1. Periodical cicadas exhibit an extraordinary capacity for self-organising spatially synchronous breeding behavior. The regular emergence of periodical cicada broods across the US is a phenomenon of longstanding public and scientific interest, as the cicadas of each brood emerge in huge numbers and briefly dominate their ecosystem. During the emergence, the 17-year periodical cicada species Magicicada cassini is found to form synchronised choruses, and we investigated their chorusing behavior from the standpoint of spatial synchrony.

2. Cicada choruses were observed to form in trees, calling regularly every five seconds. In order to determine the limits of this self-organising behaviour, we set out to quantify the spatial synchronisation between cicada call choruses in different trees, and how and why this varies in space and time.

3. We performed 20 simultaneous recordings in Clinton State Park, Kansas, in June 2015 (Brood IV) with a team of citizen-science volunteers using consumer equipment (smartphones). We use a wavelet approach to show in detail how spatially synchronous, self-organised chorusing varies across the forest.

4. We show how conditions that increase the strength of audio interactions between cicadas also increase the spatial synchrony of their chorusing. Higher forest canopy light levels increase cicada activity, corresponding to faster and higher-amplitude chorus cycling and to greater synchrony of cycles across space. We implemented a relaxation-oscillator-ensemble model of interacting cicadas, finding that a tendency to call more often, driven by light levels, results in all these effects.

5. Results demonstrate how the capacity to self-organise in ecology depends sensitively on environmental conditions. Spatially correlated modulation of cycling rate by an external driver can also promote self-organisation of phase synchrony.

Data was collected and processed as described in Materials and Methods of Self-organising cicada choruses respond to the local sound and light environment (2020) Ecology and Evolution.

Further details are provided in the Supplementary Information S2 and S3, and in the supplementary figure S26.

The 20 cicada sound volume time-series represent 0.1 second segments of data, aligned and filtered as described (the final stage in figure S26, prior to wavelet transformation; step X).

The normalised (smoothed) forest canopy brightness profile is as shown in figure 3c.

Latitude and longitude for the 20 recording locations (derived from multiple GPS devices) are as depicted in supplementary figure S22 and elsewhere.

Data is presented in csv format, output from MATLAB in Windows 10.