Figures

Fig. 1. Mobile phone users can collect acoustic data from mosquitoes characterized by the base frequency and harmonics.

A, Illustration showing the collection of mosquito acoustic data by mobile phone users in different locations. B, Methods to acquire wingbeat sounds from mosquitoes using mobile phones include lab methods like (i) collecting them in cages, and field methods like (ii) followingvmosquitoes in free-flight, or (iii) capturing them in inflated bags. C, Spectrogram for a flight trace acquired from an individual female Anopheles gambiae mosquito using a 2006 model Samsung SGH T-209 flip phone. The wingbeat base frequency at every instant is computationally identified by a simple automated algorithm and marked with a black line. (Top) The time-averaged spectrum of this flight trace shows the distribution of acoustic power among the base frequency and multiple harmonics. D, The variations in wingbeat base frequency of the mosquito during this flight trace are represented by a probability distribution of the frequency identified in each window of the spectrogram. (Top) Raw base frequency data is represented as a violin plot with an overlaid box plot marking the inter-quartile range, black circle representing mean frequency, gray vertical bar for median frequency, and whiskers indicating 5th and 95th quantiles.

Fig. 2. Mobile phones sensitively acquire high fidelity acoustic data from mosquitoes with comparable performance across models

A, Schematic of experimental setup for recording a tethered mosquito using mobile phones, with synchronized high speed cameras or high performance microphones as visual and auditory reference standards. Synchronization on the order of microseconds is achieved using a piezoelectric buzzer and LED controlled by a microprocessor. B, Overlaid spectrograms for female Culex tarsalis mosquitoes obtained independently using high speed video (magenta) and mobile phone audio (cyan), aligned to within 2 ms and showing a spectral overlap (blue) within 2 Hz across all time instances. The mobile phone data is noisy but faithfully reproduces the base frequency peak of 264 Hz and the first two overtones. C, Base frequency distributions from video and audio are indistinguishable by the 2-sample T-test (n = 165, = 1%). D, Signal-tonoise ratio (SNR) estimates over distance from a standardized sound source show that mobile phone microphone performance within a 100mm radius is superior or comparable to high performance studio microphones. E, SNR over distance for the wingbeat sound produced by a tethered female Cx. tarsalis mosquito (normalized for a source amplitude of 45 dB), provide working limits where phones can detect the audio signal – 50 mm for the low end T-209 feature phone and 100 mm for the iPhone 4S and Xperia Z3 Compact smartphones. F, Variation of the base frequency distribution sampled by 8 different phones is low compared to the natural variation within a population of about 200 lab-reared Anopheles stephensi females. Raw data  are shown with overlaid box plots marking the inter-quartile range, black circles for mean frequency, gray vertical bars for median frequency, and whiskers indicating 5th and 95th quantiles. G,H, The Jensen-Shannon divergence metric for base frequency distributions (G, lower left triangle) shows low disparity, ranging between 0:144 and 0:3, against a minimum of 0 for identical distributions. Likewise, the Bhattacharya distance (H, upper right triangle) shows high overlap, with values between 0:935 to 0:986, against a maximum of 1 for identical distributions.

Fig. 3. Mosquitoes of different species are distinguishable based on base frequency distributions and spatio-temporal metadata

A, Distribution of base frequencies for lab-reared female mosquitoes of 19 medically relevant species, for recordings obtained with the 2006 model T-209 low-end feature phone (except Cu. incidens, Cx. pipiens and Cx. quinquefasciatus, recorded using iPhone models). B(lower left triangle), Jensen-Shannon divergence metric for base frequency distributions. Distributions are spaced apart with high J-S divergence in most cases, with only four pairwise combinations having J-S divergence around 0:3 – the maximum divergence for the same species across different phones. C (upper right triangle), Classifiation of species pairs according to the possibility of distinguishing them using mobile phones —(i) no frequency overlaps, hence distinguishable by acoustics alone, (ii) overlapping frequency distributions, but not geographically co-occurring hence distinguishable using location, (iii)overlapping frequency distributions but distinguishable using time stamps, (iv) partially overlapping frequency distributions but no location-time distinctions, hence distinguishable but not in all cases, (v) indistingishable due to highly overlapping frequency distributions with co-occurrence in space and time. D,E, Variations in base frequency distribution (D) for field-recorded sounds corresponding to wild mosquitoes having a wide (about two-fold) variation in body size and wing area (E), showing small differences between individuals compared to the variation within each flight trace.

 

Fig. 4. Spatio-temporal activity of mosquitoes in the field can be mapped using crowdsourced acoustic data from mobile phone users.

A, Sample spectrograms from female Culex spp. (top) and Anopheles spp. (bottom) mosquitoes captured in the field at Ranomafana village in Madagascar. B, Frequency distributions for fieldcaught Culex spp. and Anopheles spp. mosquitoes in Ranomafana, forming a reference for identification of recordings from either species at this field site. C, Map of Ranomafana village showing distribution of female Culex spp., Anopheles spp., and Mansonia spp. mosquitoes, from mobile phone data recorded by 10 volunteers over the approximately 1 km X 2 km area. Each square represents one recording, and black circles indicate locations where volunteers reported encountering no mosquitoes. The map shows a complementary spatial gradient from riverbank to hillside in the relative proportion of Anopheles spp. and Culex spp. mosquitoes. Further, mosquito hotspots are interspersed with points having a reported lack of mosquitoes, highlighting the potential importance of micro-factors such as the distribution of water and livestock. D, Spatio-temporal activity map for female Oc. sierrensis mosquitoes in the Big Basin Park field site, using data collected by 13 hikers recording mosquitoes with their personal mobile phones, over a 3-hour period in an approximately 4:5 km X 5:5 km area. Each brown square represents one Oc. sierrensis female recording, and black dots represent sites where hikers reported encountering no mosquitoes at all. (Inset top left) Temporal distribution of the overall mosquito activity data depicted in (D) based on recording timestamps, showing the rise and fall of activity in each hour of the field study.

Figure S1: Schematic of proposed surveillance system using crowdsourced acoustic data from mobile phones

A-F. Schematic of proposed surveillance system using crowdsourced acoustic data from mobile phones System architecture showing the collection of data by individual mobile phone users, processing to identify species of interest, and compilation into a map of mosquito activity. The diagram is depicted centering around data collection at a field site designated Location X. A-D occur prior to mobile phone based data collection, and represent steps required to enable crowdsourced acoustic surveillance at the field location. A, The mosquito population in the field at Location X is sampled, either by users in Ziploc bags or by using methods such as trapping, and live specimens characteristic to the location are collected. B, Wingbeat sounds of these field collected mosquitoes are recorded, with an acoustic dataset associated with each individual specimen. C, Specimens are identified to the genus (and preferably species) level by a method such as morphological ID through optical microscopy, or molecular ID through PCR. D, Acoustic data is processed and associated with specimen IDs to yield frequency distributions characteristic of the prevalent species in that field location, forming a reference database of mosquito sounds specific to Location X. E-H represent the proposed method for mobile phone based acoustic surveillance at the field location, assuming that the reference database of mosquito sound is already in place. E, Mosquitoes are recorded in the field by a user with a mobile phone, and the audio file together with metadata is compiled into a database for processing. F, The acoustic signals are processed to extract the frequencies present in the recorded mosquito sound. G, The computed acoustic spectrum and metadata obtained from the mobile phone are compared to the reference database for that location, and the most likely species corresponding to the computed frequency is identified. H, The identified species from this observation, together with the time and location metadata, are mapped back to the field Location X. This closes the loop for mobile phone based acoustic surveillance, from crowdsourced recorded data to information on spatio-temporal mosquito activity.

Fig. S2. Synchronized recordings of tethered mosquitoes using studio and mobile phone microphones shows exact correspondence at near-field distances below 50 mm

A-E, Comparison of power spectral density for synchronized simultaneous recordings of individual Culex tarsalis female mosquitoes using the MXL 991 studio microphone, the Apex 220 reference microphone and a mobile phone, taken at varying distances. The left column corresponds to the SGH T-209 feature phone, the middle column to the iPhone 4S iOS smartphone, and the right column to the Xperia Z3 Compact Android smartphone. A, Superimposed averaged spectra show that all phones acquire wingbeat sound at a high signal-to-noise ratio at 10 mm away from the mosquito. B,C,D, Overlaid spectrograms synchronized to within 20 ms in time show a near-perfect spectral match of within 5 Hz at each time interval, for the mobile phone microphone (red channel), MXL 991 (green channel) and the Apex 200 (blue channel), shown together as RGB images with intensity of colour corresponding to variations in power spectral density. Mobile phones strongly acquire mosquito sounds at 10 mm or even 50 mm, but their sensitivity drops sharply at distances of 100 mm. E, Superimposed averaged spectra show that only the Xperia Z3 continues to acquire wingbeat sound at 100 mm away from the mosquito, albeit at low signal-to-noise ratio. The T209 feature phone picks up low frequency noise between 300 to 600 Hz that overwhelms the mosquito frequencies, the iPhone 4S has low noise acquisition throughout, and the Xperia Z3 picks up high frequency noise above 1 kHz that leaves the mosquito frequency band relatively unaffected.

Fig. S3. Mosquito species can be distinguished with mobile phone acoustics and metadata

A-F, Illustrative examples for distinguishing between medically relevant mosquito species using acoustics and metadata. All inset images of mosquito specimens are taken fromWalter Reed Biosystematics Unit mosquito ID databases. A, Distinction by acoustic data alone – Cx. pipiens and Anopheles gambiae, which co-occur in many regions, can easily be distinguished by sound alone. B, Distinction by location metadata – An. atroparvus and An. dirus have overlapping acoustic spectra, but recordings are easily distinguished from each other by metadata pertaining to their distinct spatial distributions in Europe and South-East Asia respectively. C, Distinction by time metadata – Aedes aegypti and An. gambiae can occur together in many locations and have overlapping wingbeat frequency distributions, but can be easily distinguished by time of recording based on their diurnal and crepuscular biting habits respectively. D, Partial distinction by acoustic data – Ae. aegypti and Ae. albopictus have similar appearances, geographical distributions and biting habits in many areas. Although the wingbeat frequency distributions are not completely distinct, interquartile ranges do not overlap and a significant fraction of recordings can still be classified correctly as one or the other, making acoustic identification faster and easier than microscopy. Similarly in the case of Cx. pipiens and Cx. quinquefasciatus, which have partially distinguishable frequency spectra despite being otherwise indistinguishable except using PCR. E, Partial distinction by acoustic data – An. arabiensis, An. gambiae, and An. quadriannulatus, which are members of a species complex that are identical in appearance and often overlapping in habitat, have non-overlapping interquartile ranges for wingbeat frequency distributions implying that the majority of acoustic samples can be classified correctly as one among the three. F, An. arabiensis is indistinguishable based on mobile phone acoustic data from An. merus, another members of the An. gambiae s.l. species complex, exposing a relatively rare limitation of species identification using mobile phone acoustic surveillance.

Fig. S4. Mobile phones are capable of acquiring mosquito sounds in a variety of field environments

A-F, Raw spectrograms of acoustic data acquired by various mobile phone users in different field conditions, with base frequencies of mosquito sounds highlighted by a box. The signals include sources of noise such as human speech, fire truck sirens, and birdsong, and were acquired in both urban (A-D) and forested (E,F) environments, including indoor (A,B) and outdoor (C-F) settings. Mosquitoes recorded were wither followed in free-flight (A,C,E or captured in a plastic ziploc bag prior to recording (B,D,F). All spectrograms show raw spectra without background correction or noise removal, and show the spectra from extraneous acoustic sources (speech, sirens) to distinguish the characteristics of mosquito spectra from other sounds. Spectrograms A-F correspond to sounds in Supplementary Audio SA2-7.