The Seed Bank Is a Disturbance Clock

01The evidence

• Seed bank longevity varies by orders of magnitude among co-occurring species and is correlated with habitat disturbance, which is the pattern a timing account would produce.Seed traits and seed bank longevity
• Life-cycle transitions through the seed bank have been quantified in demographic models, making the persistence term measurable rather than inferred.Oikos, seed bank transitions

02The tension

Bet-hedging explains why dormancy exists but not how long it should last. The observed variation spans years to centuries, and the usual explanation is phylogeny or seed size. Neither predicts the values, and neither explains why closely related species in different disturbance regimes differ so sharply.

03The theory

A seed bank is a timer set to the disturbance interval. If seeds persist much longer than the interval between gaps, most of them die in the soil without ever seeing light. If they persist much less, the population misses the gap and loses the site.

The optimum is therefore near the return time, with a spread that reflects the unpredictability of the interval. Species in frequently disturbed habitats should have short-lived seeds; species in rare-gap habitats should have long-lived ones, and the relation should be quantitative.

The theory makes a strong claim about timescales. Dormancy should respond to changed disturbance within decades, because the trait is under strong selection every generation. That is fast enough to test in sites with documented regime changes, such as fire suppression or altered grazing.

It also resolves an applied puzzle. Weed seed banks persist far longer than the crop rotation they evolved against, which is a mismatch created by agriculture imposing a new and much more regular disturbance cycle.

04What this is not

05Predictions

• P1Seed longevity should correlate with disturbance return time across habitats, with a common slope across taxa.
• P2Sites with documented suppression of disturbance should show shortened dormancy within decades.
• P3Weed species under regular tillage should have shorter dormancy than congeners under irregular disturbance.

06What would refute it

No relation between longevity and return time, after controlling for seed mass, would leave the timing account without support.

07What it would change

It would make seed bank management predictive, and give restoration a rule for how long a site must be managed before the bank is exhausted.