This paper provides a detailed account on the malaria vectors and transmission intensity from five years of malaria vector entomological surveillance in mainland Tanzania. The findings from the analysis indicate that based on morphological identification, out of 63226 combined mosquitoes reported from 32 councils, 62% are An. gambiae s.l, 33% are An. funestus, and 0.05% are other anopheles. Out of 41383 mosquitoes that were amplified during laboratory analysis based on samples submitted from 14 councils, 51% are An. arabiensis, 33% are An. funestus s.s., 15% are An. gambiae s.s. Based on 14 qualified councils, out of 29524 mosquito tested, 14301 48% are An. arabiensis, 34% are An. funestus s.s.,18% are An. gambiae s.s.
The estimated man biting rates (MBR) varied across the different mosquito species and strata but on average significantly higher in An. gambiae than An. funestus (p < 0.05). The MBR was significantly higher in high stratum than both moderate (p < 0.05) and low (p < 0.01) strata. There were some variations between sporozoite rates (SR) between different species and strata, however, no significant difference between either species or strata can be reported based on the data from the 14 councils. In addition, the estimated entomological inoculation rates (EIR) were not significantly different across councils as well as between species and strata. Several mosquito samples that tested positive for SR in 2019 and 2021 is likely due to an outbreak from councils in very low stratum while in most part samples tested SR negative so on average EIR values are a slightly higher in very low stratum councils than one would have expected compared to councils in low stratum. On average, the EIR was < 1 in several councils in Tanzania which according to Beier et al [26] it is an indication that the malaria transmission is interrupted in those councils but it is very important to maintain the gain through a robust surveillance system. However, the EIR was > 1 in a number of councils in moderate and high transmission areas mediated mostly by An. funestus s.s. and An. arabiensis. The NMCP and partners must maintain and strengthen indoor control interventions targeting An. funestus [28] but equally important to consider targeting outdoor transmission that is mediated by An. arabiensis. The intervention such Larval Source Management (LSM) [29, 30] targeting immature mosquitoes should be considered but with careful planning and deployment based on WHO recommendations and/or in country experiences. Fortunately, LSM is considered a priority intervention in Tanzania and the plans for its implementation are well elaborated in the National Malaria Strategic Plan [31]. In addition, NMCP in collaboration with PORALG and a partner project, Towards Elimination of Malaria in Tanzania (TEMT) is implementing the LSM as a pilot project in Tanga region. The experiences and lessons from the TEMT project and modelling approaches [32, 33] should be considered to guide the scaling up of LSM in Tanzania.
The key findings from MVES are similar to several research-based studies conducted in specific study areas in Tanzania as indicated in selected references [13, 34, 35]. Anopheles funestus s.s. and An. arabiensis are observed to be more dominant in the Lake and South-East zones respectively, mostly, high transmission stratum. In general, An. arabiensis is found in most councils in higher numbers as compared to An. funestus s.s with low numbers for An. gambiae s.s. across different councils. An. funestus s.s. is becoming a more efficient species with higher EIR values reported in recent years (i.e., 2020 and 2021) as compared to those of An. arabiensis also reported in another study [36] in Tanzania. The impact of seasonality is observed across all councils, in general, the monthly mosquito densities show strong seasonal signals with two peaks after the rainy seasons, although the precise timing of the peaks differs slightly between species and councils.
During molecular charactersation, several mosquito samples were reported as unamplified. The NMCP in collaboration with in-country research institutions and academia should consider purifying and reanalysing the DNA of ‘unamplified’ samples as a watch for other important vectors including An. stephensi. Recently, the WHO issued a vector alert calling for countries in sub-Saharan Africa to increase vigilance for this invasive vector. As Tanzania updates its national vector surveillance framework to integrate An. stephensi, as a pre-emptive action against a threat to invasion it will also be important to ascertain that the vector is not already in the country unnoticed. In Sudan, An. stephensi was first described in samples that failed in PCR for An. gambiae s.l. species identification [37].
The MVES program is designed to ensure sustainability where community volunteers at the household level are responsible for setting mosquito traps under DVCOs’ supervision. The DVCOs are expected to perform morphological identification, label and pack the samples, right after the three consecutive days of mosquito collection, ready for the national supervision team to transfer the samples to the laboratory. The national supervision team is expected to visit all councils on quarterly basis (i.e., 4 times a year), perform supervision, identify and resolve any field encountered challenges, collect samples and send them to the laboratory. During the implementation using this approach several challenges were noted including low commitment from some of DVCOs leading to poor reporting of data, misidentification in some mosquito samples by some DVCOs, mismanagement of traps and chargers, and improper sample storage in the field. Also, occasionally the national supervision was conducted three times a year instead of four due to budget constraints leading to delays in sample submission to the laboratory. In addition, there was a fuzzy linkage between field and laboratory data which made tracing back of information to the household or village level not possible.
Despite the challenges, the combined dataset from all the councils, and especially those with consistent data reporting and sample submission to the laboratory, provides an assessment on malaria vector species composition, their abundance and seasonality, place of biting, host preference (vector behavior), and entomological inoculation rates for each species by strata. These entomological indicators are important in assessing the performance of previously deployed vector control interventions over time and in providing guidance on re-deployment going forward.
As a way forward, several adjustments are being made to streamline the MVES and improve data quality with lessons from the five years of experience. As an example, change in data entry template was done to ensure that cells are locked and the template cannot be modified on the ground - starting from 2021 there is an improvement in data quality. Along similar vein, the NMCP will need to finalize its plan to deploy an electronic database system to manage both field and laboratory data with proper data linkage to the household level. Given the vast size of the country and heterogeneity in malaria transmission, an electronic database system will facilitate monitoring of data reporting progress, recording data electronically even with no internet connectivity, ensure accountability at different levels, and provide interactive dash-boards to visualize data in real time by program management. The system will also ease the data sharing with DHIS2 and/or other data repositories/platforms in line with the NMCP desire to link entomological with epidemiological data and all information related to malaria control elimination strategies in the country. The generic schema described [38] provides key principles for designing and developing entomological databases that can be used to support diverse entomological studies including routine surveillance conducted by NMCPs. One such electronic system is Mosquito Database Management System (MosquitoDB), www.mosquitodb.io, that may be adapted by NMCPs to manage both field and laboratory data.
It is important to ensure that in addition to having a robust electronic entomological system, DVCOs are constantly trained and are committed to collect and record data timely. A suitable approach and method should be deployed to make sure that the information on mosquito resting behaviours is also recorded [38]. In-line with these recommendations for improvements, NMCP should consider increasing the number of councils to ensure that it is well positioned to monitor invasive mosquito species including An. stephensi.
The MVES system in Tanzania sets a good example to other countries either struggling to maintain or planning to establish malaria vector entomological surveillance systems. The experiences to be shared are particularly on the MVES’s methodology including the criteria provided for selecting sentinel sites.