Scientific Findings of the Southern and Central Africa International Center of Excellence for Malaria Research: Ten Years of Malaria Control Impact Assessments in Hypo-, Meso-, and Holoendemic Transmission Zones in Zambia and Zimbabwe.

For a decade, the Southern and Central Africa International Center of Excellence for Malaria Research has operated with local partners across study sites in Zambia and Zimbabwe that range from hypo- to holoendemic and vary ecologically and entomologically. The burden of malaria and the impact of control measures were assessed in longitudinal cohorts, cross-sectional surveys, passive and reactive case detection, and other observational designs that incorporated multidisciplinary scientific approaches: classical epidemiology, geospatial science, serosurveillance, parasite and mosquito genetics, and vector bionomics. Findings to date have helped elaborate the patterns and possible causes of sustained low-to-moderate transmission in southern Zambia and eastern Zimbabwe and recalcitrant high transmission and fatality in northern Zambia. Cryptic and novel mosquito vectors, asymptomatic parasite reservoirs in older children, residual parasitemia and gametocytemia after treatment, indoor residual spraying timed dyssynchronously to vector abundance, and stockouts of essential malaria commodities, all in the context of intractable rural poverty, appear to explain the persistent malaria burden despite current interventions. Ongoing studies of high-resolution transmission chains, parasite population structures, long-term malaria periodicity, and molecular entomology are further helping to lay new avenues for malaria control in southern and central Africa and similar settings.

Policy Implications of the Southern and Central Africa International Center of Excellence for Malaria Research: Ten Years of Malaria Control Impact Assessments in Hypo-, Meso-, and Holoendemic Transmission Zones in Zambia and Zimbabwe.

The International Centers of Excellence for Malaria Research (ICEMR) were established by the National Institute of Allergy and Infectious Diseases more than a decade ago to provide multidisciplinary research support to malaria control programs worldwide, operating in endemic areas and contributing technology, expertise, and ultimately policy guidance for malaria control and elimination. The Southern and Central Africa ICEMR has conducted research across three main sites in Zambia and Zimbabwe that differ in ecology, entomology, transmission intensity, and control strategies. Scientific findings led to new policies and action by the national malaria control programs and their partners in the selection of methods, materials, timing, and locations of case management and vector control. Malaria risk maps and predictive models of case detection furnished by the ICEMR informed malaria elimination programming in southern Zambia, and time series analyses of entomological and parasitological data motivated several major changes to indoor residual spray campaigns in northern Zambia. Along the Zimbabwe-Mozambique border, temporal and geospatial data are currently informing investigations into a recent resurgence of malaria. Other ICEMR findings pertaining to parasite and mosquito genetics, human behavior, and clinical epidemiology have similarly yielded immediate and long-term policy implications at each of the sites, often with generalizable conclusions. The ICEMR programs thereby provide rigorous scientific investigations and analyses to national control and elimination programs, without which the impediments to malaria control and their potential solutions would remain understudied.

Literature review of the systematics, biology and role in malaria transmission of species in the Afrotropical Anopheles subgenus Anopheles (Diptera: Culicidae).

It has been 34 years since the last update of the subgenus Anopheles Meigen, 1818 in the Afrotropical Region. Eleven species occur in the region, and of these, eight occur only on the African continent, two in both Africa and Madagascar, and one exclusively in Madagascar. Three species are implicated in the transmission of malarial parasites in localised areas: An. coustani Laveran, 1900 (Cameroon, Central African Republic, Democratic Republic of the Congo, Kenya, Madagascar, Tanzania and Zambia), An. paludis Theobald, 1900 (Cameroon, Democratic Republic of the Congo) and An. ziemanni Grnberg, 1902 (Cameroon, Chad, Ethiopia and Rwanda). Several arboviruses have been isolated from An. coustani in Kenya, Madagascar and Senegal. Recent molecular studies indicate possible undescribed species within An. coustani that could be resolved with integrated molecular, morphological and cytogenetic methods.

Expanded geographic distribution and host preference of Anopheles gibbinsi (Anopheles species 6) in northern Zambia.

BACKGROUND: Nchelenge District in northern Zambia suffers from holoendemic malaria transmission despite a decade of yearly indoor residual spraying (IRS) and insecticide-treated net (ITN) distributions. One hypothesis for this lack of impact is that some vectors in the area may forage in the early evening or outdoors. Anopheles gibbinsi specimens were identified in early evening mosquito collections performed in this study area, and further insight was gleaned into this taxon, including characterizing its genetic identity, feeding preferences, and potential role as a malaria vector. METHODS: Mosquitoes were collected in July and August 2019 by CDC light traps in Nchelenge District in indoor sitting rooms, outdoor gathering spaces, and animal pens from 16:00-22:00. Host detection by PCR, COI and ITS2 PCR, and circumsporozoite (CSP) ELISA were performed on all samples morphologically identified as An. gibbinsi, and a subset of specimens were selected for COI and ITS2 sequencing. To determine risk factors for increased abundance of An. gibbinsi, a negative binomial generalized linear mixed-effects model was performed with household-level variables of interest. RESULTS: Comparison of COI and ITS2 An. gibbinsi reference sequences to the NCBI database revealed > 99% identity to "Anopheles sp. 6" from Kenya. More than 97% of specimens were morphologically and molecularly consistent with An. gibbinsi. Specimens were primarily collected in animal pen traps (59.2%), followed by traps outdoors near where humans gather (24.3%), and traps set indoors (16.5%). Host DNA detection revealed a high propensity for goats, but 5% of specimens with detected host DNA had fed on humans. No specimens were positive for Plasmodium falciparum sporozoites. Animal pens and inland households > 3 km from Lake Mweru were both associated with increased An. gibbinsi abundance. CONCLUSIONS: This is the first report of An. gibbinsi in Nchelenge District, Zambia. This study provided a species identity for unknown "An. sp. 6" in the NCBI database, which has been implicated in malaria transmission in Kenya. Composite data suggest that this species is largely zoophilic and exophilic, but comes into contact with humans and the malaria parasites they carry. This species should continue to be monitored in Zambia and neighbouring countries as a potential malaria vector.

Intensity of insecticide resistance in the major malaria vector Anopheles funestus from Chikwawa, rural Southern Malawi.

BACKGROUND: Malaria vector control using insecticide-based approaches has proven to be an effective strategy. However, widespread insecticide resistance among malaria vector populations across sub-Saharan Africa threatens to derail control efforts. This study was conducted in Chikwawa district, an area in rural southern Malawi characterised by persistent malaria transmission and reports of insecticide resistance in the local mosquito population. The aim of the was to characterise the intensity of insecticide resistance within a population of Anopheles funestus sensu lato (s.l.), a major vector of malaria in this district. METHODS: Live adult females belonging to the An. funestus group were collected from households by indoor aspiration. The CDC bottle assay was used for phenotypic quantification of resistance to deltamethrin, permethrin and alpha-cypermethrin at 1×, 2.5×, 5× and 10× the recommended diagnostic dose for each of these insecticides. WHO tube assays were used to determine susceptibility to bendiocarb, dichlorodiphenyltrichloroethane (DDT) and pirimiphos-methyl insecticides at diagnostic concentrations. RESULTS: Anopheles funestus s.l. exposed to 10× the recommended diagnostic dose was highly resistant to alpha-cypermethrin (mortality 95.4%); in contrast, mortality was 100% when exposed to both deltamethrin and permethrin at the same dose. Despite showing susceptibility to deltamethrin and permethrin at the 10× concentration, mortality at the 5× concentration was 96.7% and 97.1%, respectively, indicating moderate resistance to these two insecticides. WHO susceptibility assays indicated strong resistance against bendiocarb (mortality 33.8%, n = 93), whereas there was full susceptibility to DDT (mortality 98.9%, n = 103) and pirimiphos-methyl (mortality 100%, n = 103). CONCLUSIONS: Strategies for managing resistance to insecticides, particularly against pyrethroids, must be urgently implemented to maintain the effectiveness of insecticide-based vector control interventions in the area. Such strategies include the wide-scale introduction of third-generation synergist insecticide-treated bed nets (ITNs) and next-generation dual active ingredient ITNs. The use of effective non-pyrethroids, such as pirimiphos-methyl, clothianidin and potentially DDT, could provide a window of opportunity for indoor residual spraying across the district. This strategy would support the current Malawi Insecticide Resistance Management Plan which aims at rotating insecticides to minimise selection pressure and slow down the evolution of resistance to approved insecticides. These actions will help to prevent malaria vector control failure and improve progress towards malaria elimination.

Effectiveness and cost-effectiveness of reactive, targeted indoor residual spraying for malaria control in low-transmission settings: a cluster-randomised, non-inferiority trial in South Africa.

BACKGROUND: Increasing insecticide costs and constrained malaria budgets could make universal vector control strategies, such as indoor residual spraying (IRS), unsustainable in low-transmission settings. We investigated the effectiveness and cost-effectiveness of a reactive, targeted IRS strategy. METHODS: This cluster-randomised, open-label, non-inferiority trial compared reactive, targeted IRS with standard IRS practice in northeastern South Africa over two malaria seasons (2015-17). In standard IRS clusters, programme managers conducted annual mass spray campaigns prioritising areas using historical data, expert opinion, and other factors. In targeted IRS clusters, only houses of index cases (identified through passive surveillance) and their immediate neighbours were sprayed. The non-inferiority margin was 1 case per 1000 person-years. Health service costs of real-world implementation were modelled from primary and secondary data. Incremental costs per disability-adjusted life-year (DALY) were estimated and deterministic and probabilistic sensitivity analyses conducted. This study is registered with ClinicalTrials.gov, NCT02556242. FINDINGS: Malaria incidence was 0·95 per 1000 person-years (95% CI 0·58 to 1·32) in the standard IRS group and 1·05 per 1000 person-years (0·72 to 1·38) in the targeted IRS group, corresponding to a rate difference of 0·10 per 1000 person-years (-0·38 to 0·59), demonstrating non-inferiority for targeted IRS (p<0·0001). Per additional DALY incurred, targeted IRS saved US$7845 (2902 to 64 907), giving a 94-98% probability that switching to targeted IRS would be cost-effective relative to plausible cost-effectiveness thresholds for South Africa ($2637 to $3557 per DALY averted). Depending on the threshold used, targeted IRS would remain cost-effective at incidences of less than 2·0-2·7 per 1000 person-years. Findings were robust to plausible variation in other parameters. INTERPRETATION: Targeted IRS was non-inferior, safe, less costly, and cost-effective compared with standard IRS in this very-low-transmission setting. Saved resources could be reallocated to other malaria control and elimination activities. FUNDING: Joint Global Health Trials.

New distribution record of Anopheles rivulorum-like from Sadiola, Mali, with notes on malaria vector insecticide resistance.

BACKGROUND: The SEMOS gold mine in Sadiola, southwestern Mali, has been implementing a malaria vector control programme for 15 y using indoor residual house spraying and sporadic larval control. Periodic screening of the vector populations have been carried out over the years to provide information to the control programme, mainly on vector species present and their insecticide resistance status. The data from five entomological surveys, carried out in 2006, 2011, 2014, 2016 and 2018, are presented. METHODS: Adult mosquitoes were collected resting on walls inside houses and on verandas. Insecticide susceptibility assays were carried out and mosquitoes subsequently identified by species using molecular assays. RESULTS: The major malaria vector mosquitoes, Anopheles gambiae and Anopheles arabiensis were abundant at each sampling period with Anopheles coluzzii and Anopheles funestus being rare or absent. Anopheles rivulorum was identified in 2006 and Anopheles leesoni in 2016. The presence of Anopheles rivulorum-like, identified for the first time in 2018, was not screened for in previous surveys. Insecticide susceptibility bioassays showed resistance in both A. gambiae and A. arabiensis to pyrethroids, carbamates and dichlorodiphenyltrichloroethane over the 12 y. CONCLUSIONS: This is the first record of A. rivulorum-like west of Côte d'Ivoire. Resistance levels to the three classes of insecticides were variable but appeared to decrease after pyrethroids were discontinued for house spraying.

Modeling Temperature Variations Using Monte Carlo Simulation: Implications for Estimation of the Postmortem Interval Based on Insect Development Times*, †.

The association between insect development and temperature is well established. Thermal summation using accumulated degree-day measures is commonly used. However, the time at which evidence is collected is important in these estimates. The aim of this study was to provide a simulated model of the effect of temperatures on six dipteran species commonly associated with cadavers, from the death scene to the refrigerator, and finally at the time of autopsy. Temperatures measurements were sampled over a 16-month period from the external environment (external to the mortuary), within the mortuary refrigerator, and within the mortuary autopsy suite. Monte Carlo simulation using accumulated degree-days (ADD) was used to estimate the variations based on the mean and standard deviation of the temperature measurements. It was found that there was a negative correlation between the base temperature of the fly species (lowest temperature at which the flies will survive) and developmental likelihood. Species with high base temperatures (Chrysomya albiceps, Chrysomya chloropyga, and Musca domestica) were less likely to continue development in refrigerators than species with lower base temperatures (Lucilia sericata and Piophila casei). The findings of this study highlight the importance of recording temperature measurements and the period of refrigeration on PMI estimation especially when continued development occurs in spite of a period of cooling of the insect evidence.

Malaria Vectors and Vector Surveillance in Limpopo Province (South Africa): 1927 to 2018.

Despite the annual implementation of a robust and extensive indoor residual spraying programme against malaria vectors in Limpopo Province (South Africa), significant transmission continues and is a serious impediment to South Africa's malaria elimination objectives. In order to gain a better understanding regarding possible causes of this residual malaria, we conducted a literature review of the historical species composition and abundance of malaria vector mosquitoes in the Limpopo River Valley region of the Vhembe District, northern Limpopo Province, the region with the highest remaining annual malaria cases in South Africa. In addition, mosquito surveys were carried out in the same region between October 2017 and October 2018. A total of 2225 adult mosquitoes were collected using CO2-baited tent and light traps, human landing catches and cow-baited traps. Of the 1443 Anopheles collected, 516 were members of the An. gambiae complex and 511 An. funestus group. In the malaria endemic rural areas outside the Kruger National Park, one specimen each of An. gambiae s.s. and An. funestus and only three of An. arabiensis were collected. The latter species was abundant at a remote hot spring in the neighboring Kruger National Park. Eighteen other species of Anopheles were collected. Our survey results support the historical findings that An. arabiensis, the species widely held to be the prime malaria vector in South Africa, is a rare species in the malaria endemic Limpopo River Valley. The implications of the mosquito surveys for malaria transmission, elimination and vector control in northern Limpopo Province and neighboring regions are discussed.

Updated list of Anopheles species (Diptera: Culicidae) by country in the Afrotropical Region and associated islands.

The distributions of the Afrotropical Anopheles mosquitoes were first summarized in 1938. In 2017, an extensive geo-coded inventory was published for 48 sub-Saharan African countries, including information such as sampling methods, collection dates, geographic co-ordinates and the literature consulted to produce the database. Using the information from the 2017 inventory, earlier distribution lists, museum collections and publications since 2016, this paper presents an updated, simplified list of Anopheles species by mainland countries and associated Afrotropical islands, with comments where applicable. It is intended as a supplement to the 2017 geo-coded inventory.

Genetic differentiation and population structure of Anopheles funestus from Uganda and the southern African countries of Malawi, Mozambique, Zambia and Zimbabwe.

BACKGROUND: Anopheles funestus (s.s.) is a primary vector of the malaria parasite Plasmodium falciparum in Africa, a human pathogen that causes almost half a million deaths each year. The population structure of An. funestus was examined in samples from Uganda and the southern African countries of Malawi, Mozambique, Zambia and Zimbabwe. METHODS: Twelve microsatellites were used to estimate the genetic diversity and differentiation of An. funestus from 13 representative locations across five countries. These were comprised of four sites from Uganda, three from Malawi and two each from Mozambique, Zambia and Zimbabwe. RESULTS: All loci were highly polymorphic across the populations with high allelic richness and heterozygosity. A high genetic diversity was observed with 2-19 alleles per locus and an average number of seven alleles. Overall, expected heterozygosity (He) ranged from 0.65 to 0.79. When samples were pooled three of the 12 microsatellite loci showed Hardy-Weinberg equilibrium. Unsupervised Bayesian clustering analysis of microsatellite data revealed two clusters with An. funestus samples from Mozambique, Uganda and Zambia falling into one group and Malawi and Zimbabwe into another. The overall genetic differentiation between the populations was moderate (FST = 0.116). Pairwise differentiation between the pairs was low but significant. A weak but significant correlation was established between genetic and geographical distance for most populations. CONCLUSIONS: High genetic diversity revealed by the loci with low to moderate differentiation, identified two clusters among the An. funestus populations. Further research on the population dynamics of An. funestus in east and southern Africa is essential to understand the implications of this structuring and what effect it may have on the efficient implementation of mosquito vector control strategies.

Key to the females of Afrotropical Anopheles mosquitoes (Diptera: Culicidae).

BACKGROUND: In 1987, Gillies and Coetzee published a pictorial key for the morphological identification of adult female mosquitoes. Since then, several new species of anopheline mosquitoes have been described. METHODS: The 1987 key to adult female mosquitoes was used as the template for the current key. RESULTS: New species described in the literature over the past 32 years have been included. A list of all currently known Afrotropical species is provided. Anopheles stephensi is included for the first time as occurring on the African continent. CONCLUSIONS: An updated key for the morphological identification of Afrotropical anopheline species is presented.

Anopheline Mosquito Species Composition, Kdr Mutation Frequency, and Parasite Infectivity Status in Northern Tanzania.

The scaling-up of malaria control interventions in northern Tanzania has resulted in a decline in malaria prevalence and vector species composition. Despite this achievement, residual malaria transmission remains a concern in the area. The main aim of this study was to investigate malaria vector species composition, parasite infectivity rates, and the presence of insecticide knockdown resistance (kdr) mutations in three sites that have experienced a significant decline in malaria in northern Tanzania. Adult mosquitoes were sampled using light traps in houses and hand-aspirators in cowsheds, whereas the standard dipping method was used for sampling mosquito larvae. Adult mosquitoes identified as Anopheles gambiae s.l. and An. funestus s.l. and larval stages III and IV of An. gambiae s.l. were stored in absolute ethanol for further laboratory molecular identification. The identified species in the An. gambiae complex were An. gambiae s.s., An. merus, An. quadriannulatus, and An. arabiensis, whereas the An. funestus group comprised An. funestus s.s., An. rivulorum, and An. leesoni. For An. gambiae s.s. analyzed from Zeneth, 47.6% were kdr-East homozygous susceptible, 35.7% kdr-East heterozygous resistant, 9.6% kdr-East homozygous resistant, and 7.1% undefined, whereas specimens from Kwakibuyu were 45.5% kdr-East homozygous susceptible, 32.7% kdr-East heterozygous resistant, 16.3% kdr-East homozygous resistant, and 5.5% undefined. There were no kdr-West alleles identified from any specimen. The overall malaria parasite infectivity rate was 0.75%. No infections were found in Moshi. The findings indicate that populations of the major malaria vector mosquitoes are still present in the study area, with An. funestus taking a lead in malaria transmission.

Fine-scale spatial and temporal variations in insecticide resistance in Culex pipiens complex mosquitoes in rural south-eastern Tanzania.

BACKGROUND: Culex mosquitoes cause considerable biting nuisance and sporadic transmission of arboviral and filarial diseases. METHODS: Using standard World Health Organization procedures, insecticide resistance profiles and underlying mechanisms were investigated during dry and wet seasons of 2015 and 2016 in Culex pipiens complex from three neighbouring administrative wards in Ulanga District, Tanzania. Synergist tests with piperonyl butoxide, diethyl maleate, and triphenyl phosphate, were employed to investigate mechanisms of the observed resistance phenotypes. Proportional biting densities of Culex species, relative to other taxa, were determined from indoor surveillance data collected in 2012, 2013, and 2015. RESULTS: Insecticide resistance varied significantly between wards and seasons. For example, female mosquitoes in one ward were susceptible to bendiocarb and fenitrothion in the wet season, but resistant during the dry season, while in neighbouring ward, the mosquitoes were fully susceptible to these pesticides in both seasons. Similar variations occurred against bendiocarb, DDT, deltamethrin, and lambda-cyhalothrin. Surprisingly, with the exception of one ward in the wet season, the Culex populations were susceptible to permethrin, commonly used on bednets in the area. No insecticide resistance was observed against the organophosphates, pirimiphos-methyl and malathion, except for one incident of reduced susceptibility in the dry season. Synergist assays revealed possible involvement of monooxygenases, esterases, and glutathione S-transferase in pyrethroid and DDT resistance. Morphology-based identification and molecular assays of adult Culex revealed that 94% were Cx. pipiens complex, of which 81% were Cx. quinquefasciatus, 2% Cx. pipiens, and 3% hybrids. About 14% of the specimens were non-amplified during molecular identifications. Female adults collected indoors were 100% Cx. pipiens complex, and constituted 79% of the overall biting risk. CONCLUSIONS: The Cx. pipiens complex constituted the greatest biting nuisance inside people's houses, and showed resistance to most public health insecticides possible. Resistance varied at a fine geographical scale, between adjacent wards, and seasons, which warrants some modifications to current insecticide resistance monitoring strategies. Resistance phenotypes are partly mediated by metabolic mechanisms, but require further evaluation through biochemical and molecular techniques. The high densities and resistance in Culex could negatively influence the acceptability of other interventions such as those used against malaria mosquitoes.

Anopheles parensis contributes to residual malaria transmission in South Africa.

BACKGROUND: Understanding the contribution of outdoor-resting Anopheles mosquitoes to residual malaria transmission is important in terms of scaling up vector control towards malaria elimination in South Africa. The aim of this project was to assess the potential role of Anopheles parensis and other Anopheles species in residual malaria transmission, using sentinel surveillance sites in the uMkhanyakude District of northern KwaZulu-Natal Province. METHODS: Monthly vector surveillance was conducted at the sentinel sites from January 2017 to May 2018. Outdoor-placed clay pot resting traps were used to collect male and female adult Anopheles mosquitoes. All Anopheles gambiae complex and Anopheles funestus group specimens collected were identified to species and all females were screened for Plasmodium falciparum circumsporozoite protein (CSP) by enzyme-linked immunosorbent assay (ELISA). Samples showing infectivity for P. falciparum were further verified by a nested PCR and subsequent DNA sequence analysis. RESULTS: From a sample of 491 anophelines, Anopheles arabiensis (n = 228) and An. parensis (n = 194) were the most abundant. Other species collected included Anopheles merus (n =11), Anopheles quadriannulatus (n = 10), Anopheles leesoni (n = 29), Anopheles rivulorum (n =18), and Anopheles vaneedeni (n =1). Of the 317 female specimens screened for P. falciparum CSP, one Anopheles arabiensis and one An. parensis showed positive by ELISA and Plasmodium nested PCR. For the An. parensis specimen, confirmation of its species identity was based on sequence analysis of the ITS2 region, and the presence of P. falciparum DNA was further confirmed by sequence analysis. CONCLUSIONS: Anopheles parensis is a potential vector of malaria in South Africa although its contribution to transmission is likely to be minimal at best owing to its strong zoophilic tendency. By contrast, An. arabiensis is a major vector that is primarily responsible for the bulk of residual malaria transmission in South Africa. As all recently collected sporozoite-positive Anopheles mosquitoes were found in outdoor-placed resting traps, it is necessary to introduce interventions that can be used to control outdoor-resting vector populations while maintaining the efficacy of South Africa's indoor house spraying operations.

Linking human behaviours and malaria vector biting risk in south-eastern Tanzania.

To accelerate malaria elimination in areas where core interventions such as insecticide-treated nets (ITNs) are already widely used, it is crucial to consider additional factors associated with persistent transmission. Qualitative data on human behaviours and perceptions regarding malaria risk was triangulated with quantitative data on Anopheles mosquito bites occurring indoors and outdoors in south-eastern Tanzania communities where ITNS are already used but lower level malaria transmission persists. Each night (18:00h-07:00h), trained residents recorded human activities indoors, in peri-domestic outdoor areas, and in communal gatherings. Host-seeking mosquitoes were repeatedly collected indoors and outdoors hourly, using miniaturized exposure-free double net traps (DN-Mini) occupied by volunteers. In-depth interviews were conducted with household representatives to explore perceptions on persistent malaria and its control. Higher proportions of people stayed outdoors than indoors in early-evening and early-morning hours, resulting in higher exposures outdoors than indoors during these times. However, exposure during late-night hours (22:00h-05:00h) occurred mostly indoors. Some of the popular activities that kept people outdoors included cooking, eating, relaxing and playing. All households had at least one bed net, and 83.9% of people had access to ITNs. Average ITN use was 96.3%, preventing most indoor exposure. Participants recorgnized the importance of ITNs but also noted that the nets were not perfect. No complementary interventions were reported being used widely. Most people believed transmission happens after midnight. We conclude that insecticide-treated nets, where properly used, can still prevent most indoor exposures, but significant risk continues unabated before bedtime, outdoors and at communal gatherings. Such exposure is greatest for rural and low-income households. There is therefore an urgent need for complementary interventions, particularly those targeting outdoor-biting and are applicable for all people including the marginalised populations such as migratory farmers and fishermen. Besides, the differences in community understanding of ongoing transmission, and feedback on imperfections of ITNs should be considered when updating malaria-related communication and interventions.

Swarms of the malaria vector Anopheles funestus in Tanzania.

BACKGROUND: Anopheles funestus mosquitoes currently contribute more than 85% of ongoing malaria transmission events in south-eastern Tanzania, even though they occur in lower densities than other vectors, such as Anopheles arabiensis. Unfortunately, the species ecology is minimally understood, partly because of difficulties in laboratory colonization. This study describes the first observations of An. funestus swarms in Tanzania, possibly heralding new opportunities for control. METHOD: Using systematic searches by community-based volunteers and expert entomologists, An. funestus swarms were identified in two villages in Ulanga and Kilombero districts in south-eastern Tanzania, starting June 2018. Swarms were characterized by size, height, start- and end-times, presence of copulation and associated environmental features. Samples of male mosquitoes from the swarms were examined for sexual maturity by observing genitalia rotation, species identity using polymerase chain reaction and wing sizes. RESULTS: 581 An. funestus (98.1% males (n = 570) and 1.9% (n = 11) females) and 9 Anopheles gambiae sensu lato (s.l.) males were sampled using sweep nets from the 81 confirmed swarms in two villages (Ikwambi in Kilombero district and Tulizamoyo in Ulanga district). Six copulation events were observed in the swarms. Mean density (95% CL) of An. funestus caught/swarm/village/evening was 6.6 (5.9-7.2) in Tulizamoyo and 10.8 (5.8-15.8) in Ikwambi. 87.7% (n = 71) of the swarms were found in Tulizamoyo, while 12.3% (n = 10) were in Ikwambi. Mean height of swarms was 1.7 m (0.9-2.5 m), while mean duration was 12.9 (7.9-17.9) minutes. The PCR analysis confirmed that 100% of all An. funestus s.l. samples processed were An. funestus sensu stricto. Mean wing length of An. funestus males was 2.47 mm (2.0-2.8 mm), but there was no difference between swarming males and indoor-resting males. Most swarms (95.0%) occurred above bare ground, sometime on front lawns near human dwellings, and repeatedly in the same locations. CONCLUSION: This study has demonstrated occurrence of An. funestus swarms for the first time in Tanzania. Further investigations could identify new opportunities for improved control of this dominant malaria vector, possibly by targeting the swarms.

Correction to: Housing gaps, mosquitoes and public viewpoints: a mixed methods assessment of relationships between house characteristics, malaria vector biting risk and community perspectives in rural Tanzania.

Following publication of the original article [1], the author flagged that the clause "and competing household priorities" was missing from the second sentence of the conclusion section of the Abstract; while this clause was in the Conclusion section of the main article text.