Insecticide resistance and population structure of the invasive malaria vector, Anopheles stephensi, from Fiq, Ethiopia

Abstract

Anopheles stephensi invasion in Ethiopia poses a risk of increased malaria disease burden in the region. Thus, understanding the insecticide resistance profile and population structure of the recently detected An. stephensi population in Fiq, Ethiopia, is critical to inform vector control to stop the spread of this invasive malaria species in the country. Following entomological surveillance for An. stephensi in Fiq, Somali region, Ethiopia, we confirmed the presence of An. stephensi morphologically and molecularly in Fiq. Characterization of larval habitats and insecticide susceptibility tests revealed that Fiq An. stephensi is most often found in artificial containers and is resistant to most adult insecticides tested (organophosphates, carbamates, pyrethroids) except for pirimiphos-methyl and PBO-pyrethroids. However, the immature larval stage was susceptible to temephos. Further comparative genomic analyses with previous An. stephensi populations from Ethiopia using 1704 biallelic SNPs revealed genetic relatedness between Fiq An. stephensi and east-central Ethiopia An. stephensi populations, particularly Jigjiga An. stephensi. Our findings of the insecticide resistance profile, coupled with the likely source population of Fiq An. stephensi, can inform vector control strategies against this malaria vector in Fiq and Jigjiga to limit further spread out of these two locations to other parts of the country and continent.

Introduction

Malaria is a major global health problem, with an estimated 247 million cases and 619,000 deaths reported in 2021¹. In Ethiopia, despite past successes in reducing the malaria burden due to the use of indoor residual spraying (IRS) and long-lasting insecticidal nets (LLINs) in malaria control and prevention strategies², it remains a public health concern with an estimated 2 million cases and over 8000 deaths reported in 2021¹. The emergence of An. stephensi, a malaria vector commonly found in South Asia and parts of the Arabian Peninsula, in the Horn of Africa (HoA)^3–5, also threatens any gains against malaria in Ethiopia⁶. Indeed, a recent An. stephensi-mediated malaria outbreak was reported in Dire Dawa, an urban hub in eastern Ethiopia⁷. Since the detection of An. stephensi in the HoA and subsequently, in Sudan (2019)⁸, Nigeria (2020)⁹, Yemen (2021)^10,11, Kenya (2022)¹², and Ghana (2023)¹³, the World Health Organization (WHO) has launched an initiative¹⁴ and released an updated vector alert in 2022¹⁵ to recommend increased surveillance and research to determine the range of the invasion in order to stop the further spread of An. stephensi in Africa. In Ethiopia, An. stephensi was first detected in Kebridehar, Somali region, in 2016⁴ and has subsequently been confirmed to be broadly distributed in eastern Ethiopia¹⁶. However, it was first detected in Fiq, Somali region in June 2021¹⁷. Malaria cases in the Somali region of Ethiopia are relatively low compared to other regions of Ethiopia¹⁸. However, it remains understudied, and more research on malaria parasites and their vectors is needed.

Studies of the breeding sites and the ecology of mosquitoes are very crucial to inform mosquito control strategies such as mosquito larvicidal (temephos) and environmental control (larval habitat removal). Also, the WHO recommends larval source management as one of the immediate control strategies against An. stephensi in urban and peri-urban settings of invaded regions¹⁵. When larval source management or reduction is not feasible, such as a household or town water storage, larviciding may be considered. However, this vector control method can be costly when treating large numbers of larval habitats¹⁹. Thus, an alternative cost-effective approach is to target specific habitats that can produce large numbers of adult mosquitoes¹⁹. Thus, determining the Fiq An. stephensi susceptibility to larvicide such as temephos can help inform the decision making for this control method against the invasive malaria vector in Fiq town.

Moreover, genomic analyses can further inform control strategies against the newly detected Fiq An. stephensi. Specifically, assessing the genetic diversity and population structure of Fiq An. stephensi population in comparison to an established one in the region can give insights into its demographic history, dispersal patterns, and potential source populations.

Hence, we conducted an entomological survey after a year of the first detection of An. stephensi in Fiq town, Somali region, Ethiopia to first characterize larval habitats for An. stephensi and determine their insecticide susceptibility status, including susceptibility to the larvicide, temephos. After identifying the morphology, we confirmed it molecularly and used genomic approaches to analyze the demographic history and population structure of An. stephensi in Fiq. We compared the population structure with the previously detected An. stephensi populations in eastern Ethiopia to determine the extent of establishment in Fiq. We further assessed their genetic connectivity to those populations to uncover their potential source populations within the region.

Material and methods

Study setting

Fiq is a small town located in the Somali region of Ethiopia. It is situated at an altitude of 1200 m and 195 km west of Jigjiga, the capital city of the Somali Region. The town features sporadic mountains and small house constructions. It is common to see water stored in pits for building bricks for house construction.

Sample collection

Entomological survey was carried out in Fiq town during the rainy season from May to June 2022, by Jigjiga University. Adult mosquitoes were collected indoors and outdoors using pyrethrum spray collections (PSC) and CDC light traps (CDC LT), respectively, and Prokopack aspiration from animal shelters. Sampling of mosquito larvae was conducted from houses with water storage containers, particularly, cisterns and plastic sheet water storage. During each survey, a habitat was first visually inspected for the presence of mosquito larvae, and then twenty samples were taken with a soup ladle (350 ml capacity) from each breeding habitat (see Supplementary Fig. S1 online). The Anopheles larvae were separated from the Culicine larvae based on their position to the water surface. Collected Anopheles larvae were reared to the adult at a field laboratory for morphological species identification (see Supplementary Fig. S2 online).

Mosquito identification

Adult Anopheles were morphologically identified at the species level using the Afrotropical mosquitoes key²⁰ (see Supplementary Fig. S2 online). Polymerase chain reaction (PCR) assays were conducted on a subset of An. stephensi specimens identified by morphology. These assays targeted portions of the mitochondrial cytochrome oxidase subunit I (COI) and the nuclear internal transcribed spacer 2 (ITS2) loci for molecular identification, as previously reported⁴. For further confirmation of species identification, we performed phylogenetic analysis with the generated COI sequences and those of previously detected An. stephensi from Ethiopia retrieved from GenBank (Accession# OK663483, OK663480, OK663484, OK663481, OK663479, OK663482)²¹. These accession numbers represent unique COI haplotypes of An. stephensi previously identified across eastern Ethiopia²¹. Phylogenetic relationships were inferred using a maximum likelihood approach with RAxML GUI²² using the GTR model of nucleotide substitutions, gamma model for rate of heterogeneity (GTRGAMMA option), and one thousand replicates in one run for bootstrap analysis. Anopheles maculatus was designated as an outgroup. The tree with the highest log likelihood was visualized and formatted in FigTree²³.

WHO insecticide susceptibility tube tests

Insecticide susceptibility tests were conducted on adult female An. stephensi reared from wild larvae (see Supplementary Fig. S2 online) following standard procedures²⁴. One hundred mosquitoes from each population were tested for each insecticide using the diagnostic concentration, and 50 mosquitoes were used for controls. The insecticides used were 0.1% bendiocarb, 0.25% pirimiphos-methyl, 0.05% alpha-cypermethrin, 0.05% deltamethrin, 0.75% permethrin, and 0.15% cyfluthrin. Based on the WHO mortality criteria, resistance was determined as follows: 98–100% mortality indicates susceptibility, 90–97% mortality indicates possible resistance (requires further investigation), and less than 90% mortality confirms resistance ¹⁹.

Piperonyl butoxide synergist assays

Piperonyl butoxide (PBO) synergist assays were conducted on An. stephensi against two pyrethroids (deltamethrin and permethrin). The synergist assays were conducted by pre-exposing mosquitoes to a 4% PBO paper for 60 min. Mosquitoes were then transferred to tubes with the pyrethroid of interest for 60 min and the susceptibility was determined based on the WHO mortality criteria²⁴ described above.

Temephos susceptibility tests

Larval susceptibility tests were performed to assess the sensitivity of An. stephensi larvae to temephos, an organophosphate larvicide. According to WHO procedure²⁵, conventional techniques were employed for the testing, and 312.5 mg/l, 62.5 mg/l, 12.5 mg/l, and 2.5 mg/l concentrations were utilized to generate a final concentration of 1.25 mg/l, 0.25 mg/l, 0.05 mg/l, and 0.01 mg/l when 1 ml of each concentration was added to 249 ml of normal water. To detect resistance, an estimated diagnostic dosage of 0.25 mg/l was employed²⁶, and larvae mortality data were interpreted following the same WHO mortality criteria used for adult mosquitoes¹⁹. For each larvicide concentration, four duplicates of 25 larvae were employed, with two replicates serving as controls. To get 100 larvae examined per dosage, four beakers were utilized per dose. After 24 h all larvae without movement on the water surface were considered as dead.

Target-site insecticide resistance loci analysis

To analyze phenotypic and genotypic associations in observed pyrethroid-resistant An. stephensi, we sequenced the pyrethroid target site in the voltage-gated sodium channel (vgsc) gene and downstream intron to genotype any knockdown resistance mutation (kdr). Analyses were conducted using previously published protocols^27,28. To genotype any kdr mutations, sequences were queried to NCBI BLAST to confirm correct kdr locus, then aligned to reference sequences from Samake et al.²⁷ using CodonCode Aligner version 8 (CodonCode Corp., Centerville, MA, USA). We then generated an updated Ethiopian An. stephensi kdr mutation phylogenetic tree with kdr mutation sequences from Samake et al.²⁷ also including the only non-African An. stephensi kdr mutation sequence available in Genbank (Accession# JF304952)²⁸. Phylogenetic relationships were inferred using a maximum likelihood approach with RAxML GUI¹⁸ using the GTR model of nucleotide substitutions, gamma model for rate of heterogeneity (GTRGAMMA option), and one thousand replicates in one run for bootstrap analysis. The tree with the highest log likelihood was visualized and formatted in FigTree²³.

Nuclear population structure and genetic diversity

To assess nuclear population structure and potential source populations, we first generated double digest restriction-site associated DNA (ddRAD) sequences of Fiq An. stephensi in addition to previously published raw ddRAD sequences of Ethiopian An. stephensi ²⁹. Genomic DNA was extracted from adult An. stephensi mosquitoes using Qiagen DNeasy Blood and Tissue kit (Qiagen). DNA quality was assessed on a 1% agarose gel to ensure at least 10 kilobase DNA fragments and quantified using Nanodrop One Spectrophotometer (Thermo Fisher Scientific Inc.) to ensure a minimum concentration of 20 ng/μl. ddRAD-seq library preparation followed protocols outlined in Lavretsky et al.³⁰ (also see Samake et al.²⁹). In short, each genomic DNA was enzymatically fragmented using SbfI and EcoRI restriction enzymes and ligated with Illumina TruSeq compatible barcodes for demultiplexing purposes. Libraries were quantified using Qubit dsDNA BR Assay Kit (ThermoFisher Scientific, MA, USA), pooled in equimolar amounts, and sequenced using single-end chemistry sequencing on an Illumina HiSeq X at Novogene (Novogene CO., Ltd., Sacramento, CA, USA; see detailed methods in Supplementary Document S1 online). We then retrieved raw sequence reads of previously reported An. stephensi populations from 10 different sites across eastern Ethiopia (n = 183, BioProject PRJNA888109, Samake et al.²⁹) to generate a combined bi-allelic SNPs dataset. Raw Illumina reads were demultiplexed, processed and SNP genotyped using the computational pipeline described in Lavretsky et al.³¹ (also see Samake et al.²⁹). We used Trimmomatic³² to trim or discard poor-quality sequences using a Phred score of ≥ 30 to ensure only high-quality sequences were retained. We then used the Burrows-Wheeler Aligner³³ (bwa) to align the remaining quality reads to the An. stephensi reference genome (Accession PRJNA661063; Chakraborty et al.³⁴). Samples were sorted and indexed in Samtools and genotyped using the ‘mpileup’ function in BCFtools³⁵.

We then identified variation among samples with a principal component analysis (PCA) using the–pca function in PLINK v.1.9 ³⁶ and visualized with the R package ggplot2³⁷. Next, individual maximum-likelihood population assignment probabilities were attained across samples using ADMIXTURE v.1.3³⁸. Each ADMIXTURE analysis was run with a tenfold cross validation (CV) and with a quasi-Newton algorithm to accelerate convergence³⁹. To limit possible stochastic effects, each analysis was based on 1,000 bootstraps for each population K value of 1–10. The block relaxation algorithm for point estimation was used for each analysis and terminated once the log-likelihood of the point estimation increased by < 0.0001. The optimum population value was based on the average of CV-errors across the analyses per K value. ADMIXTURE assignment probability outputs were visualized using the R package ggplot2³⁷. Additionally, nucleotide diversity (π), counts of segregating SNPs (S), Tajima’s D, and pairwise estimates of fixation index (F_st) by site were calculated in the R package PopGenome⁴⁰.

Genetic network

To further access potential source populations of the Fiq An. stephensi population, we performed a network analysis with the combined biallelic SNPs dataset from Fiq sequences (n = 20) and Genbank retrieved An. stephensi sequences from 10 different sites across eastern Ethiopia (n = 183, Samake et al.²⁹). We used EDENetworks⁴¹, which allows network analyses based on genetic distance matrices without a prior assumption. The network consists of nodes representing populations connected by edges/links weighted by their F_st based Reynolds’ genetic distances (D)⁴² which provide the strength of connectivity between pairs of populations⁴¹. The thicker the edge/link, the stronger the genetic connectivity between the two populations. Moreover, node size is proportional to the cumulative weighted edge linkages for each population. Thus, the larger the node the higher the connectivity hub or sink. Statistical confidence of the nodes was evaluated using 1000 bootstrap replicates. Nodes that appear in the top 5 and top 1 lists of betweenness centrality (BC) values (number of shortest genetic paths passing through a node) can be considered as statistically significant⁴³.

Results

A total of 221 adult mosquitoes were collected using CDC LT, PSC, and Prokopack from animal shelters. Of these, 219 adults Culex spp, and only two adult An. stephensi were collected using Prokopack aspiration (see Supplementary Table S1 online). No other Anopheles mosquitoes were collected using PSC and CDC LT.

Anopheles stephensi larval habitats

Thirty-one mosquito potential breeding sites were inspected. Twenty-six containers including eight plastic ones, fifteen cisterns, and three barrels were positive for mosquito larvae (see Supplementary Table S2 online). There were no mosquito larvae recovered from five containers. A total of 4103 larvae of mosquito were collected from these different breeding sites. Of these, 3713 Anopheles and 390 Culex were identified (see Supplementary Table S2 online). All of the Anopheles mosquitoes were reared to adults successfully and morphologically identified as An. stephensi.

Insecticide susceptibility and synergist assays

According to the WHO bioassay procedure, 1200 non-fed female An. stephensi were tested with various insecticide classes and PBO-pyrethroids. Based on the WHO mortality criteria, Fiq An. stephensi revealed resistance to all insecticides tested, except for pirimiphos-methyl which resulted in 100% mortality (Fig. 1a; also see Supplementary Table S3 online). Pre-exposure to PBO restored complete sensitivity to deltamethrin and permethrin in the synergist tests (Fig. 1b; also see Supplementary Table S3 online).

Fig. 1.

Mortality rate of An. stephensi exposed to different insecticides in Fiq, Ethiopia. Bars represent mean values (with 95% CIs) (a) Diagnostic dose assays. (b) Synergist assays.

Temephos susceptibility

The Temephos susceptibility test against Fiq An. stephensi larvae revealed that concentrations of 1.25 mg/l and 0.25 mg/l killed 100% of the An. stephensi larvae after 24 h (Table 1). However, mortality rates for the rest of the concentrations, such as 0.05 mg/l and 0.01 mg/l, were 94 (95% CI [79, 100]) and 16 (95% CI [4, 28]) percent, respectively, below the WHO mortality criteria of greater than 98 percent for susceptibility (Table 1).

Table 1.

Mortality rate of An. stephensi larvae against WHO standard concentrations of Temephos.

Concentration	Final concentration	Mortality rate (%)				Mortality rate (%) mean [95% CI]
		R1	R2	R3	R4
312.5 mg/l	1.25 mg/l	100	100	100	100	100
62.5 mg/l	0.25 mg/l	100	100	100	100	100
12.5 mg/l	0.05 mg/l	100	96	80	100	94 [79, 100]
2.5 mg/l	0.01 mg/l	12	8	20	24	16 [4, 28]

NB: A total of 1000 adult An. stephensi identified and no other Anopheles species encountered to provide proof that the susceptibility test was done on An. stephensi larvae. R1-R4 are replicates.

Molecular identification

A subset of 154 morphologically identified An. stephensi collected from the various sites in Fiq were preserved with silica gel and sent to Baylor University for molecular and genomics analyses. Genomic DNA was extracted from 20 of those specimens for a comparative sample size with previous An. stephensi SNPs dataset from eastern Ethiopia²⁹. The 20 samples were successfully confirmed as An. stephensi based on the presence of the characteristic band in the ITS2 endpoint assay (see Supplementary Fig. S3 online), and phylogenetic analysis of the cytochrome oxidase subunit I (COI) mitochondrial marker further confirmed An. stephensi. All the analyzed Fiq An. stephensi COI sequences clustered with the most prevalent Ethiopian An. stephensi COI haplotype (Hap2)²¹ (see Supplementary Fig. S4 online).

Kdr mutation population frequency

Of the 20 An. stephensi samples analyzed for pyrethroid target site phenotypic-genotypic association in observed resistant An. stephensi, one (5%) carried the kdr L1014F mutation with a heterozygote allele (see Supplementary Table S5 online). The Kdr L1014S was not observed. Phylogenetic analysis confirmed the Fiq kdr L1014F mutation, as the analyzed sequence clustered with previously published An. stephensi kdr L1014F mutation sequences from Ethiopia²⁷ and India²⁸ (bootstrap 100) (Fig. 2).

Fig. 2.

Phylogenetic tree of kdr L1014F mutation in Fiq An. stephensi and available global An. stephensi kdr mutation sequence. The evolutionary history was inferred by using the Maximum Likelihood method based on the General Time Reversible model. The tree with final ML optimization likelihood (− 146.34) and bootstrap values > 70 are shown.

Nuclear genetic diversity and population structure

For a comparative analysis between the Fiq studied population and the previously reported An. stephensi populations in 10 different sites across eastern Ethiopia²⁹, genetic diversity and population structure assessments were based on a combined SNPs dataset of 1704 independent bi-allelic SNPs after filtering for linkage disequilibrium and non-biallelic SNPs. From the combined Ethiopian An. stephensi bi-allelic SNPs analyzed, the lowest nucleotide diversity was observed in Godey (π = 0.1979) in southeastern Ethiopia, and the highest nucleotide diversity was observed in Awash Sebat Kilo (π = 0.2300) in northeastern Ethiopia (see Supplementary Table S4 online). Fiq had the second lowest nucleotide diversity (π = 0.1986) (see Supplementary Table S4 online). Anopheles stephensi populations from southeastern Ethiopia and Fiq were also found to have the highest Tajima’s D values, 1.36, 1.48, 1.19, and 1.29 for Degehabur, Kebridehar, Godey, and Fiq, respectively, indicating a lack of rare variants relative to neutral expectations⁴⁴ (see Supplementary Table S4 online).

Assessing population structure using the combined An. stephensi independent bi-allelic SNPs dataset and based on principal component analysis (PCA), we identified three semi-discrete genetic clusters (Fig. 3b and c). Plotting the first two principal components grouped An. stephensi populations regionally with Fiq An. stephensi mostly clustering with east central Ethiopia An. stephensi populations (Fig. 3). The ADMIXTURE analysis based on an optimum population K model 5 more specifically identified five main genetic similarities across An. stephensi populations that include (1) Erer Gota, Dire Dawa, and Jigjiga, (2) Fiq, (3) Bati and Semera, (4) Gewane and Awash Sebat Kilo, and (5) Degehabur, Kebridehar, and Godey (see Supplementary Fig. S5 online). From these genetic clusters, Fiq An. stephensi displayed the lowest admixture proportions as it is composed of a single ancestral lineage that is predominantly observed in Jigjiga admixed An. stephensi population (see Supplementary Fig. S5 online). The F_st pairwise estimates also showed Fiq An. stephensi being less differentiated from Jigjiga An. stephensi from east central Ethiopia populations (F_st = 0.07) and more differentiated from Bati An. stephensi from northeastern Ethiopia populations (F_st = 0.14) (see Supplementary Fig. S6 online).

Fig. 3.

(a) Anopheles stephensi sampling locations in eastern Ethiopia. Fiq data was generated in this present study. Data from all other sites were previously generated (Samake et al.²⁹, 2023). Map was created with mapchart.net. (b) Principal component analysis (PCA) of individual An. stephensi population in eastern Ethiopia. The amounts of variation explained by each principal component (PC 1 on the x-axis, PC 2 on the y-axis) are given in percentages. The sample sites are Awash Sebat Kilo (AW18), Bati (BA18), Dire Dawa (DD18), Degehabur (DE18), Erer Gota (ER18), Fiq (FI22), Godey, (GD18), Gewane (GW18), Jigjiga (JJ18), Kebriderar (KB18), Semera (SM18). (c) Scatterplot of subgroup variation based on principal component analysis (PCA). The amounts of variation explained by each principal component (PC 1 on the x-axis, PC 2 on the y-axis) are given in percentages. Anopheles stephensi population subgroups are color-coded as described in the legend.

Genetic network

Network reconstruction was based on the same combined 1704 independent bi-allelic SNP dataset. Out of the ten Ethiopian An. stephensi populations analyzed, Fiq An. stephensi was found to be connected to two populations, Dire Dawa and Jigjiga, that were recovered as statistically significant nodes based on the bootstrapping test (Fig. 4, also see Supplementary Fig. S7 online). However, the network revealed higher genetic connectivity between Fiq and Jigjiga than Dire Dawa (Fig. 4a). This finding suggests Jigjiga is the potential source population for the studied Fiq An. stephensi.

Fig. 4.

(a) Genetic network of An. stephensi populations in eastern Ethiopia. Network nodes represent populations/hubs and links represent weighted genetic distances/genetic connectivities. The figure is produced by EDENetworks based on a genotype autosomal ddRAD-seq loci matrix by applying a single realization of bootstrapping with 0.85 percentage of nodes at each location and thresholded at 0.12. The colors and sizes of the links represent the strength of genetic connectivity from lowest (white) to highest (black). The colors and sizes of the nodes represent the cumulative weighted links from lowest (white) to highest (black). (b) Map showing Fiq road connections to other sites. The map was generated using ArcGIS Pro v 3.1 (Environmental Systems Research Institute (ESRI), Redlands, CA, USA).

Discussion

We report the presence of An. stephensi in abundance during a rainy season (May–June 2022) in Fiq, Somali region, Ethiopia. From the more than three thousand and five hundred Anopheles larvae collected, all were reared and morphologically identified as An. stephensi. Molecular identification of a subset for further molecular analyses also confirms the studied samples as An. stephensi. All An. stephensi larval habitats identified were artificial breeding sites such as plastic sheet water storages, covered and uncovered cisterns, and barrels which are consistent with other An. stephensi larval habitats reported in eastern Ethiopia⁴⁵. The fact that no other Anopheles larvae were collected suggests that An. stephensi could persist during dry seasons¹⁵ in Fiq, characteristics typically distinct from An. arabiensis, the primary malaria vector in Ethiopia^46,47. However, in Kenya, An. stephensi larvae were found in both artificial containers and a riverbed setting⁴⁸, emphasizing the potential diversity of larval habitats of this invasive An. stephensi that is worth noting for future entomological surveillance of this invasive malaria vector in Ethiopia and Africa.

The newly detected Fiq An. stephensi were found to be resistant to bendiocarb (carbamate), cyfluthrin, deltamethrin, permethrin, and alphacypermethrin (pyrethroids), but susceptible to pirimiphos-methyl (organophosphate) according to the WHO criteria of 98—100% mortality indicating susceptibility²⁴. However, restored susceptibility to deltamethrin and permethrin was observed following the synergist bioassay with piperonyl butoxide (PBO), which is a detoxifying enzyme inhibitor that bolsters the effects of pyrethroid insecticides, thus indicating a metabolic insecticide resistance mechanism^49,50. This finding is consistent with previous reports of insecticide resistance among the Ethiopian An. stephensi populations⁵¹ and suggests that PBO plus pyrethroids may be more effective against adult An. stephensi in Fiq. Additionally, Fiq An. stephensi larvae were found to be susceptible to the diagnostic dose of temephos (0.25 mg/l) with a 100% mortality rate (Table 1), which is consistent with recent larvicide studies conducted on three other An. stephensi populations in eastern Ethiopia^52,53. Thus, temephos may be an effective larvicide to control An. stephensi larvae in Fiq.

Also, our molecular analysis of the pyrethroid target site knockdown gene revealed the presence of the kdr mutation in a single pyrethroid-resistant Fiq sample (Fig. 2). The low kdr mutation frequency among the Fiq pyrethroid-resistant studied An. stephensi is consistent with previous low kdr mutation frequency in the Ethiopian An. stephensi from eastern Ethiopia²⁷. This finding confirms that kdr mutations may not be a primary mechanism of pyrethroid resistance in the invasive An. stephensi²⁷. Thus, the metabolic resistance mechanism again seems more relevant, as revealed by the positive synergist assays in pyrethroid-resistant An. stephensi in this study (Fig. 1), and previously reported in Balkew et al.⁵⁴. However, the analysis of gene expression at metabolic resistance loci is needed to fully understand the underlying molecular mechanisms of insecticide resistance development in the invasive An. stephensi.

Furthermore, comparative population structure analyses of the newly detected An. stephensi with other Ethiopian An. stephensi revealed high genetic similarities between Fiq An. stephensi and east central An. stephensi (Fig. 3b and 3c) and the recentness of Fiq An. stephensi compared to them. The recency of Fiq An. stephensi was further revealed by their low genetic diversity and high Tajima’s D compared to the east-central Ethiopia An. stephensi populations. Thus, these findings suggest that Fiq An. stephensi population is from a recent founder event, such as a bottleneck event from east central Ethiopia An. stephensi populations. Further population pairwise F_st and genetic network analyses pinpoint Jigjiga as the potential source of the founder event in Fiq. Specifically, Fiq An. stephensi were less genetically differentiated from Jigjiga An. stephensi than the other analyzed Ethiopian An. stephensi and shared higher genetic connectivity with Jigjiga than other sites (Fig. 4a). Although Fiq is located 195 km away from Jigjiga, the genetic connectivity observed could be explained by the fact that they are connected by a major road (Fig. 4b). The same major road also connects both Jigjiga and Fiq to Dire Dawa, the site of a recent An. stephensi-mediated malaria outbreak⁷. Thus, our finding highlights the potential role of roads, perhaps through human and/or good transportation, in the expansion of An. stephensi into Fiq. The role that transports routes (marines, roads, etc.) play in invasive species incursion into new geographic areas has been well documented^55–58. Thus, this potential mechanism of An. stephensi dispersal within the region warrants further investigation.

Conclusion

The present study revealed the high prevalence of the invasive An. stephensi in Fiq, its larval habitats, insecticide resistance status for both adult and immature stages, genetic diversity, population structure, and potential source populations. Our results showed that Fiq An. stephensi population is susceptible to pirimiphos-methyl, PBO-pyrethroids, and temephos.Q1 Thus, these insecticides may be effectively used in control strategies against this invasive malaria vector in Fiq. We also found that the Fiq An. stephensi population shares genetic connectivity with two major An. stephensi hubs (i.e., Jigjiga and Dire Dawa) in eastern Ethiopia, with a stronger connection to Jigjiga. Thus, heightened vector control in those areas could help prevent further An. stephensi incursions into Fiq and beyond. Overall, this study provides a comprehensive approach to investigate a recent An. stephensi expansion into a new geographical area to determine the extent of establishment, assess effectiveness of insecticides, and identify potential source populations to prevent further spread.

Author contributions

JNS, SY, and TEC conceptualized the study. SY, MA, TEC, and SZ organized field collections. SY conducted the bioassays. JNS and TEC collected and analyzed the molecular and genomic data. JNS, SY, SZ, and TEC contributed to the writing of the manuscript. All authors reviewed and approved the final manuscript.

Funding

This research was funded by the Centers for Disease Control and Prevention. SZ was funded by the U.S. President’s Malaria Initiative. The findings and conclusions in this report are those of the author(s) and do not necessarily represent the official position of the U.S. Centers for Disease Control and Prevention.

Data availability

DNA sequences: Bioproject PRJNA1042829 and Genbank accessions SAMN38322421 – SAMN38322440.

Declarations

Competing interests

The author(s) declare no competing interests.

Disclaimer

The authors alone are responsible for the views expressed in this article and they do not necessarily represent the views, decisions, or policies of the institutions with which they are affiliated.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-78072-4.