Methodological Diversity in Citizen Science Mosquito Surveillance: A Scoping Review

ABSTRACT


INTRODUCTION
Initiatives involving the community in monitoring the spread of mosquitoes, pathogens, and environmental changes are examples of successful partnerships between the general public and scientists (Roger 2019;Pawson et al. 2020;Braz Sousa et al. 2020).Such initiatives are identified as citizen science when there is an intentional community engagement and active public participation in scientific research.Citizen science is becoming more common and widespread within the scientific community as it both enhances the capacity for data collection and provides an opportunity for greater meaningful engagement between stakeholders, including community members, researchers, public health agents, industry, and policymakers (Roger 2019;Tarter et al. 2019).
The current COVID-19 pandemic has highlighted the importance of strengthened relationships between governments and public mobilisation to monitor and contain disease spread.Educational campaigns related to COVID-19 specifically have become more prevalent, and citizen science programs tracking the virus's spread are becoming popular (Katapally 2020).Public participation in scientific research can overcome logistical and spatial barriers regarding data collection and enhance outcomes by bringing local contextual problems to the discussion.Further, citizen science programs promote public understanding of science amongst participants (Bonney et al. 2016).There is potential for citizen science methods to be applied to other areas of need for public health benefit, such as the prevention of mosquito-borne diseases.Indeed, community engagement and mobilisation are part of the WHO pillars to tackle mosquito-borne diseases globally (WHO 2017).
Approximately 700 million people are infected by mosquito-borne diseases every year worldwide, and at least 1 million people die annually because of these diseases.These numbers may be much higher as diseases are still underreported in several countries (Beatty et al. 2011;Amarasinghe et al. 2011;Silva et al. 2016).Also, mosquitoborne diseases have a significant economic impact on the affected populations (WHO 2017;WMP 2021).With exotic mosquitoes introduced to new areas and lack of effective monitoring and control, such diseases are emerging and re-emerging worldwide (Bartumeus et al. 2018;Jones et al. 2021).Effective control of mosquito-borne disease requires surveillance systems that can report on the diversity and abundance of mosquitoes.Such surveillance typically makes use of specialised mosquito traps and trained professionals to identify disease-carrying species, so that control resources (such as measures to prevent mosquito bites and insecticide to reduce populations) can be deployed.
Citizen science initiatives in mosquito surveillance are expanding rapidly across the globe, enabling innovative methods to prevent and control arboviral diseases, and low-cost public participation and upscaling of traditional mosquito monitoring methods (Palmer et al. 2017;Braz Sousa et al. 2020;Carrillo et al. 2021).Advances in technology have enabled greater engagement in citizen science, and have facilitated interconnectivity amongst participants and researchers with sharing of real-time and geotagged data (Palmer et al. 2017;Caputo et al. 2020;Pawson et al. 2020).Adoption of new technologies has allowed citizen science programs to overcome some of the logistical and spatial barriers in data collection, and some methods can even provide real-time information regarding human mosquito encounter and nuisance (Bartumeus et al. 2018).
No study to date has compared the citizen science methods used for mosquito surveillance from a global perspective.Although some reviews were published in 2015, 2017, and 2018 highlighting the importance and impact of citizen science mosquito surveillance programs, they focused on the outcomes in Europe and North America (Kampen et al. 2015;Palmer et al. 2017;Bartumeus et al. 2018).A recent review of school-based citizen science projects (Abourashed et al. 2021) further highlights the potential for engaging the public throughout their life course.
Given the rapid expansion of these programs, it is also timely to review these programs and compare how other countries and continents are approaching this issue.
This paper presents a scoping review of the range, nature, and outcomes of citizen science mosquito surveillance interventions and programs globally.It further discusses the future of such programs as national implementations in their countries and how they could be integrated with formal mosquito surveillance programs.

METHODS
This review was conducted according to the Scoping Review extension of the Preferred Reporting Items for Systematic Reviews and Meta Analyses (PRISMA) statement and guidelines (Tricco 2018).

SEARCH STRATEGY AND ELIGIBILITY CRITERIA
We searched the electronic databases Scopus, Medline, PubMed, Web of Science, and Google Scholar, as well as the references lists of selected articles.This review aimed to identify peer-reviewed scientific literature and other relevant publicly available reports, published in English, describing citizen science mosquito surveillance initiatives (search strategy is presented in Supplemental File 1: Appendix 1).
We initially conducted a preliminary search on the aforementioned databases and found that interventions specifically named citizen science are relatively recent, the first one dated 2011.Indeed, the term citizen science was first introduced in the mid-1990s (Phillips 2018).In our pilot search, we found the earliest interventions involving the engagement of the public in mosquito monitoring dated from 2000, although these were not typically named as citizen science activities.For this reason, this scoping review included articles published from 2000 to March 2021, and databases were searched from February to March 2021.Keywords used in the electronic search, and inclusion and exclusion criteria, are presented in Table 1.Grey literature included newspaper, magazine, and website articles as well as other published materials and other unpublished works (e.g., reports, infographics, posters, conference abstracts).This scoping review used the PEO (population, exposure, outcome) approach to frame the search strategy.PEO is often used to determine the association between specific exposures/approaches or risk factors and outcomes (Moola et al. 2015).In brief, this entails analysing each article to determine the population (e.g., city, group, state, district, etc.), the exposure (e.g., type of citizen science activity), and outcome (e.g., results of the work, findings, etc.).The search strategy involved main categories of terms describing citizen science and mosquito surveillance, as well as targeted time frame and language of the publication.The concept of citizen science has also been characterised as community science, crowd science, public participation in scientific research, volunteer-based monitoring, and participatory science (Carr 2004;Bonney et al. 2009;Miller-Rushing et al. 2012).As this approach has been named in different ways throughout past decades, we decided to include synonyms and other historical names in the search strategy.Further, most citizen science projects related to mosquitoes focus on the prevention of mosquito-borne diseases, as well as on the outcomes and benefits to public health.Thus, we included mosquito-related topics and some disease names in the search (Table 1).

SCREENING
All references were exported to EndNote X9 ® (2019, Clarivate Analytics, Toronto, Canada) for initial screening and to remove duplicates.After removing duplicates, the references were exported to Covidence ™ (2019, Alfred Hospital in Melbourne, Australia, Instituto de Efectividad Clinica Y Sanitaria (EROS) in Buenos Aires, Argentina) for screening.We screened the articles in a two-stage process, assessing the title and abstract as a first step and reviewing the full text as the second step.Screening was conducted by two independent reviewers (LBS and KLB).Articles were screened against the inclusion and exclusion criteria, and disagreements between reviewers were resolved through discussion.Reasons for exclusion included incorrect comparator, intervention, and/or study design (Figure 1).
Data extraction was conducted using a standardised template in Microsoft Excel (version 2017, © Microsoft Cooperation 2017).Data were extracted by one reviewer (LBS) according to the following fields: author(s), year of publication, origin/country (where the study was conducted), aims/purpose, study population and sample size, methods, intervention type and duration, comparator, outcome measures, and key findings.

RESULTS
A total of 1,027 articles were identified through the database search, and an additional 8 articles were identified through the grey literature search.After removing duplicates, 332 articles were imported from EndNote ® to Covidence™.Fourteen additional duplicates were identified and removed in Covidence™, resulting in 318 studies being screened.After title and abstract screening, 269 articles were excluded, leaving 49 full-text articles for full-text review.A further 16 articles were excluded during full-text review (Figure 1).
After the final full-text screening, 33 articles were eligible for inclusion.These papers described 23 different mosquito monitoring programs involving community participation (Supplemental File 2: Appendix 2).Additional citizen science  interventions were identified through grey literature search, authors' knowledge, blogs, and citizen science platforms.A total of 29 citizen science mosquito surveillance projects were found to be operating in 16 countries, besides three programs with wide geographic coverage.A summary of these programs is presented in Table 2. Amongst these programs, 21 were explicitly identified by the authors as citizen science approaches, and seven were communitybased programs and/or interventions not identified as citizen science in the published articles but considered as citizen science interventions in this review owing to the participatory nature of the interventions.
Although some programs did not refer to or categorise their approaches as citizen science, our scoping review identified several programs or pilot projects involving community-based interventions, community participation in focus groups to discuss their preference for specific interventions for mosquito control, and/or community mobilisation in identifying and eliminating mosquito breeding sites.Nonetheless, we considered some of these strategies as citizen science, as they involve the community in participatory research, and we present them in Table 2.
The participants engaged as citizen scientists were the general public (86%) and school students (14%).Programs also combined either general public and professionals or general public and network of community leaders or village volunteers in mosquito surveillance activities.
In terms of the life stage of mosquitoes assessed, the programs focused on adult mosquitoes (55%), mosquito eggs (17%), mosquito larvae (10%), and pupae (7%).Also, programs focused on identifying and reporting mosquito breeding sites (24%), and breeding sites combined with dengue or malaria symptoms (7%).The targeted mosquitoes included Aedes mosquitoes (whether specific  or not to the Asian tiger mosquito Ae. albopictus and the yellow fever mosquito Ae. aegypti) (59%), and non-specific species (41%).
The methods used for citizen science mosquito surveillance included the use of smartphone applications to report data (28%) and traps or available containers to collect specimens (48%); one program used a dedicated website for participants to upload acoustic data of mosquito wingbeats, and seven programs surveyed breeding sites as methods (24%).One program focused on acoustic data of flying mosquitoes, and two programs also collected data of adult mosquito nuisance.As reporting strategies, participants sent photos of collected mosquitoes (17%), mailed specimens for identification (41%), and reported the findings (usually of breeding sites and larvae presence) to the local health authorities and/or researchers (17%) (list of methods and strategies per program are shown in Supplemental File 3: Appendix 3).
Amongst the fourteen programs using traps for data collection, four used ovitraps (traps in which female mosquitoes lay eggs), seven used adult traps, and three programs advised participants to use any available container at home to capture mosquitoes.Three programs used do-it-yourself (DIY) approaches, in which participants made their traps at home to collect adults or eggs.Six programs required equipment delivered to participants, provided at no cost, including a trap kit or material to make the trap.Only one program required participants to purchase the traps.
Great variation in program length was evident (Table 1).Whilst some programs are listed as ongoing (9/22.41%), the majority were conducted over a period of 1 to 3 years (13/22, 59%).The North American Mosquito Project was a fixed-term program that was antecedent to the ongoing Invasive Mosquito Project, and as such is considered a single project in the aforementioned Table .All studies in this review reported achieving their study aims, and the benefits afforded citizen scientists by their engagement can be understood through different socioeconomic and geographic contexts.Also, in the same program, households received feedback to take specific actions to manage mosquito breeding sites (Suwanbamrung et al., 2018).Also, facilitated engagement between citizens and scientists was reported in different programs and countries (Lwin 2017;Parra et al. 2020;Craig et al. 2021).
Citizen science mosquito programs have been developed in several countries, notably across Europe (Palmer et al. 2017), where six different programs exist.These programs assist local authorities in the monitoring of abundance and diversity of endemic mosquitoes as well as in detecting exotic mosquitoes (Tyson, 2018).Mosquito Alert, Mueckenatlas, ZanzaMapp and Mosquito Habitat Mapper are examples of current and ongoing projects in Spain, Germany, Italy, and the United States, respectively.These programs have shown a high level of impact, utility, and reliability when compared with traditional surveillance methods in these countries, allowing realtime vector reports, mosquito presence and abundance tracking, and exotic species detection (Bartumeus et al., 2018, Eritja et al., 2019).Also, community engagement in mosquito monitoring has brought unexpected outcomes in specific contexts, as recently observed in Spain where the occurrence of the invasive species Ae. japonicus was detected for the first time (Eritja et al. 2019).In the US, the GLOBE Observer program yielded data similar to the efforts of thirteen years of professional data collection in only three years of evaluation (Amos et al. 2020).
Mo-Buzz in Sri-Lanka focused on gathering information on community perception of mosquito density and denguelike symptoms (Lwin 2017;Lwin et al. 2019).Dengue Chat in Nicaragua and Lansaka Model in Thailand are examples of local programs that collected data on breeding site distribution and the presence of larvae in the surveyed areas (Coloma 2016;Suwanbamrung et al. 2018).
In Australia, Mozzie Monitors and Zika Mozzie Seeker have been engaging the community in tracking mosquito presence and abundance, and in detecting exotic species.Mozzie Monitors is the first citizen science mosquito surveillance program in the world that combines fixed-point traps and public reporting in adult mosquito monitoring (Braz Sousa et al. 2020).This approach engages members of the community to monitor species of medical importance and increases knowledge of mosquito fauna across Australia.Preliminary studies have demonstrated that this program is as effective at recording diversity of mosquitoes as is the formal mosquito surveillance programs, and it costs only about 20% of the total budget estimated for the traditional methods (Braz Sousa et al. 2020).
Zika Mozzie Seeker is an initiative of Metro South Health (a local area network of Queensland Health).It has been running since 2017 in Queensland, Australia, where Aedes aegypti is established in the state›s far north region.The program engages community members to set up DIY backyard mosquito egg traps, collect the eggs, and send them to researchers for molecular analysis.To detect the presence of Ae. aegypti and Ae.albopictus in sample, the program uses PCR technology developed by Queensland Health (Montgomery 2020).
Also, pilot projects were described in the Solomon Islands and in Rwanda, aiming to identify the opportunities and challenges to establish citizen science-based programs in these areas (Murindahabi et al. 2019;Craig et al. 2021).The citizen science engagement in Honiara, Solomon Islands, showed potential to upscale capacity for mosquito data collection locally.Participants also reported being motivated to participate in programs of social benefit and to take opportunities to learn new skills (Craig et al., 2021).

DISCUSSION
This scoping review found 29 citizen science programs operating across 16 countries, and three of these programs showed wide geographic coverage.Citizen science has been explored as a tool to monitor native and exotic mosquitoes, and the programs have used different approaches, including traps, smartphone apps, questionnaires, and partnership with local volunteers that inspect mosquito breeding sites in their neighbourhoods.Overall, these programs have demonstrated capacity to upscale data collection and to connect citizens to researchers and public health authorities.
Almost all the programs reported being free to join and allowing open communication before, during, and after trials.Also, programs such as Mosquito Alert and Muckenatlas, which have been running for almost ten years, have maintained an active cohort of citizen scientists that keep contributing over time.The Mozzie Monitors program has proven effective to collect mosquitoes of medical and ecological importance and has provided monthly feedback to participants through a free-to-use online platform.
The World Health Organisation (WHO) has set goals to tackle vector mosquitoes worldwide and to reduce mosquito-borne infections by at least 60% by 2030.This plan, named Global Vector Control, was established in 2016 and focuses on four main pillars: strengthening interand intra-sectoral action and collaboration; engaging and mobilizing communities; enhancing vector surveillance, and monitoring and evaluating interventions; and scaling up and integrating tools and approaches (WHO 2017).These four pillars support the development of sustainable vector control implementations.
Programs that seek partnerships with stakeholders in different sectors are aligned with the first of the WHO's fundamental pillars to tackle mosquito-borne diseases worldwide.All programs identified through articles included in this scoping review align with the second of the WHO's pillars of effective vector control; as community engagement and mobilisation are the core of this pillar, every citizen science approach described here meets this requirement.The sustainability of the programs aligns with the third pillar set by WHO; to ensure these programs are sustainable, it is necessary to evaluate the programs' long-term benefits and possible educational gains.Finally, the last pillar proposes that programs should develop methods that are appropriate to the local settings.To meet this concern, strategies of mosquito monitoring must be accessible to the broad community.Programs that rely on participants owning technological devices (such as smartphones and internet connection) could face some challenges in impoverished areas.
While there is great potential to assist local health authorities, citizen science mosquito surveillance initiatives also face many challenges in ensuring outcomes are translatable to improved public health policy and practice.Despite being a relatively new field of study, these programs have been increasing owing to the concern regarding the emergence and re-emergence of mosquitoborne infectious diseases (Palmer et al. 2017).Citizen science may prove critical in detecting exotic species and disease vectors, demonstrating that this partnership is a powerful tool to be used in mosquito surveillance programs worldwide, as it allows highly geographically refined data collection.

CONCLUSIONS
The included citizen science programs focused on different strategies and methods according to their local and national contextual needs.Most of the programs reported being free or low cost, and were designed so that participants required no specialist technical expertise.Also, citizen scientists valued the opportunity to actively contribute to a scientific activity in which they saw value.
Citizen science mosquito surveillance programs have proven successful both locally and nationally to yield information on mosquito abundance, diversity, distribution, and nuisance.Programs have been using different methods to address specific aims, and some projects have contributed to unexpected findings, such as the detection of exotic species.The programs listed in this global review could establish a long-term citizen science culture on mosquito surveillance through a hands-on approach, besides bringing additional information to formal mosquito surveillance programs.Partnerships with government departments could benefit researchers, databases, public health agents, and the broad public.They could also be used as early warning systems to predict mosquito-borne diseases risks or outbreaks in areas where some diseases are endemic.
Additionally, programs using low-cost technologies have the potential to upscale existing mosquito surveillance.However, the integration of such successful methods with the assistance and representativeness of local partners, such as village volunteers or local community leaders, could provide an opportunity to grow citizen science mosquito monitoring interventions in impoverished areas with low access to technology or low internet coverage.
Notwithstanding these potential advantages, the role of citizen science mosquito surveillance in informing disease management directly remains unclear.Ensuring that citizen science methods yielded robust metrics that could be used as triggers for vector control should be a priority for research.
Establishing a citizen science culture in mosquito surveillance through a hands-on approach is essential and aligns with the Vector Control Response pillars outlined by WHO.Although the WHO does not mention the term citizen science in its Global Vector Control Response, the 64-page document developed in partnership with global experts in mosquito monitoring and neglected diseases highlights the importance of effective community engagement and mobilisation in vector control.Further, better coordination within and between sectors and novel interventions with proven effectiveness are among the response's main purposes.Citizen science frameworks on mosquito surveillance could have the potential to address the goals and targets set by the WHO.

FUTURE ACTIONS
There is a demand for citizen science initiatives on mosquito management.This demand is supported by the WHO and numerous local initiatives that we found in the first screening of this scoping review.Also, numerous studies have been reported in African countries, presenting frameworks to establish citizen science approaches to address the challenges related to mosquito management in the future (Asingizwe et al. 2018;Ashepet et al. 2021).
Further research is necessary to address the logistical issues in establishing sustainable programs running in partnership with government agencies and public health authorities.Also, additional studies focusing on transdisciplinary approaches are needed to assess how increased data on mosquito populations through citizen science can be translatable to improved public health policy and practice.For example, researchers in Australia are running a trial named Mozzie Month.This is the first national trial of the Mozzie Monitors program in the country, and aims to answers questions about feasibility, sustainability, biosecurity, ownership, and governance.The trial expects to further investigate how integration of citizen science-derived data and professional surveillance are achieved.
The need to expand projects involving public participation in mosquito-and mosquito-borne disease-related studies is evident.Enhancing community engagement and mobilisation in mosquito monitoring through citizen science could be a tool to tackle mosquito-borne diseases in increasingly sustainable and long-term ways.

Figure 1
Figure 1 PRISMA extension for Scoping Reviews flow diagram showing the scoping review process for citizen science based mosquito surveillance programs.

Table 1
Systematic search.

Table 2
Sousa et al.Citizen Science: Theory and Practice DOI: 10.5334/cstp.469Sousa et al.Citizen Science: Theory and Practice DOI: 10.5334/cstp.469Sousa et al.Citizen Science: Theory and Practice DOI: 10.5334/cstp.469Citizen science mosquito surveillance programs worldwide Community-based programs and/or interventions non-identified as citizen science in the published articles but considered as citizen science interventions in this review owing to participatory research.*No specific name for these programs/interventions.