Mast seeding

Why do some trees produce an enormous seed crop in one year and almost none for several after?

An oak wood produces so many acorns one autumn that they crunch underfoot, and then almost none for two or three years. A steady annual crop looks obviously better: seeds every year, chances every year, and the empty years look like a defect. But the pattern is old and widespread — oaks, beeches, spruces, the great dipterocarps of Southeast Asia — so before calling it a defect, ask what a lean year might buy that a moderate one would not.

Who else has an interest in an acorn crop? Weevils that develop inside acorns, mice, jays, squirrels — animals living substantially off this one food. How do their numbers behave under a steady annual supply? They rise until it is fully used, so a dependable crop supports a population sized to take most of it every year.

What does a lean year do to them? It starves them, and their numbers fall. And what does an enormous year do to a population just cut back? It hands them more than they can eat or store, and a share of the seeds survives simply because nobody was left to take it. Neither famine nor flood works alone — the pair does. That is predator satiation, and it makes the variability the mechanism rather than the waste.

But notice a second requirement. Suppose one tree tried this alone, in a wood where every other tree fruited steadily. Its lean years would starve nobody, and its bumper year would arrive among consumers kept plentiful by its neighbours. It would simply lose. The strategy pays only if trees do it together — which turns the question from why a tree varies into how a whole region varies in step.

There is a second reward for acting together, and it hints at a mechanism. Oaks and beeches are wind-pollinated, and wind pollination is desperately inefficient for a lone flowering tree; one that flowers when its neighbours flower gets a far better return on the same investment. So whatever nudges a few to flower heavily together improves everyone's success — a self-reinforcing coupling through the air. A simpler bookkeeping account exists too: seed is expensive, so a tree may build stores over several years and spend them in one burst, which gives variability directly with no strategy required.

So which is it? Here honesty matters more than tidiness. Predator satiation is well supported as a benefit — lean years demonstrably suppress seed predators, and big years lose a smaller fraction of the crop. Resource depletion constrains what a tree can do. Pollen coupling plausibly synchronises neighbours. And shared weather — a hot summer, a drought, in dipterocarps often an El Nino event — can put whole regions on one schedule with no interaction at all. None is sufficient alone, and the leading picture combines them: weather as cue, resources as constraint, pollen coupling as the lock between neighbours, satiation as the reason it was favoured. How synchrony holds across hundreds of kilometres, and whether weather is a true cue rather than merely a limit, remain genuinely unsettled.

Three clarifications. A mast year is not a fixed cycle — intervals vary and cannot be reliably forecast, which is itself part of the defence, since a predictable rhythm could be tracked by the animals that eat seeds.

It is worth resisting the language of intent, too. No tree decides to starve a weevil; trees whose seed years happened to fall together left more descendants than trees that fruited alone, and the pattern is what that difference accumulated into.

And the consequences run beyond the wood: rodent numbers rise the year after a heavy acorn crop and their predators follow a year later still, and in eastern North America acorn masts have been linked through mouse populations to later shifts in tick numbers and Lyme-disease risk.

• Why some species mast strongly and their close relatives do not.
• How jays and squirrels, which cache far more than they recover, end up dispersing the very seeds they were meant to consume.

Every term the collection defines is gathered in the glossary.