How do fireflies sync with no leader?

Start with one firefly. Between flashes, imagine an inner clock filling up. When it reaches the top: blink — then the clock starts again.

The same child leans over a marsh railing and calmly studies one clearly winged firefly glowing on a reed.

The clock is our model — not a tiny dial inside the insect.

In our rule, a seen flash adds a small pulse of charge to another clock. Tap once and watch the receiver jump ahead.

How much does one seen flash change a clock? The receiver always starts in the same place, so every comparison is fair.

pulse strength0.040 charge

nonelarger jump

Same seed. Same 18 starting phases. Same rule, pulse strength, all-to-all visibility, and 12-beat horizon. Only the natural beats change.

same startall see allsame pulse

Commit your prediction before either field runs.

By 12 solo beats, which field will hit one all-18 flash twice in a row?

Nothing up our sleeve: the whole model recipe

18 per field
seed 0x5f3759df

all-to-all
delay 0 · noise 0

equal: T = 1.00
mixed: T = 0.75…1.25

ε = 0.005
dt = 0.001 · stop = 12

Between events: φ += dt/T. A received flash adds ε to u(φ) = ln(1 + (e³−1)φ)/3. Reaching u = 1 joins that avalanche, flashes, resets, and sends another pulse. Fired nodes stay reset during that same avalanche.

Equal beats: target reached at t = 8.817; five full avalanches by 12.

Mixed beats: target not reached; largest avalanche = 5.

But real fireflies are more complicated.

What did our ideal model leave out?

Real fireflies usually see nearby individuals, not every firefly at once. Signals have travel and response delays. Natural periods vary; wind, motion, blocked sight, and biological noise matter. Flash timing is also courtship behavior, and the patterns differ by species and by sex.

Does any tiny nudge always sync any network?

No. The classic Mirollo–Strogatz result concerns identical, globally pulse-coupled oscillators with a suitable response rule and synchronization from almost all starts. Other connection patterns, delays, variation, noise, or coupling rules can split into groups or never reach global synchrony.

The same child quietly watches many winged fireflies share one warm glowing moment over the marsh, with no leader visible.

No leader is required when the right interaction lets repeating clocks reshape one another. Always ask: which rule, which connections, how strong, and how much variation?

psst, grown-ups

Model note. This is a deterministic Mirollo–Strogatz-style integrate-and-fire demonstration, not a claim that all firefly species follow one exact algorithm. The 1990 theorem establishes synchronization from almost all initial states for its ideal identical globally pulse-coupled system under suitable assumptions. Biological firefly research instead measures species-specific courtship signals, local visibility, delays, and variability.