher reflectance in both the red and Near Infrared (NIR) bands of the satellite data, while the burnt plots have low reflectance in the red as well as the NIR bands. The jhum fields appear in a light to dark grey tone amidst the red-toned forested area. The current year jhum are normally dark grey (pre-burn) or greenish (burnt) tone, and jhum fields with second year crop are of a lighter grey tone. The current jhumland prepared for current cropping, and the current jhum with second year cropping could also be identified by referring to the vegetation cover in the preceding year’s imagery. In the same procedure, the shifting cultivation maps for the years 2020 to 2022 were also prepared to understand the dominant land use and farming system in the study area over the years. Another land use which is dominating in the study area is the rubber plantations. Since the phenology of the rubber tree makes it sometimes difficult to accurately delineate the area covered by rubber visually, we made a digital approach to map the area under rubber in the study area by taking advantage of the high reflectance in the NIR band. We used Sentinel-2 data (10 m resolution) of 2021 for mapping the area under rubber.

Mosquito collection

Adult female Anopheles mosquitoes, resting both indoors and outdoors, were collected from the specified study sites using a hand catch with a mouth aspirator and torch light during the evening hours of 6:00 to 9:00 pm. Adult mosquito samples were collected randomly from both indoor and outdoor human dwellings where malaria cases had been reported. Indoor resting collections were carried out for 1 h in selected human dwellings, all of which had adjoining cattle sheds. Outdoor 1-hour collections were conducted mainly in and around these cattle sheds, as well as at other potential resting sites, such as surrounding vegetation, bamboo fencing, and outside houses (Supplementary Fig. 1). The cattle sheds identified during the survey were simple, open-sided structures. They generally consisted of a roof made of tin or local materials, supported by wooden or bamboo posts, without any side walls. This open design is common in the region, allowing for increased airflow, easy cleaning, and convenient access for both cattle and caretakers. In KHR, 28 sampling areas were visited, with 16 from endemic areas and 12 from non-endemic areas. In GHR, 7 sampling areas were visited, i.e., 5 from endemic areas and 2 from non-endemic areas. A total of 1095 Anopheles samples were collected (508 from GHR and 587 from KHR).

Identification of mosquito species

The mosquito specimens were identified up to the species level by morphological and molecular methods. Besides morphological methods, molecular methods were used for the species belonging to a complex group whose morphological characters were almost indistinguishable.

Identification based on morphology

Collected Anopheles species were morphologically identified using the keys of Das and Rajagopal (1990)²³, using a Leica S APO Stereo Zoom microscope. Anopheles species belonging to the Maculatus Group were identified at the group level using this key. Maculatus Group identifying character involves (i) femur and tibia speckled, hind tarsomere 5 and part of 4 white and (ii) female palpi with one narrow and two broad pale bands (Supplementary Fig. 2).

Mosquito dissection for molecular studies

The head-thorax and abdomen of each mosquito were carefully dissected by separating the abdomen anterior to the junction of the thorax and abdomen (Supplementary Fig. 3) so that the abdomen part does not contaminate the head and thorax, thus showing false positivity²⁴ for Plasmodium. The middle legs of the specimens were used for species-level identification by PCR. DNA extraction was performed utilising the QiAmp DNA Extraction Kit (Qiagen), adhering to the instructions provided by the manufacturer and final elution was done in 50 µL. Finally, the quantification of DNA samples was performed using an Eppendorf BioSpectrometer^® Basic.

Identification based on molecular studies

Of the 1095 archived Anopheles collected from the GHR and KHR regions, 703 Anopheles specimens were An. maculatus s.l. Unfortunately, due to the extremely humid climate and high rainfall in the study area, fungal growth occurred in several preserved specimens, compromising their integrity and preventing further molecular processing. Therefore, among the 703 specimens of the Maculatus Group, only 215 were found in proper condition and were transported to our main laboratory at ICMR-RMRC, Dibrugarh, for molecular studies. Species belonging to the Maculatus Group were identified using previously described molecular methods²⁰.

Allele-specific PCR assays for identifying Maculatus Group species

The species belonging to the Maculatus Group were identified using the allele-specific PCR method of Walton et al. (2007)²⁵. The PCR reactions were prepared in a final volume of 25 µL, which included 12.5 µL 2X Promega master mix, 0.5 µM each of the universal forward and reverse primers, 50 ng of DNA template, and nuclease-free water to make up the remaining volume. All amplifications were carried out on a Bio-Rad C1000 thermal cycler. Using this protocol, five species of the Maculatus Group can be identified based on species-specific amplicon sizes: 180 bp for An. maculatus s.s., 203 bp for An. pseudowillmori, 301 bp for An. rampae, and 477 bp for An. dravidicus and 242 for An. sawadongporni. Regarding the thermal profile, we adhered to reference procedures used by our lab in earlier studies²⁰. Molecular analysis was used to validate morphologically identified specimens, especially because a large number of specimens were aged and lacked many distinguishing features. Supplementary Table S1 contains information about the PCR primers utilised in each molecular assay.

ITS 2 sequencing for identifying the species

The Internal Transcribed Spacer 2 (ITS2) region, a non-coding segment between the 5.8 S and 28 S rRNA gene, was amplified²⁵ and sequenced for those Anopheles species samples that didn’t get amplified for the Maculatus Group and COI PCR. Using NCBI BLAST searches, ITS sequences were compared to similar sequences retrieved from the GenBank database. MEGA X software was used to create the maximum likelihood phylogenetic hierarchy by applying the bestfit model (T92 + G) with a bootstrap value of 1000.

COI sequencing for identifying the species

Applied Biosystem 3500 Genetic Analyzer was used to amplify and analyse the mitochondrial Cytochrome Oxidase I (COI) gene²⁶. Using NCBI BLAST searches, COI sequences were compared to corresponding sequences retrieved from the GenBank data repository. The MEGA X 10.2.6 software version was used to generate the maximum likelihood phylogenetic tree by applying the bestfit model (T92 + G) with a bootstrap value of 1000.

COI haplotype network and genetic diversity analysis

DnaSP²⁷ and PopART v.1.7 software²⁸ were used to analyse the haplotype network, which was generated using the minimum spanning network method for the COI gene. This contained COI sequences of An. maculatus s.s. and sibling species obtained from GenBank. Haplotype network analysis was performed on 79 sequences, comprising 58 from this study (GenBank accession numbers (nos.): PQ333934-PQ333992). The statistics of Genetic Diversity, such number of segregation sites (S), haplotypes (h), haplotype diversity (Hd), nucleotide diversity (π), average number of nucleotide differences (k), neutrality tests of Tajima’s D, Fu and Li’s D test statistic, and Fu and Li’s F test statistic were obtained for COI sequences. DnaSP²² has been used to estimate pairwise genetic distance (F_ST) from the COI sequences of KHR and GHR for An. maculatus s.s. and sibling species at both sites compared with countries such as Thailand and Myanmar. 10 sequences from each site were taken and compared for genetic diversity. Pairwise mean genetic distances at the amino acid level (COI gene region) between KHR1, KHR2 and GHR were computed using the p-distance model in MEGA X (Supplementary Table S2). The p-distance method calculates the proportion of amino acid sites at which two sequences differ, excluding gaps and missing data. This method provides a straightforward measure of genetic divergence suitable for intraspecific and interpopulation comparisons.

Detection of Plasmodium in mosquitoes

To detect the Plasmodium sporozoites within mosquitoes, DNA was extracted from the head-thorax region of each Anopheles specimen. The cytochrome b gene was analysed using nested quantitative PCR (nqPCR)²⁹ with modifications as previously described³⁰. For the primary (N1) PCR, each 10 µL reaction contained 5 µL of 2X GoTaq qPCR master mix (SYBR Green) (Promega), 250 nM of each forward and reverse primer, 50 ng of DNA template, and nuclease-free water to bring the final volume to 10 µL. For the nested (N2) PCR, conducted for both P. falciparum and P. vivax, respectively, the reaction mixture similarly consisted of 5 µL of 2X GoTaq qPCR master mix (SYBR Green) (Promega), 250 nM of each primer, and 50 ng of DNA template prepared as a 1:10 dilution, with the final volume adjusted to 10 µL. The details of the PCR primers used for blood meal analysis are given in Supplementary Table S1. The Tm range used for distinguishing Pf was 78.6–79.6 °C, while for Pv it was 74.8–75.8 °C, and the samples’ Ct value did not exceed < 35. The PCR results were verified through melt curve analysis and subsequently confirmed by agarose gel electrophoresis. As a negative control, DNA isolated from the head-thorax of Aedes, Culex, and the abdomen of Anopheles was used, while the 3D7 strain, obtained from parasite culture, served as the positive control for Pf and Pv positive blood sample, as the Pv control. The abdomen of the Anopheles mosquito served as the negative control to validate the precision of bisection, ensuring no abdominal part remained in the head and thorax region, hence any parasite presence found in the head and thorax is exclusively from the salivary glands, indicative of vectorial potential.

Human blood meal detection

PCR was used to detect the blood of the host, including human, cattle, and dog, in the abdomen of all Maculatus Group mosquitoes, regardless of whether they were obvious blood feeders, as the samples used were stored for a long time; hence, the blood-fed status was not palpable. The multiplex PCR method described by Pitzer et al. (2011)³¹ was followed with slight modifications. Each 10 µL reaction contained 5 µL of 2X Promega master mix, 0.6 µM of the universal reverse primer, host-specific forward primers including Cattle-F (0.2 µM), Dog-F and Hum-F (0.1 µM), 50 ng of DNA template extracted from the abdomen, and nuclease-free water to make up the remaining volume. The details of the PCR primers used for blood meal analysis are given in Supplementary Table S1.

Results

Composition of Anopheles species by morphological methods in the study sites

Examination of the collected mosquito species confirmed a greater abundance of An. maculatus s.l., coupled with the absence of known vectors. There was a minimal collection from indoor resting areas in human dwellings in both GHR and KHR, with almost all collections made from cattle sheds and outdoor locations (Fig. 1, Supplementary Fig. 4). Out of the 1095 morphologically identified Anopheles, the majority were An. maculatus s.l. (n = 702, 64.10%) with the rest being other species (each species < 10%) (Table 1; Fig. 2). The KHR region had more overall An. maculatus s.l. with 358 (78.33%) in WKH, 77 (82.79%) in SWKH, 34 (91.89%) in EWKH, whereas the GHR region had 178 (57.41%) in WGH, 38 (40.42%) in SWGH, 17 (16.34%) in SGH. Morphological examination of the collected specimens revealed that the Maculatus Group of specimens was more prevalent than other anophelines (Table 1; Fig. 2).

Fig. 1.

Map showing the locations of the mosquito sampling sites in the study villages of KHR & GHR in Meghalaya. The map shows the location of the study sites composed on a GIS platform (ESRI ArcGIS Enterprise 10.8). The location points are shown with an elevation map in the background prepared using ASTER (Advanced Spaceborne Thermal Emission and Reflection Radiometer) Global Digital Elevation Model (GDEM) Version 3.

Table 1.

Total Anopheles composition of indoor and outdoor hand collection in the study villages of KHR and GHR, Meghalaya.

Species	Garo Hills			Khasi Hills			Total (%)
	West Garo Hills (%)	South West Garo Hills (%)	South Garo Hills (%)	South West Khasi Hills (%)	Eastern West Khasi Hills (%)	West Khasi Hills (%)
An. maculatus s.l.	178 (57.41)	39 (41.48)	17 (16.34)	77 (82.79)	34 (91.89)	358 (78.33)	703 (64.20)
An. vagus	36 (11.61)	27 (28.72)	30 (28.84)	1 (1.07)	0 (0)	2 (0.43)	96 (8.76)
An. annularis	33 (10.64)	2 (2.12)	6 (5.76)	0 (0)	0 (0)	31 (6.78)	72 (6.57)
An. pallidus	0 (0)	3 (3.19)	3 (2.88)	0 (0)	1 (2.70)	7 (1.53)	14 (1.27)
An. kochi	5 (1.61)	6 (6.38)	4 (3.84)	0 (0)	0 (0)	1 (0.21)	16 (1.46)
An. hyrcanus s.l.	12 (3.87)	0 (0)	26 (25)	0 (0)	0 (0)	0 (0)	38 (3.47)
An. minimus s.l.	3 (0.96)	8 (8.51)	18 (17.30)	1 (1.07)	0 (0)	7 (1.53)	37 (3.37)
An. culicifacies s.l.	4 (1.29)	0 (0)	0 (0)	0 (0)	0 (0)	9 (1.96)	13 (1.18)
An. phillipinensis/nivipes	3 (0.96)	0 (0)	0 (0)	0 (0)	0 (0)	2 (0.43)	5 (0.45)
An. baimaii	6 (1.93)	2 (2.12)	0 (0)	0 (0)	0 (0)	2 (0.43)	10 (0.91)
An. subpictus	0 (0)	0 (0)	0 (0)	1 (1.07)	0 (0)	0 (0)	1 (0.09)
An. jamesii	3 (0.96)	2 (2.12)	0 (0)	0 (0)	0 (0)	6 (1.31)	11 (1)
An. barbirostris s.l.	1 (0.32)	0 (0)	0 (0)	0 (0)	0 (0)	7 (1.53)	8 (0.73)
An. aconitus	0 (0)	3 (3.19)	0 (0)	0 (0)	0 (0)	0 (0)	3 (0.27)
An. karwari	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	3 (0.65)	3 (0.27)
An. pseudojamesii	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	2 (0.43)	2 (0.18)
Unidentified Anopheline species *	26 (8.38)	2 (2.12)	0 (0)	13 (13.97)	2 (5.40)	20 (4.37)	63 (5.75)
Total	310	94	104	93	37	457	1095

*Ambiguous species.

Fig. 2.

(a): Map showing the land use land cover pattern (LULC) of villages of GHR, Meghalaya and the proportion of Anopheles mosquitoes collected in this study. (b): Map showing the LULC of villages of KHR, Meghalaya and the proportion of Anopheles mosquitoes collected in this study. (c): Map showing the LULC of districts of GHR and KHR, Meghalaya and the proportion of Anopheles mosquitoes collected in this study, respectively. The maps show the location of the study sites composed on a GIS platform (ESRI ArcGIS Enterprise 10.8). The location points are shown with an elevation map in the background prepared using ASTER (Advanced Spaceborne Thermal Emission and Reflection Radiometer) Global Digital Elevation Model (GDEM) Version 3.

Identification of species belonging to the Maculatus Group using molecular methods

Molecular methods were undertaken using species-specific multiplex PCR, which revealed An. maculatus s.s. (~ 180 bp) and An. pseudowillmori (~ 203 bp) corresponding with their PCR band size (Supplementary Fig. 5)¹⁸. For samples that didn’t amplify in group-specific PCR, the COI and ITS genes were amplified and sequenced for identification. In addition to An. maculatus s.s. and An. pseudowillmori, An. willmori was detected in KHR; however, no band was observed in the group-specific PCR, as An. willmori–specific primers were not included in the reference protocol.

Overall, molecular identification revealed that out of 215 samples, 177 were An. maculatus s.s., 33 An. pseudowillmori and 5 An. willmori.

COI & ITS sequencing along with phylogenetic analysis

58 Anopheles samples (GenBank accession nos. PQ333934-PQ333992) amplified for the COI gene and 42 samples (GenBank accession nos. PQ350015-PQ350056) for the ITS2 region, both from the KHR and GHR study villages, were sequenced by the bi-directional Sanger sequencing method, and a phylogenetic tree was constructed to assess evolutionary relationships. Phylogenetic analysis of COI (Fig. 3) and the ITS2 (Fig. 4) region of three species of the Maculatus Group, An. maculatus s.s., An. pseudowillmori and An. willmori in the KHR and GHR regions of Meghalaya was carried out. In the COI tree, the major clade of An. maculatus s.s. is supported by a bootstrap value of 98, An. pseudowillmori by 100 and An. willmori by 96, whereas in the ITS tree, all three species’ major clades are supported by the bootstrap value of 99. An. maculatus s.s. COI tree (Fig. 3) was composed of a large, diffuse cluster containing samples from Meghalaya, Myanmar, and Thailand. Many nodes are poorly supported (bootstrap values range from 37 to 68), indicating high intra-species diversity. In the ITS2 tree (Fig. 4), all the sequences fell into the same clade, except for the Sri Lankan sequences. Similarly, in the An. pseudowillmori node of the COI tree, multiple sub-clustering was observed compared to the ITS2 tree, suggesting more mitochondrial diversity compared to the nuclear ITS2 gene. An. willmori was clearly resolved as a monophyletic group in both markers with high confidence.

Fig. 3.

COI tree of Anopheles sp. collected from Meghalaya, constructed by using the maximum likelihood method applying a bootstrap value of 1000. Red-coloured sequences are from our present study.

Fig. 4.

Phylogenetic analysis of ITS sequences of Anopheles sp. collected from Meghalaya, constructed using the maximum likelihood method applying a bootstrap value of 1000. Red-coloured sequences are from our present study.

Further phylogenetic analysis of An. maculatus s.s. using the maximum likelihood method for both KHR and GHR revealed that the An. maculatus s.s. from KHR forms a separate clade, indicating a degree of genetic divergence from those in GHR. However, a few An. maculatus s.s. from KHR closely match those from GHR. The genetic diversity of An. maculatus s.s. in both regions can be interpreted from Fig. 5. The tree shows two well-defined species clusters: a large An. maculatus s.s. clade containing KHR1, KHR2, KHR3 and GHR sequences along with Thailand and Myanmar sequences, and a clearly separate An. pseudowillmori clade forming an outgroup. Within An. maculatus s.s., KHR1, KHR2, and KHR3 form distinct clades, indicating the presence of multiple circulating haplotypes in the region.

Fig. 5.

Phylogenetic analysis of COI sequences of An. maculatus s.s. (KHR1 & KHR2 in red colour, KHR3 in yellow colour) collected from Meghalaya, constructed using the maximum likelihood method, applying a bootstrap value of 1000. Here, An. pseudowillmori has been used as an outgroup.

COI haplotype network analysis

There were 13 new haplotypes of An. pseudowillmori from KHR (Hap_1, Hap_2, Hap_3, Hap_4, Hap_5, Hap_6, Hap_7, Hap_8, Hap_9, Hap_10, Hap_11, Hap_12, Hap_13 detected. There was a shared haplotype between An. pseudowillmori sequence from this study and the sequences from Alipurduar (West Bengal) formed the Hap_14, which has been reported earlier (Rajkonwar et al.)¹⁷(Fig. 6). Similarly, 4 new haplotypes of An. maculatus s.s. from KHR (Hap_5, Hap_7, Hap_10 & Hap_11) and 3 new haplotypes from GHR (Hap_1, Hap_3 & Hap_4) were detected in the analysis. The shared haplotypes between An. maculatus s.s. from this study and the sequences from Thailand and Myanmar are Hap_2, Hap_6, Hap_8 & Hap_9 (Fig. 7). A star-like haplotype network (Fig. 8) was observed when sequences of An. maculatus s.s. from both KHR (KHR1 and KHR2) and GHR were analysed together, indicating population expansion within the region.

Fig. 6.

Haplotype network analysis of An. pseudowillmori, collected from KHR, Meghalaya.

Fig. 7.

Haplotype network analysis of An. maculatus s.s. collected from KHR and GHR, Meghalaya.

Fig. 8.

Haplotype network analysis of An. maculatus s.s. collected from KHR (KHR1 and KHR2) and GHR, Meghalaya.

Genetic diversity and tajima D test in An. maculatus s.s. species

Based on the COI gene analysis of the twenty-nine An. maculatus samples from India and the thirty-two retrieved sequences from GenBank, the molecular analysis revealed that the number of segregating sites (S) was slightly higher in the Indian population compared to those from other countries (Supplementary Table S3). Within India, the mosquito population from the KHR1 region in Meghalaya showed more segregating sites than those from the KHR2 and GHR regions (Supplementary Table S4). However, the average number of nucleotide differences (K) was found to be slightly lower in the Meghalaya population (K = 1.30542) than in other populations. Overall, the genetic diversity index showed that An. maculatus population from all geographical regions indicated higher haplotype diversity (Hd) and lower nucleotide diversity (π) among them; the Indian population represented Hd = 0.7167, and π = 0.0021, slightly lower than the other two populations from Thailand and Myanmar.

In India, specifically in Meghalaya, the genetic diversity assessment (Supplementary Table S4) among three subpopulations from the KHR1, KHR2, and GHR regions revealed that An. maculatus s.s. populations from the KHR1 region showed high haplotype diversity (Hd = 0.20). Overall, the nucleotide diversity value for all three populations is low, but in KHR1, the value is slightly higher (Supplementary Table S4). The neutrality test values among the three populations were negative for Tajima’s D, and the values for all populations were negative but not significant.

Calculation of F_ST value in An. maculatus s.s. species

The calculation for the pairwise Fixation index F_ST among the population of An. maculatus s.s. from KHR1, KHR2, GHR, Thailand and Myanmar regions based on COI sequences were depicted in Table 2. The F_ST values range from 0.02083 to 0.29508. Surprisingly, we found the highest F_ST value for the KHR1-KHR2 population, followed by KHR1-GHR, which was equal to or greater than 0.25, indicating higher genetic differentiation.

Table 2.

F_ST analysis of An. maculatus s.s. populations from the KHR and GHR regions of Meghalaya, Thailand, and Myanmar.

Locations	Meghalaya_KHR1	Meghalaya_KHR2	Meghalaya_GHR	Thailand	Myanmar
Meghalaya_KHR1	0.00000
Meghalaya_KHR2	0.29508	0.00000
Meghalaya_GHR	0.29028	0.02083	0.00000
Thailand	0.17905	0.11475	0.13844	0.00000
Myanmar	0.09380	0.18555	0.20238	−0.06667	0.00000

Plasmodium infections and blood meal positivity among the members of the Maculatus Group

Molecular studies using qPCR revealed high Plasmodium positivity for P. falciparum in both KHR and GHR, with only one P. vivax infection detected in GHR. No mixed infections were found. From both regions, 64 Plasmodium infections were found, out of which only Pf infections were found in An. maculatus s.s. (27.52%) and An. pseudowillmori (45.45%) except for one Pv infection found in An. maculatus s.s. from GHR (Table 3, Supplementary Fig. 6,7, and 8).

Table 3.

Distribution, Plasmodium positivity, and Blood meal analysis of An. maculatus s.s. and An. pseudowillmori of Maculatus Group collected from KHR and GHR for the year 2019 to 2023 (2021 excluded as no collection could be done due to the COVID-19 pandemic situation).

Endemicity	District	PHC (Block)	SC	Village	Group [Total collected] (Detected by the morphological method)	No. of individuals used for molecular analysis	Species (detected by molecular method)	Plasmodium positives (positivity%)	Blood meal positivity			Human blood positives (total tested, positivity%)
									Human	Cattle	Cattle + Human
Endemic	SWGH	Salmanpara (Zikzak)	Salmanpara National	Salmanpara	An. maculatus s.l. [39]	10	An. maculatus s.s (10)	2 (20) * 1 Pv	4	1	4	8 (9, 88.88)
	WGH	Purakhasia (Dalu)	Halchati	Sanjengpara	An. maculatus s.l. [28]	3	An. maculatus s.s (2)	2 (100)	1	0	0	1 (1, 100)
							An. pseudowillmori (1)	0	0	0	1	1 (1, 100)
		Kherapara (Dalu)	Sanknigre SC	Sanknigre	An. maculatus s.l. [25]	3	An. maculatus s.s (3)	0	0	0	3	3 (3, 100)
		Darengre (Gambegre)	Gambegre SC	Darengre	An. maculatus s.l. [18]	0	—	0	0	0	0	0 (0, 0)
	WKH	Shallang (Mawshynrut)	Riangdim	Goreng A	An. maculatus s.l. [25]	3	An. maculatus s.s (3)	0	0	0	0	0 (2, 0)
				Goreng B	An. maculatus s.l. [4]	4	An. maculatus s.s (1)	1 (100)	0	1	0	0 (1, 0)
							An. pseudowillmori (3)	3 (100)	0	1	0	0 (3, 0)
				Khalu B	An. maculatus s.l. [98]	60	An. maculatus s.s (59)	16 (27.11)	11	3	3	14 (58, 24.13)
							An. pseudowillmori (1)	0	0	0	0	0 (1, 0)
				Bokchung	An. maculatus s.l. [75]	16	An. maculatus s.s (16)	7 (43.75)	3	2	0	3 (15, 20)
				Riangdim	An. maculatus s.l. [47]	22	An. maculatus s.s (21)	11 (52.38)	2	3	1	3 (20, 15)
							An. pseudowillmori (1)	1 (100)	1	0	0	1 (1, 100)
				Riangchi	An. maculatus s.l. [10]	2	An. maculatus s.s (2)	0	0	2	0	0 (2, 0)
		Maweit (Nongstoin)	Porkhadoh	Nongshilliang	An. maculatus s.l. [78]	74	An. maculatus s.s (47)	6 (12.24)	0	9	5	5 (33, 15.15)
							An. pseudowillmori (22)	10 (45.45)	0	7	9	9 (20, 45)
							An. willmori (5)	0	0	0	2	2 (2, 100)
Non endemic	WGH	Purakhasia	Halchati	Madragre	An. maculatus s.l. [107]	6	An. maculatus s.s (6)	0	4	0	1	5 (6, 83.33)
	SGH	Baghmara	Karukolgre	Karukolgre	An. maculatus s.l. [17]	0	—	0	0	0	0	0 (0, 0)
	SWKH	Ranikor (Ranikor)	NA	Mawpyllun	An. maculatus s.l. [45]	0	—	0	0	0	0	0 (0, 0)
		Khonjoy (Ranikor)	NA	Khonjoy	An. maculatus s.l. [32]	2	An. maculatus s.s. (1)	1 (100)	0	1	0	0 (1, 0)
							An. pseudowillmori (1)	0	0	1	0	0 (0, 0)
	EWKH	Wahrit (Mairang)	NA	Mawkhap	An. maculatus s.l. [34]	9	An. maculatus s.s. (5)	3 (60)	1	0	2	3 (5, 60)
							An. pseudowillmori (4)	2 (50)	2	0	2	4 (4, 100)
	WKH	Aradonga (Mawshynrut)	NA	Aradonga	An. maculatus s.l. [21]	1	An. maculatus s.s. (1)	0	0	0	0	0 (0, 0)
Total					An. maculatus s.l. [703]	215	An. maculatus s.s. (177)	49 (27.68) *1Pv	26	22	19	45 (156, 29.22)
							An. pseudowillmori (33)	16 (48.48)	3	9	12	15 (31, 48.38)
							An. willmori (5)	0	0	0	2	2 (2, 100)

* 1 Pv from West Garo Hills.

Blood meal analysis revealed that 15% of the specimens contained human blood in their abdomen, 16% contained cattle blood, and 17% contained both cattle and human blood (Table 3).

Ecological description along with molecular distribution of Maculatus Group species

The major LULC pattern of the study sites is described as follows:

GHR

In SWGH, Salmanpara (Zikzak block) has 58% forest cover, 11% scrublands, 23% agriculture, 1% jhumlands, and 6% built-up areas. In WGH, Sanjengpara (Dalu block) has 80% forest, 7% scrubland, 9% agriculture, and 3% built-up areas. Sanknigre (Gambegre block) has 77% forest, 2% scrubland, 16% built-up, and 2% jhumland areas. Darengre (Gambegre block) has 86% forest cover and 13% built-up areas. Madragre (Dalu block) has 73% forest, 11% scrubland, 10% agriculture, 4% built-up, and 2% jhumland areas. In SGH, Karukolgre (Baghmara block) has 95% forest cover, 3% jhum lands, and 2% built-up areas (Fig. 2).

KHR

In WKH, Aradonga (Mawshynrut block) has 89% forests and 11% built-up areas. Bokchung (Mawshynrut block) has 89% forest cover, 9% built-up areas, and 2% scrublands. Goreng A (Mawshynrut block) has 93% forest, 3% scrubland, 1% jhumland, and 3% built-up areas. Goreng B (Mawshynrut block) has 95% forest cover, 3% jhum lands, and 2% built-up areas. Khalu B (Mawshynrut block) has 98% forest cover and 1% built-up area. Riangdim (Mawshynrut) has 84% forest, 9% built-up areas, 2% scrubland, and 5% jhumland. Riangchi (Mawshynrut) has 97% forest cover, 2% scrubland, and 1% jhumland. Nongshilliang (Nongstoin block) has 81% forest cover, 9% scrubland, 6% agricultural land, and 4% built-up areas. In SWKH, Khonjoy (Ranikor block) has 92% forest, 4% scrubland, 2% built-up, and 1% jhumland areas. Mawpyllun (Ranikor block) has 44% forest cover, 39% scrubland, and 4% built-up areas. In EWKH, Mawkhap (Mairang block) has 78% forest cover, 12% agriculture, 7% scrubland, and 2% built-up areas (Fig. 2).

Discussion

Our study revealed five significant findings: a greater diversity of anophelines in GHR compared to KHR; an overall and specifically in KHR, the preponderance of the Maculatus Group, albeit both observations having the limitation of only cross-sectional resting collection; genetic diversity within the Maculatus Group from GHR and KHR; the potential vectorial role of the Maculatus Group as revealed by a highly sensitive qPCR method; and the opportunistic blood-feeding status of the Maculatus Group members.

In Meghalaya, 1095 anophelines were collected during the resting mosquito collection, covering six districts of KHR and GHR, and morphologically identified to the group or species level using the identification key. Overall, the Maculatus Group emerged as the most dominant one, comprising ~ 75% of the total catch. In the KHR, An. maculatus s.l. comprised ~ 80% of the total anophelines (the rest were An. vagus, An. annularis, An. pallidus, An. kochi, An. minimus, An. culicifacies s.l., An. phillipinensis/nivipes, An. baimaii, An. subpictus, An. jamesii, An. barbirostris s.l., An. karwari, An. pseudojamesii), whereas in GHR, along with An. maculatus s.l., which constituted ~ 46% of all the species collected, ~ 12 more anophelines (An. vagus, An. annularis, An. pallidus, An. kochi, An. hyrcanus s.l., An. minimus, An. culicifacies s.l., An. phillipinensis/nivipes, An. baimaii, An. subpictus, An. jamesii, An. barbirostris s.l.) were found.

Our study was not longitudinal and was based exclusively on resting collections, which represents a limitation. However, a similar preponderance of the Maculatus Group, along with the rarity of the primary vectors and low numbers of other species, was found in another study in the same area, which was a longitudinal collection involving both light traps and resting between 2018 and 2021⁶. We recorded several additional species—An. vagus, An. pallidus, An. kochi, An. culicifacies, An. philippinensis, An. karwari and An. pseudojamesii, though in very few numbers, which were not reported in the earlier study. Conversely, species such as An. hyrcanus, An. xui, An. nivipes, An. splendidus, An. jamesii, and An. nitidus, previously detected, was absent from our collections. Despite these differences, several species were consistently observed across both studies, including An. maculatus, An. pseudowillmori, An. willmori, An. annularis, An. barbirostris s.l., and An. jeyporiensis. An exception was An. jeyporensis, which was recorded in much higher numbers in their study than in ours. This difference may be attributed to differences in methodology, as approximately 90% of their collections used light traps, whereas our study relied exclusively on resting collections. The identification of these anophelines was conducted morphologically using the identification keys of Das and Rajagopal²³ up to the species level (An. minimus, An. annularis, An. subpictus, An. vagus, An. pallidus, An. kochi, An. baimaii, An. jamesii, An. karwari, An. pseudojamesii) as well as in groups (An. maculatus s.l., An. barbirostris s.l., An. hyrcanus s.l., An. culicifacies s.l., An. phillipinensis/nivipes) where identification at the species level is not possible due to overlapping characters (Table 1).

The ecological characteristics of the two regions contrasted significantly: GHR villages were situated in low-lying areas (below 200 m) with reduced rainfall (under 300 cm) and featured mixed deciduous forests, whereas KHR was located at higher elevations (ranging from 350 to 1380 m), received greater rainfall (between 350 and over 500 cm), and boasted denser or specialised forest types, such as evergreen and hill-pine forests (Supplementary Fig. 9). The patterns of species were indicative of these environmental contrasts. Both An. maculatus s.s. and An. pseudowillmori were found in both areas, although their abundances were significantly higher in the colder, forest-rich, high-rainfall KHR; An. maculatus s.s. accounted for 78–91% of specimens, whereas An. pseudowillmori predominated in heavily forested areas such as Nongshilliang and An. willmori was furthermore occurring exclusively in KHR. On the other hand, a more varied anopheline fauna with less dominance of these important vectors was supported by the lower-elevation, more diversified GHR. There is a marked ecological contrast between KHR and GHR in terms of rainfall, elevation, and forest cover. The maps clearly show a West-to-East gradient, with the Garo Hills (West) having lower richness (14.0) but higher evenness (0.65), and the Khasi Hills (East) having higher richness (17.0) but lower evenness (0.45) (Supplementary Fig. 10). We attempted to correlate these differences with the observed species patterns, and future studies will be crucial in determining whether these ecological factors contribute to species divergence between the two regions.

We additionally conducted molecular confirmation for 284 specimens as per feasibility. The concordance between morphological and molecular identification was high (> 98%) (Supplementary Table S5); hence, we can infer the anopheline composition difference between KHR and GHR (Table 1). Importantly, molecular analysis allowed a clear distinction of species within the Maculatus Group, which is not possible morphologically^25,32. At the species level, we could distinguish An. pseudowillmori, which was found primarily in WKH and KHR, and An. maculatus s.s., which was found in both GHR and KHR within the Maculatus Group; however, molecular identification was conducted on only a subset of the total collected specimens; therefore, we cannot infer about the overall composition. Nevertheless, it was observed that within the same subset, An. pseudowillmori was absent in GHR. Thus, we have observed a distinct phenomenon between GHR and KHR, both in overall anopheline species diversity and in the species composition within the Maculatus Group.

The COI phylogeny of An. maculatus s.s. revealed a large, diffuse cluster that included samples from Meghalaya, Myanmar, and Thailand. Several internal nodes have poor bootstrap support (37–68), indicating significant intra-specific variation. In contrast, the ITS2 phylogeny grouped all An. maculatus s.s. sequences into a single clade, with the exception of those from Sri Lanka. The COI tree of An. pseudowillmori revealed considerable sub-structuring, but the ITS2 tree showed a more compact clustering pattern. This pattern suggests greater mitochondrial diversity than the nuclear ITS2 gene region. Meanwhile, both the COI and ITS2 sequence analysis consistently revealed a well-supported monophyletic clade for An. willmori. Earlier studies by Upasana et al. (2023) have reported inconsistencies between GenBank and BOLD databases, leading to species-level mismatches due to the presence of misidentified sequences of closely related taxa⁶. In contrast, the present study observed no such mismatches, and species identifications were consistent across all databases (Supplementary Table S6). However, An. willmori could not be cross-validated using the BOLD database due to the absence of reference sequences, and its identification was therefore inferred using complementary phylogenetic analysis.

Futhermore, in genetic differentiation among An. maculatus s.s., a difference was observed between the KHR and GHR regions in the phylogenetic analysis, revealing variation among the species. The genetic diversity tools (haplotype and nucleotide diversity) are the major indicators used in population genetics studies³³. In our analysis, nucleotide diversity (π) value (Supplementary Table S3 and S4) was lower than haplotype diversity (Hd) across all the populations. Notably, the nucleotide diversity in the KHR2 and GHR populations was lower compared to the KHR1, Thailand, and Myanmar populations. The result may explain an increased number of low-frequency mutations, indicating a recent population expansion, followed by a small effective population size after a bottleneck event^6,33,34, consistent with the given Tajima’s D value (Supplementary Table S4). The Tajima’s D values observed in our study were statistically nonsignificant. However, we included these values in the discussion not to imply statistical significance, but to interpret the direction and trend of the estimates in conjunction with other genetic diversity indices (such as haplotype diversity (Hd), nucleotide diversity (π), and Fu and Li’s tests), and to explore possible demographic scenarios, such as population expansion or bottleneck events.

The observed high haplotype diversity (Hd > 0.5) (Supplementary Table S3 and S4) may be consistent with an increased mutation rate of mtDNA^35,36. In the KHR1 population, elevated haplotype diversity could be associated with ecological heterogeneity^33,35,37; However, the study findings remain speculative without further ecological or demographic analysis. Similarly, the lower Hd and nucleotide diversity (π) observed in KHR2 may be compatible with a founder event or reduced population size^35,38, but this interpretation remains uncertain. While it is possible that KHR2 originated from a subset of individuals derived from a larger KHR1 population³⁵, additional demographic modeling would be required to support this scenario. The non-significant and slightly negative Tajima’s D value observed in KHR1, Thailand, and Myanmar populations indicated that these populations were at mutation-drift equilibrium, maintaining a stable large population^39,40. In contrast, a stronger negative value for KHR2 and the Garo Hills population indicates an overall population increase following a more recent bottleneck or mutational selection^33–35,41.

The haplotype network exhibited a star-like topology, defined by a singular, high-frequency centre haplotype (Hap_2) from which several low-frequency haplotypes emanated. This shape signifies recent demographic growth, characterised by a rapid increase in the frequency of an ancestral haplotype and the accumulation of newly emerged mutations in the network’s periphery. The existence of multiple peripheral, tip haplotypes encircling a predominant type is commonly acknowledged as a sign of population expansion accompanied by substantial gene flow (Avise, 2000)⁴². The minor, linear side branches (e.g., Hap_4–Hap_7 and Hap_9–Hap_10) further suggest localised differentiation within the broader expanding population (Fig. 8).

Overall, the genetic diversity in the mosquito population was multifactorial³³, encompassing both internal causes, such as mutation⁴¹, and external causes related to ecological conditions that enable natural populations to adapt and survive in changing environments^33,41. The state of Meghalaya, a part of the Indian Peninsula with its unique hilly terrain, fertile forests and ecological features demonstrating significant infection of P. falciparum and P. vivax^3–6, with its mosquito diversity showing adaptation to alteration in environmental conditions swiftly, causes the genetic diversity within the population and gene flow level^33,43.

According to the pairwise F_ST value, high genetic differentiation might be achieved due to the large geographical distances between populations⁴⁴. When spatial separation increases, genetic similarity and gene flow between populations decrease, a phenomenon known as isolation by distance⁴⁵. In contrast, our study reveals a high F_ST value between KHR1 and KHR2 (0.29508), despite both populations being from the same region (Khasi Hills, Meghalaya). This unexpectedly high F_ST value suggests limited gene flow, long-term isolation, and potentially inbreeding within each population⁴⁶. Although F_ST is not a direct measure of genetic distance, it measures the inbreeding or relative increase in homozygosity resulting from population subdivision. Hence, it is therefore considered as a measure of genetic differentiation between subsamples⁴⁶. Although we acknowledge that we do not have values for additional parameters, such as the inbreeding coefficient and allele richness, we cannot attribute this genetic differentiation to inbreeding.

Nevertheless, the high divergence of An. maculatus s.s. between KHR1 and KHR2 indicates that they are genetically distinct populations, possibly due to ecological barriers such as forest cover, land type, river or differences in breeding habitat^35,37. Interestingly, KHR1 is more genetically similar to Myanmar (distance 0.09380) than to KHR2 (0.29508), which is unusual for two nearby populations and suggests that KHR1 may share ancestry with the Southeast Asian populations rather than with KHR2 or the GHR populations. In contrast, KHR2 is more similar to GHR (0.02083), which is likely due to local gene flow or recent divergence between these two populations in Meghalaya^37,39.

F_ST analysis also revealed that, based on the An. maculatus s.s. populations of KHR and GHR; both appear to be genetically isolated, as evidenced by the haplotype network analysis. The An. maculatus s.s. populations of KHR are genetically distant from GHR, Thailand and Myanmar. Haplotype network analysis revealed new haplotypes of An. maculatus s.s both from KHR and GHR, whereas An. pseudowillmori haplotypes were found only from KHR. Some haplotypes of An. maculatus s.s. were conspecific with Thailand and Myanmar, and An. pseudowillmori with Alipurduar, India²⁰. Overall, our findings clearly emphasise that KHR and GHR are genetically isolated, as indicated by their species composition. These samples have been preserved, and additionally, more sample collections are in process for future studies, including a large-scale study of microsatellite and next-generation sequencing (NGS) analysis, to obtain a more comprehensive understanding of genetic differentiation and genetic distance between the regions.

Nonetheless, given the above-mentioned limitations, we observed three unique patterns: (i) the difference in diversity of the anophelines found in KHR and GHR regions, (ii) species-level variability within the Maculatus Group in these two regions, and (iii) genetic variations of An. maculatus s.s. between the KHR and GHR regions. By investigating ecological and meteorological variations between these landscapes, we were able to understand why these regional patterns emerge and what factors likely drive the observed divergence. These insights also highlight priority areas and considerations for future studies within the Maculatus Group.

Maculatus Group emerged as the major mosquito group in our study; it became evident that Meghalaya harbours a substantial An. maculatus s.l. population, specifically in areas where malaria cases were being reported. Despite this, the species had not previously been recognised as a vector in the region. Studies were conducted in areas with reported cases, but where the primary known vectors were almost not found. This gap formed the basis for our decision to explore and reassess the vector scenario in these areas. This is the very first time its potential vectorial role has been indicated based on our findings. In Meghalaya, the Maculatus Group of Anopheles is more abundant in almost all the regions of the KHR and the GHR^6,8. An. maculatus s.s was detected as a malaria vector in southern Thailand, the Malaysian Peninsula, and Java^47,48. Similarly, An. pseudowillmori is regarded as a secondary malaria vector in northwestern Thailand^49,50. Recently, An. maculatus s.s. was reported to exhibit the highest P. vivax infection rate (~ 5%) in a study conducted in the Kokrajhar region of Assam⁵¹.

As this study was conducted in remote field areas, the dissection method for detecting sporozoites in mosquito salivary glands, which requires fresh specimens, was not operationally feasible. In this context, quantitative PCR (qPCR) was found to be a more suitable alternative due to its higher sensitivity compared to conventional methods, particularly for detecting low-level parasite densities⁵². We have explored the use of a TaqMan probe–based assay; however, the protocol is currently under optimisation and has not yet been fully standardised. This will be further refined and evaluated in future studies. Eventually, we employed nested quantitative PCR (nqPCR) using SYBR Green, targeting the mitochondrial COI gene. This method offered superior sensitivity and cost-effectiveness in identifying Plasmodium infections, including those with both high and low concentrations of sporozoites, effectively (Supplementary Figs. 6, 8). Many recent studies have found Plasmodium positivity in anophelines using the Canier et al. method^20,53–57. To ensure specificity, melt-curve analysis with distinct Tm ranges within Ct values < 35 for P. falciparum and P. vivax was observed and confirmed by gel electrophoresis. This approach provided us with insight into sensitivity and species-specific detection, while a quantitative assessment will be addressed in future studies using standardised controls. All the other precautions were taken to ensure that it is a vector and can play a potential role (Supplementary Figs. 2, 7, 8; Supplementary Table S5). In these areas, where malaria cases were high but primary vectors were rarely detected, the detection of An. maculatus s.s. with ~ 28% Plasmodium positivity and An. pseudowillmori with ~ 48% strongly suggests their involvement in local malaria transmission. In the absence of established vectors, the species that are found in high abundance, including An. maculatus s.s., An. pseudowillmori and An. jeyporiensis can be considered a likely candidate for transmitting malaria in the region⁶. In this respect, An. maculatus s.l are of particular concern as they were the most abundant species reported in those areas where P. falciparum cases were more prevalent^8,58,59. The presence of malaria cases in these areas, coupled with parasite positivity in An. maculatus s.l. and the absence of established primary vectors, suggests that An. maculatus s.s. may function as a potential vector or even serve as the primary vector in situations where the known primary vectors are missing. In such scenarios, should we consider it as a secondary vector or a potential vector transitioning towards a primary role?

A significant number of An. maculatus s.s. engaged in double feeding, indicating opportunistic feeding behaviour of An. maculatus s.s. The analysis revealed an interesting finding, as ~ 15% of specimens contained human blood in their abdomen, ~ 16% contained cattle blood, and ~ 17% contained both cattle and human blood. On analysing the blood meal result of human or human mixed blood positivity at the species level, ~ 29% was found in An. maculatus s.s., ~ 48% in An. pseudowillmori and 40% in An. willmori. In Malaysia, this species is more attracted to cattle than to humans, and bites outdoors in large numbers. But when cattle were fewer in number, they fed on humans in large proportions⁶⁰. Also, An. maculatus was found in an approximate ratio of 3:1 biting outdoors and indoors, respectively, suggesting exophagy⁶¹. Indoor residual spray (IRS) and insecticide-treated nets (ITN) target anthropophilic and endophagic mosquito species⁶². Even after total coverage of ITN or IRS, exophilic and exophagic vectors can still sustain some transmission. It was also noticed in Vietnam that long-term use of ITN induces higher biting densities of An. maculatus s.l. outdoors than indoors⁶³. The Maculatus Group of Anopheles is exophilic in NE India, but they can bite both indoors and outdoors. With the wide-scale use of ITN, the density of endophagic and anthropophilic primary vectors, An. baimaii and An. minimus species are now decreased. In such a situation, species exhibiting exophily, exophagy, and opportunistic feeding behavior like An. maculatus s.s. can play an important role in malaria transmission. This type of vector behavior creates hurdles for malaria elimination as observed earlier (Mouchat & Hamon, 1963)⁶⁴. These vectors can sustain some level of malaria transmission, termed “residual malaria,” even after achieving good coverage of ITN or IRS (Durnez and Cossemans, 2013)⁶².

The archived samples showed consistent molecular sensitivity across years, as reflected by the successful recovery and detection of both blood meals and Plasmodium infections. DNA concentration increased in more recent samples, with corresponding improvements in blood meal and parasite detection rates, indicating better preservation and sample quality (Supplementary Table S7). Despite relatively low DNA concentrations in older samples (2019–2020), Plasmodium positivity was still detected, demonstrating that archived specimens retained sufficient DNA for reliable molecular analysis. A direct comparison with freshly collected samples was not performed; however, the high overall detection rates (≈ 30% Plasmodium positivity) support the suitability of archived samples for sensitive molecular assays.

Overall, our findings demonstrate that morphological identification alone is insufficient for accurately resolving species identification in many of the morphologically similar and closely related groups. Integrating molecular tools is therefore essential for reliable species confirmation, especially in areas with high diversity and overlapping morphological traits.

Conclusion

Our study concludes that identifying the sibling species within the Maculatus Group, by both morphological and molecular methods, is essential for understanding the species composition in malaria-endemic zones of the state. Further analysis of Plasmodium positivity and blood meal sources provides valuable insights into their feeding behaviour, vector competence, and role in malaria transmission. Studying the blood-feeding patterns of An. maculatus s.s. established the ecological dynamics between mosquitoes, humans, and livestock. Factors such as proximity to human habitation, the availability of cattle in the vicinity, and the behaviour of mosquitoes play pivotal roles in determining blood meal preferences. By integrating molecular techniques with ecological research, targeted interventions can be developed to reduce malaria transmission and promote interdisciplinary collaboration for effective disease control.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization was done by IPB, DRB, AKS, and JR. Collections were done by BK and DMS. Morphological and molecular analyses were conducted by AKS, JR, GH, and PT. Data analysis was done by JR, PA, AKS, SC, RP, KN and IPB. Manuscript was written by DRB, IPB, AKS, JR and PA. Proofreading and edits were done by DRB, IPB, AKS, JR, KB, HK, CB, and SRH. All authors read and approved the final manuscript.

Funding

This work was funded by an Indian Council of Medical Research (ICMR) extramural grant NER/78/2020/ECD-I to IPB.

Data availability

The gene sequences in this study were deposited in the NCBI database under accession numbers PQ333934-PQ333992 and PQ350015-PQ350056.

Declarations

Competing interests

The authors declare no competing interests.

Institutional review board statement

Ethical permission for the work has been obtained by the Institutional Bioethics Committee. The study protocol was approved by the Institutional Ethical Review Committee on 02.11.2021 [RMRC/Dib./IEC (Human)/2020-21/293], which is according to the Declaration of Helsinki.