December 9 2012, Samburu
A rainstorm fills the Ewaso Nyiro river. True to its name, the water is a rich orange-brown and as thick as Thai iced tea. It gurgles alongside the camp, snaking past wide sandy banks with steep sides like miniature beaches bordered by miniature cliffs.
To see the elephants, the river and its shaded woodland border is the place to be in the dry season. But these days, with water puddling up everywhere elephants are ranging further from it. The other day we came upon a vast herd on the bank. Gathered in its ranks were multiple families, but on the following days hardly anyone came down again except to cross to the other side.

A vast herd marches in unison.
A few days ago we encountered two different phenomena in elephant social structure within the span of a few hours. The first was a family unit that may be in the process of fissioning. The Royals, originally a single family several years ago, were now often found in four smaller units. Shifra calls these ‘core groups’. We came upon each little core group (consisting of less than 10 individuals in each case) by itself, although the entire ensemble was within driving range so we had the good luck to come across all of them within the space of a few hours. In the absence of any genetic data, if a researcher were to happen upon them today with no knowledge of the ten-year history, I can well imagine that she might classify them as separate families. At least one of the core groups seldom interacted with the others. I wondered if their movements remained coordinated nonetheless, as it could hardly be coincidence that we would run into all four core groups on one day, and none on any other occasion. This was just like my experience of some Asian elephant social units – with an important difference. The members of each core group appear to be stable according to Shifra. In the Asian case, some families were very cohesive whereas in others the females shuffled around almost daily even if some pairs of adult females were constant companions.
In the second half of the same day, we encountered the opposite phenomenon – two large family groups moving together. There were the First Ladies and The Winds, together with a few young bulls. Altogether we counted 58 individuals moving leisurely over a wide swathe of landscape. Apparently these two units often associate. This too we might sometimes see in Uda Walawe, but rarely – and in contrast to here, they occur in dry seasons.
Over the subsequent days we encountered a different large aggregation – this time consisting of at least five families, though some families came and went. They included the American Indians, Clouds I and II, Mountains, Artists, Spices, a few of the Virtues and Storms. A few of these consisted of oddities, where adults or young had switched families following poaching incidents that claimed the older females.

I can just imagine the days when thousands of elephants did this all across Africa.
The neatest thing about this large aggregation was that at any given time it in fact moved as a single herd – so much so that they were prone to stampeding in unison, triggered by some invisible, possibly innocent stimulus, such as a running calf. They seemed nervous and excited, changing direction spontaneously, not dallying much to graze or rest. It seemed so energetically costly – why bother? There is food, water, and mud everywhere! The little family groups off on their own, by contrast, had the leisure of sitting around in mud wallows or resting under trees as long as they liked. What were they gaining by this frantic dashing this way and that?
Elephants work in mysterious ways.