Pagina principale Animal Dispersion In Relation To Social Behaviour

Animal Dispersion In Relation To Social Behaviour

Anno: 1962
Editore: Hafner Publishing Company
Lingua: english
Pagine: 653 / 700
File: PDF, 37.78 MB
Download (pdf, 37.78 MB)
Leggi il libro online
 
You can write a book review and share your experiences. Other readers will always be interested in your opinion of the books you've read. Whether you've loved the book or not, if you give your honest and detailed thoughts then people will find new books that are right for them.
.NUNC COCNOSCO EX PARTE

THOMASJ. BATA LIBRARY
TRENT UNIVERSITY

Digitized by the Internet Archive
in 2019 with funding from
Kahle/Austin Foundation

https://archive.org/details/animaldispersionOOOOwynn

ANIMAL DISPERSION
in relation to Social Behaviour

ANIMAL DISPERSION
in relation to Social Behaviour

V. C. WYNNE-EDWARDS
Regius Professor of Natural History
University of Aberdeen

HAFNER PUBLISHING COMPANY
NEW YORK

Published in Great Britain by
Oliver and Boyd Ltd. Edinburgh and London 1962

1962

First published

© 1962

V. C. Wynne-Edwards

Printed in Great Britain by
Oliver and Boyd Ltd., Edinburgh

Preface
The theory presented in this book links together the subjects of population
and behaviour. It applies to animals in general, which gives it an exceedingly
wide scope. During the last seven years it has provided me with a novel and,
it has often seemed, commanding viewpoint from which to survey the
everyday events of animal behaviour; and some of the most familiar acti¬
vities of animals, the purpose of which has never properly been understood,
have readily been seen to have important and obvious functions. It has
turned out to be an agreeable and characteristic feature of the theory not to
keep butting against widely held, pre-existing generalisations, but to lead
instead into relatively undisturbed ground. Needless to say, the reader
is confronted at the beginning with two or three fundamental principles that,
on account of their unfamiliarity alone, he may be expected to eye with a
certain amount of scepticism, until they can by degrees be critically appraised
in the light of each succeeding chapter of evidence.
The theory has been developed more or less freely, although this has
led into some fields of knowledge where I have had to explore as a novice
as I went along. Given an apparently rational, single explanation of the
origin of social behaviour, for instance, one of the interesting corollaries has
been to interpret the zoological background of conventional behaviour in
man. This i; s not without its philosophical implications, although actually
these have only very lightly been touched upon; seldom, as a matter of fact,
could one find any clearer indication than emerges here of the closeness of
man’s kinship with his fellow animals. Our human experience as social
beings turns out to be of frequent use in gaining an understanding of sociality
in other species and, with reasonable discretion, it can be drawn upon
without any serious fear of anthropomorphism. The evident loss by man,
almost within the historic period, of the means of limiting population growth,
which he formerly possessed like other animals, stands out with disturbing
clarity.
A large book has been required in order to contain enough factual
evidence to support the theory in each of its widespread ramifications. The
literature that could and ideally should have been consulted is too vast to be
within the compass of one person in any reasonable time. Though fairly
full documentation for what has been included is generally attempted, it is
too much to hope that nowhere will key references have been overlooked;
and to those authors whose illuminating works I have inadvertently passed
over I offer my apologies. During the final revision it has been practicable
to incorporate only a few of the most recent publications, which have poured
out in an ever-growing stream since the various chapters were first written,
v

42162

VI

PREFACE

in some cases several years earlier. There must be places also where en¬
thusiasm has led me rashly into conclusions that experts with their betterdigested knowledge will regard as unwarranted. Such errors and omissions,
whether numerous or few, will I hope be found to relate in the main to inci¬
dental points, and do nothing to undermine the solidarity of the central
theory itself.
A general idea of the subject and how it is developed can be obtained
without necessarily reading the whole book from cover to cover. The first
chapter is intended to provide both the foundations of the theory and a
preliminary review of it. Thereafter all the way through the book there are
summaries at the end of the chapters. These are usually fairly full, though
their condensation is naturally bound to make them tougher meat than the
book itself. They may show the reader where to skip, where to pick out for
detailed reading the subjects likely to interest him most, and where to turn
to the main text when it seems essential to find corroborating evidence;
they may also supply a brief reminder of the chief points that have emerged at
each stage, after the chapter itself has been read. It is appropriate to note
here that a previous, very condensed five-page sketch of the theory of animal
dispersion was published in 1959 (Ibis, 101: 436-41).
I am indebted for information and discussion to many more friends and
correspondents than can be mentioned individually by name. I have abun¬
dant cause to be grateful to my colleagues in Aberdeen for their knowledge,
criticism and encouragement, especially Doctors George Dunnet, Robert
M. Neill, Philip Orkin, David Jenkins, Adam Watson and Guy Morison;
the same is true of Dr William M. Clay and my other colleagues in the
Department of Biology at the University of Louisville, Kentucky, where I
spent four rewarding months between the writing of Chapters 19 and 20.
I must expecially thank Mr Ben Feaver, of Haverfordwest, and Miss
Mabel Slack, of Louisville, for allowing me to use their coloured photographs
for plates IV and IX. The butterfly specimens in plate VIII were kindly lent
by the Royal Scottish Museum, and were selected and photographed under the
skilled supervision of Dr D. R. Gifford, of Edinburgh University. For
plate II I am grateful to Dr Robert S. Simmons, of Baltimore, Maryland;
and for plates III (upper photo) and V, to Dr Robert Carrick, of Canberra,
Australia. Acknowledgments are also due to the following for kind
permission to reproduce illustrations already published elsewhere, namely
Akademische Verlagsgesellschaft, Leipzig (fig. 6); Editor of Ardea (fig. 47);
Editor, Biological Bulletin Wood's Hole (fig. 5); Cambridge University
Press (figs. 28, 48, 49); Dieterich’sche Verlagsbuchhandlung, Wiesbaden
(fig. 7); Entomological Society of America (fig. 40); Messrs Walter de
Gruyter, Berlin (fig. 18); Editor of the Ibis (pi. XI); Messrs Macmillan &
Co. (figs. 30, 31, 45 and 46); Marine Biological Association of the United
Kingdom (figs. 33 and 36); Royal Irish Academy (fig. 32); the Royal Society
(fig. 12); Messrs Oliver & Boyd (figs. 24 and 37-8); Sears Foundation for
Marine Research, Yale University (fig. 29); Smithsonian Institution,

PREFACE

vii

Washington (figs. 13 and 14); Superintendent of Documents, U.S. Govern¬
ment Printing Office (figs. 2 and 41-4); University of California Press (fig.
34, pis. VI and VII); University of Chicago Press (fig. 50); Messrs H. F. & G.
Witherby (fig. 22); and the Zoological Society (pi. X).
Finally it is a pleasure to acknowledge the immense help I have received
from my wife.

Contents
Chap.

Page

Preface

.........

1

An outline of the principles of animal dispersion

2

The integration of social groups by visible signals .

3

Social integration by the use of sound: land animals

4

Social integration by underwater sound and low-frequency
vibrations

.

.

y
1
23

.

41

........

65

5

Social integration by electrical signals ....

82

6

Social integration by olfactory signals ....

90

7

Social integration through tactile perception .

.

.

8

The social group and the status of the individual

.

.127

9

Dispersion in the breeding season: birds

.

.

.

10

Property-tenure in other groups .

.

.

.

.165

11

Communal nuptial displays.

.

.

.

.193

12

Display characters and natural selection

.

.

.

223

13

Further consideration of castes in animal societies

.

255

14

Communal roosts and similar gatherings

.

283

15

Timing and synchronisation

.....

326

16

Vertical migration of the plankton

17

Associations between species

18

.

.

.

114\X
145

....

366

.....

389

Siblings and mimics

.......

423

19

The use of tradition

.......

449

20

Fluctuations, irruptions and emigrations

21

Recruitment through reproduction

.

466

....

484

22

Socially-induced mortality ......

530

23

Deferment of growth and maturity

....

557

........

603

.........

631

References
Index

IX

.

.

Illustrations
Plate

Facing page

I

Springbuck in ‘ pronking ’ attitudes

33

ii

Male and female bronze frogs

48

hi

Evening mass-manoeuvres of starlings at a roost
A swarm of 3-4000 whirligig beetles

iv

.

Manx shearwaters assembled on the sea

205

v

Bogong moths aestivating on the roof of a cave

.

314

vi

Acorn woodpeckers: the ‘ bent tree settlement ’

.

321

vii

vm
ix

x
xi

Acorn woodpeckers’ storage tree

....

322

Mimicry in East African butterflies

....

442

Monarch butterflies overwintering on the butterfly trees at
Pacific Grove, California

458

Sections of ovaries from two non-breeding fulmars

570

The gannetry at Cape St. Mary, Newfoundland

571

xi

.

Chapter 1

4n outline of the principles of animal dispersion
1.1 Introduction, p. 1. 1.2. Optimum density in relation to resources, p. 4.
1.3. The existence of natural conventions, p. 10. 1.4. Social organisation,
p. 13. 1.5. Social evolution and group-selection, p. 18. 1.6. Carr-Saunders’
principle of the Optimum Number, p. 21.
‘ Most of the singing and elation of spirits of that time [the amorous
season] seem to me to be the effect of rivalry and emulation; and it is to
this spirit of jealousy that I chiefly attribute the equal dispersion of birds
in the spring over the face of the country.’ Gilbert White, The Natural
History ofSelborne, Letter XI to the Hon. Daines Barrington, 8 February
1772.

1.1. Introduction
Animal dispersion may be defined as comprising the placement of
individuals and groups of individuals within the habitats they occupy, and
the processes by which this is brought about. It is something we are inclined
to take largely for granted. People who observe these things know well that
animals can be expected to occur more plentifully in habitats that contain
an abundance of the kinds of resources they need, and that they will generally
be scarcer or absent altogether where the environment is less well provided.
Experience teaches us also the sort of habitat preferred by each species; and
the inference is clear enough that animals must often have complex adap¬
tations—either in the form of innate instincts or acquired responses—to
enable them to select their habitats appropriately. Inference may carry our
thoughts a step further, suggesting to us that habitat selection and dispersion
must be essentially internal or domestic affairs that each species arranges for
itself as best it can. But seldom do we get very far beyond the point of
accepting dispersion as something that automatically happens, or even dimly
begin to realise the gigantic, universal organisation it entails. Lack (1954a,
p. 264) has rightly said that it ‘ presents a remarkable, and until now largely
unappreciated problem ’.
Lack devoted the final chapter of his book, The natural regulation of
animal numbers, to dispersion. Like Gilbert White (quoted at the head of
this chapter) he realised that it must be due, in part at least, to the responses
and behaviour of the animals themselves. ‘ Dispersion [of birds] in winter ’,
he says, ‘ presents no particular problem and seems to be adequately explained
by supposing that, within their favoured habitats, the individuals or flocks
A

2

PRINCIPLES

OF

ANIMAL

DISPERSION

1.1

avoid areas where food is short and tend to settle where it is abundant
(loc. cit., p. 264). And further on (p. 269), ‘ the problem of dispersion was
obviously in Howard’s mind when he propounded the territory theory,
though he rightly linked territory with other matters, particularly pair
formation, and, also rightly in my opinion, linked dispersion with the food
requirements of the young
We shall find as we proceed that we have to

Fig. 1. Jespersen’s correlation between the number of pelagic birds and the abundance
of plankton in the North Atlantic. In each 10° sector the upper figure gives the
average volume of macroplankton in cc in a standard haul, and the lower figure
(in larger type) the average number of birds recorded per day. The correlation
coefficient between them is +0-85. (Redrawn on an equal-area projection from the
data in Jespersen’s ‘ Dana ’ report, 1930.)

travel a long way from this initial viewpoint; but it is a very appropriate
place to start, because it was from Lack’s stimulating book and in particular
the final chapter that the title of this one and the inspiration for writing it
came.
A good illustration of what is involved in dispersion, all the better because
it is quantitative, was provided by Jespersen (1924, 1930) as a result of the
Danish oceanographic expeditions made in search of the birthplace of
freshwater eels (Anguilla) between 1913 and 1928. Standard hauls of
macroplankton were taken at scores of stations all over the North Atlantic,

1.1

INTRODUCTION

3

and these, measured by bulk, could be used as a broad index of the richness
of the surface waters, in terms of food for pelagic birds; the hauls were
grouped and averaged to give a single figure for each 10-degree sector of the
ocean. A log was kept of the number of birds seen each day; and the counts
for all the days spent in each 10-degree sector were similarly combined and
averaged, to give an index of the population-density of birds. In spite of
the collective character of the data and the differences in season at which
the different areas were sampled, the correlation that emerges between bird
density and abundance of plankton is very strong (see fig. 1): in numerical
terms it can be expressed by a coefficient of +0 85. The probability that
such a situation could have arisen by chance alone is negligible
(P^0-001).
Tiffs is a dispersion that the birds must have brought about by their own
efforts; and, as far as can be judged from the averages of grouped samples
such as we are given here, it results in an efficient pro rata allocation of birds
to the available food-resources. The population-density in fact appears to
be graded so that in every area it bears about the same constant average
relation to the amount of plankton present; and this certainly appears to be
a situation that we should no longer dismiss as ‘ presenting no particular
problem
Another correlation of the same kind, in which the relationship between
eater and food is simpler and more direct, and on a smaller geographical
scale, was demonstrated by Hardy and Gunther (1935, p. 273), between the
population-density of whalebone whales round South Georgia (determined
from the kills by commercial whalers) and that of their principal food-species,
the krill Euphausia superba.
Close adjustments between animals and their essential resources are not
confined to birds and mammals, but can be found even in the Protozoa.
There is usually of course at least a residual element of chance in dispersion,
and of the influence of extraneous factors beyond the animals’ control; but
freely-moving individuals endowed with the necessary perceptions always
have the means of combating chance and seeking to improve their situation.
If we look, for instance, at the dispersion and settlement on the floor of the
sea of free-swimming larvae, after they have drifted in the plankton appar¬
ently like seeds borne on the wind, we find it is far from being a purely
random scatter; in fact their metamorphosis can generally be postponed for
some considerable time while they search for a suitable substratum (cf.
Wilson, 1952, p. 118). Thus oyster larvae tend to be gregarious on settlement,
and prefer to attach themselves when they have found surfaces of suitable
texture on which other oyster larvae have already settled, at least up to a
certain density (Cole and Knight-Jones, 1949); and practically the same has
been found to apply to larvae of the barnacle Elminius modestus (KnightJones and Stevenson, 1951). The complexity of the responses involved in
dispersion, even in comparatively young and lowly animals, is very
apparent.

4

PRINCIPLES

OF

ANIMAL

DISPERSION

1.2

1.2. Optimum density in relation to resources
Ideally the habitat should be made to carry everywhere an optimum
density, related to its productivity or capacity, without making any parts so
crowded as to subject the inhabitants to privation, or leaving other parts
needlessly empty. It seems quite possible, on the basis of general experience,
that something approaching these conditions may in fact often be realised.
We ought therefore to consider what factors would be likely to determine
the optimum density of a population; and this brings us to one of the
fundamental principles underlying the theory of animal dispersion that we
are setting out to examine.
Before proceeding, however, it should be stated that in the great majority
of species of animals, just as in the sea birds, the critical resource, as far
as population-density is concerned, is food. There are of course many other
requirements to be satisfied in making a habitat habitable at all—it must be
accessible, and within certain limits of tolerance as far as physical conditions
are concerned; it must be able to provide whatever shelter is required, and
be free from incompatible organisms of other kinds. But, granted that every
essential need is met at least minimally, and life for a particular species is
consequently supportable at all, food so generally becomes the factor which
ultimately limits population increase that we can afford to regard any other
ultimate limiting factors—mere ‘ standing room ’, for instance, in the case
of some of the sessile barnacles (Balanus etc.) in the intertidal zone, or nestholes for breeding populations of pied flycatchers (Muscicapa hypoleuca) or
purple martins (Progne subis)—as being special exceptions to a general rule.
It is with the rule and not with the exceptions that we need at this stage to be
concerned.
Our best approach to the subject of optimum density is to study man’s
own experience in exploiting natural recurrent resources for consumption
as food and for other purposes—resources consisting of ‘ wild ’ species
whose numbers are sustained and renewed by allowing nature to take its
course. Man the fisherman, in particular, still acts essentially as a predator
exploiting natural prey. He goes away to sea with his nets and gear and
comes home with the spoils; and hitherto he has been effectively prevented
(except on a negligible scale) from undertaking any kind of cultivation or
husbandry to increase the natural rate of replenishment of the stock. His
experience is therefore closely relevant to the study of the relationship
between other species and their food-resources.
Events of the last hundred years and more have proved beyond doubt
that fishery resources are not inexhaustible. There is no more striking
illustration of this than the one presented by the history of whaling. As
early as the 16th Century the Basques had begun to make serious inroads
into the stock of the North Atlantic right whale (Balaena glacialis). By the
17th Century the centre of the fishery had shifted north to the seas and fiords
of Spitzbergen, and from there it moved still later to Baffin Bay. In these
Arctic waters the principal prey was the Greenland whale {Balaena mysticetus).

1.2

OPTIMUM

DENSITY

IN

RELATION

TO

RESOURCES

5

The northern fishery reached a high peak of prosperity in the early years of
the 19th Century: thousands of people were then directly or indirectly
deriving their living from it, and in Britain alone the ports of Hull, Whitby,
Leith, Dundee and Peterhead thrived as never before. Capital and enterprise
poured into the industry, and there can have been little or no thought of
moderating the catch and preventing overfishing. Although the invention
of steamships and the harpoon gun about 1860 made it possible to catch the
faster-moving blue and fin whales (Balaenoptera) and to stave off the ultimate
disaster for another fifty years, before 1914 the whole North Atlantic fishery
had crumbled in ruins. The stocks of the two right whales have never
recovered, and the population of Greenland-whaling men and of those who
ministered to them has become effectively extinct.
A similar though less disastrous history has followed the pursuit of all
the chief commercial fishes, such as the cod (Gadus caUarias), haddock
(G. aeglifinus), plaice (Pleuronectes platessa), halibut (Hippoglossus), and at
last apparently even the herring (Clupea harengus). What first opened our
eyes to the reality of overfishing was the spectacular recovery made by some
of these heavily exploited species in the North Sea, as a result of the respite
they got in the 1914-18 war when vessels and crews were otherwise engaged
and the Dogger Bank was covered by minefields. The ability commonly
shown by partially depleted fisheries to regenerate in this way is, incidentally,
a heartening feature in a generally depressing prospect, and ecologically it
is of the highest importance.
Since 1918 the subject of overfishing has attracted increasing study by
fishery scientists, and an understanding has been gained of the principles
underlying the conservation and optimal exploitation of fishery resources
(c/. Beverton and Holt, 1957). (An account of a very illuminating experiment
on fishery exploitation by Silliman and Gutsell (1958) will help to bring these
principles home in a later chapter; see p. 499). In its simplest terms the
empirical position is that the effort devoted to exploiting what can be thought
of at the outset as a virgin fishery can be steadily increased, with corresponding
increases in the total catch, up to a certain critical maximum. If the fishing
intensity is stepped up beyond this level, the stock begins to suffer depletion
and the annual catch in time starts getting smaller; a position is thus reached
where increasing effort produces a smaller total return.
This critical level of fishing intensity is the one that gives the greatest
possible sustained landings of fish, both absolutely and in terms of unit
effort expended, and also the greatest cumulative total catch. We are
concerned here as simply as possible with the production of food, and can
afford to neglect everything to do with variations in the details and efficiency
of exploitation, for instance through altering the mesh-sizes of nets, and
also ignore the question of market values and financial returns. The
maximum is represented graphically in figure 2 as the highest point on the
curve relating sustained yield (the same as ‘ equilibrium yield ) to fishing
intensity.
This is more likely to be a rounded summit than a sharp

6

PRINCIPLES

OF

ANIMAL

1.2

DISPERSION

culmination, so that in practice it is usually reached without anyone realising
it; and it can actually be revealed only by careful long-term analysis or
experiment.
By the time commercial fisheries unknowingly begin to exceed the optimum
level they are generally paying handsomely. Increased effort still gives
temporarily increased returns. The yield per man-hour tends of course to
drop, at first very slightly, but this is likely to be masked for some time by
natural irregularities in productivity from year to year. In any case there is a
strong tendency to shut one’s eyes to the unwelcome signs of overfishing, so
that depletion may go a long way before any pinch is actually felt or the cause
of it admitted by those directly involved.

0

.2

4

INSTANTANEOUS

.6

.8

1.0

1.2

1.4

RATE OF EXPLOITATION

Fig. 2.

Graph of the relation between the rate of exploitation (horizontal axis) and the
equilibrium yield of fish (vertical axis) in Silliman and Gutsell's experiment. Under
these particular conditions, the highest sustained yield would have been obtained by
a 35-40 per cent exploitation rate, and any exploitation above about 55-60 per cent
would have led to the extinction of the stock. (From Silliman and Gutsell, 1958.)

Having understood the problem it becomes possible to see a remedy:
this is to determine the approximate optimum yield and limit the effort to
match it. The slogan ‘ fish less and catch more ’ puts the immediate aim in
a nutshell. Politically it is extremely difficult to achieve, because it requires
all participants to reach an agreement on some workable method of limi¬
tation, and thereafter scrupulously to abide by it. What is fundamental to
conservation is that the system of free enterprise, with every man or company
for himself, must be exchanged for a common code of rules.
Tins has nevertheless been agreed to in a promising number of cases in
different parts of the world, among them that of the southern whale fishery.
There the International Whaling Commission has for some years successfully
controlled both the number of whaling expeditions taking part and the size

E2

OPTIMUM

DENSITY

IN

RELATION

TO

RESOURCES

7

of the combined catch for the year in terms of so many ‘ blue-whale units
the moment the agreed total has been reached, it declares the fishery closed
for the season. Admittedly the Commissioners’ task is extremely difficult,
and the results imperfect. For one thing, each expedition now tends to be
expensively equipped to kill whales at the maximum possible speed in order
to get the biggest obtainable share of the quota before the season is closed.
The different expeditions are thus still in intense competition, and it would
be much better if each country or outfit could be assigned an individual
quota, allowing it freedom to take its harvest economically and in its own time,
and entirely banishing the competitive element from the whaling grounds.
A century ago, indeed, in the worst days of the North Pacific fur-seal fishery,
things got so bad that vessels had to mount cannon to defend themselves
and international incidents and bloodshed occurred; this was finally
prevented and the present extremely successful scientific management made
possible, by the North Pacific Sealing Convention of 1911—perhaps the
first international treaty of the kind.
There are in fact five lessons to be learnt from this phenomenon of
overfishing, the importance of which it would be difficult to exaggerate.
The first is that overfishing reduces both the yield per unit effort and the total yield;
in some circumstances, if sufficiently severe, it can damage the stock beyond
recovery and even lead to its extermination. This actually happened in the
19th Century through the commercial exploitation in North America of the
passenger pigeon (Ectopistes migratorius)—‘ the most impressive species of
bird that man has ever known’ (Schorger, 1955, p. viii; see also below,
p. 451). Secondly, the size of the optimum catch is not self-evident, and can
be determined and adjusted only in the light of experience. Thirdly, the
participants must all come to a common agreement or convention to limit
their catch, preferably while the stock is still at or near its maximum abundance
and productivity, and must forego any immediate personal advantage in
favour of the long-term benefit of the community as a whole. Fourthly,
argument about who shall participate and how much he shall take ought
always to be transferred to some higher court, so that direct destructive and
wasteful competition is entirely eliminated from the catching of the actual
fish. Fifthly, no profitable fishery is immune from over-exploitation; the
consequences are certain to follow wherever the optimum rate of cropping
is persistently exceeded. Exactly the same is true of any other kind of profit¬
able, renewable, living resource, whether of game, fur-bearers, or—if
nutrients are not returned to the soil, or plants are killed or regeneration
impaired—even of natural vegetation.
The importance of these inferences lies, of course, in the fact that there
is no difference in principle between man exploiting fish or whales and any
other predator exploiting any other prey. All predators face the same aspects
of exactly the same problem. It is impossible to escape the conclusion,
therefore, that something must, in fact, constantly restrain them, while in the
midst of plenty, from over-exploiting their prey. Somehow or other free

PRINCIPLES

8

OF

ANIMAL

DISPERSION

1.2

enterprise ’ or unchecked competition for food must be successfully averted,
otherwise ‘ overfishing ’ would be impossible to escape: this could only
result in lasting detriment to the predators and the risk, if they persisted in
it, that the prey might be exterminated altogether.
That some such restraint often exists in nature is most easily compre¬
hended, perhaps, in a situation where a population of animals has to depend
for a period on a standing crop, which must be made to last out until the
season comes round for its renewal or until alternative foods become
available. Many small northern birds, for instance, are provisioned ahead
for four or five months in the winter by the standing crop of seeds of a few
species of trees or herbs, or by a finite stock of overwintering insects. It
would be fatal to allow birds to crowd in freely in autumn up to the maximum
number that could for the moment be supported by the superabundance of
food, without regard to later consequences. The optimum populationdensity, on the contrary, would generally be one that could be carried right
through till the spring on the same stock of food, supposing the birds were
winter-resident species such as tits (Parus), robins (Erithacus) or wood¬
peckers (Pici). To achieve this optimum it would clearly be necessary to put
a limit on the population-density from the beginning, while the resource was
still untapped.
The need for restraint in the midst of plenty, as it turns out, must apply
to all animals whose numbers are ultimately limited by food whether they
are predators in the ordinary sense of the word or not. It is commonly the
only way in which plenty can be conserved and maintained. It applies in
general as forcibly to herbivores as to carnivores. Ruminants, for instance,
are notoriously capable of impairing fertility by long-continued overgrazing;
as is well known, this has resulted, in the brief span of history, in reducing
large tracts of primeval forest and natural grassland in the Near and Middle
East and other ancient centres of pastoral civilisation almost to the bare
stones. The habitat generally contains only a finite total quantity of those
nutrient elements, such as nitrogen, phosphorus, potassium and calcium,
on the repeated circulation of which the continuity of life depends. One of
them usually sets a limit to primary productivity. It seems probable, in
closed communities at any rate, that selection will favour an optimal balance
in the rate of circulation and in the proportion of the total that is held at
any given moment by (i) the animal biomass, (ii) the plant biomass, and (iii)
the remainder of the system. The hoarding of precious nutrients within the
animal biomass, which is likely to be accentuated by over-population, may,
especially if it is followed by emigration on a large scale, seriously prejudice
the continued productivity of the system as a whole.
Where we can still find nature undisturbed by human interference,
whether under some stable climax of vegetation or proceeding through a
natural succession, there is generally no indication whatever that the habitat
is run down or destructively overtaxed. On the contrary the whole trend of
ecological evolution seems to be in the very opposite direction, leading

1,2

OPTIMUM

DENSITY

IN

RELATION

TO

RESOURCES

9

towards the highest state of productivity that can possibly be built up within
the limitations set by the inorganic environment. Judging by appearances,
chronic over-exploitation and mass poverty intrude themselves on a mutuallybalanced and thriving natural world only as a kind of adventitious disease,
almost certain to be swiftly suppressed by natural selection. It is easy to
appreciate that if each species maintains an optimum population-density
on its own account, not only will it be providing the most favourable
conditions for its own survival, but it will automatically offer the best possible
living to species higher up the chain that depend on it in turn for food.
Such prima fade argument leads to the conclusion that it must be highly
advantageous to survival, and thus strongly favoured by selection, for animal
species (1) to control their own population-densities, and (2) to keep them
as near as possible to the optimum level for each habitat they occupy.
Regarding the first of these conditions, the general hypothesis of self¬
limitation of animal numbers has been growing rapidly in favour among
animal ecologists in recent years (e.g., Kalela, 1954; A. J. Nicholson, 1954;
Errington, 1956; Wynne-Edwards, 1955; Andrewartha, 1959); so far as
it goes the evidence given already in the introductory section of course
confirms it. To build up and preserve a favourable balance between
population-density and available resources, it would be necessary for the
animals to evolve a control system in many respects analogous to the physio¬
logical systems that regulate the internal environment of the body and adjust
it to meet changing needs. Such systems are said to be homeostatic or
self-balancing, and it will be convenient for us to use the same word.
Physiological homeostasis has in general been slowly perfected in the course
of evolution, and it is thus the highest animals that tend to be most indepen¬
dent of environmental influences, as far as the inward machinery of the body
is concerned. Population homeostasis, it may be inferred, would involve
adaptations no less complex, and it might therefore be expected that these
would similarly tend to reach the greatest efficiency and perfection in the
highest groups.
We are going to discover in the concluding chapters of the book that
such homeostatic adaptations exist in astonishing profusion and diversity,
above all in the two great phyla of arthropods and vertebrates. There we
shall find machinery for regulating the reproductive output or recruitment
rate of the population in a dozen different ways—by varying the quota of
breeders, the number of eggs, the resorption of embryos, survival of the
newborn and so on; for accelerating or retarding growth-rate and maturity; j
for limiting the density of colonisation or settlement of the habitat; for
ejecting surplus members of the population, and even for encompassing
their deaths in some cases in order to retrieve the correct balance between
population-density and resources. Not all types of adaptation have been
developed in every group, though examples of parallel evolution are abundant
and extraordinarily interesting.
Indeed this newest manifestation of
homeostasis in the processes of life seems unlikely to remain long in doubt.

10

PRINCIPLES

OF

ANIMAL

DISPERSION

1.3

1.3. The existence of natural conventions
At this point, however, we must leave the subject of homeostasis
temporarily in order to consider the second conclusion of our prima facie
argument, namely that it would be advantageous to be able to keep the
population-density at or near the optimum level. We are still concerned with
those species of animals—knowing them to be in the great majority—whose
numbers are ultimately limited by the resource of food.
It should first be recognised that, unless they are on the verge of extinction,
all animals and plants have a great latent power of increase. Physiological
provision is made, even in the slowest breeders, for the production of off¬
spring in excess of what is needed merely to sustain the population. In order
to prevent a geometrical progression of increase, therefore, some kind of
brake must be applied; and if this is to allow the increase to proceed freely
when the population-density is low and economic conditions permit it, but
to prevent density going any higher once a particular limiting threshold has
been reached, then the application of the brake requires to be ‘ densitydependent
That is to say, that when the density is low, multiplication will
be relatively unhindered, but as it mounts towards a ceiling the checks on
increase will become progressively more severe, until population losses through
death and emigration catch up with the gains from reproduction and
immigration, and the increase is brought to a halt.
When an experimental population is set up in some kind of a confined
universe with a finite but regularly renewed supply of food, whether the
population consists of Drosophila, flour-beetles, Daphnia, Lebistes (guppies)
or mice—to name some of the commonest experimental animals—this is
exactly what happens. As we shall see later (see Chapter 21, p. 495), the
experiments can be replicated time and again, and if the universe is made
the same, the population will reach the same predictable ceiling of numbers
each time within a narrow margin of error. It will be a matter of special
remark when the experiments come to be described that the ceiling is never
imposed by starvation, except perhaps indirectly in mice where the mothers
often run out of milk at high population-densities, with profound effects on
the production of recruits. On the contrary, the ceiling is normally imposed,
and the level indefinitely maintained, while the members of the population
are in good health—sometimes actually fat—and leading normal lives.
Guppies, for instance, still breed and viviparously produce young, but after
the ceiling is reached and there are no vacancies in the population, the young
are gobbled up by their elders within a few minutes of their birth (see p. 543).
It is not necessary here to go into the subject of external density-dependent
checks on population—caused by agents such as predators, parasites and
pathogens that are likely to take a mounting percentage of lives as populationdensity rises and economic conditions become more severe. This will be
discussed sufficiently fully in the appropriate place later on (p. 546). It will
suffice for the present to say that these external checks, while they may some¬
times be extremely effective in preventing population-density from rising,

THE

EXISTENCE

OF

NATURAL

CONVENTIONS

11

are on the whole hopelessly undependable and fickle in their incidence, and
not nearly as perfectly density-dependent as has often been imagined. They
would in most cases be incapable of serving to impose the ceilings found
in nature; what is more, experiment generally shows that they are un¬
necessary, and that many if not all the higher animals can limit their
population-densities by intrinsic means. Most important of all, we shall find
that self-limiting homeostatic methods of density-regulation are in practically
universal operation not only in experiments, but under ‘ wild ’ conditions
also.
Towards the fringe of its range the existence and population-density of
any particular species of animal is often overwhelmingly dictated by the
physical conditions of the environment—heat, cold, drought, shelter and the
like—and by such biotic factors as the presence of better-adapted competitors
or the absence of requisite vegetational cover; and it is frequently one of
these factors that precludes any further geographical extension of its range.
If we ignore the fringe, however, and confine our attention to the more
typical part of the range, dispersion in the great majority of animals reflects
the productivity of the habitat in terms of food, in just the same way as we
saw in the special case of Jespersen’s correlation between pelagic birds and
plankton. That this is so is nowhere in serious dispute; and if we take care
to exclude the minority of species where the food correlation is weak or
absent, we can repeat once more that food is generally the ultimate factor
determining population-density, and the one that predominates far above
others.
We have already the strongest reasons for concluding, however, that
population-density must at all costs be prevented from rising to the level
where food shortage begins to take a toll of the numbers—an effect that
would not be felt until long after the optimum density had been exceeded.
It would be bound to result in chronic over-exploitation and a spiral of
diminishing returns. Food may be the ultimate factor, but it cannot be
invoked as the proximate agent in chopping the numbers, without disastrous
consequences. By analogy with human experience we should therefore look
to see whether there is not some natural counterpart of the limitationagreements that provide man with his only known remedy against overfishing
—some kind of density-dependent convention, it would have to be, based on
the quantity of food available but ‘ artificially ’ preventing the intensity of
exploitation from rising above the optimum level. Such a convention, if it
existed, would have not only to be closely linked with the food situation, and
highly (or better still perfectly) density-dependent in its operation, but,
thirdly, also capable of eliminating the direct contest in hunting which has
proved so destructive and extravagant in human experience.
It does not take more than a moment to see that such a convention could
operate extremely effectively through the well-known territorial system
adopted by many kinds of birds in preparation for the breeding season.
Instead of limiting the number of expeditions and fixing the total annual

12

PRINCIPLES

OF

ANIMAL

DISPERSION

1

catch like the International Whaling Commission, conventional behavior
in this case limits the number of territories occupied in the food-gathering
area. Birds will not, of course, submit to being overcrowded beyond a
certain density, nor to the reduction of their territories below a basic minimum
size, so that territorial behaviour is perfectly capable of imposing a ceiling
on population-density (as C. B. Moffat realised in 1903). In its action it is
completely density-dependent, allowing the habitat to fill amid mounting
rivalry up to a conventional maximum density, after which any further
intrusion is fiercely repelled. Where territories retain this simple primitive
character as feeding areas, and where enough information is available, the
evidence indicates that minimum territory size is inversely related to the
productivity of the habitat: in other words, it is closely linked with the
presumptive food-supply. Finally, our last condition is completely met, in
that the contest among the participants is all for the possession of territory,
and once they have established their claim to the ground they can do the
actual food-getting in perfect peace and freedom, entirely without interference
from rivals.
The substitution of a parcel of ground as the object of competition in
place of the actual food it contains, so that each individual or family unit
has a separate holding of the resource to exploit, is the simplest and most
direct kind of limiting convention it is possible to have. It is the commonest
form of tenure in human agriculture. It provides an effective proximate
buffer to limit the population-density at a safe level (which is obviously
somewhere near the optimum though we can only guess in animals how
nearly perfection is attained); and it results in spreading the population
evenly over the habitat, without clumping them in groups as we find in many
alternative types of dispersion.
Much space is devoted in later chapters to studying the almost endless
diversity of density-limiting conventions, and only the briefest indication of
their range can be given here. What must first be appreciated is their
artificial or slightly unreal quality. The food-territory just considered is
concrete enough—it is the very place where food is found and gathered.
But some birds have territories in which they nest but do not feed; and some
have only token territories which are the nest-sites they possess and defend
in a breeding colony. These, as we shall see later, perform exactly the same
function, because the number of acceptable nest-sites in the colony is
‘ artificially ’ limited by the birds’ traditional behaviour. In a rookery, for
instance, the fullest membership of the community, conferring the right to
breed, belongs only to those pairs that can secure a nest-site, construct and
defend a nest. Supernumerary ‘ unauthorised ’ nests are constantly pillaged,
and there are almost always in consequence a good many non-breeding
adults present. These latter, however, may be accepted members of the
rookery population, all of whom have the free right to all the resources of the
communal territory of the rookery (see p. 158). Communal feedingterritories have been closely studied also in a few other birds, such as the

SOCIAL

ORGANISATION

13

Australian ‘ magpie ’ Gymnorhina tibicen (Dr. Robert Carrick, orally), and
'in Crotophaga ani and Guira guira in Cuba and Argentina respectively
(Davis, 1940a and b); but it will appear later that the phenomenon is evidently
a general one, applying, not in the breeding season alone, to many birds
that flock and to gregarious mammals at any time. In solitary mammals
such as foxes the home-ranges of individuals freely overlap; but each bona
fide resident must have the established use of a traditional earth or den in
order to be tolerated and allowed the freedom of the local resources; and
so it goes on (cf. fig. 8, p. 100).
In most cases the personal status of the individual with respect to his
rivals assumes a great importance; and in fact conventional contests
frequently come to be completely divorced from tangible rewards of property,
and are concerned solely with personal rank and dominance. All who carry
sufficient status can then take whatever resources they require without further
question or dispute. We shall find that, although widespread and common,
these abstract goals of conventional competition are especially characteristic
of gregarious species.
The subject is evidently a complex one, that will require much ampli¬
fication. Enough introduction has perhaps been given, however, to reveal
that conventional competition really exists, and to suggest the forms the
prizes or goals can take and the way they can be made to serve as dummies
or substitutes for the ultimate goal that should never be disputed in the
open—the bread of life itself.
1.4. Social organisation
In any homeostatic system there are necessarily two component processes.
One is the means of bringing about whatever changes are required to restore
the balance when it is disturbed, or to find a new balance when this becomes
necessary through changes in conditions external to the system. It has
already been briefly indicated in the previous section that populations (among
the higher animals at any rate) do have the necessary powers to adjust their
own population-densities. The other essential component of homeostasis
is an input of information, acting as an indicator of the state of balance or
imbalance of the system, that can evoke the appropriate corrective response.
A stimulus is required that will check and reverse the trend of the system
when the balance sways in one direction, in order to bring it back into
equilibrium. A device of this kind is familiar to electronic engineers in the
design of stable electrical systems, and is described by them as negative
feed-back.
In the balance that we are considering here it is postulated that populationdensity is constantly adjusted to match the optimum level of exploitation of
available food-resources; and as food-supply ‘ futures ’ change with the
changing seasons the population-density must be adjusted to match in so
far as this is possible in existing circumstances. What is needed in the way
of feed-back, therefore, is something that will measure or reflect the demand

14

PRINCIPLES

OF

ANIMAL

DISPERSION

1.4

for food, assessing the number of mouths to be fed in terms of present and
prospective supplies; to use another analogy, it has to undertake the
instrumentation, and to respond to population-density and economic
conditions in the same general way as a thermometer is used in a thermostatic
system.
Free contest between rivals for any commodity will readily provide such
an indicator. The keener the demand the higher the price in mettle and
effort required to obtain the reward: the tension created is thus ideally
density-dependent. One of our guiding first principles, however, is that
undisguised contest for food inevitably leads in the end to over-exploitation,
so that a conventional goal for competition has to be evolved in its stead;
and it is precisely in this—surprising though it may appear at first sight—
that social organisation and the primitive seeds of all social behaviour have
their origin. This is a discovery (if it can be so described) of the greatest
importance to the theory.
Any open contest must of course bring the rivals into some kind of
association with one another; and we are going to find that, if the rewards
sought are conventional rewards, then the association of contestants auto¬
matically constitutes a society. Putting the situation the other way round,
a society can be defined for our purposes as an organisation capable of
providing conventional competition: this, at least, appears to be its original,
most primitive function, which indeed survives more or less thinly veiled
~even in the civilised societies of man. The social organisation is originally
set up, therefore, to provide the feed-back for the homeostatic machine.
It might easily be assumed that male birds competing for territories (to
return to the same illustration) were the direct opposite of a society, being
all at enmity with each other; but this would be a completely false conclusion.
As Kalela (1954) first pointed out, they are in fact strongly coordinated
together, and often conspicuously coherent with non-random contiguous
territories—essentially a static flock with a higher ‘ individual distance ’
than we commonly associate with the use of that word. They sing in
emulation of each other with the primary object of being heard and recognised
by their rivals, particularly at the purely conventional singing-hours at dawm
and dusk (this convention is dealt with in Chapter 15); and they are in
personal even though militant contact with all their neighbours. This
two-faced property of brotherhood tempered with rivalry is absolutely
typical of social behaviour; both are essential to providing the setting in
which conventional competition can develop. The remaining essential
characteristic of a society, it may be repeated, is that it is concerned with
convention: indeed conventional competition and society are scarcely
distinguishable facets of a single phenomenon.
All the higher animals—especially the vertebrates and arthropods_have
evolved organised societies that come within our definition. The lower
down we go in the evolutionary scale the less conspicuous and elaborate does
the homeostatic machinery become; but we shall find later on that wherever

1.4

SOCIAL

ORGANISATION

15

secondary sexual characters have appeared there is almost sure to be a
recognisable conventional society (cf Chapter 12). The existence of such
societies is usually most clearly revealed, however, by the possession of
methods of mutual communication and recognition, through conventional
signs and signals given and received, which form an indispensible link in
social integration. In fact the best starting point for the more detailed
development of the theory has seemed to be a study of the nature and
occurrence of methods of social communication in animals; and to this
therefore the next several chapters will be devoted.
The reader is already aware that conventions and conventional behaviour,
with which we are so pre-eminently concerned, are in the nature of artefacts,
more or less widely divorced from the real sanctions by which their observance
is ultimately enforced. The more remote they are from absolute reality,
the more do they become symbols arbitrarily endowed with a meaning (just
as are the conventional signs that make up a printed page); and the more
they are likely to differ from ordinary ‘ real ’ adaptations by being somewhat
bizarre or even extravagant in nature. It is particularly desirable to keep
this in mind at this stage. The mode of their evolution, which may already
have begun to arouse questions in the reader’s mind, will be dealt with a little
later, in the next section of the chapter.
The actual regulation of population-density is largely a matter of exercising
control over recruitment and loss in the population. In some kinds of
animals there is a fairly continual dissipation of numbers through mortality,
whereas in others discrete generations may arise to succeed each other year
by year. In either case there is usually an annual breeding season, the basic
function of which is in the first case to make good the losses of the preceding
year, and in the second to create a new generation to succeed its progenitor.
We have already briefly referred to experiments which demonstrate with
great clarity that recruitment (at least in the species experimented with) is
density-dependent; and it is obvious therefore that the feed-back part of
the machinery will have to be especially active just before breeding commences,
in order to elicit the required response from the breeding stock and produce
the quota of recruits that current economic conditions dictate.
The same will be true whenever there is about to be some kind of selfinflicted loss to the population—for instance by the voluntary or forced
emigration of part or all of its members, as in migratory birds before they
leave their summer or winter quarters. Conventional competition is likely
to be conspicuous also if the dispersionary equilibrium is upset from outside,
for example by an unforeseen failure of the food-supply. As far as birdmigration is concerned, we are going to find evidence to suggest that all the
way along the route it is subject to automatic traffic control, so that orderly
millions of individuals can complete the successive hops of their long flights
without ever encountering dangerous traffic-jams and locally exhausted
fuel-resources (see Chapter 14, p. 292).
It is well known that the territorial convention in birds tends to be most

16

PRINCIPLES

OF

ANIMAL

DISPERSION

1.4

active and vigorous at a particular one of these special junctures in homeo¬
static regulation, namely just before mating and egg-laying take place; in
practice it often begins to build up some time before, and continues until the
process of reproduction is concluded, but the climax is most often reached
just when theory would predict it, immediately before breeding begins. At
this time conventional competition can assume a variety of forms. Not only
is there direct contest over the definition of territorial boundaries, but the
males usually spend a large amount of time in aggressive display, by spirited
singing or conspicuous flight or both. These displays are, of course, abstract
conventions, endowed with a meaning that broadly conveys the threat of
physical reprisals upon any rival that dares to dispute the occupant’s claim
to possession.
Symbolic displays are largely wasted if there is no audience to receive
and interpret them; and because of the ordinary necessities of life such as
feeding and sleeping they cannot often be indulged in all day long. The
conventional, social nature of bird song is in no way more clearly revealed
than by the almost universal tendency for all the individuals concerned to
do it together, at times when, for a brief convenient period, it becomes the
dominant communal activity. This usually occurs at the two periods of
day when synchronisation is easiest, namely when the rate of change of
daylight is most rapid, at dawn and dusk. Then for some minutes at least
there is generally a chorus in which all the rival members of the local society
take part. Nothing could be more perfectly adapted to indicate the
population-density than such a synchronised vocal display.
Specially-timed communal displays (Chapter 15) occur in every group of
the higher animals—in the dancing of gnats and midges, the milling of
whirligig-beetles, the manoeuvres of birds and bats at roosting-time, the
choruses of birds, bats, frogs, fish, insects and shrimps,—even, if we are not
mistaken, in the vertical migration and surface-assembly of innumerable
species in the plankton (Chapter 16). They are very commonly synchronised
at dawn or dusk or both; but in fact they need not in certain circumstances
(as when the participants have little else to do, or when feeding is a communal
activity anyway) be confined to special hours at all. They form, as a class
a tremendously important and hitherto completely unexplainable component
of social behaviour. It is essential to have a term to designate them, and we
shall call them epideictic displays—signifying literally ‘ meant for display ’,
but connoting in its original Greek form the presenting of a sample. Though
rather a rare word in its previous technical senses it already has a place in
many English dictionaries.
Epideictic displays are especially evolved to provide the necessary feed¬
back when the balance of population is about to be restored, or may need
to be shifted, either as a seasonal routine or as an emergency measure
They generally involve conventions that have evolved away from the direct
primitive contest, which is liable to end in bitter bloodshed and even killing
of participants, and have come to assume a highly symbolic quality, not even

1.4

SOCIAL

ORGANISATION

17

directly implying threat in many cases, but producing a state of excitation
and tension closely reflecting the size and impressiveness of the display. In
many cases not only is a special time of day set aside, but also a traditional
place, to which all the participants resort for the purpose; and this is
undoubtedly the underlying cause of almost all types of communal roosting
and hibernation, and many other gregarious manifestations among normally
solitary animals {see Chapters 11, 14 and 19).
The study of epideictic phenomena leads inevitably to a reappraisal of
the epigamic displays that characterise the marital relations of the sexes and
typically culminate in fertilisation. We shall find that much that has been
regarded hitherto as epigamic is unquestionably epideictic. In the prenuptial
period, when dispersion is often organised on the basis of the mated pair,
it is normal for the male to assume the whole dispersionary task, and to
participate exclusively in epideictic displays as the representative of a mated
pair (and sometimes at a later stage of a whole family party). The male is
then the epideictic sex, and many of his conventional displays and adornments
have their primary significance for other males with whom he is in epideictic
competition, and only quite secondarily, if at all, do they affect or concern
the female. Considerable attention will be given, in Chapters 12 and 21
especially, to unravelling the epigamic and epideictic aspects of sexual
displays and the functions and evolution of sexual adornment, dimorphism,
polygamy and the like.
Situations of this kind quite commonly arise, in which conventional
competition concerns only one section of the population—the larvae, for
instance, or the breeding adults as a group, or one sex alone; and in that
case the particular category concerned is generally distinctively recognisable
as a separate social caste. A special development of this has apparently
led to the phase systems of locusts and some other species, in which the
migratory phase is uniquely involved in the intensive epideictic displays that
culminate in swarming and exodus flights. A discussion of this subject
forms part of Chapter 20; and it is in the same chapter that consideration
is given to the well-known fluctuations in numbers and density that occur in
many populations, seeming as they do to negate any sweeping assumption
of a universal evolutionary tendency towards homeostasis.
It is equally possible to discover an opposite situation, in which, instead
of having a single species divided into a number of distinctive social castes,
two or more species for the time being merge their identities and collaborate
in a common dispersionary and social system. A good example of this is
found in certain groups of weaver-birds (Ploceidae) in Africa, in which at
breeding time all the males assume a highly specific livery and each species
is more or less independently dispersed; but during the rest of the year
neither sex nor species is readily recognisable, and mixed flocks are formed
that share a common life. Sharing food-resources with other species is very
common in nature; and it is obviously of no avail for one species to evolve
a conventional code for the conservation of its resources if this is freely

18

PRINCIPLES

OF

ANIMAL

DISPERSION

1.5

violated by competing species. To an important extent in some groups, in
consequence, interspecific conventions have been built up, and a mutual
‘ understanding ’ reached between competitors for the same resource, just
as normally occurs at the intraspecific level. These are refinements of extra¬
ordinary interest from the evolutionary point of view, perhaps confined to
the highest groups but appearing here and there in birds, mammals and other
vertebrates, and in insects such as the social Hymenoptera. They present
a radical antithesis to Gause’s hypothesis that animals of similar ecology
cannot survive together in the same habitat.
For the most part, of course, the book like the theory it presents is
concerned with economic affairs within the species, at the population level,
but Chapters 17 and 18 are devoted to the kind of interspecific relationship
just mentioned and some others like mimicry, where the mingling of models
and mimics may make the assessment of the population-density of either
species alone exceedingly difficult; and in this situation it will be remembered
that there are quite a number of mimetic Lepidoptera in which the males—
no doubt as usual the epideictic sex—are non-mimetic and retain their
visually distinctive specific characters.
1.5. Social evolution and group-selection
It is part of our Darwinian heritage to accept the view that natural
selection operates largely or entirely at two levels, discriminating on the one
hand in favour of individuals that are better adapted and consequently leave
more surviving progeny than their fellows; and on the other hand between
one species and another where their interests overlap and conflict, and where
one proves more efficient in making a living than the other. Selection at the
individual level is often designated as intraspecific, and that at the higher
level interspecific. The latter covers a broad range of relationships; it is
frequently concerned not so much with ecological overlaps between closely
allied species in the same genus as with the mutually conflicting needs of two
independent predators seeking the same prey, or two unrelated contestants
for the same micro-habitat.
Neither of these two categories of selection would be at all effective in
eliciting the kind of social adaptations that concern us here. We have met
already with a number of situations—and shall later meet many more—in
which the interests of the individual are actually submerged or subordinated
to those of the community as a whole. The social hierarchy, or ‘ peckorder ’ and its equivalent, is a common and important product of conventional
competition, and its function is to differentiate automatically, whenever
such a situation arises, between the haves and the have-nots {see Chapter 8).
For those high enough in the scale the rewards—space, food, mates—are
forthcoming; but when food, for instance, is already being exploited up to
the optimum level, the surplus individuals must abide by the conventional
code and not remain to contest the issue if necessary to the death. It is in
the interests of survival of the stock and the species that this should be so,

1.5

SOCIAL

EVOLUTION

AND

GROUP

SELECTION

19

but it ruthlessly suppresses the temporary interests of the rejected individual,
who may be condemned to starve while food still abounds.
The same applies to conventions in general. They are usually in fact
extremely important adaptations, potentially of great survival value to the
species: but it takes a group of individuals to maintain a convention, and if
the individuals are isolated from one another it falls to the ground. Human
conventions as we know are characteristic of a place, a people, a creed, a
profession; and in general the characteristics and functions of a society
belong to it collectively, and cannot be completely represented or discharged
by any solitary member.
There are a great many important characters of this kind, not only in
animals but in plants also, that are in the nature of collective attributes, all
possessing the common quality of contributing to the welfare and survival
of the group as such, and when necessary subordinating the interests of the
individual. One of these is the reproductive rate. If intraspecific selection
was all in favour of the individual, there would be an overwhelming premium
on higher and ever higher individual fecundity, provided it resulted in a
greater posterity than one’s fellows. Manifestly this does not happen in
practice; in fact the reproductive rate in many species, and recruitment of
adults in others, is varied according to the current needs of the population.
Restrictions can be imposed in a number of ways, for instance by each female
laying fewer eggs, or eating more of them, or by allowing fewer females to
breed—all adaptations that cut directly across the interests of prolificity in
the individual. In the extreme case, in various social insects, it has been
possible to evolve castes of sterile individuals, something that is inconceivable
in a world where the most successfully fecund were bound to be individually
favoured by selection and the infertile condemned to extinction; it could
only have evolved where selection had promoted the interests of the social
group, as an evolutionary unit in its own right.
It has become increasingly clear in recent years, not only that animal
(and plant) species tend to be grouped into more or less isolated populations,
due very largely to the physical discontinuities of the habitat, but that this
is a very important feature from an evolutionary standpoint in the pattern
of their distribution (cf. Sewall Wright, 1938; Dobzhansky, 1941, p. 166 et
seq.; Carter, 1951, p. 142). A great weight of evidence will be forthcoming
in the course of the book to confirm this conclusion. At least in terrestrial
habitats the food-resources on which animals depend are of a strictly
localised, immobile character, depending ultimately on a stationary or very
slowly changing pattern of vegetation. The local stock of any given
animal species, exploiting its resources, consequently tends to adopt many
conventions of a strictly localised or topographical character for example
the traditional sites of breeding places. Other conventions rely equally
strongly on a procession of mutual relationships among the individual local
inhabitants. Above all, the local stock conserves its resources and thereby
safeguards the future survival of its descendents; and no such conservational

20

PRINCIPLES

OF

ANIMAL

DISPERSION

1.5

adaptation could have evolved if the descendents did not normally fall heirs
to the same ground. Thrifty exploitation today for the benefit of some
randomly chosen and possibly prodigal generation of strangers tomorrow
would make slow headway under natural selection.
Thus it is clearly of the greatest importance in the long-term exploitation
of resources that local populations should be self-perpetuating. If confirma¬
tion were needed of this conclusion, it could be found in the almost incredible
faculties of precise navigation developed in all long-distance two-way migrants
whether they are birds, bats, fish or insects, to enable them to enjoy the
advantages of two worlds, and still retain their life-long membership in the
same select local stock. Ideally, localisation does not entail complete
reproductive isolation however; we shall have to consider later the pioneering
element also—in most species relatively small—that looks after colonisation
and disseminates genes. These are matters to be discussed in Chapter 20.
Strict localisation endows each local population with potential immor¬
tality. The population can therefore undergo adaptation as such, in the
same way as the subspecies, the species, or any other group. The extent of
its differentiation from other populations will be limited by the amount of
pioneering and interchange of membership that occurs with neighbouring
populations; but we shall see that local conventions and traditions tend to
be long-perpetuated, and that social organisations, with their collective
conventions, are sometimes extremely persistent.
Some such local groups may in practice maintain their identity for centuries,
and even, as we shall see later, for thousands of years. Others are not so
fortunate and suffer a constant turnover of colonisation and extinction,
especially in the less permanent types of habitat. None of them last for
ever in the same place, on account of the secular changes in climate and
geology. Survival is the supreme prize in evolution; and there is con¬
sequently great scope for selection between local groups or nuclei, in the
same way as there is between allied races or species. Some prove to be better
adapted socially and individually than others, and tend to outlive them, and
sooner or later to spread and multiply by colonising the ground vacated by
less successful neighbouring communities.
Evolution at this level can be ascribed, therefore, to what is here termed
group-selection—still an intraspecific process, and, for everything concerning
population dynamics, much more important than selection at the individual
level. The latter is concerned with the physiology and attainments of the
individual as such, the former with the viability and survival of the stock or
the race as a whole. Where the two conflict, as they do when the short-term
advantage of the individual undermines the future safety of the race
group-selection is bound to win, because the race will suffer and decline, and
be supplanted by another in which antisocial advancement of the individual
is more rigidly inhibited. In our own lives, of course, we recognise the
conflict as a moral issue, and the counterpart of this must exist in all social
animals {see p. 131).

1.6

CARR-SAUNDERS

‘OPTIMUM

NUMBER’

21

Group-selection is not by any means a new concept, though it has never
been accorded the general recognition that its importance deserves. Parti¬
cular attention has been paid to it in the present work, because it is funda¬
mental to the dispersion hypothesis; the principal allusions to it in the
chapters that follow have all been assembled in the index for ease of general
reference. As being the only possible method of evolving sterile castes in
social insects it has been recognised by Sturtevant (1938, p. 75) and O. W.
Richards (1953, pp. 145-6); and the general action of natural selection on
integrated social units as such has been explicitly pointed out by Allee (1940).
References to the same general ideas are quite widely scattered in the
literature (cf v. Haartman, 1955, with respect to the evolution of clutch-size
in birds); and, as described in the following section, they formed an essential
part of Carr-Saunders’ principle of the Optimum Number.
1.6. Carr-Saunders' principle of the Optimum Number
It is probably plain from the context that, after reaching the end of the
book, the author turned back to this chapter and completely revised it. In
the original draft there was no sixth section, because it was only relatively late
in the day that I discovered that, albeit in a rather special and restricted
context, the theory of dispersion through conventional behaviour had actually
been published before.
This was in fact a very rewarding discovery. In its previous form it had
been conceived as applying to man—and indeed primitive man—alone;
and its author, Sir Alexander Carr-Saunders (1922), clearly perceiving that
it depended on social evolution through group-selection, had assumed its
origin as an evolutionary process to date from roughly the lower Palaeolithic,
at which time, in his view, ‘ we should look for the origin of social organiz¬
ation ’ (loc. cit., p. 239). Had he realised that social organisation goes back
probably to the lower Cambrian, and is adumbrated in the Protozoa and in
the plant kingdom, he could as easily have extended the principle as we have
done here.
In this pioneer work, The Population Problem, Carr-Saunders was at great
pains to show first that unrestricted nomadism did not exist in primitive
races, but that all of them were without exception essentially territorial;
so-called nomads simply travelled about within their tribal areas instead of
having permanent settlements (i.e. they all tended to be locally self-per¬
petuating): next that they all limited their fertility by a variety of practices,
the most important of which are abstention from intercourse, abortion and
infanticide. Later (pp. 200, 213) he comes to the point that every population
has an optimum number, or an optimum density, that enables the greatest
income per head to be earned; and above this density returns diminish.
He sees that it is fatal to depend on starvation to eliminate the surplus,
because it inevitably leads to social instability and to making useless all the
special skills in food-getting on which the healthy population depends (p. 214).
In this the wishes of the parents may have to be overridden in the interests

22

PRINCIPLES

OF

ANIMAL

DISPERSION

1.6

of the community; and strict conformity with social practices is enforced
by social pressure (p. 216).
Finally, this cannot be brought about without group-selection. ‘ Those
groups practising the most advantageous customs will have an advantage in
the constant struggle between adjacent groups ’ (p. 223); but group-evolution,
and the optimum regulation of numbers in this way ‘ are clearly only appli¬
cable to races among whom social organization has become established ’
(p. 239).
There the matter rests. The several foundations of the present theory
are all comprehended. Nothing could have given me greater reassurance
than the knowledge that so distinguished a student had earlier pioneered
the road. A fuller discussion will be found in Chapter 21 (p. 493).

Chapter 2

The integration of social groups by visible signals
2.1. Signals used for social integration and other purposes, p. 23.
2.2. Visual signals and perception, p. 26. 2.3. Bioluminescence and visual
signalling, p. 35. 2.4. Summary, p. 39.

2.1. Signals used for social integration and other purposes
A social organisation that is evolved to provide the basis for conventional
competition presupposes that its members will be able first of all to recognise
one another, and then to comprehend the bond of interest that unites them—
a bond that arises from seeking the same conventional rewards, and turns
under stress into rivalry. Recognition and comprehension imply communi¬
cation, and this is a two-way process of transmitting and receiving, of signal
and perception. The signals may evoke one or sometimes more than one of
the senses, commonly either visual, auditory, chemical or tactile. They tend
to be conventional signs, for the most part arbitrary and without directly
intelligible reference to the response they elicit in the percipient. To take
an example not connected with dispersionary behaviour, nestling birds may
beg for food by stretching their necks and opening their mouths as wide as
possible, or in some species by gripping the parent's bill, their signals in
either case being closely related to the response released; but they may
equally well solicit food by making vocal sounds that are purely symbolic,
with a meaning understood only by an established convention, and not
inherent in the sounds themselves. The same applies to the warning colourpatterns of black with yellow or red that occur in insects, amphibia and snakes,
for instance; the meaning of these, if common opinion is not mistaken,
relates conventionally to the distasteful or venomous qualities of the bearers.
The use of signals for social purposes naturally demands the possession
of sense-organs capable of perceiving them. Animal groups containing
many highly-coloured species are in some cases—among the insects, bony
fishes and birds, for example—known to possess colour-vision; and similar
correlations often hold in respect of other senses, as in the amphibia where
sound production and the development of the tympanic membrane go to¬
gether, or in the mammals where the multiplicity of scent glands coincides
with a highly developed olfactory sense. But there are, of course, many other
animals with brilliant colour patterns which they themselves apparently
cannot perceive, such as sea-anemones, pennatulids, corals, tubicolous and
other polychaete worms, and echinoderms; so that it is always necessary

24

INTEGRATION

BY

VISIBLE

SIGNALS

2.1

when considering the part that any of these potential signals could play in
social communication to take account of their reception as well as their
emission.
It is noteworthy that the types of signal that can be instantly perceived
at a distance are on the whole commoner, and probably therefore more
effective under most conditions, than those that rely on chemo-reception
and the tactile sense.
Considerably greater attention has been given to studying the development
and uses of colours, patterns and visual signals generally than to those
involving any of the other senses; but in all cases the functions to be dis¬
charged in communication are essentially the same, and it is not particularly
difficult to devise an analytical scheme applicable to signals in general along
the same sort of lines as the familiar one drawn up by Poulton (1890) for the
uses of animal colours.
From our point of view the primary subdivision in the functions of
communication should clearly fall betw-een signals directed at other members
of the same species and those (like warning colours) directed at other kinds
of animals; these divisions are respectively intra- and inter-specific. It is
naturally with the former that we are generally concerned in the study of
social integration. Intraspecific signals are obviously varied in purpose;
those that are essentially social in function are capable of expressing a wide
range of intentions, needs, moods and emotions that could if necessary be
systematically analysed. They include expressions of threat, warning, fear,
pain, hunger, and—at least in the highest animals—such elemental feelings
as defiance, well-being, superiority, elation, excitement, friendliness, sub¬
mission, dejection, and solicitude.
Poulton’s classification was hardly
concerned with these at all, and out of Ins various categories of coloration
that can reasonably be regarded as signals, only the ‘ episematic ’ (recognition
marks) and ‘ epigamic ’ (courting colours) belong to our intraspecific division.
It suits our purpose to continue to recognise epigamic signals, which can with
advantage be restricted to communications between members of opposite
sexes concerned more or less directly with mating and marital life. For the
rest, however, it would lead us unnecessarily deep into ethology to attempt
any detailed analysis of the purposes of social and familial communications
at this stage. Some of the special purposes will emerge as we go along, but
for the present all intraspecific signals that are not epigamic can be uncritically
lumped togethei as social, since there appear to be none which this term cannot
reasonably be stretched to cover.
/nto'-specific signals, on the other hand, have no primary message for
members of the signaller’s own species, and are not directly concerned there¬
fore with the subject of dispersion; but it is desirable to consider them briefly,
in order to be clear later on regarding the different purposes for which signals
are emitted. They seem to fall into four main categories, not rigidly
separated, including (a) warning or ‘ keep-off’ signals, intended to save any
assailant or disturber-of-the-peace from the formidable consequences of

2.1

SIGNALS

USED

FOR

SOCIAL

INTEGRATION

25

molesting an animal protected by unpleasant taste, sting or odour- (b)
intimidating signals, intended to frighten or shock the assailant into leaving
the owner alone, or giving the latter a moment’s pause in which to effect its
escape, to this Poulton s ‘ pseudaposematic ’ category corresponds in partand (c) decoying signals, intended to draw the approach or rivet the attention
of a member of another species, either to enable it to be seized for food (as
in the angler-fishes belonging to the order Pediculati), or to divert it from
attacking the eggs, the defenceless offspring, or the vital parts of the signaller
itself. A fourth category (d) has developed from the fact that it is greatly
to the advantage of most animals to suppress entirely the emission of signals
most of the time, in order that their presence and identity may be overlooked
by other animals; and to this end some animals have not merely adopted
negative or cryptic colours and attitudes that enable them to disappear into
their surroundings, but they give out actively dissembling or masking signals,
which cause them, for example, to be mistaken by predators for unpleasant
species when they are not, as in the case of Batesian mimics, or by their prey
for innocuous species when they are in fact dangerous, as in the case of some
ant-like spiders (for fuller details, see Cott, 1940; Tinbergen, 1953, Chap. 6).
Animals that are permanently protected by their nauseating taste, poison
and defensive weapons may emit continuous warning signals, through their
conspicuous permanent colour patterns, which are effective to visual predators
whenever there is light enough for them to be seen. But the majority of
signals made either for intra- or inter-specific communication are produced
ad hoc in the appropriate context. This almost always applies to categories
(b) and (c) in the preceding paragraph. Shock signals in particular owe
their effect to the element of surprise; thus the cryptically-coloured eyed
hawk-moth (Smerinthus ocellatus), resting on the bark of a tree-trunk, when
touched by a sharp object such as a bird’s bill, suddenly spreads its fore¬
wings to reveal a pair of dark staring ‘ eyes ’ on the hind-wings, which are
waved slowly back and forth (Tinbergen, loc. cit., p. 94). The pistol-shrimp
(Alphaeus califomiensis) is supposed for similar reasons to emit its unexpected
loud detonation (MacGinitie, 1949, p. 276), and the tube-dwelling polychaete
worm Chaetopterus to turn on its floodlights (Nicol, 1952, p. 429). Decoy
signals may similarly require special postures and movements, such as the
‘ broken-wing act ’ or injury-feigning which is widely developed among birds.
There is not always any certain criterion for distinguishing inter- from
intra-specific signals: warning and threat, for example, are common to both,
and may use the same conventions in each case, although the ulterior object
is totally different. Usually, however, it is the circumstances that make it
clear to what kind of audience the signals are being directed.
It is generally recognised that, instead of the visual colours and patterns
so often evolved for communication, any of the other media for signalling
can be effectively substituted. Thus some species such as rattlesnakes
(Crotalus) apparently make use of warning sounds, and some of frightening
sounds, like the pistol-shrimps already mentioned or the common dor-beetle

INTEGRATION

26

BY

VISIBLE

2.2

SIGNALS

(Geotrupes) which crepitates when seized; others use intimidating or
repellant odours, as do the devil’s coach-horse (Staphylinus) and the groundbeetle (Carabus), or the caterpillars of swallowtail butterflies (Papilio) which
have a foul-smelling forked horn or osmeterium that can be conspicuously
everted during the threat display.
Table I
Summary analysis of animal signals

1. Intra-specific
a. social
b. sexual
2. Inter-specific
a. warning
b. intimidating
c. decoying
d. dissembling or
masking

Poulton's nearest
equivalent

Directed at

Category of signal

other members of own species
or social group
other sex

predators and other assailants
ff

ff

ff

If

either prey or predators
ff

ff

ff

ff

episematic
epigamic

aposematic
(none)
pseudepisematic
pseudaposematic

The various media of communication—visual, acoustic, electrical,
chemical and tactile—are considered in turn in the following chapters: but
it should be noted that the more highly organised animals, including the
insects and vertebrates, seldom rely on communications of a single sensory
type. They usually combine the use of alternative methods, for example
visual and acoustic, acoustic and olfactory, or even three or four of them,
although one often tends to predominate.
2.2. Visual signals and perception
Signals intended for visual perception depend either on reflected light, in
that case being effective only in sufficiently well-lighted environments, and
consequently commonest in diurnal animals; or on bioluminescence,
effective only in the dark. The animal must in either case be living in a
medium transparent to light, and not underground, in dense vegetation, or
in turbid water, where sound or scent is likely to be a more effective means of
making contact. Visual methods, however, tend on the whole to be favoured
wherever possible above all other means of communication; for not only
can animals display infinitely varied patterns, but these can be rendered
especially conspicuous, and their meaning altered or emphasised, by
movements associated with their exhibition.
Social integration, as already mentioned, requires the recognition of
other members of the group and species; and among animals constantly in

2.2

VISUAL

SIGNALS

AND

PERCEPTION

27

motion each recognition-contact may depend more on the characteristic
general visual appearance of the moving object than on any particular mark
or pattern. The movement itself may supply the chief component of the
signal. In the Introduction (p. 1), for example, the dispersion of pelagic
birds was shown to be closely correlated with the abundance of plankton in
the surface waters of the ocean. In such a population of constantly moving
members, density adjustment must be a continuous process, and it would
appear likely to demand the frequent attention of each particular species of
bird concerned, in order that they could govern as efficiently as they do their
dispersion in relation to the food available. The most abundant and charac¬
teristic oceanic birds, the order Tubinares or petrels, appear to be largely
nocturnal in their feeding habits, taking their food when it becomes accessible
to them at the surface during the hours of darkness, as a result of its diurnal
vertical migration. They are nevertheless for the most part solitary rather
than gregarious, and silent, at sea; and their olfactory powers, while perhaps
greater than in most birds, are almost certainly insufficient for maintaining
contact among thinly-scattered populations such as theirs. Social integration
can therefore fairly safely be assumed to be visual; and this accords well
with the remarkable (and not otherwise explained) amount of time spent by
albatrosses, shearwaters and petrels, in apparently aimless cruising in the
daylight hours. Numerous voyagers have noted their tireless flight and
ceaseless rise and fall, as they roam the waves hour after hour; they alight
rather seldom for any purpose (unless in the wake of a ship), and generally
appear not to be seriously engaged in feeding. The function of this diurnal
flight could well be ‘ dispersive ’, providing the individual with a frequency
sample of visual recognition contacts, which could be integrated with the
current abundance of food, and result in compensatory movements and the
adjustment of density. Without some such adaptive mechanism, and on a
basis merely of uncoordinated chance alone, it would clearly be impossible
to achieve the high correlation between density and food-supply that has
been shown to exist.
This type of population density determination, the method of samphng
by travel and encounter, is perhaps rather general in free-moving animals
living in a continuous medium, especially those in, or on the surface of,
large bodies of water. Individuals cannot return to, or defend as their own,
a fixed position in an unlimited uniform fluid environment, because they have
no access to permanent solid points of reference. Either they can each
travel independently, achieving the appropriate average density through their
variable degree of tolerance or intolerance of the presence of others; oi else
all those in a given volume or area of water can unite into a school or flock,
and, if desired, maintain an ordered spatial pattern with respect to one another
as they travel in a body, in due course encountering and reacting to other
similar bodies. This amounts to the same thing, only the cruising unit is not
an individual but a gregarious group. The two systems are not mutually
exclusive and can easily be intermixed. Social integration must be postulated

28

INTEGRATION

BY

VISIBLE

SIGNALS

2.2

as occurring in all cases through mutual recognition and response; but of
course it need not be maintained visually. Examples will be given later of
tactile, acoustic and lower-frequency vibration signals employed for the same
purpose, and it is likely that chemical signals (including the so-called
‘ extrametabolites ’) are of importance also in these environments.
Travel and encounter can be used by very mobile animals in any medium
when they are free to range over an undivided territory or habitat. Butterflies,
for instance, may avail themselves of it, through their bright colours and
patterns, which are often concealed except during their conspicuous and
leisurely flight. The frequency of contacts will depend on three factors—
the population-density, the speed of movement and the maximum range of
recognition; so that the slower an animal moves the less frequent contacts
it will make; and there is a tendency among slower-moving species, such as
terrestrial mammals, or silk-spinning insects and spiders, or snails, for
example, to leave a persistent chemical or tactile signal wherever they go,
which must greatly increase the frequency of ‘ contacts ’; but further
consideration of this must be deferred to the appropriate place (p. 98). It
is sufficient to point out that inverse correlations would be expected, in an
unlimited fluid medium, between the population-density, the cruising speed,
and the maximum range of recognition; that is to say that slow-moving
animals must remain close together, or else have far-carrying or persistent
signals, in order to maintain integration, whereas those which travel at higher
speed can be spaced at greater intervals without losing the necessary contact.
Visual recognition within the species appears to be made possible in the
great majority of cases by the display of conspicuous patterns and colours.
Where there is nothing to be lost by continuous advertisement, as in large
birds like swans, storks, albatrosses or gannets, which can protect themselves
adequately from predators when adult, and need no concealment in hunting,
we find both sexes with their plumage as conspicuous as possible. Probably
on account of the high degree of efficacy of their visual advertisement, a
number of species in these families have dispensed almost entirely with
alternative methods of recognition-signalling; and, unlike the majority of
birds which as a group are the most vocal of animals, they tend to be silent,
except in sexual or social-dominance displays at close quarters (e.g. Cygnus
olor, Ciconia ciconia, Sula bassana and many other Steganopodes, almost all
Tubinares, etc.). The same limitation applies equally to some of the great
raptorial birds, which render themselves highly conspicuous by soaring hour
after hour; a good many are entirely silent in flight, and some may be virtu¬
ally mute at all times (including the California condor Gymnogyps californianus, certain eagles, e.g. Aquila chrysaetos, Haliaetus albicilla, and especially
vultures, e.g. Aegypius monachus, Gyps spp., Pseudogyps spp., Torgos tracheliotus, Neophron percnopterus, etc.).
The visual recognition signals typical of most birds are of much value to
ornithologists as a means of species identification, and even where there is
considerable sexual dimorphism they may be equally developed in both

2.2

VISUAL

SIGNALS

AND

PERCEPTION

29

sexes and even in the immature stages. Examples of this are the specula of
the surface-feeding ducks, produced by lustrous, coloured secondaries in the
wings; or the wing-bars and coloured rumps of various finches, such as the
pine grosbeak (Pinicola enucleator), white-winged crossbill (Loxia Ieucoptera),
chaffinch (Fringilla coelebs) and brambling (F. montifringilla). It is very
common for visual recognition signals to appear only when the bird moves
or flies or otherwise deliberately exposes them. As long as it remains still
its identity is concealed visually, alike from predators and members of its
own kind. When it moves it becomes a relatively conspicuous object anyway,
and this conspicuousness need not always be much increased by the
recognition-label. Identity-marks include the white and coloured rumps on
many passerines including the finches just mentioned; special wing-patterns,
for instance in the lapwing (Vanellus vanellus) and many other Limicolae;
white, coloured or patterned tail-feathers, as in the pipits (Anthus spp.),
wagtails (Motacilla spp.), redstarts (Phoenicurus) and related forms. The
number of examples which could be given is almost unlimited, because
visual recognition signals are found in almost all birds.
Similar visual signals occur in most other groups of animals with highly
developed eyes, and it is unnecessary to give more than representative
examples. The combination of cryptic or concealing coloration in the
motionless or resting animal, and recognition marks that appear as soon as
it moves, just noted in birds, is exceedingly common. In orthopteran insects
it is to be seen among those widely distributed grasshoppers that have their
hind-wings resplendent with black and yellow, red or blue; very similar
patterns are found in phasmids and various Hemiptera. The Carolina locust
(Dissosteira Carolina) may suffice as an example. When stationary the insect
generally passes unnoticed. When it flies, it flashes the oversize hind-wings,
coloured black bordered with brilliant yellow, and at the same time often
emits a crackling sound. (This is not nearly so loud, however, as that of its
relative and compatriot the cracker-locust Cercotettix verruculatus, which,
with much less conspicuous yellowish-green hind-wings, instead relies
principally on the sound signals consisting of a series of loud clicks like a
burst from a police-rattle.) ‘ It is generally believed that so-called “ flash
colours ” serve to confuse or misdirect an enemy in the pursuit of prey. . . . ’
(Cott, 1940, p. 376). This is a theory that has been attributed originally
to C. Hart Merriam (cf Chapman, 1928, p. 96); and Cott himself has no
better alternative to suggest. But he says lower down: ‘ It must be admitted
that in the present state of our knowledge the precise biological meaning of
flash colours is not clearly understood.’
There are fairly convincing reasons for thinking that misdirection of
pursuers is at most a secondary effect. In D. Carolina none of the nymphal
stages would be so protected, since the coloured wings and the ability to
stridulate appear only in the adults. Flights are taken spontaneously, and
the insects may frequently be seen to hover in the air, apparently displaying
themselves. ‘ They rise at first about three or four feet making a light

30

INTEGRATION

BY

VISIBLE

SIGNALS

2.2

purring or beating sound and then, rising higher, change the motion of the
wings when a curious, sharp, seesawing sound is produced ’ (Townsend,
1891). ‘ The flight consists of two positions, one with a very fast wing-flutter
while hovering in the air. Then a change of position usually occurs and a
further, longer period of hovering during which the flutter of the wings is
at a distinctly slower rate ’ (Pierce, 1948, p. 258-259). Furthermore, red¬
headed woodpeckers (Melanerpes erythrocephalus) have been observed in
southern Ontario perching on poles, from which they were making short
flights to snatch flying insects; ‘ the most conspicuous insects taken were
adults of the Carolina locust. . . . The locusts would . . . hover in the air
with rapidly beating wings or dance up and down in flight above a particular
spot. . . . During this hovering flight the locusts were easily snapped up
by the foraging woodpecker.’ On another occasion two or three wood¬
peckers were seen similarly engaged, and ‘ in some cases the bird would
follow the locust down to the ground and then carry it to a pole before
devouring it ’ (Judd, 1956). The flash colour theory is clearly contradicted
in this case, since the woodpeckers succeeded in following the locusts even
when they ‘ disappeared ’ by folding their wings and dropping to the ground.
It seems probable that the locusts are actually performing song-flights,
and their dispersion is somewhat akin to territory in birds, as in the case of
many other Orthoptera. Male rivalry and fighting have actually been
observed in close association with the flights in this species, Dissosteira
Carolina (Snodgrass, 1925, p. 413) (see also p. 48). Professor Pierce of
Harvard, the physicist who undertook important pioneer work in analysing
the sounds produced by insects, has the following comment (1948, p. 262):
‘ These aerial evolutions in [Z>. Carolina] take place prominently at the height
of the mating season, and the flight... is a hovering over a nearly fixed spot
.... It is described by various entomologists as a “ sexual maneuver by
the male in the presence of one or more females on the ground.” I have not
been able to determine that there was a female present to admire the male’s
antics. In fact, it seemed to me equally plausible to assume that the flight
of the male takes place for the benefit of other males, rather than as an
effort to charm an admiring female.’ This detached observation by a scientist
unhampered by preconceived biological ideas should not pass unnoticed.
In the Lepidoptera visual recognition marks revealed only by flight or
deliberate movement are very common also, and often there could be no
serious contention that they are developed as ‘ flash colours ’ to confuse
enemies. More or less cryptic patterns on the undersides of the wings can
be found in most of those families of butterflies which fold the uppersides
together when at rest, and thus hide their bright signal-patterns; they are
extremely common in the Satyridae (graylings, browns, arguses, etc.),
Nymphalidae (fritillaries, tortoiseshells, etc.), Lycaenidae (hairstreaks,
coppers and blues) and Pieridae (whites and sulphurs). We may note here
in passing a North American satyrid, the pearly eye (Lethe portlandica) that
has been found, like the Carolina locust, to exhibit a type of territorial

2.2

VISUAL

SIGNALS

AND

PERCEPTION

31

behaviour; the ‘ males are quite likely to adopt a particular tree trunk as a
“ territory ”, return to it day after day and chase other butterflies away from
it. Combats between males are thus frequent ’ (Klots, 1951, p. 66).
In moths, the forewings usually conceal the underwings in the resting
insect; bright patterns of white, red or yellow offset with black, resembling
those of the Orthoptera, are found on the upper surface of the underwings
in a relatively small number of genera, including especially the well-known
handsome species of Catocala (Plusiidae), and various others among the
Caradrinidae, Brephidae, Arctiidae and Sphingidae. Rather few of these,
other than the tiger-moths (Arctiidae), are abroad voluntarily in daylight,
but some of them at least appear to have a well-synchronised period of flight
at dusk, whilst the light intensity is still adequate for sight-recognition.
Specific distinctions of colour and pattern are, as usual, the rule in these
genera, though Catocala contains several groups of species based on rather
minute differences.
Turning again to the vertebrates, it is well established that various teleost
fishes can distinguish colours, and that many have developed visual recog¬
nition marks. Leaving aside luminescent signals for the present, we find
visual patterns best developed in species in shallow water where illumination
reaches a high intensity. Most of the trout, such as Salmo trutta and Salvelinus
fontinalis, have brightly-coloured ocelli on the sides of the body. In the
North American freshwater family Centrarchidae, the numerous species of
sunfishes have developed a special ear-like flap on the operculum, covered
by pigmented integument and differing sharply in size and colour-combin¬
ations in the various species. Even in some apparently ‘ difficult ’ families,
such as the freshwater Cyprinidae or the freshwater and marine Gobiidae,
it usually does not take very long to learn to recognise the species at a glance
by external characters alone. As we shall see