Systematic review of Anopheles abundance and meta-analysis of nonhuman primate malaria infection in mosquitoes in Thailand

Abstract

Background

Thailand has achieved significant progress in malaria elimination, with a reduction in annual parasite incidence from 0.53 to 0.22 per thousand in 2014 and 2024, respectively. Given the high diversity of Anopheles mosquito species, elimination efforts must be precisely targeted, taking into account the varied behaviors and vectorial capacities of different vector species. This study aims to systematically review and update the distribution, identification, bionomics, behavior, and a meta-analysis of nonhuman parasite infectivity among mosquitoes.

Methods

A comprehensive literature search was conducted in PubMed, Scopus, EBSCOhost, and Google Scholar (2013–2025) to identify studies on Anopheles species diversity, distribution, and zoonotic malaria infection in mosquitoes. The meta-analysis followed Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and was carried out using the metafor package in R.

Results

A total of 92 relevant papers were included from 811 accessed articles. Of these, most documented geographical distribution, followed by mosquito behaviors, molecular identification, and mosquito infectivity. The pooled mosquito infection prevalence for the present meta-analysis was 0.01 (95% confidence interval (CI) 0.00–0.03), indicating low nonhuman primate (NHP) malaria parasite infectivity. Using standard nested polymerase chain reaction (PCR), wild-caught Anopheles sawadwongporni, An. minimus, and An. dirus were incriminated as zoonotic malaria vectors, with a pooled infection prevalence of 0.2%.

Conclusions

This review highlights a critical need for targeted, context-specific vector control interventions that address the unique opportunistic feeding and resting behaviors of Anopheles species complexes and the bionomics of local vector species, alongside the high mobility of high-risk populations. Research involving hosts and vectors across different micro-spatial and temporal scales, observations of human behavior, and monkey–human interactions is needed to improve understanding of zoonotic malaria transmission.

Graphical Abstract

Background

Mosquitoes of the genus Anopheles are the only known vectors capable of transmitting malaria-causing Plasmodium parasites to humans. In Thailand, 79 Anopheles taxa have been reported, comprising 77 formally described species and two unnamed presumptive species [1]. Although many taxa occur within morphologically similar species complexes, only a limited number are recognized as primary or secondary malaria vectors [1]. These include species within the Dirus complex (i.e., An. dirus Peyton & Harrison and An. baimaii Sallum & Peyton), the Funestus group (i.e., An. minimus Theobald and An. aconitus Doenitz), and the Maculatus group (i.e., An. maculatus Theobald, An. sawadwongporni Rattanarithikul & Green, and An. pseudowillmori (Theobald)). Additional taxa within the Barbirostris and Sundaicus complexes and the Hyrcanus group are regarded as secondary or suspected vectors, reflecting uncertainty in their epidemiological roles and highlighting the importance of accurate species identification and evidence synthesis [1–3].

Thailand is one of eight countries identified by the World Health Organization (WHO) as having complex malaria transmission while remaining on a trajectory toward elimination by 2025 [4]. Although indigenous cases declined substantially from 32,480 in 2010 to 22,426 in 2021, recent increases from 6263 in 2022 to 9169 in 2023 have been observed, particularly along internationbal borders [5]. This resurgence has been driven by renewed Plasmodium falciparum transmission along the Thai–Myanmar border and a rising incidence of zoonotic Plasmodium knowlesi along the Thai–Malaysian and Thai–Cambodian borders [5]. Confirmed P. knowlesi cases increased from a single case in 2016 to 259 cases in 2023, with an additional 78 cases reported in 2024 [6].

The emergence and spread of NHP malaria, particularly P. knowlesi and P. cynomolgi, in Southeast Asian countries are strongly influenced by landscape change. Human-driven land-use alterations, including deforestation and conversion of forested areas to agricultural plantations, are major drivers of NHP malaria transmission, especially of these two zoonotic malaria species [7–9]. Such environmental changes create favorable conditions for Anopheles vectors by increasing vector abundance and facilitating closer spatial overlap between humans, macaque hosts, and mosquito populations [10]. Modified environments, particularly plantations, generate microclimatic conditions characterized by increased humidity, reduced wind, and shaded breeding sites that enhance vector survival and productivity [10–12]. Together, these dynamics highlight the importance of integrated One Health approach that address ecological, entomological, human, and simian behavioral drivers of NHP malaria in rapidly changing landscapes [7].

Human-driven landscape modifications, particularly deforestation, create ecological interfaces in which Anopheles mosquitoes and macaque monkeys, the natural reservoir hosts of P. knowlesi, coexist with human populations [13]. These altered environments promote zoonotic spillover by increasing spatial and temporal overlap among vectors, nonhuman primate hosts, and humans. In densely populated areas or intensively used areas adjacent to macaque habitats, this overlap increases mosquito–macaque contact rates and facilitates the transmission of malaria parasites from macaques to humans [10, 14].

Previous research on malaria in Thailand has primarily described epidemiological patterns, including prevalence or incidence, morbidity, and mortality associated with human Plasmodium parasites and the zoonotic parasite P. knowlesi [15]. Parallel advances in Anopheles systematics have underscored the importance of accurate vector identification for interpreting transmission dynamics and informing control strategies [1, 16]. Consequently, Anopheles taxonomy in Thailand has evolved from reliance on morphological characters alone to the incorporation of molecular and genetic tools. Techniques such as PCR-based assays (including AS-PCR, SCAR-PCR, RFLP-PCR, and multiplex PCR), DNA sequencing of nuclear and mitochondrial markers (ITS2, D3, and COII), and microsatellite markers have become essential for distinguishing morphologically similar sibling species that differ in vectorial capacity [1, 16, 17].

Thailand has a long history of malaria vector research, initially based exclusively on morphological identification [18–20]. A comprehensive survey conducted in 1975 documented 32 species and one subspecies [21]. Subsequent cytogenetic studies revealed the presence of isomorphic or cryptic species within several major Anopheles complexes and groups, including the An. dirus complex, the An. maculatus complex [22–25], the An. hyrcanus group [26], and the An. barbirostris and An. umbrosus species groups [27]. Building on this work, Rattanarithikul et al. [20] expanded the recognized Anopheles fauna in Thailand to 73 species, including newly described members within the Leucosphyrus group, the Sundaicus complex, and the Minimus complex based on morphological characters [20]. Molecular diagnostic tools, particularly multiplex PCR-based assays, were subsequently developed to facilitate routine identification of sibling species within these complexes [28–32].

Despite these advances, evidence relevant to NHP malaria transmission remains fragmented and the relative roles of different Anopheles species are incompletely synthesized. In particular, there has been a lack of a comprehensive, systematic integration of molecularly confirmed vector records, entomological infection data, and species-specific ecological information relevant to zoonotic malaria in Thailand. This systematic review therefore collates and critically appraises published evidence on the distribution and diversity of Anopheles species in Thailand, with emphasis on sibling species confirmed using molecular methods since 2013. In addition, a meta-analysis is conducted to estimate mosquito infectivity with NHP malaria parasites where data permit. By integrating these findings with recent evidence on vector bionomics, surveillance, and control, this study aims to clarify current knowledge, quantify available evidence, and identify priority gaps to support malaria elimination and management of residual and zoonotic transmission in Thailand.

Methods

Strategy of the literature survey

A systematic literature review was conducted following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [33, 34].

Databases used and search strategy

Peer-reviewed, published literature was reviewed using four electronic bibliographic databases: PubMed, Scopus, EBSCOhost, and Google Scholar. In addition, manual Google searches and screening of reference lists were performed to identify additional articles (Fig. 1).

Fig. 1.

PRISMA flow diagram showing database search results with numbers of studies included/excluded at each stage

Articles published from 2013 to 2025 were searched using the term “Anopheles” in combination with the Boolean operators AND/OR and the following keywords: “diversity,” “composition,” “Plasmodium,” “vector infectivity,” “malaria,” “simian malaria,” “knowlesi malaria,” and “Thailand” or “Thai” or “Greater Mekong Subregion.” Studies that only cited another article’s collections were excluded. All articles were exported as .csv files for Microsoft Excel, and duplicates were removed manually. EndNote version 21.5 was used to manage and cite references.

All studies were screened for relevance to the meta-analysis of mosquito sporozoite infections and scored as follows: (1) studies meeting all inclusion criteria (see below), (2) studies meeting all inclusion criteria except for some components required for the meta-analysis, or (3) irrelevant studies. The final numbers of studies included in the meta-analysis and the systematic review were 6 and 92, respectively (Fig. 1).

Scientific literature screening and data extraction

All selected articles were in English. Documents not fulfilling the following criteria were excluded: original research addressing bionomics, malaria parasite infectivity, insecticide resistance, or prevention and control of Anopheles vectors, with full-text availability. Reviewed documents included review and research articles, conference abstracts, mini-reviews, and short communications.

Data extraction was performed on articles that met these inclusion criteria: malaria parasite detection in Anopheles natural populations and infection or transmission in laboratory-reared mosquitoes. The following variables were compiled from each article: main author, publication date, study type (field, semi-field, or laboratory), study location, collection date, Anopheles species studied, mosquito sex, number of mosquitoes analyzed, human biting or landing rates, resting sites, number of mosquitoes per pool, number of pools positive for malaria parasites, parasite detection method, Plasmodium species, and taxonomic assignment.

Statistical analysis

All analyses were conducted in R version 4.5.1. A random-effects model meta-analysis was used to estimate the prevalence of P. knowlesi in Anopheles mosquitoes. The “metaprop” function in the “meta” package was used to calculate logit-transformed pooled prevalence and 95.0% confidence intervals, and results were visualized using the “forestplot” package in RStudio [35]. Heterogeneity was assessed using the I² statistic; higher values indicate greater heterogeneity across studies. Because individual study weights were not calculated, publication bias was not assessed [36].

Results

Molecular biosystematics and population genetic structure of Anopheles sibling complexes

Multiplex PCR assays targeting ITS2 are widely used to identify cryptic species within complexes or groups, including the Dirus complex (An. dirus, An. baimaii, An. cracens, An. nemophilous, and An. scanloni) [32, 37], the Maculatus group (An. maculatus s.s., An. dravidicus, An. pseudowillmori, An. rampae, and An. sawadwongporni) [37, 38], the Funestus group (An. minimus, An. harrisoni, An. aconitus, An. pampanai, and An. varuna) [29, 37, 39], and the Barbirostris complex (An. barbirostris, An. dissidens, An. saeungae, An. wejchoochotei, and An. barbirostris A3) [30, 40, 41]. A new primer set for Dirus complex species identification PCR (DiCSIP) has been developed to improve the specificity, operational range, and sensitivity for five complex members in the Greater Mekong Subregion. The assay was validated using field-collected specimens morphologically identified as the An. dirus complex, with ten specimens from each of four Thai provinces: Kanchanaburi, Prachinburi, Ranong, and Sisaket [32]. DiCSIP provides advantages over older AS-PCR assays, including improved specificity, efficiency, and sensitivity, yielding more reliable and reproducible results. Older AS-PCR assays may not represent the full genetic diversity or evolutionary changes in ITS2 sequences, whereas DiCSIP uses primers designed from up-to-date sequence information [32].

DNA sequencing studies are crucial for understanding mosquito genetic diversity, population structure, and gene function through assessments of variation within and between populations. Nucleotide diversity varies among An. epiroticus, An. minimus, and An. maculatus populations [37]. Another study suggested high genetic distances among species within a clade of the Hyrcanus group [42]. Sequencing of specific genes such as voltage-gated sodium channel enables detection of mutations associated with insecticide resistance; however, no evidence of kdr mutations was found in sampled Anopheles populations in P. knowlesi-endemic areas in Thailand, suggesting metabolic resistance [37, 43].

Studies using mitochondrial DNA markers such as COI and COII revealed that current populations of An. minimus in Thailand are divided into two lineages, A and B [37]. Lineage A is characterized by high haplotype diversity and signals of population expansion, suggesting a high degree of adaptability [44]. A recent analysis of the mitochondrial COI gene in island populations of An. baimaii in Thailand showed low nucleotide diversity but high haplotype diversity [45]. Analysis of the COI gene of An. epiroticus revealed two distinct genetic clades in Ranong Province, Thailand, indicating variation potentially linked to local adaptation [37]. Mitochondrial DNA markers are effective for species identification owing to conserved mitochondrial genes with high copy numbers and a distinct “barcoding gap,” which enables clear species differentiation and supports the identification of unknown, ambiguous, or damaged specimens when morphology is unreliable [41].

Primary vectors: dirus and minimus complexes and Maculatus group

Anopheles (Cellia) dirus Peyton & Harrison species complex (Dirus complex)

Mosquito species diversity was highest along Thailand’s international border regions with Myanmar, Laos, and Cambodia. This review highlights provinces including Saraburi, Trat, Ubon Ratchathani, Mae Hong Son, Tak, Ratchaburi, and Kanchanaburi as having particularly diverse Anopheles fauna (Fig. 2).

Fig. 2.

Distribution of molecularly confirmed species of the Anopheles dirus complex (diamonds), Funestus group (triangles), and Maculatus group (circles) based on studies from 2013–2025

During the wet season, An. dirus populations showed peak activity in Ubon Ratchathani Province in July 2014 and September 2015 [46]. Similarly, in Umphang Valley, Tak Province, the total number of An. dirus collected (n = 24) was high in the wet season (from August to October), with densities dropping to as low as two specimens by December [47]. The seasonal patterns of An. dirus s.l. populations in Thailand remain consistent with the previous review [15], with highest abundance during the wet season. In these areas, adult densities of An. dirus s.l. were positively associated with increased rainfall from July to August/September, as previously observed [15, 48].

Changes in land use due to anthropogenic activities, such as expansion of human settlements, forest fragmentation through logging, and deforestation for rice cultivation, cassava, sugar, and rubber, have created conditions favorable for the proliferation of primary malaria vectors in Thailand [10, 49]. Human activities within these altered environments increase An. dirus breeding opportunities; for example, puddles on muddy roads act as temporary freshwater larval habitats after rainfall [10]. These changes contribute to increased malaria transmission risk, especially in southern Thailand [50]. An. dirus demonstrates adaptability to land-use changes, such as the conversion of rice paddies to mature rubber plantations [46]. This adaptability allows the species to persist and proliferate in altered environments, particularly in hilly areas adjacent to natural forests [11, 51]. In Sisaket Province, An. dirus occurred in both rubber–forest and village ecotypes, with higher biting density in forest sites. Rubber tappers in plantations near forests, as well as forest goers, are therefore at greater risk [52]. Biting activity was also slightly higher during the rainy and cool-dry seasons compared with the hot-dry season [52]. The larval habitats of this species are typically small, shallow, shaded bodies of fresh, stagnant, or slow-flowing water, remaining consistent since early reviews [53]. During the dry season, these habitats contracted to subterranean seepage streams and wells, as first noted by Rosenberg et al. (1990) [54] and more recently by Saeung et al. (2025) [52]. This shift suggests that females face difficulty locating suitable breeding sites during the dry season owing to their scarcity or unsuitability for larval development [54].

Distributions of molecularly identified An. baimaii were recorded at varying densities in six provinces: Trat, Mae Hong Son, Tak, Kanchanaburi, Surat Thani, and Ranong (Fig. 2). Additionally, An. baimaii was first reported from four islands, including Chang and Kood islands in Trat Province, along the Thai–Cambodian border, and Chang and Phayam islands in Ranong Province, adjacent to the Thai–Myanmar border. This extends the known range of An. baimaii to the easternmost islands [55]. While earlier studies reported An. baimaii in western Thailand, current data confirm its presence along both the western and eastern borders [15, 56].

The presence of An. baimaii on these border islands exemplifies “anophelism without malaria” [57], underscoring the need for continuous monitoring. Geographical isolation of island populations appears to influence morphological and genetic variation, with a mean genetic variation of 0.73 between populations on Kood Island (Trat Province) and Chang Island (Ranong Province) [45]. The southern region of Thailand and the Thai–Myanmar border are locations of high sympatry for several Dirus complex species [56], particularly An. baimaii in the western regions [17].

In a survey in Surat Thani Province, molecular analysis of An. dirus s.l. specimens revealed the exclusive presence of An. baimaii (100.0%, n = 348). This was the first record of the species in this region, suggesting prior misidentification using morphology alone [51]. These and other recent molecular studies have improved understanding of An. baimaii distribution. The species has been confirmed in northwestern Thailand, including Mae Sot and Sop Moei [58] and Umphang Valley [59], although often at low densities. This represents a new record in Surat Thani, southern Thailand, despite the occurrence of other Dirus complex species in nearby southern provinces [51].

Anopheles (Cellia) minimus Theobald species complex (Minimus complex)

An. minimus, previously identified as An. minimus species A, is recognized as a major malaria vector throughout the Greater Mekong Subregion (GMS), including Thailand [44]. In contrast to An. dirus, An. minimus has a more widespread distribution, with habitats that include forest-edge areas [17, 37, 60], hilly forested regions [61], and rubber plantations, where it can be dominant [11]. These findings suggest that An. minimus habitats have changed over the past few decades, shifting from traditional breeding sites such as slow-moving streams, rice fields, and ground pools [53] to new habitats. Rubber plantations often contain a mosaic of temporary pools, drainage ditches, and other manmade water sources that are ideal for An. minimus and An. dirus [11]. This species is more commonly found in areas of high malaria transmission, particularly in high-risk areas along Thailand’s western and eastern international borders [44]. Its occurrence aligns with the current or recent malaria microstratification by the Bureau of Vector-Borne Diseases, Ministry of Public Health [6] in areas that overlap with locations where annual average cases exceed 56 per year and with the reported distribution of An. minimus from the northeastern border (Ubon Ratchathani, Chanthaburi, and Trat provinces), the northwestern region (Kanchanaburi, Tak, Mae Hong Son, and Ratchaburi), and the south-western region in Ranong Province, especially in the northern and southern parts of the province (Fig. 2).

In northwestern Thailand, An. minimus was the most abundant Anopheles species recorded in Tak and Mae Hong Son provinces during a 2-year collection period (February 2011 to January 2013), representing 43.0% of total Anopheles captures [58]. Similarly, in southern Thailand (Surat Thani Province), An. minimus s.l. was the most abundant Anopheles species across three districts (Phanom, Khiri Rat Nikhom, and Vibhavadi), representing 69.9–93.8% of collections per district [51]. A study in western Thailand detected native Wolbachia pipientis in An. minimus collected from a low-transmission area in Umphang Valley, Tak Province, along the Thai–Myanmar border [59]. The discovery of a naturally occurring Wolbachia strain in wild-caught Anopheles necessitates further research to identify the specific strain, determine its prevalence in the wild population, and assess its potential impact on malaria transmission [59, 62].

An. harrisoni is a cryptic sibling species within the An. minimus complex and a member of the An. minimus subgroup [17, 63]. It is geographically distributed in Southeast Asia [17, 64] and has been reported in Kanchanaburi Province, Thailand [65]. Morphologically, An. harrisoni is difficult to distinguish from An. minimus, and the two frequently occur in sympatry in western Thailand [66]. While An. minimus is an opportunistic feeder, An. harrisoni tends to be more zoophilic, and its precise role as a malaria vector is unclear [60]. Accurate identification requires specialized molecular methods; multiplex AS-PCR targeting ITS2 has been developed for this purpose [29]. In addition, geometric morphometrics based on wing venation geometry can differentiate An. harrisoni from An. minimus with high accuracy (up to 95.0%), although size overlap exists [66].

Anopheles (Cellia) maculatus Theobald species group (Maculatus group)

Thailand is home to seven species of the Maculatus group, of which three are considered primary malaria vectors: An. maculatus Theobald, An. sawadwongporni Rattanarithikul & Green, and An. pseudowillmori (Theobald) [1]. The complex taxonomy of this group requires ongoing revision based on available molecular techniques [38, 67] for reliable identification of sibling species, notably including An. willmori and the species formerly known as An. maculatus Form K, formally described as An. rampae [68].

The distribution of five of the seven An. maculatus group species in Thailand is shown in Fig. 2. Field studies confirm that the group’s distribution is consistent with high-transmission areas. A sporozoite rate of 0.1% for P. vivax in An. maculatus (n = 778) collected by indoor light traps supports evidence of malaria resurgence in Tak Province, western Thailand [69]. This resurgence saw a sevenfold increase in cases in 2023 (~ 17,000) compared with 2021, and twice the number reported in 2022 [6, 70, 71]. Research in northwestern Thailand (Mae Sot District, Tak Province, and Sop Moei District, Mae Hong Son Province), along the Thai–Myanmar border, found molecularly identified An. maculatus (n = 1436) to be the second most abundant species (20.1%) among 7129 Anopheles collected, after An. minimus (42.8%) [58]. Studies in Surat Thani and Narathiwat provinces in southern Thailand also reported the presence of An. maculatus s.l. (1.0% prevalence) among a total of 60 Anopheles captured by human landing catch [51].

In Ranong Province, a highly endemic area for P. knowlesi [6], molecular analysis confirmed the presence of An. maculatus (13.2%; n = 15). However, none of the collected mosquitoes were positive for Plasmodium [37]. Members of the An. maculatus group, including An. maculatus [n = 164, human landing catch (HLC) = 1.72 bites per person per night], An. sawadwongporni (n = 400, HLC = 1.32 bites per person per night), and An. pseudowillmori (n = 100, HLC = 0.86 bites per person per night), were found along the Thailand–Myanmar border [72]. Another study at the Thai–Lao border in Ubon Ratchathani Province (northeastern Thailand) found very few An. sawadwongporni (n = 3) and relatively high numbers of An. rampae, a nonvector (n = 112), suggesting a complex interplay between low-density villages that contribute to relatively low transmission and neighboring forest areas where the human biting rate of An. sawadwongporni is low [73].

Despite a 2014 malaria outbreak with an incidence of 3.94 per thousand linked to rosewood smuggling in Ubon Ratchathani [74] and ongoing transmission in 2018 of 0.28 per thousand [6, 75], no Plasmodium-infected An. maculatus were detected [76]. This may reflect low abundance of primary vectors, temporal mismatch between the study and the transmission season, and limitations in PCR sensitivity and specificity [76]. Transmission most likely occurred in forested, hilly areas outside villages rather than within village collection sites. Extensive entomological surveys of zoonotic malaria in Narathiwat Province, southern Thailand, found An. maculatus to be the most prevalent species (37.1%) among captured mosquitoes (n = 136), with a high human biting rate of 3.40 per person per night [77].

Secondary vectors: Barbirostris, Sundaicus, Hyrcanus, and Subpictus groups

Anopheles (Anopheles) Barbirostris van der Wulp species complex (Barbirostris complex)

Molecular identification within the An. barbirostris subgroup, which comprises 11 species including 8 in the Barbirostris complex, confirmed nine cryptic species: An. barbirostris (0.3%, n = 3), An. dissidens (17.9%, n = 199), An. saeungae (15.2%, n = 168), An. wejchoochotei (45.4%, n = 503), An. campestris (15.5%, n = 172), An. barbirostris species A3 (0.5%, n = 6), An. donaldi (2.9%, n = 32), An. hodgkini (0.2%, n = 2), and An. pollicaris (2.1%, n = 23) [2, 41]. The P. vivax infectivity rate of An. barbirostris s.l. was 1.7%, with two positive specimens detected from indoor and outdoor collections along the Thai–Myanmar border [78]. This finding highlights the importance of the complex as a local vector and a significant hurdle for elimination due to its contribution to residual transmission.

The Barbirostris subgroup and complex are widely distributed in Thailand, with the exception of the north-central and northeastern regions. Population density and seasonal abundance vary geographically (Fig. 3A). Peak abundance generally occurs during the rainy and cool-dry seasons, whereas populations along the Thai–Myanmar border peak during the transition from the wet to the dry season [79]. Distribution patterns of sibling species align with high-risk malaria areas, including Mae Hong Son, Tak, Kanchanaburi, and several southern provinces [6], suggesting a strong correlation between the presence of Barbirostris species and transmission risk.

Fig. 3.

Distribution of secondary and suspected malaria vectors in Thailand based on studies from 2013–2025: A Anopheles barbirostris subgroup; B Anopheles epiroticus (triangles) and Anopheles hyrcanus group (circles)

An. wejchoochotei, a species within the Barbirostris complex, is considered a probable malaria vector [80]. Molecular studies confirm its wide distribution across Thailand [81], and a comprehensive survey identified it in 19 of 23 provinces sampled [30]. Two species within the Barbirostris subgroup, An. hodgkini (n = 2) and An. pollicaris (n = 23), and one within the Barbirostris complex, An. donaldi (n = 32), were molecularly identified for the first time in Surat Thani Province, suggesting prior misidentification based solely on morphology [51]. An. wejchoochotei was the predominant species of the complex in recent surveys across numerous provinces, including Ayutthaya, Kanchanaburi, Ratchaburi, Phetchaburi, Nan, Chanthaburi, Chiang Mai, Chaiyaphum, Chumphon, Kamphaeng Phet, Khon Kaen, Mahasarakham, Mukdahan, Prachuap Khiri Khan, Sa Kaeo, and Udon Thani [41, 78, 82], and in Ranong Province [37] (Fig. 3A).

Within the Barbirostris complex, An. saeungae was reported from Lampang, Phetchaburi, Ratchaburi, Sa Kaeo, Trat, Ubon Ratchathani, and Ranong provinces [37, 41] (Fig. 3A). An. dissidens occurred in western and southern Thailand [30], mostly in mountainous areas, with distributions reported in Chiang Mai, Mae Hong Son, Sa Kaeo, Tak, and Trat provinces [41]. An. campestris and An. barbirostris were reported as rare in recent studies [41].

Anopheles (Cellia) sundaicus (Rodenwaldt) species complex (Sundaicus complex)

As a member of the An. sundaicus complex [83], An. epiroticus is a malaria vector occurring along the coastline of Southeast Asia. Its immature stages develop in habitats with varying salinity, from low and brackish to seawater concentrations [84], and it is widely distributed throughout eastern and southern Thailand [85]. A recent study in Ranong Province, a region highly endemic for P. knowlesi [86], identified An. epiroticus as a dominant species, representing 19.3% of Anopheles females collected [37]. In a separate study on Chang Island in Trat Province (eastern Thailand), An. epiroticus was also the most dominant coastal species and a likely important vector [87]. Owing to its anthropophilic and exophagic behavior [17], this species is considered a secondary vector in eastern Thailand [1].

Anopheles hyrcanus group

The An. hyrcanus group in Thailand comprises two subgroups, Lesteri and Nigerrimus [88], and at least eight species: An. argyropus, An. crawfordi, An. nigerrimus, An. nitidus, An. lesteri, An. peditaeniatus, An. pursati, and An. sinensis [89, 90]. A multiplex PCR assay has been developed to identify eight Thai Hyrcanus group species accurately; among these, An. nigerrimus, An. peditaeniatus, and An. sinensis are suspected vectors of P. vivax [89]. Furthermore, An. peditaeniatus has been identified as a potential vector for the goat malaria parasite, P. bubalis type I, in Thailand [91]. The distribution of molecularly confirmed Hyrcanus group sibling species is shown in Fig. 3B.

Studies on the An. hyrcanus group in Thailand reveal significant cytogenetic diversity. An. nitidus exhibits five metaphase karyotypes, designated as form A (X1, Y1), form B (X1, Y2), form C (X2, Y3), form D (X1, X3, Y4), and form E (X1, X2, X3, Y5) based on sex chromosome variations [92]. Forms D and E were localized in Ubon Ratchathani Province, while forms A, B, and C were found in Phang Nga Province [92]. Similarly, An. crawfordi demonstrated four metaphase karyotypes, of which forms A (X1, X2, X3, Y1) and B (X1, X2, X3, Y2) occur in both Thailand and Cambodia, whereas forms C (X2, Y3) and D (X2, Y4) are restricted to Trang Province [93] (Fig. 3B). In Khon Kaen Province, An. nigerrimus was used as an outgroup species for a phylogenetic comparison of the Barbirostris complex [41]. The Hyrcanus group is notably predominant along the Thai–Lao border in Ubon Ratchathani Province [73]. Within this group, An. peditaeniatus is the most widely distributed species in Thailand and was most common in Ubon Ratchathani [3]. Other members, including An. nigerrimus, An. nitidus, An. argyropus, and An. sinensis, were also present at varying densities across villages in the same area [3]. An. peditaeniatus and An. pursati have been recorded in several provinces, including Kanchanaburi, Ratchaburi, Phetchaburi, and Nan [82]. Seasonal abundance of the Hyrcanus group is typically highest during the rainy season [73].

Anopheles subpictus group

From 2013 to 2025, surveys across Thailand identified members of the An. subpictus complex [61, 94], which is widely distributed, including northwestern Thailand [58] and Surat Thani Province in the south [51]. These mosquitoes are often found near livestock; for example, one study reported greater abundance on cows than on humans along the Thai–Myanmar border [72]. In Ratchaburi Province, An. subpictus s.l. occurred at higher abundance than An. barbirostris s.l. [95], and it was collected at goat farms in Kanchanaburi and Ratchaburi provinces in sympatry with An. vagus [82]. Recent field collections confirmed An. subpictus in Mae Sot District (Tak Province) and Sop Moei District (Mae Hong Son Province) [58] and in Surat Thani Province, southern Thailand [51].

Role of vectors in human and zoonotic malaria transmission

In Southeast Asia, P. knowlesi is the most widely recognized and prevalent zoonotic malaria parasite infecting humans [14, 96–99]. Sporadic human infections with other simian Plasmodium species have also been reported, including P. cynomolgi in neighboring Cambodia [100] and in Thailand [101], and mixed infections involving P. inui and P. fieldi in Thailand [102]. These zoonotic malarias are primarily transmitted by mosquitoes of the Leucosphyrus group, which serve as bridge vectors between simian reservoir hosts (primarily macaques) and humans [103].

The Leucosphyrus group comprises key vectors for P. knowlesi transmission in Southeast Asia, including An. dirus and An. cracens of the Dirus complex, and An. latens, An. introlatus, and An. balabacensis of the Leucosphyrus complex [77, 103]. An. latens and An. introlatus were identified as natural vectors on the basis of detection of NHP P. knowlesi DNA and implicated in zoonotic transmission in southern Thailand [77]. Further research implicated An. latens in natural transmission of P. inui and P. fieldi. In southern Thailand, An. introlatus has been identified as a vector for P. cynomolgi and P. inui, and a potential vector for P. hylobati, as DNA of these parasites was detected in its salivary glands [77] (Table 1).

Table 1.

Infectivity of anthropophilic and simiophilic Anopheles species reported in Thailand, 2013–2024

Reference	Location/province	Plasmodium species	Infectivity rate	Parasite infection group	Parasite detection method	Anopheles mosquito	Habitat
[29]	Pak Tho District, Ratchaburi Province	P. vivax	0.4%	Human	Single-round multiplex PCR	An. minimus	Village
[58]	Mae Sot District, Tak Province	P. vivax	0.1%	Human	Real-time PCR	An. minimus	Rubber, fruit, and agricultural crop lands
[78]	Thasongyang District, Tak Province	P. vivax	0.7%	Human	ELISA	An. annularis s.l	Village
[78]	Thasongyang District, Tak Province	P. vivax	1.7%	Human	ELISA	An. barbirostris s.l	Village
[173]	Bo Rai District, Tart Province	P. vivax	2.8%	Human	Nested PCR	An. dirus	Rubber plantation
[78, 172]	Thasongyang District, Tak Province	P. vivax	0.4%	Human	ELISA	An. maculatus s.l	Village
[78, 172]	Thasongyang District, Tak Province	P. vivax	0.8%	Human	ELISA	An. minimus s.l	Village
[173]	Thong Pha Phum District, Kanchanaburi Province	P. vivax	28.8%	Human	Nested PCR	An. aconitus	Near rubber plantation
[46]	Na Chaluai District, Ubon Ratchathani	P. vivax	3.5%	Human	ELISA and nested PCR	An. dirus	Wood hut near national park
[46]	Na Chaluai District, Ubon Ratchathani	P. vivax	1.0%	Human	ELISA	An. philippinensis	Wood hut near national park
[69]	Thasongyang District, Tak Province	P. vivax	0.1%	Human	ELISA	An. maculatus	Village
[174]	Mae Sot, Tak Province5	P. vivax	0.5%	Human	PCR	An. minimus	Village
[91]	Chachoengsao Province	P. bubalis	5.7%	Buffalo	PCR and sequsencing	An. campestris or An. wejchoochotei	Buffalo farm
[91]	Chachoengsao Province	P. bubalis	2.5%	Buffalo	PCR and sequencing	An. peditaeniatus	Buffalo farm
[124]	Not specify	P. caprae	1.4%	Goat	PCR and sequencing	An. aconitus	Goat farm
[77]	Narathiwat Province, Southern Thailand	P. cynomolgi	2.8%	Simian	Species-specific PCR and Sequencing	An. introlatus	Village
		P. hylobati	2.8%
		P. knowlesi	2.8%
		P. inui,	8.1%
		P. spp.	2.8%
[77]	Narathiwat Province, Southern Thailand	P. knowlesi	1.5%	Simian	Species-specific PCR and sequencing	An. latens	Village
		P. inui	1.5%
		P. fieldi	3.1%
		P. coatneyi	1.5%
		P. hylobati	1.5%
		P. juxtanucleare	1.5%
		P. spp.	6.2%
[77]	Narathiwat Province, Southern Thailand	P. spp.	50.0%	—	Sequencing	An. macarthuri	Village
[77]	Narathiwat Province, Southern Thailand	P. caprae	0.7%	Goat	Species-specific PCR	An. maculatus	Village
[124]	Not specified	P. caprae	0.9%	Goat	PCR and sequencing	An. subpictus	Goat farm
[7]	Kaeng Khoi District, Saraburi	P. cynomolgi	20.0%	Simian	Ultrasensitive ddPCR1	An. dirus	Forest
[7]	Kaeng Khoi District, Saraburi	P. cynomolgi	6.3%	Simian	Ultrasensitive ddPCR1	An. minimus	Forest
[7]	Kaeng Khoi District, Saraburi	P. cynomolgi	21.8%	Simian	Ultrasensitive ddPCR1	An. sawadwongporni	Forest

The major human malaria vector An. dirus is widespread in Thailand (Fig. 2) and transmits human as well as NHP malaria parasites. Its adaptability to anthropophilic, simiophilic, and zoophilic hosts may contribute to escalating P. knowlesi incidence by bridging infection between monkeys and humans. An. maculatus has recently been identified as a potential vector for P. knowlesi after experimental infection in Thailand [103] and for the ungulate malaria parasite P. caprae in southern Thailand [77] (Table 1).

A recent study of the vectorial capacity of An. dirus to acquire P. knowlesi from humans demonstrated a human-to-mosquito transmission cycle under laboratory conditions [103]. This finding highlights the risk of a potential sustained human–mosquito–human transmission cycle in nature. Vector feeding preferences and habitats are crucial to zoonotic transmission, particularly because P. knowlesi occurs where macaque reservoirs are present and forest-related activities increase risk [104]. Human activities such as hunting or logging in forest-fringe areas increase exposure to vectors and facilitate contact among humans, infected monkeys, and mosquitoes [105].

Investigation of local vector ecology is essential because vector behavior varies regionally; a malaria vector in one area may not be a vector elsewhere [37]. Local studies are needed to identify the specific Anopheles species transmitting P. knowlesi in each location, and to monitor behavioral changes, including potential host shifts from macaques to humans, to understand transmission dynamics [103].

New molecular and epidemiological approaches are needed to track the complex transmission cycles of NHP malaria, which involve spillover from NHP to humans and spillback to wild NHP populations [106]. According to Masse et al. (2025), a significant knowledge gap concerns human–vector overlap and behaviors that drive zoonotic malaria transmission [107]. Limited understanding of how human activities, such as forest-related work, alter vector biting and resting behaviors hampers prediction of when and where humans are most at risk. Moreover, the lack of integrated surveillance that simultaneously monitors human infection, NHP parasite prevalence, and vector abundance and behavior leads to data fragmentation, complicating accurate modeling of transmission and the implementation of targeted interventions. This gap underscores the need for a One Health approach that integrates entomological, epidemiological, and ecological research to understand these interactions. Environmental change increases interactions between humans and wildlife, and it remains unclear how future environmental and anthropogenic changes might elevate NHP malaria risk [106].

Anopheles bionomics

Life history traits

External stressors can significantly influence the survival of An. minimus populations [44]. Higher temperatures, within an optimal range, increase vectorial capacity by accelerating parasite development and prolonging mosquito longevity [108]. This enhanced vectorial capacity, in turn, contributes to greater complexity in mosquito ecology and leads to heightened malaria transmission [15].

Intraspecific variability in host feeding varies among populations of An. dirus in Thailand and across the GMS and is likely influenced by environmental factors and host availability [15, 109]. Feeding on sheep blood significantly reduced the survival of An. dirus, An. cracens, and An. minimus compared with human, chicken, or mouse blood [110]. This negative effect extended to the next generation: offspring of An. dirus and An. cracens whose parents fed on sheep blood had significantly shorter lifespans [110]. A shorter lifespan lowers the entomological inoculation rate and reduces progeny fitness, suggesting that the presence of certain nonhuman hosts can act as a dead end for vector populations and may naturally reduce transmission risk to humans.

A study of An. dirus experimentally infected with P. knowlesi detected oocysts in midguts at day 7 and sporozoites in salivary glands at day 14 post-infection, demonstrating competence to transmit P. knowlesi [103]. This is notable because An. dirus is a confirmed zoonotic vector in Vietnam [15, 111, 112] and occurs where P. knowlesi is an increasing concern. These results reinforce the need to investigate the role of An. dirus in P. knowlesi transmission in Thailand.

A blood meal triggers physiological changes in An. dirus females, including upregulation of genes involved in oogenesis [113], confirming the link between blood feeding and reproductive capacity. Anopheles mosquitoes, including An. dirus, commonly take a first blood meal before mating and use ingested nutrients to support egg development [113]. This behavior supports oogenesis and contrasts with autogeny, in which eggs develop without a blood meal [114]. Using a forced oviposition technique, overall oviposition rates were similar between laboratory strains (36.7%) and field populations (36.9%), although species-level variation was marked. Success rates for laboratory-reared An. dirus and An. cracens were 66.0% and 12.0%, respectively. In contrast, only 8.0% of wild-caught An. dirus females from Pu Teuy Village, Sai Yok District, Kanchanaburi Province produced eggs [115]. The mean number of eggs per female induced by forced oviposition was higher for laboratory-reared An. dirus (93.61 ± 46.37) than for An. cracens (39.33 ± 17.75), and differed from field populations of An. harrisoni (36.34 ± 38.22) and An. barbirostris s.l. (147.58 ± 85.07) [115]. This reproductive performance indicates biological plasticity and suggests that colony females can retain eggs and, when prompted by constraints such as a lack of suitable sites or a physiological trigger from the forced technique, deposit an entire batch. Although this may not reflect natural behavior, it demonstrates high reproductive potential, a key factor in vector effectiveness.

Breeding habitat selection

Breeding-site preference is a critical determinant of mosquito distribution. Favorable habitats for anopheline larvae include permanent water bodies with aquatic vegetation and shaded areas, such as those near marginal streams [105]. Since 2013, research has uncovered diverse and often cryptic breeding-site preferences for Anopheles vectors in Thailand. For instance, An. dirus, typically associated with forested habitats, was found breeding in an unusual location, namely a dark limestone cave in western Thailand [116]. Another study identified larvae in stagnant water seeping from grooves in a large rock within shaded primary forest [52]. The highly cryptic and dispersed nature of An. dirus breeding sites poses a significant challenge for larval source management because they do not meet WHO’s “fixed, few, and findable” criteria for effective vector control [117]. One survival strategy employed by the An. dirus complex mosquitoes during the dry season involves the exploitation of cryptic habitats, including manmade structures such as underground water wells in Chanthaburi Province [118]. This adaptation underscores the difficulty of controlling An. dirus populations, as traditional larval source management often targets more visible sites and can miss these hidden reservoirs.

The discovery of An. baimaii in Koh Kood District, Trat Province, near Cambodia, which is the easternmost occurrence of this species in Thailand, challenges earlier records that confined it to central, southern, and western forest regions, particularly along the Thai–Myanmar border [45, 55]. Typical breeding habitats of An. baimaii are small, temporary forest-fringe water bodies such as ditches, elephant footprints, cattle hoof marks, and wheel tracks [119]. Larvae of An. minimus have been observed in unusual sites, namely crab holes and swamps in Doi Inthanon National Park, Chiang Mai Province [120], as well as a domestic water tank in suburban Hanoi, Vietnam [121]. These findings demonstrate adaptability beyond its more typical streambed habitats.

In Samut Songkhram Province, the coastal vector An. epiroticus showed a significant larval preference for brackish water containing green algae [122]. Entomological surveys in southern Thailand identified breeding habitats of forest-dwelling Leucosphyrus complex species, including An. introlatus and An. latens, with larvae found in ground pools, animal footprints, wheel tracks, and small, shallow running streams [77]. This baseline entomological information is critical for understanding simian malaria transmission and for developing control strategies.

Feeding behavior

Recent studies provide detailed insights into host preference and selection patterns of Anopheles mosquitoes, which are shaped by species-specific bionomics and environmental factors. Many Anopheles species in Thailand show a strong preference for animal hosts, such as cattle, over humans; this zoophily significantly impacts transmission dynamics [79, 123]. Buffalo bait is a potent attractant for species such as An. epiroticus in coastal regions of Thailand [87]. Most species of the An. hyrcanus group are highly zoophilic and feed on cattle. Along the Thai–Myanmar border, An. peditaeniatus showed a high preference for cattle over humans (zoophilic index 98.7%, n = 7345) and is therefore considered highly zoophilic [3]. An. dirus, a primary malaria vector, is characterized as an exophilic, anthropophagic feeder, with a tendency to feed and rest outdoors, particularly where indoor residual spraying (IRS) is used, which drives host seeking outdoors [11]. Depending on location and environment, An. minimus is often exophilic in Thailand, with a high exophagic rate and preferential outdoor feeding [73]. In other countries, such as Vietnam, this species can exhibit both anthropophilic and zoophilic tendencies, as well as endophagic behavior [39, 109]. An. minimus displayed strong zoophagy on the Thai–Myanmar border, with ratios of 77.0% and 87.0% [123]. Its peak biting activity typically occurs from early evening until midnight. An. harrisoni is primarily zoophilic, preferring animal hosts, whereas the closely related sibling An. minimus is an opportunistic feeder that bites both humans and animals [66].

Mosquitoes in the Maculatus group, specifically An. maculatus, exhibited strong zoophagy and a high zoophilic index [72, 123]. Within this group, An. sawadwongporni and An. pseudowillmori also showed high zoophilic indices, indicating a preference for animal hosts [72]. Similarly, species within the Barbirostris complex such as An. dissidens and An. saeungae are mainly zoophilic [30], in contrast to the more anthropophilic An. wejchoochotei and An. campestris [61].

During the review period, 60.7% (n = 49,569) of primary vectors were identified using HLC compared with 0.8% (n = 136) using cattle-baited traps (CBT) (Table 2). For secondary vectors, HLC and CBT recorded 24.4% (n = 4,784) and 20.5% (n = 4,021), respectively (Table 3).

Table 2.

Abundance of primary Anopheles malaria vectors by collection method

Collection method	An. dirus s.l.	An. dirus s.s.*	An. minimus s.l.	An. minimus s.s.*	An. maculatus s.l.	An. maculatus*	Total collection by collection method	Reference
BG-Pro CDC-style traps	308	66	208	48	254	27	911	[37, 40, 41, 45, 148]
BGS and CDC LT	0	0	1581	0	0	0	1,581	[142]
CBT only	1	5	35	8	66	21	136	[79, 181]
CDC light trap	81	43	8045	222	2806	16	11,213	[7, 44, 66, 69, 78, 85, 91, 94, 105, 124, 147, 172, 175–178]
HLC only	1478	366	40215	35	7447	28	49,569	[46, 51, 52, 77, 81, 153, 174]
HLC, HDT, and HDNT	0	2	0	419	0	3	424	[141]
HLC and CBT	15	323	174	13666	139	3309	17,626	[29, 30, 47, 58, 72, 76, 87, 169]
UV, LED, and fluorescent	0	4	0	5	0	0	9	[143, 144, 179]
Not specified	49	0	37	0	14	0	100	[173]

CBT cattle baited trap, HLC human landing catch, HDNT human double net trap, HDT human-baited host decoy trap, CDC LT CDC light trap

^*Confirmed by molecular identification technique

Table 3.

Abundance of secondary and suspected Anopheles malaria vectors by collection method

Collection method	An. aconitus1	An. aconitus1*	An. pseudaowillmori1	An. pseudaowillmori1*	An. sawadwongoni2	An. sawadwongoni2*	An. barbirostris s.l.2	An. barbirostris s.s.2*	An. campestris2	An. campestris2*	An. philippinensis2	An. philippinensis2*	Total collection by collection method	Reference
BG, Mosquito Magnet, CDC UV, CDC backpack	0	0	0	0	0	0	160	0	0	0	0	0	160	[180]
BG-Pro CDC-style traps	0	23	0	2	200	7	308	0	0	0	9	4	553	[37, 40, 41, 45, 148]
BGS and CDC LT	0	0	0	0	162	0	0	0	0	0	0	0	162	[142]
Black light trap and Kelambu Trap	0	0	0	0	0	0	0	0	0	153	0	0	153	[61]
CBT only	0	2	0	0	0	16	3497	0	0	0	479	27	4021	[79, 164]
CDC light trap	733	0	73	11	339	183	1339	0	51	0	18	0	2747	[7, 44, 66, 69, 78, 85, 91, 94, 105, 124, 147, 172, 175–178]
HLC only	3	2	0	35	0	19	4695	0	0	19	11	0	4784	[51, 52, 77, 81, 153, 174]
HLC and CBT	28	286	0	160	0	989	3887	3	1	0	1325	144	6823	[29, 30, 46, 47, 58, 72, 73, 76, 87, 169]
Mosquito magnet	0	0	0	0	0	0	149	0	0	0	0	0	149	[145, 182]
UV, LED, and fluorescent	0	1	0	0	0	8	4	0	0	0	17	0	30	[143, 144, 179]
Not specified	0	7	0	0	0	0	0	0	0	0	0	0	7	[173]

CBT cattle baited trap, HLC human landing catch, HDNT human double net trap, CDC LT CDC light trap

Multiple Anopheles species, including the An. minimus, An. dirus, and An. barbirostris complexes; the Maculatus group; the An. hyrcanus group (notably An. peditaeniatus); the Annularis group (An. annularis, An. nivipes, and An. philippinensis); and species such as An. aconitus, An. epiroticus, and An. subpictus, have been frequently collected using animal-baited traps (Tables 2 and 3). Collections at buffalo farms in Chachoengsao Province yielded An. wejchoochotei, An. campestris, and An. peditaeniatus, species implicated as vectors for P. bubalis (Table 1) [91]. Surveys at goat farms using aspirators or light traps in Kanchanaburi, Nan, Ratchaburi, and Phetchaburi identified probable vectors for P. caprae [124]. Further investigations using blood-meal PCR assays are warranted to determine the anthropophilic, zoophilic, or zoo-anthropophilic origins and preferences of Anopheles vectors [37].

Geographic variation in feeding behavior among primary vectors was evident in paired HLC and CBT collections between distant locations in Tak and Trat provinces separated by 600 km, and also within smaller areas in Tak Province (Table 4). An. aconitus was an exception, showing consistent zoophagy in Tak. Such differences can produce distinct epidemiological situations and influence the effectiveness of control measures. Although deploying both animal-baited traps (ABT) and HLC at the same site would provide a more complete picture of vector behavior, animal-baited traps are rarely used owing to logistical challenges [125]. As a complement to HLC, animal-baited traps could broaden the range of observed behaviors, in line with WHO recommendations for a wider array of surveillance methods [125, 126]. Expanding entomological data from both animal-baited traps and HLC is crucial for a comprehensive understanding of vector behavior.

Table 4.

Proportion of Anopheles mosquitoes collected from paired outdoor cattle bait trap (CBT) and human landing catch (HLC) at three sites in Thailand

Province (Ref.)	Proportion in percentage and (numbers) of mosquitoes caught in								Total
	Primary vectors			Secondary vectors			Other Anopheles
	Species	CBT	HLC§	Species	CBT	HLC§	CBT	HLC§
Trat [87]	An. dirus	0.15 (n = 8)	2.44 (n = 132)	–			82.29 (n = 4443)	14.15 (n = 764)	5399
	An. minimus	–	0.16 (n = 9)	–
	An. maculatus	0.80 (n = 43)		–
Tak [72]	An. dirus	0.02 (n = 1)	0.70 (n = 30)	An. aconitus	0.65. (n = 28)	0.07 (n = 3)	33.73 (n = 1451)	3.74 (n = 161)	4301
	An. minimus	9.70 (n = 417)	33.85 (n = 1,456)	–
	An. maculatus	10.49. (n = 451)	7.04 (n = 303)	–
Tak [47]	An. dirus	0.05 (n = 9)	0.10 (n = 17)	An. aconitus	1.32 (n = 218)	0.09 (n = 15)	32.38 (n = 5333)	1.07 (n = 176)	16,468
	An. minimus	38.68 (n = 6369)	14.86 (n = 2447)	An. sawadwongporni	0.98 (n = 161)	0.04 (n = 6)
	An. maculatus	10.00 (n = 1647)	0.38 (n = 62)	An. pseudowillmori	0.04 (n = 7)
			–	An. baimaii	0.00 (n = 1)
Prachuap Khiri Khan [75]*		0	0		0	0	25.12 (n = 52)	74.88 (n = 155)	207

^*Only Barbirostris complex species have been reported

^§Conducted indoors and outdoors

Although mosquito surveillance and distribution studies exist for Thailand, monkey-baited traps are more commonly reported in Malaysia [127–133], indicating a knowledge gap in Thailand regarding mosquito species that feed on nonhuman primates in natural settings. This limits understanding of zoonotic malaria transmission dynamics in Thailand and needs to be addressed.

Among Anopheles species confirmed by molecular techniques, only An. dirus, An. baimaii, An. minimus, An. harrisoni, and An. maculatus showed indoor biting activity (Fig. 4A). An. dirus and An. baimaii exhibited the highest indoor biting in the early evening (1900–2100) and again from midnight to 0400. An. dirus biting patterns show marked geographic variation: in Tak Province where most indoor mosquitoes were caught between 2200–2300 and 0400–0500, with outdoor peaks at 2100–2200 and 2300–0000 [123]. In Ubon Ratchathani, biting occurred indoors and outdoors from 2000–2300 and again from 0100–0300 [73]. On Chang Island, Trat Province, the indoor peak was 0000–0100, while the outdoor peak occurred earlier, 1900–2300 [87].

Fig. 4.

Peak biting patterns of molecularly confirmed major malaria vectors in indoor (A) and outdoor (B) habitats. Low biting indicates less than a 0.1 biting rate (bites per person per night), and peak biting indicates a biting rate greater than or equal to 1.0 biting rate (bites per person per night)

In areas where IRS was used, indoor biting peaks shifted toward earlier or outdoor biting [109]. While the outdoor peak remained unchanged, a higher proportion of mosquitoes began biting earlier in the evening, suggesting that IRS altered behavior and pushed mosquitoes to seek blood meals outdoors or earlier. In contrast, An. minimus exhibited a consistent peak in indoor biting throughout the night (Fig. 4A). DDT use had no detectable effect on the already early-biting An. minimus s.l. population [134]. However, widespread IRS altered An. minimus s.l. behavior [135], likely reflecting a species shift from An. minimus to An. harrisoni, a phenomenon also observed in Vietnam with extensive use of insecticide-treated nets [39]. Although An. harrisoni showed early evening indoor peaks (1900–2000) and another after midnight, a causal link with insecticide-treated nets has not been established. The bimodal biting of An. maculatus, peaking at 1900–2000 and again at 0200–0300, is well documented.

An. minimus biting patterns vary geographically. In rubber plantations in Trat Province, peaks occurred at 1900–2100 and 0100–0300 [11]. In Ubon Ratchathani, indoor biting was highest at 1900–2100, while outdoor peaks were 1900–2000 and 2300–0000 [73]. In farm huts in Tak Province, activity began as early as 1700–1800, peaked at 2000–2100, decreased in the early morning, then resurged from 0300 and remained relatively high until 0600–0700 [123].

In Thailand, An. maculatus outdoor biting also varies geographically. In Tak and Surat Thani provinces, peak outdoor activity occurred at 1900–2000 [51, 58, 123]. In Narathiwat, abundance peaked at 2100–2300 and declined after midnight [77]. Indoor biting generally peaked at 2100–2200, with most indoor biting occurring before 2200 [123]. Early morning activity was also reported, with some bites after 0500 [123] (Fig. 4B).

Given the early biting times of Anopheles vectors, the first human behavior observations on the Thai–Myanmar border demonstrated the role of key drivers and modulators in sustaining infection prevalence [123]. By integrating human behavior observations with mosquito collections in villages and forested farm huts used for subsistence farming, exposure was shown to occur predominantly outdoors, particularly among nonusers of long-lasting insecticidal nets. Risk was exacerbated by suboptimal bed net coverage. Early waking hours, when people had left the safety of their nets, coincided with later morning biting peaks. This adaptive mosquito behavior, an evolutionary response to indoor insecticidal interventions such as insecticide-treated nets, enables mosquitoes to obtain a blood meal before people are under nets, thereby avoiding insecticide contact [69]. This is a major driver of residual transmission, where malaria persists despite high coverage of conventional control measures.

Infection rates and Plasmodium species in Anopheles mosquitoes

Six studies from Thailand (n = 3) and Vietnam (n = 3) reporting Plasmodium infections in Anopheles mosquitoes were included in the meta-analysis (Fig. 5; Table 5). Overall mosquito infection prevalence was low (pooled prevalence = 0.01; 95% CI 0.00–0.03), but estimates varied markedly between settings, with substantial heterogeneity (I² = 97.6%, P < 0.0001). The highest infection prevalence (0.2%) among 152 mosquitoes from six anopheline species in Thailand was attributed to a droplet digital PCR (ddPCR), a highly sensitive PCR method as shown in Fig. 5 [7]. In Vietnam, An. dirus showed a very low pooled prevalence (0.02%; 95% CI: 0.00–0.00%), consistent with sporadic detection of infected mosquitoes, with the lowest estimate in Khanh Hoa Province (n = 1; pooled prevalence = 0.00) [136]. Nevertheless, molecular evidence of co-infection with P. vivax, P. falciparum and P. knowlesi in a single wild-caught An. dirus indicates that vectors in these settings can harbor both human and zoonotic malaria parasites simultaneously [136], underscoring the potential for cross-species transmission even where overall infection prevalence is low. These findings suggest that low average infection prevalence may mask focal transmission hotspots where ecological conditions favor intense enzootic circulation and increased human exposure.

Fig. 5.

Forest plot of the prevalence of nonhuman primate Plasmodium infection in Anopheles mosquitoes in Vietnam [111, 112, 136] and Thailand [7, 77, 103]. Individual study estimates and pooled prevalence values (random-effects model) are shown for five Anopheles species across three studies from Vietnam and three studies from Thailand. Error bars represent 95.0% confidence intervals (CI). ES estimated prevalence

Table 5.

Prevalence of Plasmodium infection and parasite species detected in Anopheles mosquitoes included in the meta-analysis

No	Reference	Country	Province	Mosquito infectivity (%)	Year(s) of study	Number of mosquitoes with Plasmodium species identified by PCR		Anopheles species incriminated
						Pk	Others
1	[136]	Vietnam	Khanh Hoa	0.02	2007–2008	1	Pf (n = 1), Pv (n = 1), Pm (n = 1)	An. dirus
2	[111]	Vietnam	Khanh Hoa	1.00	2008–2010	31	Pf(n = 36), Pv (n = 36), Pm (n = 4)	An. dirus
3	[112]	Vietnam	Khanh Hoa	0.59	2008–2010	33	Pf (n = 33), Pv (n = 35), Pm (n = 3)	An. dirus
4	[77]	Thailand	Narathiwat	2.99	2018–2019	2	Pin (n = 6), Pcy (n = 1), Pfi (n = 2), Pco (n = 1)	An. latens, An. introlatus
5	[103]	Thailand	Yala	0.21	2023	3	-	An. dirus
6	[7]	Thailand	Saraburi	16.44	2023	-	Pcy (n = 25)	An. sawadwongporni, An. minimus, An. dirus

Data are summarized by country, province, study period, mosquito infection prevalence, Anopheles species incriminated, and Plasmodium species identified by PCR. Studies from Vietnam are presented as comparators

Pf P. falciparum, Pv P. vivax, Pk P. knowlesi, Pcy P. cynomolgi, Pm P. malariae, Pin P. inui, Pfi P. fieldi, Pco P. coatneyi

In contrast, substantially higher infection prevalences were reported in Thailand, particularly in Saraburi Province, where P. cynomolgi infection reached 16.4% (n = 25) in An. sawadwongporni (n = 17), An. minimus (n = 5), and An. dirus (n = 3) [7]. This finding represents the first detection of simian malaria parasite DNA in wild-caught mosquitoes with infection rates of 11.2%, 3.3%, and 2.0% in An. sawadwongporni, An. minimus, and An. dirus, respectively [7], and highlights the capacity of multiple vector species to sustain NHP malaria transmission. Additional studies from southern Thailand in Narathiwat Province documented diverse NHP Plasmodium species, including P. knowlesi (n = 2), P. cynomolgi (n = 1), P. inui (n = 6), P. fieldi (n = 2), and P. coatneyi (n = 1) within Anopheles populations, reflecting complex enzootic transmission cycles [77]. Using a random-effects model, the pooled mosquito infection prevalence across Vietnam and Thailand was 0.01 (95% CI 0.00–0.03). Between-study heterogeneity was substantial (I² = 97.6%, Cochran’s Q-test P < 0.0001), indicating considerable variability in infection prevalence estimates across studies. The high level of heterogeneity likely reflects differences in vector species composition, ecological context, sampling intensity, and molecular diagnostic approaches, rather than random variation alone. Together, these results reinforce the need for surveillance strategies that extend beyond human case detection to include entomological and reservoir-host monitoring, and they support the application of One Health frameworks to identify and mitigate zoonotic malaria risk in forest-fringe and plantation-dominated landscapes.

Despite a much larger population at risk of malaria (13,606,254 inhabitants; ~ 20.0% of the total population), Thailand has fewer documented vector incrimination studies than Malaysia, which has an extensive body of research on NHP malaria [127–133], although its at-risk population is much smaller (1,405,051 inhabitants). This research gap is a concern for elimination in the GMS, particularly with the rise in indigenous P. knowlesi from 22 cases in 2020 to a maximum of 259 cases in 2023 (83 cases in 2024), and the potential emergence of other zoonoses. Limited entomological data on vector incrimination, interactions with humans and monkeys, mosquito longevity, and vector competence hinder understanding of transmission dynamics. A similar challenge is reported in Malaysia, where continued increases in P. knowlesi and other NHP cases have delayed malaria certification [137]. Cambodia and Thailand reported 11 and 239 indigenous P. knowlesi cases, respectively, in 2023, highlighting a regional challenge [5].

The marked increase in P. knowlesi cases in Thailand underscores the need for integrated knowledge of NHP malaria prevalence in humans, macaques, and Anopheles vectors [77]. Current entomological studies often rely on dissections to detect parasites and parous rates but lack direct mosquito longevity measurements [116, 138]. Identifying sporozoites in field-collected mosquitoes is also challenging. Our meta-analysis highlights the difficulty of incriminating NHP malaria vectors due to low sporozoite rates. In recent years, only two field studies [7, 77] and one laboratory study [103] have successfully detected NHP malaria parasites (P. knowlesi, P. cynomolgi, and P. inui) in mosquitoes (Table 1).

Secondary and suspected vectors in Thailand, such as An. aconitus, An. epiroticus, and members of the An. barbirostris complex, including An. campestris and An. wejchoochotei,, were infected with Plasmodium in both laboratory and field studies [58]. Notably, An. wejchoochotei exhibits high anthropophily and is considered a likely vector in Sa Kaeo Province, eastern Thailand [31]. An. campestris also shows anthropophily and has been implicated in transmission of P. falciparum and P. vivax [61]. The combination of these vectors and the mobility of migrant workers, especially along international borders, underscores the need for expanded surveillance and control strategies [61].

According to recent studies, two members of the Leucosphyrus complex, An. latens and An. introlatus, are native vectors of P. knowlesi in Narathiwat Province [77]. They also show potential to transmit other simian and avian Plasmodium species [77]. In central Thailand (Phra Phothisat Temple), An. sawadwongporni was found to carry P. cynomolgi [7], which is notable given the high prevalence of NHP malaria among wild macaques in that region.

In Ranong Province, an area endemic for P. knowlesi, three potential vector species, An. baimaii, An. hodgkini, and An. epiroticus, were identified [37]. These species warrant special surveillance for their vector potential in Thailand and elsewhere.

Beyond NHP malaria, vector–parasite interactions in Thailand are expanding. For the first time, An. subpictus and An. aconitus have been identified as probable vectors of P. caprae, a malaria parasite of goats, suggesting an emerging interface beyond human malaria [124]. In addition, An. maculatus has been identified as a vector of P. caprae in domestic goats in southern Thailand [77].

Advanced mosquito surveillance and nonmolecular identification

The human-baited double net trap [139, 140] and human-baited host decoy trap [141] are considered as alternatives to the HLC as they can reduce human exposure to infective bites during mosquito surveillance (Tables 2 and 3). Efficacy varies by location: both traps showed promise as substitutes for conventional outdoor HLC in Vietnamese forests, but performance appeared doubtful in Thai forests [141].

The CDC miniature light trap (CDC LT) is a dependable, efficient method for collecting Anopheles; effectiveness is enhanced when combined with CO_₂, an incandescent light, and a chemical lure (BG-lure) [142]. Studies on light sources indicate that traps fit with fluorescent lights and various LEDs are effective for collecting Anopheles [143]. Selecting the optimal wavelength maximizes trap performance: a UV fluorescent light was most efficient, followed by UV LED, blue LED, and green LED. In Nakhon Ratchasima Province, a 365-nm LED was the most effective light source overall, capturing 790 mosquitoes (23.7% of 3,339), but Anopheles captures were notably low (8 specimens; 0.2%) [144].

The Mosquito Magnet uses a combination of attractants and performs particularly well in coastal areas [145]. It is reliable where An. epiroticus, Culex sitiens, and Culex quinquefasciatus are prevalent. In Malaysia, Mosquito Magnet efficacy for collecting Anopheles (vectors of NHP Plasmodium) was comparable to HLC, and it provides an ethical, safer alternative [146].

Geometric morphometrics is a rapid, inexpensive method to discriminate species and sibling complexes via wing-venation shape. Its utility has been demonstrated across multiple taxa: within the An. dirus complex, geometric morphometrics distinguished An. dirus and An. baimaii with 92.4% accuracy [104]; within the An. minimus complex, An. minimus and An. harrisoni were separated with 90% accuracy [66]; in the Maculatus group, An. maculatus, An. sawadwongporni, and An. pseudowillmori were differentiated with 93% accuracy [147]; and within the Barbirostris complex, landmark-based geometric morphometrics distinguished An. dissidens, An. saeungae, and An. wejchoochotei with 74.3% accuracy [41]. Variation in accuracy reflects differing interspecific morphological distances. Geometric morphometrics does not replace genetic analysis but is a fast, low-cost alternative when molecular methods are not feasible, particularly in field settings where specimens are intact.

Artificial intelligence is being developed to support surveillance and vector-borne disease management. A recent study used a self-supervised vision transformer (DINOv2) to classify mosquito species from mobile-phone images, enabling rapid, accurate field identification [148]. The approach successfully classified 16 female mosquito species from across Thailand, including vectors of malaria and dengue [148]. Artificial intelligence is also being applied to automated egg counting for population monitoring and program evaluation.

Biomarkers are valuable for assessing human exposure to Anopheles bites and pathogen-transmission risk. They support evaluation of human–vector contact and control efficacy, identification of hotspots, and measurement of exposure in low-transmission settings [149–151]. In Thailand, the gSG6-P1 salivary peptide is a key biomarker [152, 153]. High sequence identities have been shown for An. minimus (86.9%) and An. aconitus (86.9%), and for An. maculatus (82.6%) [153]. Antibody intensity correlates positively with Anopheles exposure and entomological inoculation rate. For An. dirus, identity is lower (47.8%), indicating the need for further research [153].

New vector control methods

Newer interventions for anopheline control include volatile pyrethroid spatial repellents, such as transfluthrin and metofluthrin. These products create a “vector-free space” through noncontact irritancy, preventing landing, inhibiting feeding, and reducing mosquito–human contact in outdoor or semi-enclosed spaces [140, 154, 155]. Other methods include insecticide-treated clothing with an active ingredient such as etofenprox, which provides personal protection via contact irritancy and requires mosquitoes to land on treated fabric [140]. Topical repellents such as picaridin form a surface barrier against bites. These products are evaluated for their capacity to disrupt feeding cycles, thereby reducing vector competence and lowering disease risk [154].

Mass drug administration of ivermectin in humans or livestock is being explored as a tool for malaria control and elimination. Ivermectin is lethal to multiple Anopheles species, including An. minimus, An. campestris, An. sawadwongporni, An. dirus, and An. epiroticus [156]. Beyond effects on survival, ivermectin inhibits the sporogony of P. vivax in An. dirus and An. minimus at human-relevant concentrations [156, 157]. Primary metabolites (e.g., M1, M3, M6) also exhibit mosquito-lethal effects, potentially extending mortality beyond that of the parent compound [157]. However, recent research indicates that the B1a and B1b components of ivermectin have no effect on mosquito mortality [158].

Genetic control is a key frontier in eradication, with research focused on molecular targets to block transmission in An. dirus [113]. The suppressor of cytokine signalling (SOCS) gene is activated in the abdomen 12–24 h after an infectious P. vivax blood meal, indicating involvement in the immune response [159]. The conserved HPX15 gene may facilitate parasite invasion in Anopheles, making it a potential target [160]. Targeting the yellow-g gene in An. dirus reduces P. vivax infection without shortening lifespan, a desirable profile for control strategies [161]. Carboxypeptidase B in the midgut plays a vital role in Plasmodium sexual development; compounds such as NSC-1014, NSC-332670, and aminopterin significantly reduced carboxypeptidase B activity in An. minimus, leading to reduced parasite infection [162].

Monitoring insecticide resistance is central to sustaining the effectiveness of chemical vector control [163] within a One Health framework for zoonotic malaria. Longitudinal molecular and phenotypic evidence from Thailand indicates that, among primary malaria vectors implicated in P. knowlesi transmission, target-site resistance to pyrethroids remains rare. Molecular screening for target-site resistance in eight Anopheles species (An. baimaii, An. minimus s.s., An. epiroticus, An. jamesii, An. maculatus s.s., An. barbirostris A3, An. sawadwongporni, and An. aconitus) from Ranong Province, a P. knowlesi-endemic area in Thailand, detected only the wild-type L1014 allele in the voltage-gated sodium channel gene [37]. Consistent with these findings, WHO susceptibility bioassays conducted across multiple regions demonstrated high levels of pyrethroid sensitivity in key vector species, An. minimus s.l. and An. maculatus s.l. from Tak Province, An minimus and An aconitus from Kanchanaburi and Ratchaburi provinces [105], An dirus from Trat Province [11], and An epiroticus from coastal provinces (Trat, Songkhla, Surat Thani, and Phang Nga) [183], with 24-h mortality frequently reaching 98–100% following exposure to deltamethrin (0.05%), permethrin (0.75%), lambda-cyhalothrin (0.05%), cyfluthrin (1.5%), and etofenprox (0.5%) [164].

Despite this overall susceptibility, emerging resistance patterns were observed in selected secondary vectors and for nonpyrethroid insecticide classes. In Tak Province (western Thailand), reduced mortality to 0.1% bendiocarb was detected in An. minimus s.l. (88.8% mortality) and An. maculatus s.l. (89.9% mortality), while susceptibility to propoxur (0.1%) and organophosphates (fenitrothion 1.0% and malathion 5.0%) remained intact [164]. While field populations of An. dirus s.l. and An. maculatus s.l. from northeastern Thailand in Ubon Ratchathani Province including An. harrisoni and An. dirus from Kanchanaburi were fully susceptible to deltamethrin and transfluthrin (discriminating concentrations (0.06% for An. minimus; 0.02% for An. dirus) [184], secondary vectors (An peditaeniatus and An barbirostris s.l.) in Ubon Ratchathani exhibited resistance or suspected resistance to DDT and, in some cases, pyrethroids (e.g., deltamethrin 0.05%, permethrin 0.75%) [3, 43], indicative of metabolic rather than target-site mechanisms [43]. Synergist assays using piperonyl butoxide restored pyrethroid susceptibility in An. peditaeniatus, implicating cytochrome P450-mediated detoxification, and kdr mutations were absent in secondary vector populations examined to date [43]. An. nivipes and An. philippinensis remained fully susceptible to these insecticides [43].

Behavioral responses further complicate vector control measures. Strong contact irritancy and moderate noncontact repellency to pyrethroids (deltamethrin, permethrin, and alpha-cypermethrin) were documented in An. epiroticus from Chang Island (Trat Province), potentially reducing intervention effectiveness despite physiological susceptibility [183]. Similarly, exophagic and outdoor-resting behaviors observed in An. dirus and An. minimus populations associated with rubber plantation landscapes may limit the impact of indoor-based interventions such as IRS and ITN/LLINs [11].

Laboratory studies also indicate that mosquito age (3–5 days versus 20–23 days) and batch size (5, 10, or 20 individuals) significantly influence mortality and knockdown outcomes in WHO tube assays, particularly for volatile pyrethroids such as transfluthrin [165]. Adaptations to testing protocols, including the use of smaller batch sizes with increased replication, can improve data reliability when wild mosquito numbers are limited [165], while remaining consistent with WHO recommendation of 25 nonfed females that are 3–5 days old [166].

Consistent with a One Health approach, these findings highlight the need to align resistance surveillance and vector control strategies with landscape-driven transmission dynamics. In settings where zoonotic malaria is sustained at the interface of humans, NHP reservoirs, and forest-associated vectors, effective control cannot rely on single interventions or narrowly defined indicators of success. Addressing P knowlesi and other simian malaria parasites requires a One Health approach that recognizes how land-use change, macaque ecology, vector behavior, and insecticide susceptibility interact to shape transmission risk. Strengthening integrated surveillance across these domains, and translating this evidence into adaptive, context-specific vector control strategies, will be essential for mitigating spillover risk and sustaining malaria control gains in rapidly changing landscapes.

Conclusions

Spatial and behavioral heterogeneity among Anopheles vectors, including variation in host preference, biting time, and resting behavior, strongly shapes malaria transmission dynamics and the effectiveness of control interventions. These traits are further modified by environmental change and human activities, underscoring the need for adaptive, evidence-based surveillance within national malaria control programs. Although major vectors account for most current transmission in Thailand, the potential contribution of secondary vectors should not be overlooked. Accordingly, the use of complementary collection and monitoring approaches is essential to characterize human and vector behavioral diversity, detect early behavioral shifts. and sustain control effectiveness [167].

A persistent challenge is the incomplete understanding of sibling species diversity, geographic distribution, and sympatry within major vector complexes, particularly the Barbirostris, Minimus, Maculatus, and Hyrcanus groups. Closely related species can differ markedly in vectorial capacity, host preference, biting and resting behavior, and insecticide susceptibility, such that interventions effective against one species may be less effective against another[ 168, 170, 171]. Progress is constrained by limitations of existing molecular identification tools, which may yield ambiguous or incorrect results, while newer assays such as Dirus Complex species identification PCR (DiCSIP) with improved resolution remain underutilized [32]. These gaps are further compounded by limited ecological data on climate variability, larval habitats, forest microclimates, land-use change, and the effects of geographic isolation on gene flow, local adaptation, and resistance emergence.

Within this context, this review synthesizes recent advances in molecular biosystematics, bionomics, and vector incrimination of Anopheles complexes in Thailand to clarify their roles in human and zoonotic malaria within a One Health framework. ITS2- and mitochondrial DNA-based diagnostics have resolved cryptic sibling species, updated the distributions of primary vectors (Dirus, Minimus, and Maculatus) and key secondary groups (Barbirostris, Sundaicus, Hyrcanus, and Subpictus), revealing substantial genetic and ecological diversification across forest—agricultural mosaics and international border regions.

Evidence indicates highly heterogeneous infection prevalences, expanding assemblages of NHP malaria Plasmodium maintained predominantly by Leucosphyrus-group vectors, and marked behavioral plasticity—including early, outdoor, and zoophilic biting—that sustains residual and zoonotic transmission despite generally high pyrethroid susceptibility. Emerging nonpyrethroid resistance, cryptic larval habitats, and ongoing land-use change further complicate vector control and expose critical gaps in human–vector–macaque contact patterns and vector competence.

Overall, addressing residual and zoonotic malaria transmission in Thailand will require integrated, species-specific surveillance that links molecular identification, vector behavior, host ecology, and environmental change. Such an approach, consistent with a One Health framework, is essential to inform adaptive, landscape-specific vector control strategies as Thailand progress toward malaria elimination and confronts the growing challenge of P. knowlesi and other zoonotic malaria threats.

Abbreviations

AS-PCR

Allele-specific polymerase chain reaction

CBT

Cattle-baited traps

CDC

Centers for Diseases Control and Prevention

DiCSIP

Dirus complex species identification polymerase chain reaction

DDT

Dichlorodiphenyltrichloroethane

GMS

Greater Mekong Subregion

HLC

Human landing catch

IRS

Indoor residual spraying

ITS2

Internal transcribed spacer 2

NHP

Nonhuman primate

PCR

Polymerase chain reaction

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

RFLP-PCR

Polymerase chain reaction with restriction fragment length polymorphism

SCAR-PCR

Sequence characterized amplified region-polymerase chain reaction

WHO

World Health Organization

Author contributions

S.N.W., J.H., T.C.: conceptualization and design the study methodology; S.N.W.: conducting systematic literature investigation, screening, statistical analysis, and writing-original draft; J.H.: quality assessment of the included studies, writing-reviewing and editing; J.N., C.S., U.A., P.M., R.N.K., S.T., I.V., S.M.: writing—reviewing and editing; T.C.: supervision, funding acquisition, writing—reviewing and editing.

Funding

This research received funding from the National Research Council of Thailand (NRCT) through the High-Potential Research Team Grant Program (contract no. N42A670406) as well as the Kasetsart University Research and Development Institute (KURDI), Bangkok, Thailand (grant no. FF (KU) 52.69). Also, this work is financially supported by the office of the Ministry of Higher Education, Science, Research and Innovation; and the Thailand Science Research and Innovation through the Kasetsart University Reinventing University Program 2025.

Data availability

The data supporting the contents of this article are included within the article.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.