Seed Bank Longevity

2D companion of the biodiversity cryovault: dry the seeds, cool them and see how long a seed bank keeps them alive.

Switch to 3D

Presets

Playback

Apparatus

Model checks

The science

What is modelled

Seeds in a store die at random times that follow a normal distribution. The viability equation of Ellis and Roberts gives the spread of those times from the seed moisture and the temperature, so the whole survival curve follows from two numbers you can set in a freezer.

Equations

  • Viability in probit units v = Ki − t/σ, and the fraction alive is Φ(v).
  • Time scale log10 σ = KE − CW log10 m − CH T − CQ T2 with the barley-type constants KE = 9.983, CW = 5.896, CH = 0.040, CQ = 0.000428 (σ in days, shifted by the species offset).
  • Median life Ki σ; time to a given viability σ (Ki − Φ−1(q)).

Checks run in the page

  • 20000 simulated seeds against the normal-distribution prediction.
  • Simpson integration of the bell curve against the error-function value.
  • The drying rule (2 to the power CW) and the cooling rule.
  • The time to 85 % viability from step-by-step integration against the closed form.

Honest simplifications

The constants are the published barley-type values used as an illustration. Real species differ a lot, the equation is only reliable for moisture between about 5 and 14 percent and for temperatures from about −20 to 90 degrees, and values far outside that range (very cold, very dry) are extrapolations.

What to try
  • Press Open shelf and then Seed vault and compare the median life.
  • Halve the moisture and see the life grow by a factor of about sixty.
  • Lower the species offset to see a short-lived seed type.
Seed tray survival against the probit theory
Median life against storage temperature
Median life against seed moisture
Viability probit against time
Made with 🤖 ChatGPT