gangs and best behaviour Ð scientists find that our actions
could be a little fishy
do you behave in a crowd? Do you stand meekly behind the
next person, shuffling forward when they do, or push past
other people to get through faster? You might think that
you act in an individual way, and there's no pattern to
your movements. But researchers at Leeds have found that
our behaviour bears striking resemblances to the behaviour
of animals, following certain rules that can be mapped,
explained and even predicted.
Iain Couzin (pictured left)
explains: "Ants follow chemical trails
to food sources along which they follow set paths to avoid
congestion. Similarly human beings in large crowds, moving
in both directions along a walkway, for example, will
spontaneously form 'lanes' travelling in one direction
or another. By following local rules Ð in this case the
easiest route through a crush of people Ð we create larger
scale patterns which on the ground, with our restricted
field of vision, we are unaware of."
key to understanding crowd behaviour is computer simulation.
While it's sometimes possible to see wider patterns among
a large group, it's harder to track one individual, or
quantify their relationship with their companions.
a computer can do all of these things if provided with
the right information. Biologist Dr Jens Krause and his
research team focus their work on fish, but maintain that
the findings can be applied to other animals, and even
Krause (pictured right)
: "Fish are useful for our work because they are easy
to study. We can maintain laboratory populations, and
make observations of their behaviour. We then feed this
into complex computer simulations, to try and determine
patterns, and the final stage is to test this on wild
populations in the field."
scientists film fish in the laboratory and then digitise
the film, feeding this into a computer. They're just beginning
a project with a research centre in Kenya, where herds
of zebra have been filmed from the air. Just as with the
films of fish, this information will be put into the computer
model to analyse the animal's behaviour.
Couzin handles the computer modelling: "When you're dealing
with shoals, it is impossible to track each fish with
the naked eye. Instead I have programmed a PC to automatically
track the fish and determine their patterns of behaviour.
From this information we can develop computer simulations
of fish schools (pictured
below). Then we can change the parameters
in the model Ð distance between fish, degree of alignment
or attraction Ð to see how that affects behaviour."
behaviour is basically ruled by balancing out costs and
benefits: how great a risk or advantage does taking an
action involve, in terms of safety from or exposure to
predators and losing or gaining access to mates or food?
of animals, flocks of birds or shoals of fish stay together
for safety, but the researchers have found that far from
being a stable group, fish move between groups, and change
their position within the group to satisfy their needs,
generally in terms of food. When hungry, fish will move
to a smaller group; when fed, they will move back to the
larger. And the fish are able to tell how big a group
is, by the distance between the members: smaller groups
are more spaced out; bigger groups, closer.
findings have wider implications, as Dr Krause explains:
"The commercial fishing industry needs to know how fish
shoal. In the open ocean, with fish like tuna, for example,
it's impossible to study them in this much detail. But
understanding how the group functions is very important,
for estimating population levels, setting fishing levels,
and for conservation.
group interaction and how much movement goes on between
groups is also useful for predicting the spread of disease,
particularly in fish farms."
as with humans, there is a debate as to whether animal
behaviour is determined more by nature or nurture.
zebra fish, Dom Wright (pictured
left) has been trying to test whether genetics
or environment has the most influence on behaviour. Using
two populations, one free population collected from Nepal
which has a high tendancy to come together in shoals and
a laboratory bred population which doesn't, he has been
selectively mating and testing the offspring's shoaling
behaviour. His results seem to show that patterns of behaviour
are 50% determined by genetics, and 50% by other factors.
The genome of zebra fish has nearly been sequenced, so
Dom Wright is now trying to link the genes with the inherited
Wright said: "When you cross the lines, you can work out
the genetic architecture of the trait. It's hard to find
an exact gene for 'shoaling behaviour' but should be possible
to work out roughly how many are involved, and their general
aspects also come into play, such as gender differences
in behaviour, and even metabolic rate. Fish with a faster
metabolic rate require food more often, and so are willing
to take more risks in order to get it.
many other aspects of behaviour clearly relate to environment,
such as level of predators, attacks by parasites or even
temperature Ð which again affects the amount of food required.
human beings might not face the necessity of finding food,
or fear predators, the reason why we form groups, and
how we behave in large crowds could be analysed using
the same methods. It might even explain how we choose
Krause: "We've noted a species in Canada which seems to
form 'gangs'. In experiments in the field in Canada, we
marked fish, followed their movements, and saw that they
came back to form the same groups. We believe they must
find it safer to stay with fish they know because they
know how they will react."
among fish it's worth knowing who your mates are.