A razorbill, it is reported today, has lived for 41 years, in spite of being buffeted by the winter waves in the Atlantic for a good half of every one of those years. Is that normal? In fact how long do birds generally live?

We know that there have been many other long-lifers. The oldest bird known is a sulphur-crested cockatoo, which was over 80 when it died in London Zoo in 1982. Fulmars, another seabird, have been found alive after 40 years. A golden eagle that spent its life winging over the Highlands reached 32.

Until recently we hardly knew anything about birds' ages. It is ringing them, usually when they are still in the nest, that has taught us that some of them live for so long. Their date of ringing is recorded, and when they are found again with the little numbered ring still on their leg - washed up dead on the shore, or re-ringed when they return to their colony to nest - we know their age.

But really these razorbills and cockatoos are exceptions. All they tell us is how long birds have the capacity to live; the truth is, very few of them get there.

We can see this if we stop and realise that in most summers, under stable conditions, there are just the same number of birds around as there were the previous year. That means that a number of birds equivalent to the whole baby boom of last year must die. And they do. Most of the new birds born in any year die. So each year the population is made up of many old birds, and a small number of new ones.

Birds have a hard life of it, especially young birds, which are less skilled at dodging cats, sparrowhawks, peregrine falcons, great black-backed gulls and buzzards all of which kill them at every possible opportunity. The young birds are also less skilled at finding food, and vast numbers of them starve in the winter.

For birds that manage to survive till they are a year old, it is a quite different story. What ornithologists calculate for them is what insurance companies calculate for humans - the expectation of life.

A little bird like a robin can live up to at least eight years, as we know because one has been found as old as that. But in any summer, its expectation of life is one and a half years. It is very likely to be eaten, to fall ill, or to starve. But if it does survive for those one and a half years its expectation of life is once again one and a half years again. It may live for a day, it might even live for a decade. But its expectation of life remains constant. That is all that we can really say about how long birds live. Of course, if you have a recognisable bird in your garden, such as a blackbird with a white patch, you can see if it comes back for two, three or more years. That tells you about that bird. But for the rest, considered as individuals, we cannot tell.

Seabirds, like that aged razorbill, do seem to have longer lives in general. The colourful little puffins that nest in clifftop rabbit holes have a life expectation of almost 20 years. So do herring gulls. Gannets, however, have a life expectancy at any time of only ten years. One might have thought that seabirds have these longer lives because they have fewer predators. Yet puffins live long although they are chased by great skuas; while one cannot imagine anything attacking the gigantic white gannets, their lives appear to be shorter.

And where do birds go when they die? We do not often find a dead one. The answer is that they are nearly all eaten. The predators eat them alive - and burying beetles eat them dead. A sick or wounded bird will go to die under a bush. Then the beetles will work together to make a hollow - a kind of grave - under it, cover the corpse up again when it falls into the hollow and share it out between them. A macabre end, but a clean one.

Derwent May writes the Feather Report column in The Times

Times Recommends

• In the real world, the public sector must pay
• Playing politics in a land of double Dutch
• Be like the plinth people: not set in stone