The Divorce Rate of a Faithful Bird

Black-browed albatrosses pair for very long periods, in some cases for decades, and separation while both birds are alive is genuinely rare.

A study of around 15,500 breeding pairs in the Falkland Islands, followed for fifteen years, put the average annual rate of what the researchers call divorce at about 3.7 per cent. That is low for any species that has been counted, and it is the reason albatrosses are the standard example whenever anybody wants to say that lifelong partnership exists in nature.

Then came the finding that made the study interesting.

In the year with the warmest sea surface temperatures, the rate roughly doubled, to about 7.7 per cent. And the increase was not concentrated where you would expect it. Females that had bred successfully, whose partnership had just worked, were among those most likely to leave.

That detail removes the obvious explanation. If separation were simply a response to failed breeding, only the failures would separate, and the environmental effect must therefore be acting through some other route.

The likely mechanism is not mysterious and it is not about the birds' feelings at all. Warmer water means poorer foraging, which means longer trips, which means a bird arriving late at the colony. Albatross pairing depends on both partners being in the right place at roughly the right time. A partner who does not appear on schedule is, from the other bird's position, indistinguishable from a partner who is not coming.

So a bond that is among the most durable in the animal world can be broken by conditions that never touch the bond itself. Nothing about the relationship has to have weakened for the relationship to end.

Now the word, because it is the methodological lesson of this whole book.

Divorce here is a technical term. It means both birds were alive and one of them took a new partner. Researchers use it because there is no better short word available, and every alternative is either clumsy or wrong. The moment it appears in a headline, readers supply intention, disappointment and blame, none of which is in the study and none of which the finding requires.

This is the permanent difficulty of the subject. Science needs a vocabulary, the only vocabulary available for relationships is a human one, and borrowing it imports a mental life the data cannot support. The alternative, inventing neutral jargon, makes the findings unreadable and is usually abandoned within a decade.

There is no solution, only a discipline, and it is the same one the next two chapters need. Read what was measured. Notice which word is doing work the measurement cannot do. Hold the question of what the animal experiences open rather than answering it in either direction because one answer is more comfortable.