WHEN I did A-level biology in the late 1950s, many textbooks still classified
living things into two kingdoms, Animalia and Plantae, as Carl Linnaeus
suggested 200 years earlier. Protozoans were rammed in with the animals, fungi
and even bacteria thrust in among the plants.

In 1959 R. H. Whittaker of the University of Cornell tidied things up by
proposing five kingdoms: animals, plants, fungi, protoctists (including
protozoans and seaweeds) and monerans meaning bacteria. Plantae now included
various green algae, “bryophytes” (mosses and liverworts), “pteridophytes”
(ferns and horsetails), “gymnospers” (conifers, ginkgos, and cycads) and
angiosperms (flowering plants) with oddballs such as club mosses lurking
vaguely.

Many biologists still cling to this workmanlike classification, but three
revolutions have changed the picture entirely since the 1950s. The first was the
technique of cladistics, devised by the German entomologist Willi Hennig in the
1960s. This set the standard by which we classify creatures using the special
features they hold in common, and not just according to general impression or
presumed history. The second was molecular biology, revealing underlying DNA
and RNA. The third is the huge advance in our knowledge of what is out there:
great leaps in palaeontology have brought many more creatures to our attention,
while forays into new environments, including hot springs and deep rocks, show
that life is far more varied and extraordinary than anyone had guessed.

So now there are no longer five kingdoms but three domains: Bacteria, Archaea
and Eucaryota (the creatures like us, with proper cells with nuclei). There are
at least a dozen eukaryote kingdoms, of which animals, plants and fungi are just
a sampling.

Peter Crane of the Field Museum in Chicago and Paul Kenrick of the Swedish
Museum of Natural History have been among the first to apply these new
techniques to plants. Their conclusions to date presented in The Origin and
Early Diversification of Land Plants are a revelation. This book is a must
for professional botanists, but weekend naturalists should read it as well, just
to get a feel for the true wonders that lie out there, behind what is
immediately visible. Seek it out and buy it—it will set you onto a new
plane of inquiry.

THE first North Americans—or at least, the first who were any good at
big game hunting—arrived from Asia around 12 000 years ago. They came via
Beringia, the land bridge that sprang up between Siberia and Alaska as the sea
level fell during the last ice age. Archaeologists call them the Clovis people
after the New Mexican town where their remains first turned up.

It is now clear that they spread through both the Americas, reaching the
extreme south within 1000 years. As they went they killed virtually all the big
mammals. They hunted native herbivores to extinction. And, along with the
mammoths, mastodons and camels went the carnivores that preyed on them—the
giant running bears, a North American lion, the fearsome dire wolf and the
sabre-tooth tiger.

This is the “overkill” hypothesis, promulgated since the 1960s by Paul Martin
of the University of Arizona. But did the overkill really happen? In The
Call of Distant Mammoths, Peter Ward admirably summarises the
arguments.

The raw statistics certainly show that something violent happened. In North
America, no fewer than 33 genera of large mammals disappeared between 12 000 and
10 000 years ago, as opposed to only 20 in the preceding 3 million years.

Some studies suggest that the demise of the large mammals was not as sudden
as Martin has suggested. Others argue that there is little direct evidence for
overkill, adding that so much killing is simply implausible because there were
so few people around. Still others argue that climate change was responsible.
Yet, Ward relates, most blame the Clovis people.

Don Grayson of Washington University is among those who argue that modern
carbon-14 dating techniques do not show a sudden die-off. True, says Ward,
but—and this is his personal contribution to the debate—modern
statistical analysis shows that mass extinctions always seem to be drawn out
even when they are instantaneous, simply because the palaeontological and
archaeological records are so patchy.

No conceivable data could show a more sudden die-off, because the extinctions
were too rapid to be captured in the physical records. Nor would we expect to
find many of the characteristic Clovis spearheads among the mammoth bones, as
some critics demand. They were much too precious for hunters to leave
behind.

Of course, the climate did change rapidly as the last ice age ended and so
did the vegetation. Against this, we have to consider that the climate has
changed rapidly at least twenty times in the past million years without any
extinctions. The animals simply migrate.

So the latest theories firmly support overkill, with a little help from
climate. Traditionally, ecologists supposed that population is determined simply
by the balance of births and deaths, and that these are constant for a given
species. Now they deal in stochastic models, adding the effects of
disasters—harsh winters, epidemics and fragmentation of habitat.

Change of climate would have broken up the mammoth feeding grounds at the end
of the ice age both in North America and Eurasia, leaving small populations that
were individually vulnerable. Models developed by Steve Mithen of Reading
University also show that hunting pressure of a mere two per cent per year would
finish the mammoths off in no time.

The full story is far more intricate and difficult to summarise. Ward tells
it well. It is also highly pertinent to our time, because present-day large
mammals are at least as precarious as the late Pleistocene mammoths. And we are
many times more destructive than the Clovis people. I have only one beef with
this book: Ward has obviously been told that popular science should be presented
as combative and diluted with travelogue. It shouldn’t. The science itself, well
told, will carry the day.

The chemical substance is loosely classed as an alarm signal. However, other such signals, such as the white tails of fleeing rabbits, tend to be emitted before the animal has been injured, rather than after. The signal warns other potential prey and shows the predator that the signaller is too fit and alert to be worth chasing.

But the reason why many animals, including snails, frogs and many fish, give off signals after they are injured is less obvious. Alicia Mathis and her colleagues tested the effects of such signals on predators and prey in the laboratory. They extracted the alarm substance from the skins of non-breeding fathead minnows (Pimephales promelas), which are common in North America. An extract from the skins of breeding male minnows served as a control, as they lack the cells that produce the substance.

They released the two extracts at either side of a tank containing 11 northern pike (Esox lucius). The results were conclusive: the pike spent significantly more time on the side of the tank containing the alarm substance (American Naturalist, vol 145, p 994).

But why should a minnow that is being attacked want to attract more predators? In practice, as the Canadian biologists point out, the worst enemy of a small pike is a larger pike. Small pike are more likely to injure the minnow’s skin during capture, and thus allow the release of the alarm substance, whereas big pike simply swallow minnows whole. So the tactic is, when caught by a small predator, send out a signal to lure bigger ones. A big pike would rather eat a small pike than a minnow any day.