1. Introduction
Several on-line platforms, such as Zooniverse, eBird, GLOBE Observer, Mosquito Alert and iNaturalist facilitate citizen science activities and data collection [
1,
2,
3,
4,
5,
6]. The iNaturalist platform (
https://www.inaturalist.org, last accessed on 20 March 2022), either as a smartphone application (app) or web-based, is amongst the most commonly used tools used to find and organise biodiversity findings observed by citizen scientists worldwide [
7,
8]. A joint initiative of the California Academy of Sciences and the National Geographic Society, iNaturalist facilitates an established social network of scientists and citizen scientists who share observations of biodiversity across the globe. As of March 2022, over 2 million people contributed to iNaturalist by sharing their observations (including photographs or audio recordings), and more than 200,000 people assisted with identifying organisms included in those observations. Users can either take part in specific groups of studies or conversations about species identification [
4,
8]. Several programs that used iNaturalist to identify species and provide a real-time and geographic location were successful in local and international biodiversity monitoring, empowering volunteers with little formal expertise to make observations of a variety of species, with expert verification of data provided via the iNaturalist platform [
4,
7,
9,
10,
11].
Recent literature has investigated several projects worldwide using iNaturalist to record biodiversity data, focusing on species distribution, conservation studies, migration patterns, distribution of invasive species and community engagement [
12,
13,
14,
15]. The Australasian Fishes project on iNaturalist is an example of successful engagement between museum curators, taxonomic professionals, researchers and several citizen scientists that expand scientific knowledge in Australia and New Zealand through data shared on the platform [
13]. Recently, the effectiveness of iNaturalist was also demonstrated by researchers in Australia in assessing the impacts of bushfires, with the platform providing scientific data on fire severity and biodiversity response, yielding data relevant to understanding future recovery [
16].
The utility and relevance of iNaturalist has mainly been assessed either for biodiversity research or in projects with socio-environmental impact [
7,
11,
17,
18]. There is a gap in understanding of how iNaturalist could be used to engage the public in data collection to enhance public health. Citizen science has begun to be incorporated into projects designed to improve the health and wellbeing of the community and resources such as iNaturalist may hold great potential to assist with managing the risks of mosquito-borne disease. It has already been applied to the surveillance of exotic mosquitoes (e.g.,
Aedes aegypti,
Aedes albopictus,
Aedes japonicus, and
Aedes koreicus) in Europe where observations uploaded to iNaturalist identified
Ae. albopictus as the most abundant species observed in 14 countries, strongly aligned with their known seasonality [
19]. There may be great utility in applying iNaturalist to enhance or expand mosquito surveillance programs in other regions of the world.
The use of a citizen science approach in mosquito monitoring has been explored since 2011 [
20,
21,
22]. It has provided opportunities to upscale geographic coverage of the traditional methods and gather real-time information related to human–mosquito encounters [
23]. This novel approach to mosquito surveillance has been growing in recent years due to the concern about mosquito-borne diseases spreading and the acknowledgement that it can be financially and operationally challenging for health authorities to sustain wide reaching professional mosquito monitoring programs [
23,
24]. These citizen science-based programs have shown success in engaging communities to assist monitoring species of medical importance through low-cost methods [
2,
22,
25]. A critical outcome of such programs has been the contribution of citizen scientists to the first detection of invasive species in Europe through the Mosquito Alert phone app in Spain [
26].
In Australia, the Mozzie Monitors program (
https://mozziemonitors.com/, accessed on 20 March 2022) was established in 2018 and has engaged participants in using a readily available and easily operated mosquito trap (i.e., BG-GAT (Gravid Aedes Trap)) to collect mosquito specimens and then to send digital images to researchers for identification, an approach termed “e-entomology”. The program has demonstrated that this approach can record a similar diversity of mosquitoes when compared to a traditional program in South Australia, costing about 25% of the total annual expenses for a professional mosquito surveillance program in the state [
24].
Concomitant with participants operating traps and submitting photos of collected specimens, Mozzie Monitors has been promoting the use of iNaturalist to expand the reach of the program to include uploaded images by citizen scientists not directly involved in the trap program. The number of observations and observers has been increasing every year since the Mozzie Monitors (
https://www.inaturalist.org/projects/mozzie-monitors-australia, accessed on 20 March 2022) project page was established in 2018. The total number of mosquitoes species recorded on the Mozzie Monitors project on iNaturalist was greater than those recorded by the mosquito trap component of project [
27].
To better understand how iNaturalist could be used for public health purposes, by tracking vector mosquitoes and interactions with humans, we reviewed a suite of attributes of the iNaturalist Mozzie Monitors project, specifically (1) distribution of mosquito fauna across Australia; (2) frequency of use and data sharing; (3) ecological association and species interaction through photos and behaviour description (e.g., mosquito biting, mosquito floral feeding); (4) perception and profile of the iNaturalist users’ network.
4. Discussion
For the first time, we report here the utility of iNaturalist as a resource to record the diversity of Australia’s mosquito fauna. This platform has proven useful in engaging communities for gathering scientific data, either for biodiversity research, or in projects with socio-environmental impact (e.g., monitoring bushfire recovery in Australia). However, there is a gap in exploring whether iNaturalist could be used to engage the public in data collection aiming to enhance public health, especially by a better understanding of the relative abundance and distribution of endemic, and potentially, exotic vector species.
Australia has a diverse mosquito fauna with distinct geographic distributions. Under the influence of a changing climate, there is often debate about shifting geographic distributions of mosquitoes, either directly resulting from a change in temperature and rainfall patterns or from human facilitated movement [
37]. Our analysis demonstrates that the known regional differences in mosquito fauna are reflected in the observations to iNaturalist from individuals in each state and territory and may, consequently, be useful to identify extended geographic ranges or novel introductions. There were no observations that differed markedly from the known or suspected distributions of the mosquito species in Australia [
38]. However, there were some noteworthy differences in observations of key species between states and territories. These differences were somewhat surprising given the known distribution and relative abundance of mosquitoes and the relatively low number of observations. This may be due to a number of factors including the number of active iNaturalist users within regions where these mosquitoes are most active or morphological attributes of the individual mosquito species. For example,
Cx. quinquefasciatus and
Cx. annulirostris are known to be highly abundant and widespread throughout the country except for Tasmania, [
38] but they were not observed by iNaturalist users in all states and territories. These mosquitoes, especially
Cx. quinquefasciatus, are often active in urban areas and around human habitation but were perhaps less likely to be photographed due to their non-distinctive appearance. Large mosquitoes with bright colours and notable distinguishing patterns, such as patches and stripes, and males with feathery antennae, tended to be over-represented, as observed for
Tx. speciosus,
Ae. alternans,
Ae. vittiger and
Cq. xanthogaster. This phenomenon has also been found for other taxa, such as observations of birds on iNaturalist [
39], with larger, more conspicuously coloured species observed more often. However, while
Tx. speciosus was the second most observed species recorded on the project, it was only observed in Queensland and New South Wales and not in the Northern Territory where it is also known to be present [
38].
The relative frequency of observations was not necessarily representative of the expected relative abundance of key mosquitoes of pest and public health concern in the environment. Key species known to be nuisance-biting pests or vectors of mosquito-borne pathogens were not necessarily the most commonly observed. The large number of observations of
Ae. notoscriptus,
Ae. vigilax, and
Ae. camptorhynchus was to be expected as these mosquitoes are either active within suburban areas or exceptionally abundant. However, other species such as
Cx. annulirostris,
Ae. procax, and
Ve. funerea, that have been associated with either nuisance-biting or arbovirus [
38,
40,
41], were not commonly reported or reported at disproportionately lower frequency than would otherwise be expected. The relatively low number of observations of
Cx. annulirostris was especially surprising given that this species can be highly abundant and is considered a nuisance-biting pest. As a consequence, it may not be possible to correlate mosquito observations submitted to iNaturalist with pest and public health risks but further research is required to elucidate this relationship and its application to assisting the assessment of mosquito-borne disease risk.
It is important to note some bias in the species richness of mosquito observations by individual iNaturalist users. There is a wide range of expertise among iNaturalist users, both with regard to experience in photographing insects as well as methods of observing mosquitoes. While some photos were taken with smartphones, others were taken with more advanced cameras and lens better suited to macrophotography. The differences in quality of photograph will play an important role in determining the ability to confidently identify the specimen. The entomological experience of the photographer, irrespective of quality of photographic equipment, will also determine the likelihood that the specimens will be identified given that key characteristics of mosquitoes may not necessarily be clearly evident in some photographs due to perspective. While many observations uploaded to iNaturalist were of serendipitous encounters with mosquitoes, others were the results of specific efforts (e.g., mosquito trapping) to document mosquitoes by professional entomologists (e.g., formal mosquito surveillance programs; citizen science projects). As a consequence, observations of some mosquito species (e.g., Ae. wattensis, Cx. edwardsi) through these efforts should not be expected to be replicated through casual observations by the general public.
Observations on the platform demonstrate a seasonal pattern, utility for identifying human-mosquito encounters (especially for vector species), and ecological associations with particular habitats and plants. The photos shared could also allow for studies of floral visitation, given that little is known about the role of mosquitoes as pollinators [
42]. Similarly, anthropophilic behaviour of lesser observed mosquitoes may provide useful insights into their potential role in pathogen transmission or pest impacts. The highest number of observations and species reported on the platform match with the known seasonal population dynamics of mosquitoes in Australia, also reported in previous studies [
43,
44,
45,
46]. The peak of observations in April could be related to the seasonality of rainfall on the eastern coast in Australia, increase of the platform popularity and the ‘Mozzie Month’ campaign that took place between late February and March in 2021 [
47], where users were invited to share their observations of mosquitoes (
https://www.inaturalist.org/projects/mozzie-monitors-australia/journal/46613-mozzie-month-challenge, accessed on 20 March 2022). While participants were not specifically asked about whether publicity surrounding ‘Mozzie Month’ influenced their likelihood of submitting observations of mosquitoes, it highlights that specific events of this nature may have potential to increase the number of individuals making observations of mosquitoes or their frequency of observations.
iNaturalist popularity has been increasing, as it is becoming more well known among people interested in biodiversity, researchers and citizen science facilitators. Bioblitz events (e.g.,
https://www.inaturalist.org/projects/great-southern-bioblitz-2021-umbrella, accessed on 20 March 2022) can also play a role in the increased participation on the platform, as could be observed in Australia in 2020 after the Great Southern Bioblitz when the observations on iNaturalist surpassed 100,000 in a single month for the first time [
13]. This raised participation on the platform is also observed in the increasing number of observers on the Mozzie Monitors project. Expanding participation could lead to sustainable engagement in the Mozzie Monitors projects, as was affirmed by 83% of the respondents being moderately to extremely likely to share observations of mosquitoes, and at least 65% of them had already shared observations of this taxon.
Caution is advised when extrapolating the data available on iNaturalist to public health risks. As has been the case with ecological and biodiversity assessments based on iNaturalist observations [
48], an acknowledgement of bias is required for a better understanding of the potential application to mosquitoes and mosquito-borne disease. To explore the likelihood of users sharing observations of species related to public health (e.g., mosquitoes), we were able to conduct an assessment of active users’ profile in the platform. Similarly to previous studies on citizen science [
49,
50,
51], users on the iNaturalist platform reported being predominantly male, middle-aged, highly educated and with a strong interest in science. However, they differed from other citizen science studies where female participation was higher [
24,
52,
53]. Research has shown that, depending on the nature of the citizen science activity (whether it is an environmental-oriented recreational hobby, or competitively-driven), it could attract different segments of the population, including males, females, youths, and older adults [
54]. As previously observed for the Mozzie Monitors program using the BG-GAT trap, the first demographics assessment showed that the majority of participants were female [
24]. Thus, campaigns and recruitment that focus on the public health importance of mosquito observations sharing on iNaturalist, as well as mosquito diversity monitoring, could increase the likelihood of attracting males and females from different ages to participate in the Mozzie Monitors program.
Most respondents showed higher interest in learning about biodiversity than being aware of disease risks or preventing mosquito nuisance. This shows that the profile of most active users in the platform is more biodiversity-oriented, and this interest could be linked to mosquito species monitoring. Indeed, learning about species that occur in their local areas could lead to long-term awareness about these species’ ecological and medical importance.
Although data collected by Mozzie Monitors on iNaturalist has potential for the study of mosquito community composition in Australia, some challenges were identified, such as engaging new users and expanding participation between the established network in order not to over-rely on a few observers, as shown in the user frequency histogram where only 4% of users contributed to 50% of all data shared. Increased participation and users’ retention could improve the identification tool accuracy for mosquitoes in the platform and upscale the geographic coverage of citizen science mosquito surveillance in Australia. Encouraging professional entomologists to join in the platform and assist with identifications could enhance the connectivity between mosquito observers on iNaturalist and mosquito experts, and improve the identification tools based on artificial intelligence and crowdsourcing.
Although many mosquito surveillance citizen science programs have their own customised apps, the use of iNaturalist to monitor mosquitoes could provide a scalable mechanism to engage people with different skills and interests. Designing, maintaining and updating a custom app also requires extra costs and resources. As the iNaturalist user perceptions questionnaire shows, most active users shared varied interests in wildlife, not limited to mosquitoes or public health. Thus, citizen science participation in mosquito monitoring through the iNaturalist platform could contribute to real-time species reporting and identification and increased connectivity between the general community and researchers.
This study also identified some biases and limitations. Results showed that participation was not evenly spread geographically, with a strong location bias. Observers uploaded substantially more photos around the Australian big cities and capitals. Responses to the electronic survey also showed that most active users were highly educated. Perhaps campaigns and educational workshops on using iNaturalist could encourage more people to contribute to citizen science data collection in areas with low participation. This would also be valid in settings with a higher abundance of mosquito-borne diseases; in that case, researchers should consider ethical concerns, such as the geoprivacy of photos showing human–mosquito interaction to track disease progression. Connectedness between users and researchers could increase public health awareness about species vector status and management.
In terms of privacy and security, users can obscure the GPS information of their photos. As an open platform, iNaturalist shows the geographic location of observations. If people or families feel uncomfortable sharing the species observed in their backyards, they have the option to obscure the location. Obscured observations display a 0.2 × 0.2 degrees rectangular area around the hidden coordinates (~500 km2 at the equator). When obscured, researchers can still see the specimen’s overall location (state and or suburb). Researchers can approach the observer via iNaturalist chat to ask for more details if a potential vector species is identified in an obscured area. Additionally, the electronic survey did not explore users’ training or background. Further studies could investigate whether there is a correlation between trained scientists participating in iNaturalist and observations and identification accuracy.
As previously demonstrated, online crowdsourcing information on vector arthropods collected data on the seasonality and distribution of vector species [
2,
26,
55]. The potential of iNaturalist for detecting vector species was also verified in Europe, North America, North Africa and the Middle East [
19], and the results indicated that the iNaturalist platform could complement existing vector surveillance data. Citizen science initiatives on mosquito surveillance are also emerging in tropical countries and areas where dengue and malaria are endemic [
22,
56,
57], and ethical issues were discussed in a community-based program in Nicaragua and Mexico [
58]. As ethical discussions underlining community-based research have shown, researchers should ensure that individuals and communities have the autonomy to monitor and control the vectors in their properties and have a safe space for a dialogue focusing on mutual respect and community health [
58]. Future studies regarding the utility of iNaturalist for public health should consider the ethical issues of vector surveillance in impoverished locations where diseases such as dengue and malaria are a major problem.
Research has shown that iNaturalist users were able to register critically endangered species, as well as extremely rare species and behaviours [
13]. Similarly, the Mozzie Monitors project on iNaturalist could provide an early-warning system for detection of invasive species, as in detecting the first record of
Aedes (Downsiomyia) shehzadae in Queensland [
59]. It could also facilitate early detection of growing vector mosquito populations through citizen science.
Finally, it is notable that participation in iNaturalist has been increasing over 2021/2022. In only seven months, from July 2021 to February 2022, 877 new observations were added to the Mozzie Monitors project (an increase of 41% compared to data extracted before July 2021). In addition, 111 new observers contributed to the project, and six new species were added to the list, including Ae. biocellatus, Ae. gahnicola, Ae. rupestris, Ae. subbasalis, Cq. variegata and Cx. postspiraculosus. It is important to note that these species were added by an expert entomologist using surveillance traps. This growing engagement in the platform should be investigated in future studies to explore connectedness between overall users and identifiers and to assess an updated inventory of species list collected through citizen science.