Population Ecology: A Unified Study of Animals and Plants

CHAPTER 6 : POPULATION REGULATION 191

Fig. 6.9 Population fluctuations in a sand dune annual monitored by regular census over long periods of time;
Androsace septentrionalis: S, seed population; G, seeding; E, and on the other hand, deliberate alterations in
established plant; V, vegetative mature plants; F, flowering population size have been made to enable population
plants; R, fruiting plants. Breaks in horizontal axis response to density to be assessed.
represent the period from May to March (10 months). (Data
from Symonides, 1979.) Symonides (1979) monitored Androsace septentriona-
lis in sand dunes in Poland over an 8-year period and
food shortage and competition for territorial space examined the fluctuations in the numbers of individ-
andlor nest sites are the main causes when density- uals throughuot the life cycle (Fig. 6.9). The life cycle
dependence is recorded (see also section 6.10). Marine in Androsace is short; seedlings emerge in early March
mammals were included in the table by Sinclair (1989) and plants have flowered and set seed by late May.
who assumed that food supplies affecting fertility and Annually, and with considerable consistency during
the growth of juveniles were the predominant causes the period 1968-74, over 100000 seeds were pro-
of regulation. However, the number of studies is so duced per m2. Yet at fruit dispersal in May of each
small that no clear conclusions can be drawn. Even year, populations fell in the restricted range of 100-
less data are available for fish. 300 flowering plants per m2. Key-factor analysis (see
Silvertown, 1982 for details) revealed that the 'key'
It is clear that there are no generalizationsabout the cause of mortaIity was seed loss in the 10 months
causes of density-dependent mortality in vertebrates. intervening between seed shed in May and germina-
One of the main reasons for uncertainty is that most tion in the following March. The magnitude of this
studies have not included all the likely contenders. In loss, however, was independent of seed density. On
particular, predation is not considered in most studies average only 0.4OIo of the seed crop was present as
anld the effects of parasites and pathogens have hardly seedlings in early spring and whilst the exact causes of
begun to be considered in field studies. mortality are not known, this is a clear reflection of
the hostility of the sand dune environment for dor-
6.8 Populationregulationin plants mant Androsace seed. The only other significant period
of death during the life cycle was in the establishment
Generally two approaches have been taken to exam- of young plants from seedlings. Here density-
ining regulation in natural plant populations. On the dependence regulated the number of young plants in
one hand, populations fluctuating in size have been an undercompensatory fashion (Fig. 6.10). Such

192 PART 3: SYNTHESIS

Fig. 6.10 Effect of increasing density on seedling mortality. variation in rainfall frequency (as measured by days of
Slope of dashed line = 0.2 (r2= 0.58). Fitting equation 3.4, b recorded rainfall) was slight and, as Figs 6.1lb-d
(final slope)= 18.5 (r = 0.81). (Data from Symonides, 1979.) show, population regulation occurred both by in-
creased mortality and reduced fecundity in relation to
undercompensation is important to the persistence of density. The pattern of adult plant survivorship to
Androsace in sand dunes. We can see from Fig. 6.9 that flowering was found to be strongly dependent on
during the latter 5 years of Symonides' study, mortal- seedling density (the pattern of survivorship changing
from Deevey type 1through to type 3 with increasing
,ity of dormant seeds from primarily abiotic causes density) and the mean fecundity of individual plants
was also dependent on the density of survivors. The
increased (k,,,, rises), and in 1975, 98% of seeds next 2 years provided very differing habitat conditions
with respect to water, 1973 being wetter and 1974
were lost after dispersal and the population had dryer than average. This resource fluctuation was
declined to 36 plants per m2. Had density-dependence
at the seedling stage been stronger (e.g. exact compen- directly reflected in altered peak seedling populations,
sation) it is possible that local populations would have survivorship to flowering and mean plant fecundity
been very close to extinction. but even so the same processes of density-dependent
regulation were evident.
Regulation in this species, then, is substantial in the
juvenile stages (seed and seedling), although we might Both of these examples come from studies of natural
have expected regulation of seed number at the populations by continued census over long periods of
flowering stage as well. k-Value analysis of the data time. An alternative approach to exposing regulatory
revealed no evidence of this but only because the processes is to alter experimentally the density of
densities of flowering plants had already been re- natural plant populations. In an endeavour to expose
stricted. This illustrates the inherent limitation of all possible sources of density-dependent regulation in
simply taking population estimates by census and. the grass Vulpia fasiculata in sand dunes, Watkinson
seeking (albeit with a powerful tool) the causes of and Harper (1978) studied natural populations in
population regulation. If, for some reason, the range of which they deliberately established (by addition of
densities of flowering plants had been greater, density- seeds or removal of very young seedlings) a range of
dependent regulation of seed fecundity may well have densities from 100 to 8000 plants per 0.25 m2. This
been detected. species contrasts with Androsace in having a shorter
period between seed dissemination and germination;
Sand dune environments offer a particularly hostile seeds 'over-summer' from July through to October, a
habitat for plant growth not least with respect to water period in which there is usually less than 1% loss of
as is illustrated particularly clearly by another sand the annual seed crop. Mortality between seedling
dune annual Erophila vema, also studied by Symonides emergence and flowering varied between 7 and 41%
(1983). Over the period 1968-72, mean spring seed- but was not density-dependent.A range of abiotic and
ling abundance varied no more than 8% of the mean biotic factors were the cause of these deaths, including
over the 5 years (Fig. 6.lla). During this period, rabbit grazing, drought and seedling removal by wind
drag. This mortality was visited on populations up to
and including early flowering. Vulpia plants may bear
up to four seeds, but as Fig. 6.12a shows, the actual
numbers of seeds borne per plant was dependent on
the density of plants at flowering time. (Note, though,
that below 100 plants per 0.25 m2, seed number is
independent of density.) To examine the stabilizing
properties of this density-dependence, Watkinson and

CHAPTER 6: POPULATION REGULATION 193

Harper proposed a model in which the density of cally recorded since the beginning of the nineteenth
flowering plants, N,, was a function of the number of century (Watkinson & Harper, 1978).

-,seeds per unit area produced in the previous genera- The time at which density-independent thinning
takes place in the growth of a population of plants may
tion, S, and the probability, p, of an individual well be important: early thinning may offer a greater
surviving from seed production (at time t - 1)through period of time than late thinning for the increased
to maturity (at time t). Thus, growth of survivors compensating for density reduc-
tions. The extent of compensation, however, will
But, as Fig. 6.12a illustrates, the average number of depend on the time when the various components of
seeds borne per plant, S, was linearly related to yield are formed (section 2.5.2). Watkinson (1983)
flowering plant density N (above 100 plants per assessed the importance of this by varying the time
0.25 m2).Hence, during the life cycle when populations were thinned
by a half (a figure chosen arbitrarily). Figure 6 . 1 2 ~
(K and C are the constants describing the straight line gives a summary of these effects on the calculated
relationship) and the seed yield per unit area of these equilibrium population densities. Thinning during the
N plants is seed phase of the life cycle has no influence on
equilibrium density, since surviving plants compen-
Substituting this term with the appropriate subscripts sate by elevated seed production (Fig. 6.12a). How-
into equation 6.1 gives ever, as the time of thinning was successively delayed
(dotted line, Fig. 6.12c), equilibrium population densi-
a model describing the changes in the number of ties fell as the compensatory response became increas-
flowering plants from generation to generation. ingly muted. The individual yield components in
Vulpia-number of fertile tillers, number of spikelets
If density-independent mortality is constant from per flower and number of seeds per spikelet-were
one generation to another we may calculate an determined in accordance with the density perceived
at their formation, and in consequence compensatory
,equilibrium population density (N = N, = Nt - ,; N,l responses to density reductions were limited by the
yield components already formed. However, the ex-
Nt _. = l ) as tent to which this occurred was dependent on the
extent to which competition had occurred prior to the
when populations exceed 100 flowering plants per time of thinning. Plants sown at a later date (Novem-
0.25 m2. (Below this density, there is no density- ber, Fig. 6.12~)had not achieved a sufficient size to
dependent regulation and hence no equilibrium pop- interfere with each other's growth by January and
ulation size.) Figure 6.12b shows the predicted February, and thinning of populations at this time
equilibrium density for a range of p values and we can depressed population equilibria to a lesser extent than
see that it becomes increasingly sensitive to lowered in populations sown 1month earlier. We can appreci-
chances of survival. When p falls below 0.31, the ate, then, that the model developed earlier has the
population will decline; this is because the chance of generality to encompass all situations where popula-
an individual Vulpia plant replacing itself in the next tion regulation is achieved; yet it subsumes within it a
generation is less than 1. Pleasingly, the values of p density-dependent term which itself is determined by
measured for Vulpia in the dunes fell in the range the level of density-independentregulation.
0.34-0.59 supportingthe prediction of the model that
populations would persist in the dunes, a fact histori- This approach to assessing possible population equi-
libria may be used in exploring the reasons for the
abundance of a species along an environmentalgradi-
ent. Keddy (1981) sowed seeds of the annual sea

194 PART 3: SYNTHESIS

.W-

V1

-.0C-

W(0

a3
0a

8

.--C(0.'

CQT)

Class o-f-d-e-n-s.itIy

" 81420261 7131922284101622 " 8 1420 26 1 713 1922 28 4 10 1622
March I April , May
I March I April I May
Date Date
I I

Fig. 6.11 Population responses in Erophila vema over a Density-dependence was only noticeable in two sites
7-year period. (a) Peak population sizes given are seedlings and in each, at different stages in the life cycle
m-2 and the number of days of rainfall in the growing (Fig. 6.13a). In the landward site survivorshipof plants
season are circled; (b)survivorship curves for successive to fruiting was density-dependent whereas in the
cohorts in each year. seaward location fecundity decreased with density.
Characteristically, plants were large in the seaward
rocket (Cakile edentula) into a sand dune at three sites site where populations were largely monospecific in
(seaward, middle and landward) in Nova Scotia, Can- contrast to the much smaller plants at the landward
ada. A range of densities were sown at each site and edge of the dune in the presence of other vegetation.
plant mortality and fecundity were determined. Watkinson (1985) used the approach described in

CHAPTER 6 : POPULATION REGULATION 195

Fig. 6.11 (continued) (c)Seed production per plant according seed in the landward direction which is likely as a
to density class; (d)mean plant fecundityin relation to result of both wind and wave action (Keddy, 1981). It
density averaged over the 7 years. Density classes (numbers was found that allowing 10%of surviving seeds at the
0.01 m-3: I, 1-2; 11, 5-10; 111, 15-30; IV, 35-50, V, > 55. seaward site to migrate inland was sufficient to
(From Syrnonides, 1983.) maintain populations at all three sites and migration
levels of 50-80% produced a population distribution
section 3.2.3 to predict equilibrium population sizes with highest abundance in the middle of the dune.
that might be expected if density responses remained The suggestion is that plants exist at the middle and
constant over generations. Parameter estimates were landward sites only because of the high annual
obtained by fitting equation 3.7 to the observed data dispersal of seeds inland.
(Fig. 6.13b), assuming that self-thinning in popula-
tions was largely negligible, that plant mortality was In the species just discussed generations do not
density-independent (the inclusion of density- overlap, either as living plants or as seeds (there is no
dependence for the landward site did not qualitatively persistent seed bank)and individuals in the population
alter the conclusions) and that 60%of the seed crop in enter each stage of growth largely synchronously
each site were lost in the period after dispersal and owing to birth (germination) at very much the same
before seed germination. By iterating the model for time. However, this life history is but one-and
several generations it was found that populations perhaps the simplest-that plant species display. Many
would only persist on the seaward end of the gradient. species possess seed banks from which seedlings may
This is in direct contradiction to the observed relative be recruited over a protracted period of time. Episodic
abundance of sea rocket which was most abundant in germination from a bank of seeds will result in an
the middle of the gradient. The suggests that relative age-structured population which, as we have already
abundance cannot be explained simply in terms of seen (Chapter 3), requires a more sophisticated math-
survival and fecundity, assuming of course that the ematical description. Yet, despite this complexity we
assumptions of the model are correct. Watkinson can still unravel sources of population regulation in
therefore investigated the influence of migration of some instances by straightforward methods.

Figure 6.14 illustrates the density relationships

196 PART 3: SYNTHESIS

occurring within populations of the grass weed Avena
fatua (wild oat) infesting a crop of wheat. In the UK,
seed germination in this weed occurs in autumn
during crop sowing as well as in spring when the crop
is growing. Populations then become age-structured
according to the range of seed germination times.
Mature plants disseminate seed in mid-summer and
die before crop harvest. Manlove (1985)sowed A. fatua
over a wide range of seed densities on its own or in the
presence of a constant density of wheat plants. He
then tagged the wild oat plants as they emerged and
followed their survivorship through to seed produc-
tion. By plotting the recorded densities of seeds versus
seedlings and seedlings versus adults we can look: for
regulation at these two stages in the life cycle
(Fig. 6.14a & 6.14b). Density-dependent regulation is
suggested at the seedheedling stage: the scatter of
points in Fig. 6.14a is below the line of unit slope.
Once seedlings became established mortality during
growth to plant maturity was slight and not related to
seedling density either in the presence or absence of
the crop (Fig. 6.14b). Further experimental evidence
showed that regulation of seedling number was pri-
marily the result of seed loss from the soil surface in
late summer and early autumn, when seed predators
(birds and small mammals) foraged in the plant
stubbles after harvest. The intensity of this predation
increased in a density-dependent manner.

A second source of density-dependentregulation in
wild oat was in seed production per plant (Fig. 6.14~).
This was inversely related to mature plant density,
and it was also depressed eight-fold on average by
wheat, across the entire density range. We can exam-
ine the stabilizing properties of these two regulatory
phases by calculating the net reproductive rate, R, of
populations at different starting seed densities

Fig. 6.12 Population regulation in Vulpia fasiculata. See text Fig. 6.13 (Facing page) The influence of sowing density on
for details. (After Watkinson and Harper, 1978; Watkinson, fitness in Cakile edentula on a sea shore in Nova Scotia,
1983.) Canada. (a)Fitness components-the proportion of the
population surviving to reproduce; the n ~ a fnecundityof a

surviving plant; and the relative abundance of plants as
observed on the sea shore. (After Keddy, 1981.)(b) The
relationship between mean fecundity of a plant and the

density of reproducing plants. Lines have been fitted using
equation 3.7. (After Watkinson, 1985.)

C H A P T E R 6: POPULATION REGULATION 197

198 PART 3: SYNTHESIS

Fig. 6 14 Population regulation in Avena fatua in serves to reduce the net reproductive rate uniformly
monoculture and in a crop of wheat. (a) Density-dependent across the wild oat density range; and a density-
regulation of seedling emergence. (b)Density-independent independent control (herbicide) depresses these rates
survivorship of adult plants. (c) Density-dependent even further. We must also remember that the analy-
regulation of seed production. Solid lines are of unit slope. sis subsumesthe effect of age-structure in the wild oat
(Data from Manlove, 1985.) population. Whilst wild oats that emerge in the
autumn contribute a greater number of seeds per
(Fig. 6.15). R Declines monotonically with density in plant to the next generation than late emergers, tlhese
monoculture and in wheat (square symbols) down to differences are absorbed within the overall density
densities of c. 34 600 and 7240 seeds per m2 respec- response. Late emerging plants enter the hierarchy of
tively. These are equilibrium population densities-so exploitation(section 2.5.2) later in the growing season
long as all other factors remain constant from one
generation to the next. Figure 6.15 also demonstrates and are proportionately disadvantaged for doing so.
the effect of a density-independent control measure-a Nevertheless, even individuals lowest in the size
herbicide selective against wild oat which was applied hierarchy usually contributed one or two seeds to1 the
in early summer before flowering. Its mode of action is next generation.
to cause flower abortion in the weed (and hence seed
loss) rather than plant mortality. In consequence, Part of this study also involved an examination of
reproductive rates are depressed, the line moves the dynamics of buried seed populations. Loss of seeds
towards the origin, and the combination of crop in the soil (through seedling germination and death)
competition and herbicide reduced the equilibrium occurred at a constant rate (a type 2 survivorship
population density of wild oat to 470 seed per m2. curve) regardless of density. Slightly more than half
(0.55) of the seed population in the soil survived from
This example, whilst supportive of what we already one generation to the next. We can easily appreciate
know regarding population regulation, illustrates par- the role of the seed bank in maintenance of plant
ticularly clearly the various roles of diierent compo- populations by reconsidering the wild oats growing in
nents of regulation. The underlying cause of wheat and sprayed with herbicide. Populations arising
regulation is intraspecific. This arises in part from from seed densities greater than the 470 seeds per. m2
density-dependent seed predation and in part from had net reproductive rates less than l-competition
intraspecificcompetition determiningthe seed yield of and herbicide generally rendering plants at these
mature plants. Interspecific competition (from wheat) densities barren. Yet the persistence of dormant seeds

CHAPTER 6: POPULATION REGULATION 199

Fig. 6.15 Density-dependent and density-independent ,reproductive-rate is then SJS, - + 0.55. 0.55 is added as
regulation in Avena fatua. In a generation the number of
seeds produced (S3 is the arithmetic product of: the density this is the fractionof dormant seeds surviving over
of seeds in the soil in the autumn S, - ,; the probability of generations. The dashed line indicates the (constant) rate of

seeding emergence (see Fig. 6.14a);the probability of decline of the buried seed population; the solid line shows
seedling survival (Fig. 6.14b); and the number of seeds the rate of increase required to keep a population at
equilibrium. The slopes of the lines are not significantly
produced per plant (Fig. 6.14~)T. he rate of increase or net different from one another. (Data from Manlove, 1985.)

over generations enables recruitment of individuals in and trampling of grazing animals. In grassland, estab-
the succeeding generation and a rapid return to the lishment of new plants from seed in both species is
equilibrium density. rare. Yet as Fig. 6.16 shows, natural populations of
both species undergo considerable turnover whilst
Identifying the precise causes of population regula- maintaining relatively static population sizes. Ramet
tioin in species with a clonal growth form is compli- populations (tillers or rosettes) experienced a constant
cated by the necessity of taking into account vegetative death risk (type 2 survivorship curve), the life expect-
propagation by ramet fragmentation (as well as the ancies of ramets in both species being 11-18 months
practical problem of identifying rarnets). Two species (Sarukhan & Harper, 1973; Weir, 1985). At the same
in which vegetative propagation is common are the time there was recruitment of new ramets, the process
creeping buttercup Ranunculus repens and the grass occurring more or less continuously in Holcus but with
Holcus lanatus, both occurring in grasslands in the UK. noticeable gain and loss dominated periods in Ranun-
In the buttercup, daughter rosettes are produced on culus. Sarukhan and Harper were able to detect
stolons 10-12 cm apart, and in late summer (July- density-dependent regulation in the buttercups as
August) these become detached from the parent as populations accommodated to the pulse of recruit-
interconnecting stolons decay. In Holcus, tillers are ment (Fig. 6.17) whilst in Holcus this process was
borne on shoot complexes (-c5 cm apart) which con- absent because of the fine scale periodicity in ramet
tinually become fragmented through natural decay

200 PART 3: SYNTHESIS

Fig. 6.16 Population turnover in (a) Ranunculus repens, and phenomenon amongst clonal plants. Others may sim-
(b) Holcus lanatus. (After Sarukhan & Harper, 1973; Weir, ply increase in size with age and retain structural and
1985.) physiological integrity. Certainly it has been shown
that in some species clones are of considerableage and
replacement. This comparison leads us to a final that they occupy large areas. For instance, Harberd
important conclusion on population regulation in (1961) found clones of a grass Festuca ovina in Scottish
plants. In species where there are pulses of recruit- pastures up to 10 m in diameter and Oinonen (1967)
ment (as exemplified by unitary species and some has dated clones of bracken, Pteridium aquilinum over
clonal ones) there is the opportunity for density- 450 years old extending in size over 100m. 'The
dependent regulation to occur. Conversely, where extent to which these clones are intact plants or
there is very rapid turnover at the modular level there populations of ramets remains an open question.
may be little opportunity for such regulation because
of inherent morphological and growth constraints 6.9 Genetic change
within the growth form. In this case density-
dependent regulation may be most likely at the genet Discussion of population regulation is usually framed
level during seedling establishment and occupation of in terms of ecological time scales; and although it is
the regeneration niche (section 4.12). It would be generally accepted that the individuals concerned
unwise to assume that fragmentation is a ubiquitous are the products of natural selection genetic change

CHAPTER 6 : POPULATION REGULATION 201

tant, the difference was inherited and remained until

the F, and F, generations. Clearly, demographic
characters can respond to selection on an 'ecological'
time-scale. In this case they did so in a way that made
reproductive-rate inversely related to population size,
and thus tended to regulate the population.

Fig. 6.17 Life expectancy (L) in weeks of vegetative 6.10 Territoriality
propagules of Ranunculus repens as affkcted by its own
species' density. Densities are average values for number of One topic which is intimately tied up with population
plants (m-2) observed at each site in April 1969, 1970 and regulation is territoriality. But as Davies (1978a) has
1971. (After Sarukhan & Harper, 1973.) pointed out, in a much fuller review of the subject
than can be given here, there are a number of
occurring, by definition, on an evolutionary time questions pertinent to territoriality which must be
scale, is usually negIected as a reguIatory mechanism. kept quite distinct. We can distinguish initially be-
There are some dissenting ecologists, however-and tween 'What causes territorial behaviour?'and 'What
Pimentel(1961)is the most frequently quoted of these are the consequences of territoriality?';but even the
who would suggest that such neglect is unwarranted, first of these is itself the confusion of two quite
andl that any discussion of population regulation is separate questions, namely 'What is the ultimate
incomplete if it fails to take account of contemporary cause or driving-force, i.e. what is the selective advan-
adaptive genetic change. There is, however, rather tage associated with territoriality?' and 'What is the
little concrete evidence in favour of this notion, and it proximate cause or mechanism through which terri-
is probable that the ecological and evolutionary time tories are established?' We shall restrict ourselveshere
scales are usually dissimilar. Nevertheless, there is to considering the consequences of territoriality, and
some supporting evidence, and a particularly impres- the selectiveadvantages associatedwith it. But first we
sive example is provided by the work of Shorrocks must define what is meant by a territory and by
(19'70). Shorrocks maintained populations of the fruit-
fly Drosophila melanogaster in the laboratory, and wteirlrlirteocroiaglnbizeehaavteiorurirt.orFyo'l.lo.w. winhgenDevaveire[sin(d1i9v7id8uaa)l,s]woer
obtained regular four-generation cycles in abundance,
unrelated to any regular change in an environmental groups are spaced out more than would be expected
variable. Pairs of flies were removed from the popula- from a random occupation of suitable habitats'. Note,
tions and classified as 'peak' or 'non-peak', depending therefore,that territoriality can be ascribed not only to
on the type of abundance their parents experienced, conventional cases like breeding great tits (Fig. 6.18a),
and[the numbers of offspring produced by these pairs but also to barnacles (Fig. 6.18b) and many plants.
when maintained under identical, uncrowded condi- The rather special case of plants will be discussed in
tions were noted. The pairs in the 'peak' category the next section.
produced significantlyfewer offspring than their coun-
terparts in the 'non-peak' category, and, more impor- The most important consequence of territoriality is
population regulation. Territorial behaviour is closely
allied to contest competition, and this, as we saw in
section 2.4, leads to exactly compensating density-
dependence. The contest nature of territoriality is
demonstrated by the fact that when territory owners
die, or are experimentally removed, their places are
rapidly taken by newcomers. Thus, Krebs (1971)
found that in great tit populations, vacated woodland
territories were reoccupied by birds coming from

202 PART 3: SYNTHESIS

Fig. 6.19 Although the size of the territories of the golden
winged sunbird Nectarina reichenowi varies enormously,
each territory contains approximately the same number of
Leonotis flowers (Gill & Wolf, 1975).(After Davies, 1978a.)

Fig. 6.18 Illustrations of territoriality.(a) Great tits (Krebs, (Clarke, 1WO), dragonflies (Moore, l964), butterflies
1971),and (b) barnacles (Crisp, 1961)spaced out more than (Davies, 1978b)and limpets (Stimson,1973).It must be
would be expected from a random distribution on the realized, however, that the exact number of territories
available suitale habitats. (After Davies, 1978a.) is usually somewhat indeterminate in any one year,
and certainly varies from year to year depending on
hedgerow territories where reproductive success was environmental conditions(Fig. 6.19);and it is, perhaps,
noticeable suboptimal: Watson (1967)found that with for this reason that life-table analyses for the great tit
red grouse the replacements were non-territorialindi- and the red grouse fail to provide clear-cut evidence of
viduals living in flocks, which would not have bred, density-dependenceat the appropriatestage(Podoler&
and would probably have died in the absence of a Rogers, 1975).
territory.In both cases,therefore,overall fecundityand
population size were limited by territorial behaviour: Wynne-Edwards (1962) felt that these regulatory
by a 'contest' for a limited number of territories. Re- consequences of territoriality must themselves be the
moval experiments have demonstrated similar phe- root causes of territorial behaviour. He suggested that
nomena in mammals (Healey, 1967; Carl, 1971), fish the selective advantage accrued to the population as a
whole : that it was advantageous to the population not
to over-exploit its resources. There are, however,
powerful and fundamental reasons for rejecting this
'group selectionist' explanation-essentiaIly, it
stretches evolutionary theory beyond reasonable lim-
its (MaynardSmith, 1 9 7 6 t a n d Wynne-Edwards him-
self (Wynne-Edwards, 1977) has subsequently
recognized these reasons and accepted the rejection of
his ideas. Thus, if we wish to discover the ultimate
cause of territoriality, we must search, within the
realms of natural selection,for some advantage accru-
ing to the individual.

CHAPTER 6 : POPULATION REGULATION 203

It must be recognized that in assessing individual fore, favours the spacing out of nests: it pays each
advantage, we must demonstrate not merely that individual to be territorial.
there are benefits, but that these exceed the costs
associated with territoriality. This has been done in We can suggest then, from this very limited number
the case of the golden-winged sunbirds examined in of examples, that territoriality has evolved as a result
Fig. 6.19. Gill and Wolf (1975) demonstrated that of the advantages accruing to territorial individuals.
although the size of territories may vary enormously, But as an essentially independent consequence of this,
the nectar supply defended is suited to support an there is competition approaching pure contest, and
individual's daily energy requirements. They were therefore powerful (though not, of course, absolute)
able to measure the time that territory owners spent in regulation of populations. Figure 6.21 shows a key-
various activities (including territory defence), and factor analysis for the tawny owl population in
they showed that, as a result of the exclusion of other Wytham Woods near Oxford (Southern, 1970) and
sunbirds, nectar levels per flower inside a territory demonstrates how territoriality can be translated into
were higher than in undefended flowers. Thus, Gill the currency of population dynamics through this
and Wolf found that territory owners could obtain method. The number of territory-holding adults
their daily energy requirements relatively quickly,and changed very little over a 13-year period (see
that, overall, the energetic costs of territorial defence Fig. 5.lf), but there was a great deal of variation in the
were easily offset by the benefits of the energy saved numbers attemptingto breed (rangingfrom 22 pairs in
by ,a shortened daily foraging time. 1959 to none in 1958). Failure to breed each year (k,)
was the key factor in this study. Poor small mammal
A different type of individual advantage resulting years corresponded with years in which the owls did
from territoriality was demonstrated by Krebs (1971) not attempt to breed. Prey availability was indepen-
for the great tit population of Wytharn Woods dent of owl density so k, was a density-independent
(Fig. 6.20). There, the major predators of nestlings are factor. Losses outside the breeding season were, how-
weasels, Mustela nivalis, which may rob up to 50% of ever, found to be density-dependent.Tawny owls are
the nests in some years (Dunn, 1977).But, as Fig. 6.20 territorial all year round and these losses were prima-
shows, the closer a nest is to another nest, the greater rily of young birds which were unable to establish
the chance of predation. Individual selection, there- territories and which consequently starved.

Fig. 6.20 The influence of territory size on the risk of 6.11 'Space capture' in plants
predation in the great tit. (After Krebs, 1971.)
Although territoriality,as such, is not normally asso-
ciated with plants, there is, in plants, a phenomenon
which is at least analogous to territoriality. The
phenomenon can be caricatured in the proverb:
'Possession is nine points of the law', and has been
referred to and discussed by Harper (1977) as 'space
capture'. In fact we have already discussed it briefly in
section 2.5.2 as an explanation for skewed frequency
distributions of plant weights.

Figure 6.22a shows that in experimental popula-
tions of the grass Dactylis glomerata, the comparatively
low weights exhibited by late-emerging plants are
lower than would be expected from the reduction in
growing period alone (Ross & Harper, 1972). The
reason for this is indicated by the data in Fig. 6.22b

204 PART 3: SYNTHESIS

Fig. 6.21 Key-factor analysis for the tawny owl Strix aluco. small mammal years is the key factor. (b)k,, losses outside
the breeding season plotted against numbers of young
(a) Total generation mortality and individual k-values fledged, illustrating a strongly density-dependent mortality.
(After Southern, 1970.)
plotted against time in years, with the regression coefficient
of each individual k-value plotted against total generation
mortality shown in brackets; k,, failure to breed in poor

(Ross & Harper, 1972). Plants were grown from seed (Tamm, 1956). Despite the crop of seedlings entering
either under 'unrestricted' conditions: alone at the the population between 1943 and 1956, it is quite
centre of a 7.4-cm diameter pot; or 'restricted' apparent that the most important factor determining
conditions: in a bare zone, 4.2 cm in diameter, sur- which individuals were established in 1956 was
rounded by seeds sown at a density of 2.5 cm-2. After whether or not they were established in 1943. Of the
initially growing at the same rate the 'restricted' plants 30 specimens that had reached large or intermediate
grew at a slower rate than the 'unrestricted' plants, size by 1943, 28 survived until 1956, and some of
and maintained this difference for at least 3 weeks. It these had ramified. By contrast, of the 112 plants that
appears that the growth achieved by a plant, and thus were either small in 1943 or appeared as seedlings
the size and fitness it attains, is determined early in its subsequently, only 26 survived until 1956, and not
life by the pre-emption or 'capture' of space (or the one of these was sufficiently well established to h~ave
resources implied by that space). Space is then un- flowered.
available (or, at least, relatively unavailable) to other
plants, and these grow more slowly and attain a lower Similar patterns are obvious from simple obsenva-
fitness as a consequence. tions of tree populations. The survival-rate of the few
established adults is high: that of the many seedlings
Presumably, genetic predisposition to early emer- and saplings is comparatively low. In all such cases it
gence, and a chance association with favourable is clear that the major prerequisite for high (or indleed
microsites, both play some part in determining which positive) fitness is the capture of space. We can think
plants actually capture space. In either case, however, of this as the plant equivalent of animal territoriality,
the result is to push competition towards the contest and the regulatory consequences are essentially the
end of the scramble-contest continuum. The plants same.
capture what is, in effect, a territory, and as a
consequence there is a more exact regulation of plant The significance of pre-emptive space capture for
numbers. the fitness of individual plants and the fact that plants
are sedentary emphasizes the need to consider the
Essentially, the same phenomenon is shown by the spatial arrangement of plants in the regulation of plant
herbaceous perennial Anemone hepatica in Fig. 6.23 populations. The structure of a population of a peren-

CHAPTER 6: POPULATION REGULATION 205

Fig. 6.22 (a)The influence of emergence time on the dry ship) of an individual are described in relation to the
weight per plant of Dactylis glomerata. The dashed line number of neighbouring plants occurring with in-
shows the weights that would have been achieved had the creasing distance (Pacala & Silander, 1985). Whilst
weights of late emergers been attributable only to their these approaches provide an important way of linking
reduced growth period. (b)The growth of D. glomerata the performance of individuals to overall population
seedlings with and without neighbours. (After Ross & performance, if only spatial interrelations are consid-
Harper, 1972.) For further explanation, see text. ered then considerable variation in the size of individ-
uals is unaccounted for. As we have seen above, the
nial plant at any particular point in time will be a relative time of emergence of a plant is crucial to an
reflection of past phases of successful recruitment and individual's success, and methods of analysis have still
the outcome of contest competition for resources, to be developed that adequately account for fitness in
pairticularly light. The age- and size-structure arising relation to both local spatial variation in neighbour-
from recruitment and the process of asymmetric hood density and time of plant 'birth' (Firbank &
competitionwill have a major bearing on the changing Watkinson, 1990).
size distribution of individual plants (Hutchings,
1985). Recognition of the potential importance of the Whilst considerable size inequalities may exist
size of immediate neighbours to the fitness of any one amongst individuals in plant populations for very
individual has led to the use of neighbourhood analyses. many reasons, their influence on population dynamics
may be not nearly so marked. Crawley (1993) has
In these the fitness components held and survivor- pointed out that the effects of size differences are often
strongly mediated by plasticity in form and size as
plants are inherently modular. The linear relationship
between size and fecundity (see Fig. 2.11) means that
the tendency for overcompensating density-
dependenceoften evident in insect populations will be
much reduced and in consequence plant populations
are much less likely to show cyclical and chaotic
dynamical behaviour. An apparent exception to this is
Erophila verna which may exhibit cyclical dynamics
(Silvertown & Lovett Doust, 1993).

6.12 Chaos in ecological systems

In section 3.4 we saw how time-lags, high reproduc-
tive rates and highly overcompensating density-
dependence are capable of provoking many types of
fluctuations in population density in single-species
population models, without invoking any extrinsic
cause. The most surprising type of population fluctu-
ation that can be produced from such models is that
which has been described as chaotic. One of the more
surprising discoveries of population ecology in the
1970s was that many of the simple models of animal
population dynamics were seen to display chaotic
fluctuations (see May, 1974, 1976; May & Oster,
1976). Chaos has generated much interest because it

206 PART 3: SYNTHESIS

Fig. 6.23 The behaviour of Anemone hepatica in a forest. populations. This search has progressed along three
Each line represents one individual: straight for unrarnified fronts. The first involves fitting population models
ones; branched where the plant has ramified;bold where capable of predicting a range of dynamics from stalble
the plant flowered; and broken where the plant was not to chaotic depending on the values of one or more
seen that year. Group A were alive and large in 1943; parameters. The second is concerned with the exam-
group B alive and small; group C appeared first in 1944; ination of data from laboratory populations, usually of
group D in 1945; and group E thereafter, presumably from insects, that have been maintained over a large
seedlings. (From Tamm, 1956; after Harper, 1977.) number of generations. The third involves the devel-
opment of new analytical techniques from non-linear
had previously been supposed that the irregular mathematics designed to detect chaos in time-series
population fluctuations typically seen in real data data. We shall consider each of these techniques
must be due to random environmentalfluctuations or below.
sampling errors. That such fluctuations may arise in
deterministic models in which all the parameter Hassell et al. (1976) were the first to fit models to
values are known exactly is quite remarkable. One of field data in this context. They estimated the pararne-
the aims of population ecology is to be able to ters of a simple single-species discrete-generation
understand the dynamics of populations to such an model (equation 3.4) capable of showing chaos, using
extent that future dynamics can be predicted. If 28 sets of insect population data (24 from the field =and
chaotic behaviour prevents us from being able to do four from the laboratory). Their results are shown1 in
this with deterministic, simple models, then it will be Fig. 6.24 (see also section 3.4).With two exceptions all
extremely difficult to achieve it in real populations. the field life-table data fell in the region of monotonic
damping, with a stable equilibrium point. Of the
The discovery of the potential for chaos in popula- remaining field studies, one fell in the region of
tion models stimulated a search for chaos in real

CHAPTER 6: POPULATION REGULATION 207

Fig. 6.24 Stability boundaries between the density- themselves point out. The most serious of these is that
dependent parameter, b, and population growth rate, R, there are no truly single-species populations in the
from equation 3.4, together with estimated parameters real world, and it is unclear how valid it is to abstract
the dynamics of any population from its natural
from field (a)and laboratory(0)studies. (After Hassell community to a single-species population model. In
addition the use of a single-species model biases the
et al., 1976.) results in favour of stability since complex dynamics
are more likely in complex systems. We have already
damped oscillations and the other in the region of seen how delayed density-dependence, one phenome-
stable limit cycles. This last study was of the Colorado non likely to arise as a result of biotic interactions, is
potato beetle Leptinotarsa decemlineata a well-known found in many insect populations (Turchin, 1990)but
outbreak pest in present day agroecosystems (see remained obscure from those looking for density-
section 6.6). Of the four studies of laboratory popula- dependence from life-table studies (see section 6.7.1).
tions, one fell in the region of monotonic damping, two
fell in the region of damped oscillations, and the fourth A number of other workers have looked for evi-
fell in the chaotic region. Figure 6.24 could be inter- dence of chaos in laboratory populations, mostly of
preted as indicating a tendency for natural popula- Drosophila (e.g. Thomas et al., 1980; Mueller & Ayala,
tions to have stable equilibria while laboratory 1981),but failed to find it. However, by the late 1980s
populations show cyclic or chaotic behaviour. There the laboratory approach seemed to be more promising
were, however, a number of important assumptions
implicit in this study, many of which the authors

208 PART 3 : SYNTHESIS

Fig. 6.25 (a)Predicted dynamics for the interaction between for (a),but for Callosobruchusmaculatus and Lariophagus
Callosobruchus chinensis ( ) and the pteromalid parasitoid distinguendus. (c)Observed dynamics of populations of
Lariophagus distinguendus (-- -). The stimulations were Callosobruchuschinensis (0)and pteromalid Anisopteromalus
started with two adult hosts on day 1and 0.02 parasitoids calandrae (0)from the laboratory system of Utida (1950).
were added on day 17 of the seventh generation. (b)As (After Bellows & Hassell, 1988.)

with the discovery of chaos in a series of population for distinguishing chaos from random noise in the
models, particularly host-parasitoid models, with pa- analysis of time-series. One such scheme is that of
rameters within the range of biological realism. For Sugihara and May (1990), who used a library of past
example, Bellows and Hassell (1988) investigated an patterns in a time-series to make short-term predic-
age-structued host-parasitoid model which exhibited tions about future patterns. By then plotting the
intriguing irregular fluctuations using experimentally correlation coefficient of prediction and actual trajec-
determined parameter values (Fig. 6.25a & 6.25b). tories against prediction time interval they considered
The irregular cycles seen in a study of Utida (1950) of it possible to make the distinction between dynamical
a similar host-parasitoid system, the bruchid beetle chaos and measurement error. This is based on one of
Callosobruchus chinensis and its pteromalid parasitoid the 'signatures' of chaos, namely that in a chaotic
Anisopterorntrlus calandrae were qualitatively so similar system, the accuracy of forecast falls off with predic-
to the results of the simulation model of Bellows and tion time, whereas in a system that is merely 'noisy'
Hassell that they, too, may arise from the internal the forecastingaccuracy is roughly independent of the
dynamics of the interaction (Fig. 6.25c). prediction interval. These new mathematical tech-
The third approach in the search for chaos in real niques are particularly demanding of data and there
populations has been the development of techniques are few biological data sets of sufficient quality and

CHAPTER 6: POPULATION REGULATION 209

length to be suitable for the application of the tech- from exponential stability to chaos can be found in
niques. However, data from childhood diseases, not- natural systems. (The single case of chaos has, how-
ably measles and chicken pox have been analysed by ever, been questioned by Perry et al., 1993 on the
Sugihara and May (1990). The non-linear forecasting basis of more extensive data.)Although to some extent
technique seems to indicate that prediction error in the same criticisms can be levelled at this work as that
measles occurs in a manner consistent with chaotic of Hassell et al., in that they assume a particular type
dynamics but that this is not the case with chicken pox of underlying model, the model and fitting techniques
where prediction error is characteristic of pure addi- are more robust than those used in the earlier study.
tive noise (actually superimposed on a seasonal cycle). The significant similarity between the two studies is
However, the conclusions of the childhood diseases that both suggest that chaos is rare in field populations
analysis and the technique itself have been challenged of insects.
by Hastings et al. (1993).
One study which adopts both the first and third
Another approach to detecting chaos in time-series approaches to investigating chaos, fitting population
data is that of Turchin and Taylor (1992). Their models and detection from time-series, is that of
approach in some ways follows that of Hassell et al. Hanski et al. (1993), who explored the population
(1976),but deals with multi-species systemsby includ- oscillations of microtine rodents in boreal and Arctic
ing in the model delayed population density, which regions. Their population model was a predator-prey
reflects the interactions with other species. Unlike model with seasonality in which the predators were
Hassell et al. (1976), Turchin and Taylor did not mustelids. The model was able to generate stable
specify a particular equation, but fitted a response dynamics, limit cycles and chaos and when parame-
surface, the logarithm of yearly population change, as terized with field data predicted dynamics that closely
a fiunction of lagged population densities (see the resemble the dynamics of boreal rodent populations.
Turchin & Taylor paper for details of the technique). The analysis of extensive time-series data of microtine
Turchin and Taylor analysed time-series data for 14 rodents in Fennoscandia was achieved by using the
insect and 22 vertebrate populations. Their results technique of non-linear analysis developed by Turchin
were different from those of Hassell et al. (1976). They and Taylor. Both the model prediction and the ob-
found exponentially stable point equilibria in only served dynamics are chaotic. This result suggests that
three out of 14 insect populations, compared with 21 the multi-annual oscillations of rodent populations in
out of 24 field studies reported by Hassell et al. (1976). Fennoscandia are due to delayed density-dependence
The remaining populations displayed unregulated be- imposed by mustelid predators, and they are chaotic.
haviour (one), damped oscillations (six), limit cycles
(one),quasi-periodic oscillations (two)and chaos (one). How to distinguish chaos from measurement error
The vertebrate data exhibited a similar range of and how to find it in natural populations are some
behaviours, though none was chaotic. Thus the con- way from being resolved. The search for chaos is likely
clusion that Turchin and Taylor reached was that the to be a particularly vigorous area for research in the
complete spectrum of dynamical behaviours ranging 1990s (see Godfray et al., 1991 and Hastings et al.,
1993 for reviews).

Chapter 7

7.1 Introduction 7.2 Metapopulationdynamics

This chapter is divided into two parts, each concerned The term metapopulation was coined by Levins
with entities larger than populations. In the first part (1970), though Andrewartha and Birch (1954) were
(section 7.2) we go beyond the traditional boundary of among the first ecologists to acknowledge that models
population ecology to consider the concept of a of single, isolated populations were inadequate for
metapopulation, a set of local populations which describing the dynamics of species in which local
interact via individuals moving among populations populations frequently become extinct.
(Hanski & Gilpin, 1991). In Chapter 6, we examined
the central issue of population ecology, namely popu- The metapopulation concept has taken some time to
lation regulation. However, it is probable that some develop and the reasons may be that ecologists were
populations are not tightly regulated; they do not unclear that Levins had said anything new, and that
persist in one location for long periods of time at the literature on colonization and extinction at the
densities between positive limits. Sets of such popula- time was dominated by MacArthur and Wilson's
tions may nevertheless persist in the same region, and (1967) island biogeography text. This book was con-
therefore their regulation and long-term persistence cerned with explaining the equilibrium number of
also depends on extinction and colonization events of species on islands. It was dealing with the same
local populations rather than only with the birth and parameters of colonization and extinction as Levins,
death processes (section l.l)occurring within the local but always with colonization from a source area
populations. We shall consider the elementary models (mainland)to an island. Another real difficulty for the
of metapopulation dynamics, an example of a species development of the metapopulation concept is that
with a metapopulation structure, and look at other metapopulations function on temporal and spatial
areas of ecology to which the metapopulation concept scales that are not convenient for ecologists to study;
is likely to have applications in the future. local populations may persist for longer than the life
span of the average research grant and metapopula-
The subject matter of this book is the ecology of tions last longer than that. The growing awareness by
populations, and it would be unreasonable, in the ecologists of the role of spatial distributions in popula-
second part of this final chapter (section 7.3), to tion ecology (as seen in each of Chapters 4-6) in the
attempt to cover the ground of some other book 1980s and early 1990s has undoubtedly hastened the
concerned with the ecology of communities. Never- developmentof the metapopulation concept,though it
theless, it will be valuable to consider, briefly, the roles is worth stressing that there is a diierence between a
that the various processes considered in the earlier metapopulation and a patchily distributed local popu-
chapters play in determining community structure. lation.
Specifically, we shall consider, in turn, the roles that
interspecific competition, predation, disturbance, in- 7.2.1 Metapopulation models
stability, and habitat size and diversity can play; and
then draw what conclusions we can regarding their Levins (1969) distinguished between the dynamics of
general importance in actual communities. single populations and a set of local populations. He
assumed that individual local populations were either
at carrying-capacity (K) or extinct, thereby ignoring

CHAPTER 7: BEYOND POPULATION ECOLOGY 211

local dynamics except for colonization and extinction there is some density-dependence in local dynamics
events. The variable p, in his model denotes the (constant K in the Levins model).
fraction of habitat patches occupied by a species at
time t. The spatial arrangement of patches is assumed The next level of complexity in single-species
to be of no consequence; individuals in one patch are metapopulation models comes with the incorporation
equally likely to disperse to any of the other patches. of local population dynamics into the models.
The basic Levins model has the following form: Figure 7.1 shows some examples which suggest that

dpldt = colonization rate - extinction rate. (7.1)

Here, dpldt is the rate of change in the fraction of
occupied patches and the equation is therefore analo-
gous to population models in which rate of change of
abundance is the difference between birth- and death-
rates. In his first model, Levins (1969) assumed that
the rate of colonization was proportional to p, the
fraction of patches from which potential immigrants
were available to colonize empty patches, and to 1-p,
the fraction of unoccupied patches, or targets for
colonization. Since all local populations are assumed

to be at their carrying-capacity,they all have the same
extinction probability and the model becomes:

dpldt = mp( l - p) - ep,

where m is the rate of recolonization of empty patches
and e is the rate of local extinction of patches. Thus the
amount of recolonization increases both with the
number of empty patches prone to colonization(1 -p)
ancl the number of occupied patches able to provide
colonizers, p. Extinctions simply increase with p. The
equilibrium value of p (obtained by setting dpldt = 0)is:

This simple model makes one important point: a Fig. 7.1 Relationship between the probability of local
rnetapopulation can only persist if, when small, the extinction and p in (a) mangrove island insects (Simberloff,
rate of establishment of new local populations exceeds
the rate of local extinctions. This conclusion, while in 1976), (b) leafhoppers (Konthanen,1950), and
itself unremarkable, nevertheless provides us with a
means of relating metapopulation dynamics to the (c)freshwater molluscs living in small ponds (Boycott,
structure of the environment,particularly the size and 1930). (After Hanski, 1991.)
isolation of habitat patches (Hanski, 1989). In addition
the model contains the three basic elements necessary
for population regulation at the metapopulation level.
They are that colonizationand extinction events are at
least to some extent uncorrelated among local popu-
lation~t,here is dispersalbetween habitat patches, and

212 PART 3: SYNTHESIS

extinction probability is not constant between patches, 7.2.2 Examples of metapopulations
but decreases as the fraction of patches occupied (p)
increases. The reasons for this are two well-known In order for species to conform to a metapopulation
phenomena: (i) the increase in extinction probability structure there must be a tendency for local popula-
with declining population size (e.g. Williamson, 1981); tions to become extinct, and a relatively poor ability to
and (ii) the observed increase in local population size disperse and colonize new habitats. Waterfleas (Daph-
as p increases, (e.g. Hanski, 1982; Gaston & Lawton, nia spp.)in rockpools (Bengtsson, 1991)and pool frogs
1990). This is contrary to the assumptions of the in successional ponds (Sjogren, 1991) provide close
Levins model, which Hanski (1991) then modified approximations to the Levins concept. However, some
accordingly: of the most persuasive examples come from studies of
butterflies (e.g. Harrison et al., 1988; Thomas &
dpldt = mp(1 -p) - eo e-aPp Harrison, 1992).

where eo and a are two extinction parameters. If eo is Thomas and Harrison studied nine metapopulations
greater than m, the model may have two alternative of the silver-studded blue butterfly Plebejus argus in
stable equilibria (Fig. 7.2), separated by an unstable North Wales. This species is patchily distributed
equilibrium, a threshold value for metapopulation wherever it has been studied. It typically occurs in
persistence (Hanski, 1991). The importance of alterna- temporary habitats, but despite this, most adults move
tive equilibria lies in the possible use of metapopula- only short distances. In North Wales, P. argus lays its
tion models to aid the re-establishment of species in eggs along the margins between bare ground and the
fragmented habitats. If alternative equilibria exist the vegetation on which the larvae feed, usually on south
propagule (or size of the introduction) must be suffi- facing slopes. The larvae are relatively polyphagous
cient to push the metapopulation beyond its threshold (hosts from the Ericaceae,Leguminosae and Cistaceae)
value. and it is oviposition site rather than host plant that
limits persistence. The larvae and pupae are also
Further developments in metapopulation models, associated with Lasius ants, with which it has an
both single and multi-species, are likely to involve the extremely close, and probably mutualistic relationship
incorporationof stochastic processes and a closer link (Jordano & Thomas, 1992).The metapopulations stud-
between regional and local dynamics (see papers in ied by Thomas and Harrison occurred in two main
Gilpin & Hanski, 1991).

Fig. 7.2 Alternative stable equilibria

in a modified Levins model, equation
7.5, which takes into account the

empirically observed (Fig. 7.1)
negative relationship between

extinction probability and p. The
thick line gives the colonization and
extinction rates divided by p. In
(b)the two functions have only one
intersection point, which is stable,
but in (a)there are two intersection

points, a stable one (a)and an

unstable one (0).In this case the
trivial solution (y = 0)is also stable.
(After Hanski, 1991.)

CHAPTER 7: BEYOND POPULATION ECOLOGY 213

biotopes, heathland and limestone grassland. In On a regional scale the distribution of P. argus was
heathland the successional nature of the habitat is limited by its low dispersal tendency. Suitable habitat
maintained by burning, cutting, grazing or distur- only a few kilometres from existing metapopulations
bance, which all serve to create oviposition sites. In remained uninhabited. Evidence that at least some of
limestone grassland, the suitable microhabitat of veg- these habitats are suitable (not just in the minds of
etation and bare ground can be created and main- human observers) comes from the successfulintroduc-
tained by grazing indefinitely, and succession is tion into the Dulas Valley. Within metapopulations
extremely slow in any case on similar limestone local butterfly distribution closely followed that of
habitats, such as screes, crags, quarries, etc. Thomas suitable habitat with most patches within 1 km of
and Harrison (1992)described surveys of suitable sites existing populations being occupied. As might be
in North Wales in 1983 and 1990. Examples of the expected turnover was greater in heathland, where
distribution of Plebejus argus are shown in Fig. 7.3a transient successional habitats remained for relatively
and 7.3b for a heathland and a limestone metapopu- short times, than in limestone grassland, where the
lation. The Dulas Valley limestone metapopulation is habitat can be maintained in the suitable successionaI
of particular interest as it maps the colonization of a state for long periods by grazing. In addition heath-
network of habitat patches following the introduction land habitat patches were generally smaller than
in 1942 of 90 females to one patch. those on limestone grassland, and small patches

Fig. 7.3 (a) The distribution of Plebejus
argus in the South StacksIHolyhead
Mountain heathland metapopulation
in 1983 and 1990. Filled outlines
indicate areas where P. argus was
present in both 1983 and 1990. Empty
outlines indicate areas where P. argus
was not present in both time periods:
e, presumed extinction (P. argus was
present in 1983 only); c, presumed
colonization(present 1990 only);
v, vacant but apparently suitable

habitat in 1990. (b)Colonization of the
Dulas Valley limestone site by P. argus.
following the introduction of 90
females in 1942. The 1973 signs give

the outermost boundaries of patches
occupied in 1971-73. (After Thomas
& Harrison, 1992.)

214 PART 3: SYNTHESIS

showed higher turnover than large patches in any islands. The equilibrium value of p is:
case.
p = m/(m + e), (7.6)
Whether the individuals in a series of patches
should be regarded as a series of separate populations, which is always positive for islands with any turnover.
a metapopulation or as one population with a number Equation 7.5 describes the frequency of occurrence of
of resource patches depends on the distribution of a species when neither immigration nor extinction
dispersal distances relative to patch size and the probabilities are affected by regional occurrences.
distance between patches. If dispersal distances are Most metapopulations probably function somewlhere
extremely low relative to the distances between between that described by Levins and the modified
patches, then the patches can be said to hold a series MacArthur-Wison model. It is almost inconceivable
of separate populations. Conversely, dispersal dis- that all habitat patches will be so similar that each has
tances are extremely high in relation to distances the same carrying-capacity, an assumption of the
between patches, then the patches are simply part of Levins model. Those local populations which are
the heterogeneity experienced by a single population. much larger than average will have reduced exlinc-
The populations of P. argus in North Wales described tion probabilities and thus function as small main-
by Thomas and Harrison, particulary those on heath- lands. They may be so large that they are extremely
land, are intermediate in these respects and would long-lived and the appropriate metapopulation struc-
seem to fit the metapopulation concept reasonably ture is one of an almost permanent patch with
well. transient satellite patches. A good example of such a
structure is provided by Harrison et al. (1988)for the
7.2.3 Applications of the rnetapopulation concept bay checkerspot butterfly in California (Fig. 7.4). In
1987 the metapopulation consisted of a 2000-ha
The equilibrium theory of island biogeography pro- habitat patch (M~rganHill) containing of the order of
posed by MacArthur and Wilson (1967)is, as we have 106 adult butterflies and nine small populations con-
seen, concerned with the same basic processes as the taining between 10 and 350 adults on patches of
metapopulation models, namely colonization and ex- between 1 and 250 ha. Of some 27 small habitat
tinction. The theory attempted to account for the patches in the region that were suitable for the
number of species on real islands or habitat islands, butterfly, only those closest to Morgan Hill were
and proposed that the species number was an equilib- occupied. Such a pattern of patch occupancy is not
rium between immigration and extinction rates, them- explicable in terms of habitat quality, but it rather
selves determined by island area and distance from reflects the butterfly's limited powers of dispersal. The
the source of colonists. The difference between the large population at Morgan Hill acts as the dominant
MacArthur-Wilson model and the metapopulation source of colonists to the smaller patches. In this case
models is that in the former there is a mainland which the persistence of the metapopulation is relatively
is an inexhaustible source of colonists, and in the unaffected by population turnover on the smaller
latter all colonists must come from existing local patches.
populations. The MacArthur-Wilson model is con-
cerned with numbers of species, the metapopulation In the 1970s island biogeography theory was linked
models are concerned with populations of single with the design of nature reserves (Diamond & May,
species. However, Hanski and Gilpin (1991) and 1981). How many Amazonian plant and animal
Gotelli (1991) provide a single-species version of the species will survive if only 1% of the Amazonian
MacArthur-Wilson model analogous to equation 7.1 rainforest survives?At what rate will species be lost?Is
for changes in the fraction of islands occupied: it best to allocate a certain total area to one large or
several small reserves? Species area curves (like
dp/dt = m(1 -p) - ep, (7.5) Fig. 7.10) offered insights into the first question and
the MacArthur-Wilson equilibrium theory held out
where m is now migration from the mainland to the the promise of a solution to questions like the second

CHAPTER 7 : BEYOND POPULATION ECOLOGY 215

Fig. 7.4 Metapopulation of the bay
checkerspot butterfly Euphydryas
editha bayensis. The black areas
represent patches of the butterfly's

serpentine grassland habitat. The
2000-ha patch labelled 'Morgan Hill'
supported a population of in the
order of 106 adult butterflies in
1987. The nine smaller patches
labelled with arrows supported
populations of in the order of
between 101 and 102 butterflies in
that year. Eighteen other small

patches were found to be suitable
but unoccupied. (After Harrison,

1991.)

ancl third. As environments become increasingly frag- Like Hanski (1990)we are tempted to speculatethat
mented, many species, which previously had continu- metapopulation regulation may prove to be as impor-
ous spatial distributions, may find themselves with a tant for many species as traditionalpopulation regula-
rather more discontinuous one. Species with very poor tion. Tree species in tropical forests have a wide
powers of dispersal in a newly fragmented habitat will diversity and patchy distributions. There is much less
simply exist in smaller populations with each one of a tendency in the tropics for species to occur in solid
facing an increased probability of extinction. Others stands than in temperate forests. This means that
with better powers of dispersal are likely to exist as there can be a considerable nearest-neighbour dis-
metapopulations. Thus the controversy alluded to tance between conspecific trees. Tree canopies of
above about whether or not an area set aside for particular species thus effectively become patches of
nature conservation should be one large or several suitable habitat surrounded by a sea of potentially
small reserves (see Simberloff & Abele 1976)is really a unsuitable (or even downright poisonous) habitat. In
metapopulation rather than an island biogeographic the canopies of these trees live a high percentage of
problem, particularly when the reserve is being de- the metazoan species on this planet (Erwin & Scott,
signed for one or a few key species. Metapopulation 1980; May, 1988), many of which are monophagous
dynamics provides a natural framework for consider- and oligophagous herbivorous insects. These species
ing the survival of species in a network of reserves
(Hanski, 1989). may conform to the metapopulation concept, particu-
larly those that have relatively poor dispersal abilities.

216 PART 3: SYNTHESIS

This is an exciting scenario for future research and
one which cannot, for obvious reasons, be delayed
long.

7.3 Community structure Fig. 7.5 (a)The response of a population of Rumex acetosella

7.3.1 The role of interspecific competition in a mixed grassland sward to the removal of certain

Good evidence that interspecific competition can play components of the sward, expressed as a percentage of the
an important role in determining community struc-
ture is provided by an experiment carried out by population in June 1965. (b) Numbers of seedlings per plot.
Putwain and Harper (1970) on a hill grassland site in Key to treatments: (-) control; 1-( R. acetosella
North Wales that was closely grazed by sheep. The --removed; (- ---) all dicots except R. acetoseflaremoved;
species with which they were most concerned was (. . . . .) grasses removed; (- -) all species except
sorrel, Rumex acetosella, which was second in abun-
dance in the community to a grass, sheep's fescue R. acetosella removed. (After Putwain & Harper, 1970.)
(Festuca ovina). Galium saxatile (heath bedstraw) was
also abundant, and 1 4 other species of grasses and It appears that the growth of mature sorrel plants
herbs were present in varying numbers. Specific was unaffected by the removal of dicots, was increased
components of the community were experimentally only slightly by the removal of grasses, but was very
removed with herbicides by setting up plots of the significantly increased by the removal of both grasses
following types: and dicots. The rate of seedling establishment was
1 plots sprayed with Dalapon to remove all grasses; increased by the removal of grasses, or of dicots, or of
this does little and only temporary damage to Rumex the mature plants of R. acetosella itself.
acetosella and other non-gramineous species (i.e. di-
cots); The probable explanation is illustrated in Fig. :7.7,
2 plots in which individual plants of all dicots except which is a diagrammatic representation of R. acetasel-
R. acetosella were killed by the combined application of la's niche relationships within the community. R.
the herbicides 2,4-D and Tordon 22K; acetosella, since it exists within the sward, obviously
3 plots sprayed with Paraquat to remove all species has a realized niche; but it is competitively excluded
except R. acetosella which, despite having its above- from a substantial portion of its fundamental niche in
ground parts scorched, regrows rapidly from buds at
the base of the stem; and
4 plots in which R. acetosella plants alone were killed
by spot treatment with Tordon 22K.
There were also control plots that were not sprayed at
all.

Spraying took place on 2 June 1965 and then, to
distinguish between the effects of treatment on vege-
tative and seedling growth, seeds were sown in parts
of the plots on 20 September 1965. Abundance of R.
acetosella was monitored throughout the year follow-
ing treatment, and its dry weight under each regime
determined on 5 July 1966. The results are shown in
Figs 7.5 and 7.6.

CHAPTER 7: BEYOND POPULATION ECOLOGY 217

Fig. 7.7 The diagrammatic fundamental niches of Rumex
acetosella (R),grasses (G)and dicots (D). Light-shaded area:
the realized niche of R. acetosella;dark-shaded areas: the

fundamental niche of R. acetosella seedlings. (Modified from

Putwain & Harper, 1970.)

Fig. 7.6 The dry weight (g) of Rumex acetosella per plot at We have, remember, seen several similar examples
the end of the period of observation. For key to sward in Chapter 4: seed-eating desert ants showing differen-
treatment, see Fig. 7.5. LSD indicated at p = 0.05. (After tiation in size and foraging strategy; Panicum and
Putwain & Harper, 1970.) Glycine showing niche differentiation with respect to
nitrogen; barnacles partitioning space; bumblebees
this community by the combined action of the dicots, specializing on flowers of different corolla lengths;
and from a similar but even larger portion by the plants showing temporal heterogeneity in resource
combined action of the grasses. It is, therefore, only utilization, and so on. In the present context, the
when dicots and grasses are both absent that signifi- importance of these examples is that, in all such cases,
cant competitive release occurs. It appears, moreover, the communities are structured, and species diversity
that the fundamental niche of the sorrel seedlings lies increased, by resource partitioning based on competi-
largely within the combined realized niche of the tive exclusion.
grasses, though other, smaller portions lie within the
realized niches of the dicots and the R. acetosella 7.3.2 The role of predation
adults.
The most famous piece of evidence supporting the
This experiment shows clearly that the distribution importance of predators in determining community
and abundance of sorrel is determined to a significant structure is provided by the work of Paine (1966), and
extent by the interspecific competitive interactions their role is succinctlystated (as a hypothesis)by Paine
occurring within the grassland community. Of course, himself: 'Local species diversity is directly related to
the precise design of the niches in Fig. 7.7 is quite the efficiency with which predators prevent the mo-
arbitrary and their important dimensions are not even nopolization of the major environmental requisites by
dimly understood. Nevertheless, the figure does serve one species'. Paine presented some correlational sup-
to illustrate how communities must often be struc- port for this hypothesis, but his most persuasive
tured by species being competitively confined to small, evidence was experimental.
realized portions of their fundamental niches.
On the rocky shores of the Pacific coast of North
America the community is dominated by a remark-
ably constant association of mussels, barnacles and
one starfish; and Fig. 7.8 illustrates the trophic rela-

218 PART 3: SYNTHESIS

Number- 4/ \ \ ~ 0 1 ~ 0 . 0 3 sured. The appearance of the control area did not
Calories alter. Adult Mytilus californianus, Balanus cariosus (an
X=cO.OI 0.03- acorn barnacle) and Mitella polymerus (a goose-necked
barnacle) formed a conspicuous band in the mid-
chitons limpets Mytilus acorn Mitella intertidal; while at lower levels the diversity increased
2 spp. 2 spp. (Bivalve) barnacles (goose abruptly, with immature individuals of the above
1 sp. 3 SPP. barnacle) species, Balanus glandula in scattered clumps, a few

Fig. 7.8 The feeding relationshipsby numbers and calories anemones of one species, two chiton species (brows-
of the Pisaster dominated subweb at Mukkaw Bay. The ers), two abundant limpets (browsers), four macro-
specific composition of each predators diet is given as a pair scopic benthic algae (Porphyra, Endocladia, Rhodomela
and Corallina), and the sponge Haliclona (often
of proportions:numbers on the left, calories on the right. browsed upon by Anisodoris,a nudibranch)all present.
(After Paine. 1966.) 1 calorie (non-S1unit) = 4.186 joules.
Where the Pisaster were excluded, however, the
tionships of this portion of the community as observed situation changed markedly. Balanus glandula settled
by Paine at Mukkaw Bay, Washington. The data are successfully throughout much of the area, and by
presented both as the numbers and as the total September 1963 it had occupied 60-80010 of the
calories consumed by the two carnivorous species in available space. By the following June, however, the
the subweb: the starfish Pisaster ochraceus and a whelk Balanus themselves were being crowded out by small,
Thais emarginata. Apparently this food web is based on rapidly growing Mytilus and Mitella; and this process
a barnacle economy, with both major predators con- of successive replacement by more efficient occupiers
suming them in quantity. Note, however, that in terms of space continued, leading eventually to an experi-
of calories the barnacles are only about one-third as mental area dominated by Mytilus, its epifauna, and
important to Pisaster as either Mytilus californianus, a scattered clumps of adult Mitella. The benthic algae,
bivalve, or the browsing chiton Katherina tunicata. with the exception of Porphyra, tended to disappear
due to a lack of space, while the chitons and larger
For several years from June 1963, Paine excluded limpets tended to emigrate because of an absence of
all Pisaster from a 'typical' piece of shoreline at space and a lack of appropriate food.
Mukkaw Bay about 8 m long and 2 m in vertical
extent. An adjacent control area was left unaltered; Interpretation of Paine's experiment must be tem-
and line transects across both areas were taken pered by the admission that the altered system may
irregularly, and the number and density of resident not have reached an equilibrium (Paine, 1966). Nev-
macroinvertebrate and benthic algal species mea- ertheless,it is clear that the removal of Pisaster led to a
marked decrease in diversity, despite an actual in-
crease in the size of the standing crop. There was a
change from a 15-species system to a trophically
simpler eight-species system, and of the species that
disappeared, some were and some were not in the
normal diet of Pisaster. It seems, then, that the
influence of Pisaster on the community is at least
partly indirect; by eating masses of barnacles and the
competitively dominant Mytilus, and thus keeping
space open, Pisaster enhances the ability of other
species to inhabit the area. When space is available,
other organisms, for instance chitons, settle or move
in, and form major portions of Pisaster's nutrition.
Thus, in the absence of predation there is an increased

CHAPTER 7: BEYOND POPULATION ECOLOGY 219

Table 7.1 Effects of density on seed mortality amongst tropical trees. All published observations and experiments known
from tropical forests are included (After Connell, 1979)

tendency for competition at lower trophic levels to go individuals from an area. As such it may take a wide
to completion, driving species to extinction; but by its variety of forms: e.g. lightning, storms, land-slips or
presence, Pisaster keeps many of these populations even indiscriminate predation. In general terms, its
well below their carrying-capacity. Competitive exclu- effectwill be to prevent communities from reaching an
sion is, therefore, commonly avoided, and the diver- equilibrium; parts of them, at least, will be repeatedly
sity of the community enhanced. In short, it seems returned to early, colonizing stages of succession. Its
that, in the present case at least, Paine's hypothesis is more specific effect on species diversity, however, will
correct: predation prevents competitive exclusion and, depend on the nature of the equilibrium community
therefore, increases community diversity. itself; and this, in turn, will depend on the various
processes-competition, predation and so on-also
In theory, predation can have an even more potent discussed in this chapter. Nevertheless, a plausible,
effect on species diversity when it is frequency- general reationship between disturbance and diversity
dependent, i.e. when there is predator switching has been proposed by Connell(1979)-the 'intermedi-
leading to a 'type 3' functional response (section ate disturbance hypothesis'--and the role of distur-
5.7-4). Prey species will then be disproportionately bance in determining community structure can be
affected when they are common, and this should lead usefully discussed in this context.
to a large number of rare prey species. Unfortunately,
in practice, there is little positive evidence that this Connell recognized, essentially, three levels of dis-
occurs (Connell, 1979).Nevertheless,by examining all turbance (Fig. 7.9a). Where disturbances are frequent
the available data on seed and seedling mortality in and large, the community will tend to be dominated
tropical trees, Connell(1979)was able to show that, in by opportunistic, fast-colonizing species, with, per-
many cases, there is decreased survivorship (i) when haps, a few individuals of intermediate, secondarily
density is high (Table 7.1), and (ii) in the immediate colonizing species, probably present as juveniles. Such
presence of established adults (Table 7.2). Conversely, a community will have a simple structure and a low
these tables also show that there were several other diversity (left-hand side of Fig. 7.9a). At the other
cases when this was not so. Overall, while it is clear extreme, where disturbances are rare and small, the
that such frequency-dependent predation does occur diversity will depend on the importance of what
in nature (leading, no doubt to increased diversity), it Connell calls 'compensatory mechanisms', i.e. preda-
is equally clear that its occurrence is by no means the tion, resource partitioning, and so on. As we have
general rule. seen, where these are prevalent diversitywill be high.
In their absence, however, only highly competitive,
7.3.3 The role of disturbance late-succession species will be able to survive, and
diversity will be low (right-hand side of Fig. 7.9a).
Following Connell(1979), we shall take 'disturbance' Conversely,at intermediate levels of disturbance, even
to be the indiscriminate, catastrophic removal of all in the absence of compensatory mechanisms, there

220 PART 3: SYNTHESIS

Table 7.2 Survivorshipof seeds or seedlings either near or far from adult trees of the same species. All known published

field experiments or observations in tropical forests are listed, but in some cases typical, rather than total, results are
presented. (After Connell, 1979.)

will probably be a few adults of fast-colonizing species, exhibiting very high diversity, areas that are largely
many individuals of mid-succession, secondarily colo- undisturbed (like the Bindongo forest in Uganda)come
nizing species, and even some individuals, possibly to be dominated by a single species of tree (in this case,
juveniles, of late-successionspecies. Overall,therefore, ironwood).
diversity will be high (centre of Fig. 7.9a); in the
comparative absence of compensatory mechanisms, Overall, therefore, we can accept that, in some
the species diversity of communities will be highest at cases, diversity will be highest at intermediatelevels of
intermediate levels of disturbance. disturbance;and that generally, large frequent distur-
bances will tend to decrease diversity. Conversely, the
Some of Connell's evidence in support of this effects of disturbance on diversity will be much less
hypothesis is shown in Fig. 7.9b, where the data come clear-cut whenever compensatory mechanisms are
from observations on a coral reef off the coast of sufficiently potent to ensure that stable climax com-
Queensland, Australia. Disturbance, resulting either munities exhibit a high degree of diversity themselves.
from hurricane damage or from the effects of anchor-
ing boats on the reef, is measured as the percentage of 7.3.4 The role of instability
a site that is devoid of any live coral, and it is indeed
apparent that diversity (number of species per sample) All populations are, to a greater or lesser extent, liable
is highest at intermediate levels of disturbance. As to become extinct; and whenever this occurs, the
further evidence, Connell points out that even in structure of the community containing that popula-
tropical rainforests, which we tend to think of as tion will obviously change. However, this liability is

CHAPTER 7 : BEYOND POPULATION ECOLOGY 221

Fig. 7.9 (a)Connell's 'intermediate
disturbance' hypothesis, involving
(A) opportunistic species,
(B) secondarily colonizing species,
and (C)climax species. (b) Data in
support of the hypothesis from
Heron Island, Queensland from
damaged (A) and undamaged (0)
sites. (After Comell, 1979.)

bound to be greater in some communities than others, has been the province of theoretical ecologists, and, to
and in this sense some communities must be more paraphrase May (1979),two intertwined conclusions
unstable than others. Yet the communities with struc- have emerged.
tures conferring stability are the ones most likely to be 1 In 'randomly constructed' model ecosystems,
observed, because they persist. Structural instability an increase in the number of species in a community
must, therefore, be an important determinant of is associated with an increased dynamical fragility and
observed community structure. a diminished ability to withstand a given level of
environmental disturbance. Thus, relatively stable or
The search for what, inherently, leads to instability

222 PART 3: SYNTHESIS

predictable environments may permit fragile, species- are in the barren Arctic (where energy input is low).
rich communities to exist; while relatively unstable or An alternative explanation,however, was provided by
unpredictable environments will support only a dy- Pimm and Lawton themselves. By studying the stabil-
namically robust, and therefore relatively simple, ity properties of various Lotka-Volterra models, they
ecosystem. argued that long food chains may typically result in
2 Real ecosystems are not assembled randomly. They population fluctuations that are too severe for top
are the products of long-running evolutionary pro- predators to exist. In other words, only relatively short
cesses. We are therefore bound to ask: what special food chains are sufficiently stable to be observed in
structural features of real ecosystems may help to natural communities.
reconcile community complexity with dynamical sta-
bility? In other words, since instability will tend to Finally, Pimm and Lawton (1978)have explored the
simplify communities, what observable features of relationship between omnivory and stability by study-
community structure can be deemed to exist by virtue ing model ecosystems based on Lotka-Volterra equa-
of the stability they confer on complex, species-rich tions. Broadly speaking, they conclude that omnivory
systems?The proposed 'role of instability' will then be and overall dynamical stability are easier to reconcile
the 'selection' of these features. if the omnivores and their prey are of similar size and
population density, a situation that most commonly
Attempting to discover what these features might be pertains to insect parasitoids. As May (1978b, 1979)
has also been the province of theoretical ecologists; suggested,this may account for the diversity of insects
and, as yet, these attempts have been largely specula- in general and the diversity of parasitoids in particu-
tive. Nevertheless, there are several interesting possi- lar.
bilities (May, 1979). May (1972),for instance, and Goh
(1978) have suggested (from the analysis of models) Overall, then, while its precise role remains largely
that ecosystems will be more robust if they consist of the subject of theoretical speculation, it is quite clear
'loosely coupled subsystemsT.This term describes a that instability can play a crucial part in determining
situation in which a community consists of several the structure of natural communities.
parts ('subsystems'), within which there is consider-
able biological interaction, but between which there is 7.3.5 The role of habitat size and diversity
very little interaction. This, according to Lawton and
Pimm (1978)and Beddington and Lawton (1978),is at As Gorman (1979) has pointed out, Great Britain has
least consistent with the observation that most insect 44 species of indigenous terrestrial mammals, extant
herbivores are monophagous or oligophagous, giving or recently extinct, but Ireland, just 20 miles farther
rise to relatively discrete food chains even in species- into the Atlantic, has only 22; and while this might
rich plant communities. However, empirical evidence conceivably reflect the difficulties the mammals have
generally fails to give positive support to the hypothe- in crossing water, it actually affects bats as much as
sis (Pimrn & Lawton, 1980). any other group: only seven of Britain's 13 species
breed in Ireland. Furthermore, of Britain's 171species
Another feature of natural communities possibly of breeding birds, only 126are recorded as breeding in
subject to selection by instability is the length of food Ireland, and 24 of these do so only occasionally. For
chains, which rarely consist of more than four or five example, there are no woodpeckers in Ireland (though
trophic levels. The conventional explanation for this is there are plenty of trees), no little or tawny owls, and
that length is limited by the inefficiencyof energy flow no marsh or willow tits.
from one trophic level to the next (there is insufficient
energy left to support the higher trophic levels). Yet, as The most likely explanation is that Great Britain is
Pirnrn and Lawton (1977) have pointed out,, this far larger than Ireland. But size can exert its effects in
cannot, by itself, explain why food chains are about as two quite separate ways. Perhaps the most obvious
long in the tropics (where energy input is high) as they explanation is that differences in habitat size are
important because large habitats are more diverse.

CHAPTER 7: BEYOND POPULATION ECOLOGY 223

Fig. 7.10 The number of amphibian and reptile species Fig. 7.11 The species area curve for the number of

living on oceanic West Indian islands of various sizes. (After floweringplants found in sample areas of England
MacArthur & Wilson, 1967.) (Williams, 1964).(After Gorman, 1979.)

But there is a second explanation that applies when- mechanisms, however, is provided by data in Fig. 7.11
ever habitats can be thought of as islands (either real (Williams, 1964). This, too, is a plot of log species
islands, or 'habitat islands' of one type surrounded by
a 'sea' of another habitat type). Larger islands support number against log habitat size, but size in this case
larger populations that have a relatively low probabil- pertains to arbitrary sampling areas within a main-
ity of becoming extinct. In addition, larger islands land. Once again, on this log-log plot, the number of
represent a larger 'target' for colonization by species flowering plants rises linearly with the size of sampling
not already present (MacArthur & Wilson, 1967). On areas in England. But the slope-around 0.1-is no-
two counts, therefore, extinction and immigration, we ticeably lower than those from the island examples,
can expect larger islands (i.e. larger habitats) to and falls near the range typical for mainland studies:
support more species. Note, too, that this is an 0.12-0.17 (MacArthur & Wilson, 1967). The crucial
explanation for the fact that (small) islands generally point is that habitat size can only act via habitat
support fewer species than a nearby (larger) mainland. diversity in such cases. These arbitrary areas are
continually exchanging organisms with surrounding
A typical relationship between the number of spe- areas, and they are not, therefore, subject to the
cies living on an island and the island's area is considerations of extinction and colonization that
illustrated in Fig. 7.10, for the amphibiansand reptiles apply to isolated islands. Thus, mainland slopes from
living on oceanic islands in the West Indies (Mac- 0.12 to 0.17 reflect the effects of habitat diversity,
Arthur & Wilson, 1967). The logarithm of species while the increased slopes on islands react the addi-
number rises with the logarithm of island area in a tional size effects peculiar to island biogeography.

remarkably linear fashion, and the slope (0.30)is very Overall we can see that an increase in habitat size
much in line with those obtained in other examples. will lead to an increase in speciesnumber, and thus to
For organisms ranging from birds to ants to land an increase in the complexity of community structure.
plants, in both real and habitat islands, the slopes of This may result from the indirect effects of habitat
such log-log plots mostly fall within the range 0.24- diversity, or from effects peculiar to the island nature
0.34 (Gorman, 1979). The role of island (i.e. habitat) of many habitats; and while it is often difficult to
size as a determinant of species number (and thus partition the total effect into these two components,
community structure) is therefore, well established. there is no doubt that both are of very widespread
importance.
An indication of the fact that size acts through two

224 PART 3 : SYNTHESIS

7.3.6 Conclusions cause of the inherent differences. Note, too, as another
aspect of this reinforcement, that those effects of
We have seen that a variety of factors can influence habitat size that are attributable to habitat diversity
community structure. Yet, in truth, it has to be (section 7.3.5) will themselves be influenced by in-
admitted that precise statements as to their relative creases in the diversity of the biotic aspects of a
potencies must await further advances in our knowl- habitat.
edge and understanding. Nevertheless, certain tenta-
tive conclusions can be drawn. In short, there is good reason to believe that the
constancies, predictabilities and productivities of abi-
Perhaps the most significant of these is that a good otic environments are crucial, underlying determi-
case can be made for the constanc&,predictability and nants of community structure: and that competition,
productivity of the abiotic environment being of abso- predation and ecosystem instability are mechanisms
lutely crucial importance in determining community through which they exert their influence. This is
structure. This conclusion stems from a number of almost certainly the explanation for the single most
considerations. important cline of increasing diversity: from the poles
1 In general terms, diverse, basically fragile ecosys- to the tropics.
tems appear to be relatively stable in constant, predict-
able environments (section 7.3-4). This view is opposed, to some extent, by Connell's
2 More specifically,niches can be stably packed more 'intermediate disturbance hypothesis' (section 7.3.3),
tightly in predictably productive environments (sec- since a constant, predictable environment is likely to
tion 4.15). This suggests that interspecific competition be one with a low level of disturbance. Conversely,
will be most potent as a mechanism promoting diver- Connell's 'disturbance' requires the indiscriminate
sity under such circumstances (section 7.3.1). removal of species from an area, and an environment
3 Equally specifically, the stable existence of top can be inconstant and unpredictable without this
predators will be favoured in predictably productive happening. It is, therefore, possible that this hypothe-
environments (Paine, 1966). It is, therefore, in such sized mechanism acts independently of the other
cases that they will be most potent in keeping poten- factors considered, and influences diversity in a wide
tial competitors below their carrying-capacities, and range of environments. Finally, the 'island' effects of
thus promoting diversity still further (section 7.3.2). habitat size (section 7.3.5) are likely to superimpose
4 Finally, these other mechanisms will tend to rein- their influences on community structure wherever
force one another. High diversity at a lower trophic they occur; and, to the extent that all environments
level will certainly provide for niche diversification are patchy, they are likely to occur everywhere.
(and thus increased diversity) at the next highest
trophic level; and it is possible that this will lead to an To summarize,then, we know a reasonable amount
increased intensity of predation, and thus a further about the potentialities of the various factors deter-
increase in diversity, at the lower trophic level. Small mining community structure, but rather less about
'inherent' differences in community structure are, their actual potencies and the patterns of their action
therefore, likely to become exaggerated. Note, how- in nature. Discovering the rules through which com-
ever, that this reinforcement will occur whatever the munities are constructed from populations is one of
the many exciting challenges that confronts popula-
tion ecology today.

Abrams P. (1976) Limiting similarity and the form of the Modelluntersuchungen uber die Ursachen der Popu-
competition coefficient. Theoretical Population Biology, 8, lations-bewegung des grauen Larchenwicklers Zeiraphera
356-75 diniana, Gn. ( =Z. griseana Hb). in Oberengadin 1949166.
Zeitschriftfuer Angewandte Entomologie, 62, 202-35
Ackert J. E., Graham G. L., Nolf L. 0. & Porter D. A. (1931) Ayala F. J., Gilpin M. E. & Ehrenfeld J. G. (1973) Competition
Quantitative studies on the administration of variable between species: theoretical models and experimental
tests. Theoretical Population Biology, 4, 331-56
numbers of nematode eggs (Ascaridia lineata) to chickens.
Transactions of the American Microscopical Society, 50, Bakker K. (1961) An analysis of factors which determine
206-14 success in competition for food among larvae of Drosophila
melanogaster. Archives Neerlandaises de Zoologie, 14, 200-81
Agricultural Research Council (1965)The Nutritional Require-
ments of Farm Livestock. 2. Ruminants. Agricultural Re- Bakker K. (1964)Backgrounds of controversies about popu-
search Council, London lation theories and their terminologies. Zeitschrift fuer
Angewandte Entomologie, 53, 187-208
Akinlosotu T. A. (1973)The role of Diaeretiella rapae (McIn-
tosh) in the control of the cabbage aphid. Unpublished PhD Banks C. J. (1957) The behaviour of individual coccinellid
thesis, University of London larvae on plants. Animal Behaviour, 5, 12-24

Allee W. C. (1931) Animal Aggregations. A Study in General Baskin J. M. & Baskin C. C. (1980) Ecophysiology of second-
Sociology. University of Chicago Press, Chicago ary dormancy in seeds of Ambrosia artemisifolia. Ecology,
61,475-80
Anderson R. M. (1974) Population dynamics of the cestode
Caryophyllaeus laticeps (Pallas 1781) in the bream (Abramis Baskin J. M. & Baskin C. C. (1983) Seasonal changes in the
brama L.). Journal of Animal Ecology, 43, 305-21 germination response of buried seeds of Arabidopsis thaliana
and ecologicalinterpretation.Botanical Gazette, 144,540-3
Anderson R. M. (1979)The influence of parasitic infection on
the dynamics of host population growth. In Population Beddington J. R. (1979)Harvesting and population dynamics.
In Population Dynamics (Anderson R. M., Turner B. D. &
Dynamics (Anderson R. M., Turner B. D. & Taylor L. R. Taylor L. R. eds),pp. 307-20. Blackwell ScientificPublica-
eds), pp. 245-81. Blackwell ScientificPublications, Oxford tions, Oxford
Anderson R. M. (1981) Population ecology of infectious
Beddington J. R., Free C. A. & Lawton J. H. (1975)Dynamic
disease agents. In Theoretical Ecology: Principles and Applica- complexity in predator-prey models framed in difference
tions, 2nd edn (R. M. May ed.), pp. 318-55. Blackwell equations. Nature, 225, 58-60
Scientific Publications, Oxford
Beddington J. R., Free C. A. & Lawton J. H. (1978)Modelling
Anderson R. M. & May R. M. (1978)Regulation and stability biological control: on the characteristics of successful
of host-parasite population interactions: 1. Regulatory natural enemies. Nature, 273, 513-19
processes. Journal of Animal Ecology, 47, 219-47
Beddington J. R. & Lawton J. H. (1978)On the structure and
Andrewartha H. G. & Birch L. C. (1954) The Distribution and behaviour of ecosystems. Journal de Physique, 39(c), 5-39
Abundance of Animals. University of Chicago Press, Chicago
Begon M. (1984)Density and individual fitness: asymmetric
Andrewartha H. G. & Birch L. C. (1960) Some recent competition. In Evolutionary Ecology (Shorrocks B., ed),
contributions to the study of the distribution and abun- pp. 175-194. Blackwell Scientific Publications, Oxford
dance of insects, Annual Review of Entomology, 5, 219-42
Begon M., Harper J. L. & Townsend C. R. (1990) Ecology.
Antonovics J. & Levin D. A. (1980)The ecologicaland genetic Individuals, Populations and Communities (2nd edn). Black-
consequences of density-dependent regulation in plants. well Scientific Publications, Oxford, 958 pp
Annual Reviews of Ecology and Systematics, 11,411-52
Bell A. D. & Tomlinson P. B. (1980)Adaptive architecture in
Archer S. & Tieszen L. L. (1983)Effects of simulated grazing rhizomatous plants. Botanical Journal of the Linnean Society,
on foliage and root production and biomass allocation in 80, 125-60
m arctic tundra sedge (Eriophorum vaginatum). Oecologia,
58.92-102 Bellows, T. S. & Hassell, M. P. (1988) The dynamics of
age-structured host-parasitoid interactions.Journal of Ani-
Atkinson W. D. & Shorrocks B. (1981) Competition on a mal Ecology, 57, 259-68
divided and ephemeral resource: a simulation model.

Journal of Animal Ecology, 50, 461-71
Auer C. (1968) Erste Ergebnisse einfacher stochasticher

226 REFERENCES

Belsky A. J. (1986)Does herbivory benefit plants?A review of Car1 E. (1971)Population control in arctic ground squirrels.
Ecology, 52, 395-413
the evidence. American Naturalist, 127, 870-92
Bengtsson J. (1991) Interspecificcompetition in metapopula- Caswell H. (1989) Matrix Population Models. Sinauer Associ-

tions. Biological Joumal of the Linnean Society, 42, 219-37 ates, New York
Beverton R. J. H. & Holt S. J. (1957) On the dynamics of Caughley G. & Lawton J. H. (1981) Plant-herbivore systems.

exploited fish populations. Fishely Investigations, London In Theoretical Ecology: Principles and Applications, 2nd edn
(Series II), 19, 1-533 (May R.M. ed.), pp. 132-66. Blackwell Scientific Publica-
Birkhead T. R. (1977) The effect of habitat and density on tions, Oxford
breeding success in the common guillemot (Uria aalge). Charles-EdwardsP. A. (1984)On the ordered developmentof
Journal of Animal Ecology, 46, 751-64 plants 2. Self-thinning in plant communities. Annals of
Blackman G. E. (1968) The applications of the concepts of
growth analysis to the assessment of productivity. In Botany 53, 709-14
Functioning of Terrestrial Ecosystems at the Primaly Produc- Chesson P. L. & Murdoch W. W. (1986)Aggregation of risk:
tion Level (Eckardt F. E. ed.), pp. 243-59. Unesco, Liege
relationships among host-parasitoid models. American
Boycott A. E. (1930) A re-survey of the Freshwater Mollusca Naturalist, 127, 696-715
of the parish of Aldenham after ten years with special Clark L. R. (1963) The influence of predation by Syrphus spp.
reference to the effect of drought. Transactions of the on the numbers of Cardiaspina albitextura (Psyllidae). Aus-

HertfordshireNatural History Society, 19, 1-25 tralian Journal of Zoology, 11,470-87
Bradshaw A. D. (1965) Evolutionary significance of pheno- Clark L. R., Geier P. W., Hughes R. D. & Morris R. F. (1967)

typic plasticity in plants. Advances in Genetics, 13, 115- The Ecology of Insect Populations in Theory and Practice.

55 Methuen, London
Branch G. M. (1975) Intraspecific competition in Patella Clarke T. A. (1970) Territorial behaviour and population

cochlear Born. Journal of Animal Ecology, 44,263-81 dynamics of a pomacentrid fish, the Garibaldi, Hypsypops
Brougham R. W. (1955) A study in rate of pasture growth. rubicunda. Ecological Monographs, 40, 189-212
Clements F. E., Weaver J. E. & Hanson H. C. (1929)
Australian Journal of Agricultural Research, 6, 804-12 Competition in cultivated crops. Carnegie Institute of Wash-
Brougham R. W. (1956)The rate of growth of short-rotation ington Publications, 398, 202-33
Comins H. N. & Hassell M. P. (1979) The dynamics of
ryegrass pastures in the late autumn, winter and early optimally foraging predators and parasitoids. Journal of
spring. New Zealand Journal of Science and Technology, A38, Animal Ecology, 48, 335-51
78-87 Connell J. H. (1961)The influence of interspecificcompetition
Brown J. H. & Davidson D. W. (1977) Competition between and other factors on the distribution of the barnacle
seed-eating rodents and ants in desert ecosystems. Science, Chthamalus stellatus. Ecology, 42, 710-23
196,880-2 Connell J. H. (1970) A predator-prey system in the marine
Brown M. W. & Cameron E. A. (1979)Effects of clisparlure and intertidal region. I. Balanus glandula and several predatory
egg mass size on parasitism by the gypsy moth egg parasite,
Ooencyrtus kuwani. Environmental Entomology, 8, 77-80 species of Thais. Ecological Monographs, 40, 49-78
Bryant J. P. (1981) Phytochemical deterrence of snowshoe Connell J. H. (1971) On the role of natural enemies in
hare browsing by adventitious shoots of four Alaskan trees.
Science, 213, 889-90 preventing competitive exclusion in some marine animals

Bulmer, M. G. (1975) The statistical analysis of density and in rain forest trees. In Dynamics of Populations
dependence. Biometrics, 31, 901-1 1 (den Boer P. J. & Gradwell G. R. eds), pp. 298-312. Centre
for Agricultural Publishing and Documentation, Wagenin-
Burnett T. (1954) Influences of natural temperatures and
controlled host densities on oviposition of an insect para- gen
site. Physiological Zoology, 27, 239-48 Connell J. H. (1973) Biology and Geology of Coral Reefs. Vol. 2.

Burnett T.(1956)Effects of natural temperatures on oviposi- (Jones,0. A. & Endean, R. eds), Academic Press, New York
tion of various numbers of an insect parasite (Hy- Connell J. H. (1979) Tropical rain forests and coral reefs as
menoptera, Chalcididae, Tenthredinidae). Annals of the
Entomological Society of America, 49, 55-9 open non-equilibrium systems. In Population Dynamics
(Anderson R. M., Turner B. D. & Taylor L. R. eds),
Burnett T. (1958)Dispersal of an insect parasite over a small pp. 141-63. Blackwell Scientific Publications, Oxford
plot. Canadian Entomologist, 90, 279-83 Connolly J. (1986) On difficulties with replacement series
methodology in mixture experiments. Joumal of Applied
Burrows F. M. (1991) Biomass production, structural defor-
Ecology, 23, 125-37
mation, self-thinning and thinning mechanisms in mono- Cook R. M. & Cockrell B. J. (1978)Predator ingestion rate and
cultures. Philosophical Transactions of the Royal Society of
London B, 333, 119-245 its bearing on feeding time and the theory of optimal diets.
Journal of Animal Ecology, 47, 529-48

REFERENCES 227

Cousens R. (1985)A simple model relating yield loss to weed Davy A. J. & Jeffries R. L. (1981) Approaches to the
density Annals of Applied Biology, 107, 239-52 monitoring of rare plant populations. In The Biological
Aspects of Rare Plant Conservation (Synge H. ed.), pp. 219-
Crawley, M. J. (1983)Herbivory. The Dynamics of Animal-Plant 232. John Wiley & Sons, New York
Interactions. Blackwell Scientific Publications, Oxford
DeBach P. & Smith H. S. (1941)The effect of host density on
Crawley M. J. (1985) Reduction of oak fecundity by low- the rate of reproduction of entomophagous parasites.
density herbivore populations. Nature, 314, 163-4 Journal of Economic Entomology, 34, 741-5

Crawley M. J. (1989)Insect herbivores and plant population Dempster, J. P. (1983) The natural control of populations of
dynamics Annual Review Entomology, 34, 531-64 butterflies and moths. Biological Reviews, 58, 461-81

Crawley M. J. (1993) The population dynamics of plants. de Wit C. T. (1960) On competition. Verslagen van Landbou-
Philosophical Transactions of the Royal Society of London B, wkundige Onderzoekinigen,66, 1-82
330,125-40
de Wit C. T., Tow P. G. & Ennik G. C. (1966) Competition
Crawley M. J. & Pacala S. (1991)Herbivores, plant parasites between legumes and grasses. Verslagen van Landbouwkun-
and plant diversity. In Parasitism: Conflict or Coexistence dige Onderzoekingen,687, 3-30
(Toft C. ed.), pp. 157-74. Oxford University Press, Oxford
Diamond J. M. (1975) Assembly of species communities. In
Crisp D. J. (1961) Territorial behaviour in barnacle settle- Ecology and Evolution of Communities (Cody M. L. & Dia-
ment. Journal of Experimental Biology, 38, 4 2 9 4 6
mond J. M. eds), pp. 342-444. Harvard University Press,
Crofton H. D. (1971)A model of host-parasite relationships. Cambridge
Parasitology, 63, 343-64 Diamond J. M. & May R. M. (1981)Island biogeography and
the design of nature reserves. In Theoretical Ecology: Princi-
Crombie A. C. (1947) Interspecific competition. Journal of ples and Applications. (R.M. May ed.), pp. 228-52. Blackwell
Animal Ecology, 16, 44-73 Scientific Publications, Oxford
Dixon A. F. G. (1971a)The role of aphids in wood formation.
Curio E. (1976) The Ethology of Predation. Springer-Verlag, I. The effect of the sycamore aphid Drepanosiphum pla-
Berlin tanoides (Schr.) (Aphididae) on the growth of sycamore,
Acer pseudoplatanus (L.). Journal of Applied Ecology, 8,
Davidson D. W. (1977a) Species diversity and community 165-79
organizationin desertseed-eatingants. Ecology, 58,711-24 Dixon A. F. G. (1971b)The role of aphids in wood formation
11. The effect of the lime aphid Eucallipterus tiliae L.
Davidson D. W. (1977b) Foraging ecology and community (Aphididae), on the growth of the lime, Tilia X vulgaris
organization in desert seed-eating ants. Ecology. 58, Hayne. Journal of Applied Ecology, 8, 393-409
725-37 Donald C. M. (1951) Competition among pasture plants.
I Intraspecfic competition among annual pasture plants.
Davidson D. W. (1978)Size variability in the worker caste of Australian Journal of Agricultural Research, 2, 355-76
a social insect (Veromessorpergandei Mayr) as a function of Donald C. M. (1961) Competition for light in crops and
the competitive environment. American Naturalist, 112, pastures. In Mechanisms in Biological Competition
523-32 (MilnthorpeF. L. ed.), Symposium of the Society of Experi-
mental Biology, No. 15, pp. 283-313
Davidson J. (1938)Ori the growth of the sheep population in Dunn E. (1977) Predation by weaseIs (Mustela nivalis) on
Tasmania. Transactions of the Royal Society of South Austra- breeding tits (Parus spp.) in relation to the density of tits
lia, 62, 342-6 and rodents, Journal of Animal Ecology, 46, 634-52

Davidson J. & Andrewartha H. G. (1948a)Annual trends in a EberhardtL. (1960)Michigan Department of ConservationGame
natural population of Thrips imaginis (Thysanoptera).Jour- Division Report No 2282
nal of Animal Ecology, 17, 193-9
Ehler L. E. (1987) Patch-exploitation efficiency in a torymid
Davidson J. & Andrewartha H. G. (1948b) The influence of parasite of a gall midge. Environmental Entomology, 16,
rainfall, evaporation and atmospheric temperature on 198-201
fluctuations in the size of a natural population of Thrips
irnaginis (Thysanoptera). Journal of Animal Ecology, 17, Elner R. W. & Hughes R. N. (1978) Energy maximisation in
200-22 the diet of the shore crab. Carcinus maenas (L.). Journal of
Animal Ecology, 47, 103-16
Davies N. B. (1977) Prey selection and social behaviour in
wagtails (Aves: Motacillidae).Journal of Animal Ecology, 46, Ennos R. (1985)The significance of genetic variation for root
37-57 growth within a natural population of white clover (Trifo-
lium repens). Journal of Ecology, 73, 615-24
Davies N. B. (1978a) Ecological questions about territorial
behaviour. In Behavioural Ecology: an Evolutionary Approach
(Krebs J. R. & Davies N. B. eds), pp. 317-50. Blackwell
Scientific Publications, Oxford

Davies N. B. (1978b)Territorial defence in the speckled wood
butterfly (Pamrgeaegeria),the resident always wins. Animal
Behaviour, 36, 138-47

228 REFERENCES

Errington P. L. (1946)Predation and vertebrate populations. Gause G. F. (1934) The Struggle for Existence. Williarns and
Quarterly Review of Biology, 21, 144-77 Wilkins, Baltimore. (Reprinted 1964, by Hafner, New

Erwin T. L. & Scott J. C. (1980) Seasonal and size patterns, York)
trophic structure, and richness of Coleoptera in the tropical Gill F. B. & Wolf L. L. (1975)Economics of feeding territorial-
arboreal ecosystem: the fauna of the tree Luehea seemannii
Triana and Planch in the Canal Zone of Panama. Coleopter- ity in the golden-winged sunbird. Ecology, 56, 3 3 3 4 5
ists' Bulletin, 34, 305-22 Gilpin M. & Hanski I. (1991) Metapopulation Dynamics:

Esau K. (1953)Plant Anatomy. Wiley, New York Empirical and Theoretical Investigations. Academic Press,

Fenchel T. (1975) Character displacement and coexistence in London
mud snails (Hydrobiidae). Oecologia, 20, 19-32 Godfray H. C. J., Cook L. M. & Hassell M. P. (1991)Population

Fernando M. H. J. P. (1977) Predation of the glasshouse red dynamics, natural selection and chaos. In Genes in Ecology.
spider mite by Phytoseiulus persmilis A.-H. Unpublished (Berry R. J., Crawford T. J. & Hewitt G. M. eds), Symposium
PhD thesis, University of London of the British Ecological Society No. 33, pp. 55-86 Black-
well Scientific Publications, Oxford
Finlayson L. H. (1949)Mortality of Laemophloeus (Coleoptera.
Cucujidae) infected with Mattesia dispora Naville (Protozoa, Goh B. S. (1978) Robust stability concepts for ecosystem
Schizogregarinaria).Parasitology, 40, 261-4 models. In Theoretical Systems Ecology (Halfon E. led.),

Firbank L. G., Cousens R., Mortimer A. M. & Smith R. R. Academic Press, New York
(1990) Effects of soil type on crop yield-weed density Gold W. G. & Caldwell M. M. (1989)The effects of the spatial
relationships. Joumal of Applied Ecology, 27, 308-318
pattern of defoliation on regrowth of a tussock grass. I1
Firbank L. G. & Watkinson A. R. (1985) On the analysis of Canopy gas exchange. Oecologia, 81, 437-42
competition within two-species mixtures of plants. Journal Goldberg D. E. (1987) Neighborhood competition in an
of Applied Ecology, 22, 503-17 old-field plant community. Ecology, 68, 1211-33
Goldberg D. E. (1990)Componentsof resource competitionin
Firbank L. G. & Watkinson A. R. (1990) On the effects of plant communities. In Perspectives on Plant Competition
competition: from monocultures to mixtures. In Perspec- (GraceJ. B. & Tilman D. eds), pp. 27-50. Academic Press,
tives on Plant Competition (Grace J. B. & Tilman D. eds), San Diego
pp. 165-92. Academic Press, San Diego Gorman M. L. (1979) Island Ecology. Chapman and Hall,
London
Ford E. D. (1975) Competition and stand structure in some
even-aged plant monocultures. Joumal of Ecology, 63, Gotelli N. J. (1991)Metapopulation models: the rescue eflkct,
311-33 the propagule rain, and the core-satellite hypothesis.
American Naturalist, 138, 768-76
Free C. A., Beddington J. R. & Lawton J. H. (1977) On the
inadequacy of simple models of mutual interference for GottleibL. D. (1984)Genetics and morphological evolution in
parasitism and predation. Journal of Animal Ecology, 46, plants. American Naturalist, 123, 681-709
543-54
Griffiths K. J. (1969)Development and diapause in Pleolophus
Fujii K. (1967) Studies on interspecies competition between basizonus (Hymenoptera: Ichneumonidae). Canadian Ento-
the azuki bean weevil Callosobruchus chinensis and the mologist, 101, 907-14
southern cowpea weevil, C. maculatus. 11. Competition
under different environmental conditions. Researches in Grubb P. J. (1977) The maintenance of species richness in
Population Ecology, 9, 192-200 plant communities: the importance of the regeneration
niche. Biological Reviews, 52, 107-45
Fujii K. (1968) Studies on interspecies competition between
the azuki bean weevil and the southern cowpea weevil: 111, Gulland J. A. (1962) The application of mathematical models
some characteristicsof strains of two species. Researches in to fish populations. In The Exploitation of Natural Animal
Population Ecology, 10, 87-98 Populations (Le Cren E. D. & Holdgate M. W. eds), Sympo-
sium of the British Ecological Society No. 2, pp. 204-17.
Blackwell Scientific Publications. Oxford

Garwood N. C. (1989) Tropical soil seed banks: a review. In Haines B. (1975) Impact of leaf cutting ants on vegetation
Ecology of soil seed banks (LeckM. A., Parker V. T. & Simpson development at Barro Colorado Island. In Tropical Ecologi-
R. L. eds), pp. 149-210. Academic Press, San Diego cal Systems-Trends in Terrestrial and Aquatic Research (Gol-
ley F. B. & Medina E. eds), pp. 99-111. Springer-Verlag,
Gaston K. J. & Lawton J. H. (1990)Effects of scale and habitat
New York
on the relationship between regional distribution and local Haldane J. B. S. (1949)Disease and evolution. Symposium sui

abundance. Oikos, 58, 329-35 fattori ecologici e genetici della speciazone negli animali.
Gatsuk, L. E., Smirnova,0.V., Vorontzova, L. I., Zaugolnova, Ric. Sci. 19 (suppl.). 3-11
Hancock D. A. (1979)Population dynamics and management of
L. B. & Zhukova, L. A. (1980)Age states of plants of various

growth forms: a review. Journal of Ecology, 68, 675-96

REFERENCES 229

shellfish stocks. ICES Special Meeting on Population Assess- growth form in organisms with modular construction. In
Population Dynamics (Anderson R. M., Turner B. D. &
ment of Shellfish Taylor L. R. eds), pp. 29-52. Blackwell Scientific Publica-
Hanski I. (1981) Coexistence of competitors in patchy envi- tions, Oxford
Harrison S. (1991) Local extinction in a metapopulation
ronments with and without predation. Oikos, 3 7 , 306-12
Hanski I. (1982) Dynamics of regional distribution: the core context: an empirical evaluation. Biological Journal of the
Linnean Society, 4 2 , 73-88
and satellite hypothesis. Oikos, 3 8 , 210-21 Harrison S., Murphy D. D. & Ehrlich P. R. (1988)Distribution
Hanski I. (1987) Colonisation of ephemeral habitats. In of the bay checkerspot butterfly, Euphydqas editha
bayensis: evidence for a metapopulation model. American
Colonization, Succession and Stability. (Gray A. J., Crawley Naturalist, 132, 360-82.
M. J. & Edwards P. J, eds), Symposium of the British Hassell M. P. (1971a)Mutual interferencebetween searching
Ecological Society No. 26, pp. 155-85. Blackwell Scientific insect parasites. Journal of Animal Ecology, 4 0 , 473-86
Publications, Oxford Hassell M. P. (1971b)Parasite behaviour as a factor contrib-
Hanski I. (1989) Metapopulation dynamics: does it help to uting to the stabilityof insect host-parasite interactions.In
have more of the same? Trends in Ecology and Evolution, 4 , Dynamics of Populations, (den Boer P. J. & Gradwell G. R.
1l3-14 eds), pp. 366-79. Centre for Agricultural Publishing and
Hanksi I. (1990) Density dependence, regulation and vari-
ability in animal populations. Philosophical Transactions of Documentation, Wageningen
Hassell M. P. (1975) Density-dependence in single-species
the Royal Society of London B, 330, 141-50
Hmski I. (1991) Single-species metapopulation dynamics: populations. Journal of Animal Ecology, 44,283-95
Hassell M. P. (1976) The Dynamics of Competition and Preda-
concepts, models and observations. Biological Journal of the
Linnean Society, 4 2 , 17-38 tion. Edward Arnold, London
Hatnski I. & Gilpin M. (1991) Metapopulation dynamics: brief Hassell M. P. (1978)The Dynamics of Arthropod Predator-Prey

history and conceptual domain. Biological Journal of the Systems. Princeton University Press, Princeton
Linnean Society, 4 2 , 3-16 Hassell M. P. (1985) Insect natural enemies as regulating

Hanski I. & Kuusela S. (1977)An experiment on competition factors. Journal of Animal Ecology, 54, 323-34
and diversity in the carrion fly community. Annales Zoo- Hassell M. P. & Comins, H. N. (1976)Discrete time models for

logici Fennici, 4 3 , 108-15 two-species competition. Theoretical Population Biology, 9 ,
Hanski I., Turchin P., Korpimaki E. & Henttonen H. (1993)
202-21
Population oscillations of boreal rodents: regulation by Hassell M. P. & Comins H. N. (1978) Sigmoid functional
mustelid predators leads to chaos. Nature, 364, 232-5
Hanski I. & Woiwod I. P. (1991)Delayed density dependence responses and population stability. Theoretical Population
Nature, 350, 28 Biology, 14, 62-7
Hara T. (1988) Dynamics of size structure in plant popula- Hassell M. P., LattoJ. & May R. M. (1989)Seeing the wood for
tion~T. rends in Ecology and Evolution 3, 129-33 the trees: detecting density dependence from existing
Harberd D. J. (1961)Some observations on natural clones in
Festuca ovina L. New Phytologist, 6 0 , 184-206 life-table studies.Journal of Animal Ecology, 58, 883-92
Harcourt D. G. (1964) Population dynamics of Leptinotarsa Hassell M. P., Lawton J. H. & Beddington J. R. (1977)Sigmoid
decemlineata (Say) in eastern Ontario. 11. Population and
mortality estimation during six age intervals. Canadian functional responses by invertebrate predators and parasi-
Entomologist, 9 6 , 1190-8 toids. Journal of Animal Ecology, 4 6 , 249-62
Hassell M. P., Lawton J. H. & May R. M. (1976) Patterns of
Harcourt D. G. (1971) Population dynamics of Leptinotarsa
decemlineata (Say)in eastern Ontario. 111. Major population dynamical behaviour in single species populations. Journal
of Animal Ecology, 45, 471-86
processes. Canadian Entomologist, 103, 1049-61 Hassell M. P. & May R. M. (1973) Stability in insect host-
Harper J. L. (1977)The Population Biology of Plants. Academic parasite models. Journal of Animal Ecology, 42, 693-736

Press, London and New York Hassell M. P. & May R. M. (1974) Aggregation of predators
Harper J. L. (1981) The concept of population in modular and insect parasites and its effect on stability. Journal of
Animal Ecology, 4 3 , 567-94
organisms. In Theoretical Ecology (May R. ed.), Blackwell
Scientific Publications, Oxford Hassell, M. P. & Pacala, S. W. (1990)Heterogeneity and the
Harper J. L. (1985) Modules, branches and the capture of dynamics of host-parasitoid interactions. Philosophical
resources. In Population Biology and Evolution of Clonal Transactions of the Royal Society of London, B, 330, 203-20
Organisms (Jackson J. B. C., Buss L. W. & Cook R. E. eds),
pp 1-34. Yale University Press, New Haven Hassell M. P., Pacala S. W., May R. M. & Chesson P. L. (1991)
Harper J. L. (1989)The value of a leaf. Oecologia, 8 0 , 53-8 The persistence of host-parasitoid associations in patchy
Harper J. L. & Bell A. D. (1979)The population dynamics of environments. I. A general criterion. American Naturalist,
138, 568-83

230 REFERENCES

Hassell M. P. & Rogers D. J. (1972)Insect parasite responses Huffaker C. B. & Kennett C. E. (1959) A 10 year study of
in the development of population models. Journal of Animal vegetational changes associated with biological control of
Klamath weed species. Journal of Range Management, 12,
Ecology, 41, 661-76 69-82
Hassell M. P., Southwood T. R. E. & Reader P. M. (1987) The
Huffier C. B., Shea K. P. & Herman S. G. (1963)Experirnen-
dynamics of the viburnum whitefly (Aleurotrachelus tal studies on predation. Hilgardia, 34, 305-30
jelinekii): a case study of population regulation. Journal of
Animal Ecology, 56, 283-300 Hughes T. P. (1984)Population dynamics based on individual
Hassell M. P. & Varley G. C. (1969)New inductive population size rather than age: a general model with a reef coral
model for insect parasites and its bearing on biological example. American Naturalist, 123, 778-95
control. Nature, 223, 1133-6
Hastings A., Hom C. L., Ellner S., Turchin P. & Godfray H. C. Hutchings M. J. (1985) The structure of plant populations In
Plant Ecology (Crawley, M. J. ed.), pp. 97-136. Blackwell
J. (1993) Chaos in ecology: is mother nature a strange Scientific Publications, Oxford
attractor?Annual Review ofEcology and Systematics,24,l-33
Hatto J. & Harper J. L. (1969)The control of slugs and snails Hutchinson G. E. (1957) Concluding remarks. Cold Spring
in British cropping systems, specially grassland. Interna- Harbour Symposium on Quantitative Biology, 22, 415-2'7
tional Copper Research Association Project 115(A), 1-25
Heads P. A. & Lawton J. H. (1983) Studies on the natural Iles T. D. (1973)Interaction of environmentand parent stock
enemy complex of the holly leaf-miner: the effect of scale size in determing recruitment in the Pacific sardine as
on the aggregative responses and the implications for revealed by analysis of density-dependent 0-group growth.
biological control. Oikos, 40, 267-76 Rappouts et Procesverbaux, Conseil Internationalpour I'Explor-
Healey M. C. (1967) Aggression and self regulation of tion de la Mer, 164, 228-40
population size in deermice. Ecology, 48, 377-92
Heed W. B., Starmer W. T., Miranda M., Miller M. W. & Phaff' Inouye D. W. (1978) Resource partitioning in bumblebees:
H. S. (1976)An analysis of the yeast flora associated with experimental studies of foraging behaviour. Ecology, 59,
cactiphilic Drosophila and their host plants in the Sonoran 672-8

Desert and its relation to temperate and tropical associa- Ives A. R. (1991)Aggregation and coexistence in a carrion fly
community. Ecological Monographs, 61, 75-94
tions. Ecology, 57, 151-60
Hodgson G. L. & Blackman G. E. (1956) An analysis of the Ives A. R. (1992a)Continuous-timemodels of host-parasiltoid
interactions. American Naturalist, 140, 1-29
influence of plant density on the growth of Viciafaba 1.The
influence of density of the pattern of development.Journal Ives A. R. (1992b) Density-dependent and density-
of experimental Botany. 7, 147-65 independent parasitoid aggregation in model host-
Holling C. S. (1959) Some characteristics of simple types of parasitoid systems. American Naturalist, 140, 912-37
predation and parasitism. Canadian Entomologist, 91,
385-98 Janzen D. H. (1966)Coevolution of mutualism between ants
Holling C. S. (1965) The functional response of predators to and acacias in Central America. Evolution, 20, 249-75
prey density and its role in mimicry and population
regulation. Memoirs of the Entomological Society of Canada, Janzen D. H. (1971) Escape of juvenile Dioclea megacarpa
45.43-60 (Leguminosae)vines from predators in a deciduous tropical
Holling C. S. (1966) The functional response of invertebrate forest. American Naturalist, 105, 97-112
predators to prey density. Memoirs of the Entomological
Society of Canada, 48, 1-86 Janzen D. H. (1972a) Association of a rainforest palm and
Hoppensteadt F. C. (1982)Mathematical Methods of Population seed-eating beetles in Puerto Rica. Ecology, 53, 258-61

Biology. Cambridge University Press, Cambridge Janzen D. H. (1972b) Escape in space by Sterculia apetala
Horn H. S. (1971) The Adaptive Geometry of Trees. Princeton seeds from the bug Dysdercus fasciatus in a Costa Kcan
deciduous forest. Ecology, 53, 350-61
University Press, Princeton
Horton K. W. (1964)Deer prefer JackPine. Journal ofForestry, Janzen D. H. (1975) Interactions of seeds and their insect
predatorslparasitoids in a tropical deciduous forest. In
62,497-9 Evolutionary Strategies of Parasitic Insects and Mites (Price
Hubbard S. F. (1977)Studies on the natural control of Pieris P. W. ed.), pp. 154-86. Plenum, New York

brassicae with particular reference to parasitism by Apan- Jonsdottir I. S. & Callaghan T. V. (1988) Interrelationships
teles glomeratus. Unpublished D Phi1 thesis, University of between different generations of interconnected tillers of
Oxford Carex bigelowii. Oikos, 52, 120-8
Huffaker C. B. (1958) Experimental studies on predation:
dispersion factors and predator-prey oscillations. Hilgar- Jordano D. & Thomas C. D. (1992) Specificity of an =ant-
lycaenid interaction. Oecologia, 91, 431-8
dia, 27, 343-83
Kaban R. & Myers J. H. (1989) Induced plant responses to
herbivory. Annual Review Ecology and Systematics, 20,
331-48

REFERENCES 231

Kays S. & Harper J. L. (1974)The regulation of plant and tiller Lawton J. H. & McNeill S. (1979) Between the devil and the
density in a grass sward. Journal of Ecology, 62, 97-105 deep blue sea: on the problem of being a herbivore. In
Population Dynamics (Anderson R. M., Turner B. D. &
Keddy P. A. (1981) Experimental demography of the sand Taylor L. R. eds),pp. 2 2 3 4 4 . Blackwell Scientific Publica-
dune annual Cakile edentula growing along an environmen- tions, Oxford
tal gradient in Nova Scotia. Journal of Ecology, 69, 615-30
Lawton J. H. & Pimm S. L. (1978) Population dynamics and
Keith L. B. (1963) Wildlije's Ten-year Cycle. University of the length of food chains. Nature, 272, 190
Wisconsin Press, Madison
Leslie P. H. (1945) On the use of matrices in certain
Kira T.. Ogawa H. & Shinozaki K. (1953) Intraspecific population mathematics. Biometrika, 33, 183-212
competition among higher plants. I. Competition-density-
yield interrelationshipsin regularly dispersed populations. Levins R. (1968)Evolution in Changing Environments. Prince-
Journal of the Polytechnic Institute, Osaka City University, ton University Press, Princeton
4(4), 1-16
LevinsR. (1969)Somedemographicand geneticconsequences
Klomp H. (1964)Intraspecificcompetition and the regulation of environmentalheterogeneity for biological control. Bul-
of insect numbers. Annual Review of Entomology, 9, 17-40 letin of the EntomologicaI Society of America, 15, 237-40

Kluyver H. N. (1951)The population ecology of the great tit, Levins R. (1970) Extinction. In Some mathematical problems
Parus m. major L. Ardea, 38, 1-135. in biology. (Gerstenhaber M. ed.), pp. 77-107, American
Mathematical Society, Providence, R.I.
Konthanen P. (1950) Quantitative and seasonal studies on
the leafhopper fauna of the field stratum on open areas of Lonsdale W. M. (1990)The self-thinning rule: dead or alive?
North Karelia. Annales Zoologici Society 'Vanamo', 13, 1-91 E c o ~ o7~1~, 1, 373-88

Krebs J. R. (1971)Territory and breeding density in the great Lonsdale W. M. & Watkinson A. R. (1982) Light and
tit, Parus major L. Ecology, 52, 2-22 self-thinning. New Phytologist, 90, 431-45

Krebs J. R. (1978) Optimal foraging: decision rules for Lotka A. J. (1925) Elements of Physical Biology. Williams &
predators. In Behavioural Ecology: an Evolutionary Approach W i l h s , Baltimore
(Krebs J. R. & Davies N. B. eds), pp. 23-63. Blackwell
Scientific Publications, Oxford Lowe V. P. W. (1969) Population dynamics of the red deer
(Cervus elaphus L.) on Rhum. Journal of Animal Ecology, 38,
Krebs J. R., Kacelnik A. & Taylor P. J. (1978)Test of optimal 425-57
sampling by foraging great tits. Nature, 275, 27-31
MacArthur R. H. (1965) Patterns of species diversity. Biolo-
Kuchlein J. H. (1966)Mutual interference among the preda- gical Reviews, 40, 1510-33
cious mite Typhlodromus longipilus Nesbitt (Acari,Phytosei-
idae). I. Effects of predator density on oviposition rate and MacArthur R. H. (1972) Geographical Ecology. Harper Row,
migration tendency. Mededelingen Rijksfaculteit Landbouw-
wetenschappen te Gent, 31, 740-6 New York
MacArthur R. H. & Levins R. (1967) The limiting similarity,
Lack D. (1971) Ecological Isolation in Birds. Blackwell Scien-
tific Publications, Oxford convergenceand divergence of coexisting species.American
Naturalist, 101, 377-85
Lanciani C. A. (1975) Parasite-induced alterations in host MacArthur R. H. & Pianka E. R. (1966) On optimal use of a
reproduction and survival. Ecology, 56, 689-95 patchy environment. American Naturalist, 100, 603-9
MacArthur R. H. & Wison E. 0. (1967) The Theory of Island
Langer R. H. M. (1956) Growth and nutrition of timothy Biogeography. Princeton University Press, Princeton
(Phleum pratense). The life history of individual tillers. MacLulick D. A. (1937) Fluctuations in the numbers of the
Annals of Applied Biology, 44, 166-87 varying hare (Lepus americanus). University of Toronto Stud-
ies, Biology Series, 43, 1-136
Law R. (1975)Colonisation and the evolution of life histories Manlove, R. J. (1985)On the population ecology of Avena fatua
in Poa annua. Unpublished PhD thesis, University of Liver- L. Unpublished PhD thesis, University of Liverpool
pool Marshall D. R. & Jain S. K. (1969) Interference in pure and
mixed populations of Avena fatua and A. barbata. Journal of
Law R. (1979)Harvest optirnisation in populations with age Ecology, 57, 251-70
distributions. American Naturalist, 114, 250-9 May R. M. (1972) Will a large complex system be stable?
Nature, 238, 13-14
Law R. & Watkinson A. R. (1987)Response-surfaceanalysisof May R. M. (1973)Stability and Complexity in Model Ecosystems.
two-species competition: an experiment on Phleum arenar- Princeton University Press, Princeton
ium and Vulpia fasiculata. Journal of Ecology, 75, 871-86 May R. M. (1974) Biological populations with non-
overlapping generations: stable points, stable cycles and
Lawrence W. H. & Rediske J. H. (1962) Fate of sown chaos. Science, 186, 645-7
douglas-fir seed, Forest Science, 8, 211-18 May R. M. (1975) Biological populations obeying difference

Lawton J. H., Beddington J. R. & Bonser R. (1974)Switching
in invertebrate predators. In Ecological Stability (Usher M.
B. & Williamson M. H. eds), pp. 141-58. Chapman & Hall,
London

232 REFERENCES

equations: stable points, stable cycles and chaos. Journal of Moore N. W. (1964) Intra- and interspecific competition
Theoretical Biology, 49, 511-24 among dragonflies. Journal of Animal Ecology, 33, 49-71

May R. M. (1976) Simple mathematical models with very Morris R. F. (1959) Single-factor analysis in population
dynamics. Ecology, 40, 580-8
complicated dynamics. Nature, 261, 459-67
Mortimer A. M., Sutton J. J. & Gould P. (1989) On robust
May R. M. (1977)Thresholds and breakpoints in ecosystems weed population models. Weed Reseach, 29, 229-3 8
with a multiplicity of stable states. Nature, 269, 471-7
Mueller L. D. & Ayala F. J. (1981) Dynamics of single-species
May R. M. (1978a) Host-parasitoid systems in patchy population growth: stability or chaos? Ecology, 62,
environments: a phenomenological model. Journal of Ani- 1148-54
mal Ecology, 47, 833-43
Murdoch W. W. (1969) Switching in general predators:
May R. M. (1978b) The dynamics and diversity of insect experiments on predator specificity and stability of prey
faunas. In Diversity of Insect Faunas. (Mound L. A. & Waloff populations. Ecological Monographs, 39, 335-54
N. eds), pp. 188-204. Blackwell Scientific Publications,
Murdoch W. W. & Oaten A. (1975)Predation and population
Oxford stability. Advances in Ecological Research, 9, 2-131

May R. M. (1979)The structure and dynamics of ecological Murdoch W. W. & Stewart-Oaten A. (1989) Aggregation by
communities. In Population Dynamics (Anderson R. M., parasitoids and predators: effects on equilibrium and
stability.American Naturalist, 134, 288-310
Turner B. D. & Taylor L. R. eds), pp. 385-407. Blackwell
Murphy G. I. (1967) Vital statistics of the Pacific sardine
Scientific Publications, Oxford (Sardinops caerulea)and the population consequences. Ecol-
ogy, 48, 731-6
May R. M. (1988) How many species are there on Earth?
Science, 241, 1441-9 Murton R. K. (1971) The significance of a specific search
image in the feeding behaviour of the wood pigeon.
May R. M. & Anderson R. M. (1978)Regulation and stability Behaviour, 40, 10-42

of host-parasite population interactions. 11. Destabilizing Murton R. K., Westwood N. J. & Isaacson A. J. (1964) The
processes.Journal of Animal Ecology, 47, 249-68 feeding habits of the wood pigeon Columba palumbus, stock
May R. M. & Oster G. F. (1976) Bifurcations and dynamic dove, C. oenas and the turtle dove, Streptopelia turtur. Ibis,
complexity in simple ecological models. American Natural- 106,174-88
ist, 110, 573-99
Maynard Smith J. (1976)Group selection. Quarterly Review of Nault A. & Gagnon D. (1993) Ramet demography of Allium
Biology, 51, 277-83 tricoccum a spring ephemeral,perennial forest herb. Journal
McClure M. S. (1977) Parasitism of the scale insect, Fiorinia of Ecology, 81, 101-19
externa (Hornoptera: Diaspididae)by Aspidiotiphagus citrinus
(Hymenoptera: Eulophidae), in a hemlock forest: density Newsome A. E. (1969a) A population study of house-mice
dependence. Environmental Entomology, 6, 551-55 temporarily inhabiting a South Australian wheatfield.
McNaughton S. J. (1983) Physiological and ecological irnpli- Journal of Animal Ecology. 38, 341-60
caitons of herbivory. Encyclopaedia of Plant Physiology:
Physiological Plant Ecology 111 Responses to the chemical and Newsome A. E. (1969b) A population study of house-]mice
biological environment (Lange 0. L., Nobel P. S., Osmond
permanently inhabiting a reed-bed in South Australia.
C. B. & Zeigler H. eds),pp. 657-77. Springer-Verlag, Berlin Journal of Animal Ecology, 38, 361-77
McNeill S. & SouthwoodT. R. E. (1978)The role of nitrogen in Nicholson A. J. (1933) The balance of animal populations.
Journal of Animal Ecology, 2, 131-78
the development of insectlplant relationships. In Biochemi- Nicholson A. J. (1954a) Compensatory reactions of popula-
cal Aspects of Plant and Animal Coevolution (Harborne J . B. tions to stress, and their evolutionary significance. Austra-
lian Journal of Zoology, 2, 1-8
ed.), pp. 77-98. Academic Press, London and New York Nicholson A. J. (1954b)An outline of the dynamics of animal
Mead-Briggs A. R. & Rudge A. J. B. (1960) Breeding of the populations. Australian Journal of Zoology, 2, 9-65
Nicholson A. J. (1957) The Self-adjustment of Populations to
rabbit flea, Spilopsyllus cuniculi (Dale): requirement of a Change. Cold Spring Harbour Symposium in Quantitative
'factor' from a pregnant rabbit for ovarian maturation. Biology, No. 22, pp. 153-72
Nature, 187, 1136-7 Nicholson A. J. (1958) Dynamics of insect populations.
Annual Review of Entomology, 3, 107-36
Mertz R. W. & Boyce S. G. (1956) Age of oak 'seedlings'. Nicholson A. J. & Bailey V. A. (1935)The balance of animal
Journal of Forestry, 54, 774-75 populations. Proceedings of the Zoological Society of Lorndon,
3, 551-98
Michelakis S. (1973) A study of the laboratory interaction Noyes J. S. (1974) The biology of the leek moth, Acrolepia
between Coccinella septempunctata larvae and its prey

Myzus persicae. Unpublished MSc thesis, University of
London

Monro J. (1967) The exploitation and conservation of re-
sources by populations of insects. Journal of Animal Ecology,

36, 531-47

REFERENCES 233

assectella (Zeller). Unpublished PhD thesis, University of compartments?Journal of Animal Ecology, 49, 879-98
London Piper E. L. & Weiss A. (1993) Defoliation during vegetative
Noy-Meir I. (1975)Stability of grazing sytems: an application
of predator-prey graphs. Journal of Ecology, 63, 459-83 growth of corn-the shoot-root ratio and yield implica-
tions. Field Crops Research, 31, 145-53
Obeid M., Machin D. & Harper J. L. (1967) Influence of Podoler H. & Rogers D. (1975) A new method for the
density on plant to plant variations in Fiber Flax, Linum identification of key factors from life-table data. Journal of
usitatissimum. Crop Science, 7, 471-3 Animal Ecology, 44, 85-114
Pollard E., Lakhani K. L. & Rothery P. (1987)The detection of
Oinonen E. (1967)The correlation between the size of Finnish density dependence from a series of annual censuses.
bracken (Pteridium aquilinum (L.) Kuhn) clones and certain Ecology, 68, 2046-55
periods of site history. Acta Forestalia Fennica, 83, 1-51 Poole R. W. (1978) An Introduction to Quantitative Ecology.
McGraw-Hill, New York
Osawa A. & Sugita S. (1989)The self-thinning rule: another Porter J. R. (1983a)A modular approach to analysis of plant
interpretation of Weller's results. Ecology, 70, 279-83 growth. I. Theory and principles. New Phytologist, 94,
183-90
Pacala S. W. & Hassell M. P. (1991) The persistence of Porter J. R. (1983b) A modular approach to analysis of plant
host-parasitoid associations in patchy environments. 11. growth. 11. Methods and results. New Phytologist, 94,
Evaluation of field data. American Naturalist, 138, 584-605 191-20
Pratt D. M. (1943) Analysis of population development in
Pacala S. W., Hassell M. P. & May R. M. (1990) Host- Daphnia at different temperatures. Biological Bulletin, 85,
parasitoid associations in patchy environments. Nature, 116-40
344,150-3 Price P. W. (1980)Evolutionary Biology of Parasites. Princeton
University Press, Princeton
Pacala S. W. & SilanderJ. A. (1985)Neighbourhood models of Puckridge D. W. & Donald C. M. (1967)Competition among
plant population dynamics. I. Single-species models of wheat plants sown at a wide range of densities. Australian
annuals. American Naturalist, 125, 385-41 1 Journal of Agricultural Research, 17, 193-21 1
Putwain P. D. & Harper J. L. (1970)Studies on the dynamics
Paine R. T. (1966)Food web complexity and species diversity. of plant populations. 111.The influence of associated species
American Naturalist, 100, 65-75 on populations of Rumex acetosa L. and R. acetosella L. in
grassland. Journal of Ecology, 58, 251-64
Palrnblad I. G. (1968) Competition studies on experimental
populations of weeds with emphasis on the regulation of Radovich J. (1962) Effects of sardine spawningstock size and
population size. Ecology, 49, 26-34 environment on year-class production. California Depart-
ment Fish and Game Bulletin, 48, 123-40
Park T. (1954)Experimental studies of interspecificcompeti-
tion. 11. Temperature, humidity and competition in two Randolph S. E. (1975) Patterns of distributions of the tick
species of Tribolium. Physiological Zoology, 27, 177-238 Ixodes trianguliceps Birula, on its host. Journal of Animal
Ecology, 44, 451-74
Pearl R. (1927)The growth of populations. QuarterlyReview of
Biology, 2, 532-48 Raup M. J. & Tallamy, D. W. (1990). Phytochemical Induction
by Herbivores. John Wiley & Sons, New York
Pearl R. (1928) The Rate of Living. Knopf, New York
Perrins C. M. (1965)Population fluctuationand clutch-size in Redfern M., Jones T. H. & Hassell M. P. (1992)Heterogeneity
and density dependence in a field study of a tephritid-
the great tit (Parus major L.). Journal of Animal Ecology, 34, parasitoid interaction. Ecological Entomology, 17, 255-62
601-47
Perry J. N., Woiwod I. P. & Hanski I. (1993) Using response- Rees M. & Crawley M. J. (1989) Growth, reproduction and
surface methodology to detect chaos in ecological time population dynamics. Functional Ecology, 3, 645-53
series. Oikos, 68, 329-39
Pilbeam C. J., HebblethwaiteP. D., Ricketts, H. E. & Nyongesa Rice B. & Westoby M. (1988)Evidence against the hypothesis
T. E. (1991)Effects of plant population density on determi- that ant dispersed seeds reach nutrient-enriched micros-
nate and indeterminate growth forms of winter field beans ites. Ecology, 67, 1270-4
(Vicia faba) 1. Yield and yield components. Journal of
Agricultural Science, 116, 375-84 Richards 0.W. & Waloff N. (1954)Studies on the biology and
Pimentel D. (1961) On a genetic feed-back mechanism population dynamics of British grasshoppers. Anti-Locust
regulating populations of herbivores, parasites and preda- Bulletin, 17, 1-182
tors. American Naturalist, 95, 67-79.
Pimm S. L. & Lawton J. H. (1977)Number of trophic levels in Richman S. (1958)The transformation of energy by Daphnia
ecological communities. Nature, 268, 329-3 1 pulex. Ecological Monographs, 28, 273-91
Pimm S. L. & Lawton J. H. (1978) On feeding on more than
one trophic level. Nature, 275, 542-4 Ricker W. E. (1954) Stock and recruitment. Journal of the
Pimm S. L. & Lawton J. H. (1980) Are foodwebs divided into

234 REFERENCES

Fisheries Research Board of Canada, 11, 559-623. of two competing annual plant species. Ecology, 54,
Rigler F. H. (1961) The relation between concentration of 72340
Shinozaki K. & Kira T. (1956) Intraspecific competition
food and feeding rate of Daphnia magna Straus. Canadian
Journal of Zoology, 39, 857-68 among higher plants. VII. Logistic theory of the C-D eflrect.
Roberts E. H. (1972) Dormancy: a factor affecting seed Journal of Institute Polytechnic of Osaka City Universitgl, 7 ,
survival in the soil. In Viability of seeds (Roberts E. H . ed.),
pp. 321-59. Chapman Hall, London 35-72
Roberts H. A. & Dawkins P. A. (1967) Effect of cultivation on Shorrocks B. (1970) Population fluctuations in the fruit fly
the numbers of viable weed seeds in the soil. Weed
Research, 7, 290-301 (Drosophila melanogaster) maintained in the laboratory.
Journal of Animal Ecology, 39, 229-53
Root R. B. (1967) The niche exploitation pattern of the
blue-gray gnatcatcher. Ecological Monographs, 37, 317-50 Shorrocks B. & Rosewell J. (1987) Spatial patchiness and
community structure: coexistence and guild size of Droso-
Rosen B. R. (1979) Modules, members and communes: a phila on ephemeral resources. In The Organization of
postscript introduction to social organisms. In Biology and Communities: Past and Present. (Gee J. H. R. & Giller P. S.
Systematics of Colonial Organisms (LarwoodG. & Rosen B. R.
eds), Academic Press, London eds), Symposium of the British Ecological Society No. 27,

Rosenzweig M. L. & MacArthur R. H. (1963) Graphical pp. 29-51. Blackwell Scientific Publications, Oxford
representation and stability conditions of predator-prey
interactions.American Naturalist, 97, 209-23 Shorrocks B., Rosewell J. & Edwards K. (1990) Competition
on a divided and ephemeral resource. 11. Association.
Rosewell J.. ShorrocksB. & Edwards K. (1990)Competition on Journal of Animal Ecology, 59, 1003-7
a divided and ephemeral resource: testing the assumptions.
I. Aggregation. Journal ofAnima1 Ecology, 59, 977-1001 Sillirnan R. P. & GutsellJ. S. (1958) Experimentalexploitation
of fish populations. Fishery Bulletin, Fish Wildlife Service.
Ross M. A. & Harper J. L. (1972) Occupation of biological
space during seedling establishment.Journal of Ecology, 60, US., 58,215-41
77-88 SilvertownJ. W. (1982)Introduction to Plant Population Ecology

Rotheray G. E. (1979) The biology and host searching Longmans, London
behaviour of a cynipoid parasite of aphidophagous syrphid Silvertown J. W. & Lovett Doust J. (1993)Introduction to Plant
larvae. Ecological Entomology, 4, 175-82
Population Biology. Blackwell ScientificPublications, Oxford
Sagar G. R. & Mortimer A. M. (1976) An approach to the
Simberloff D. S. (1976)Experimentalzoogeographyof islands:
study of the population dynamics of plants with special effects of island size. Ecology, 57, 6 2 9 4 8
reference to weeds. Applied Biology, 1,1-43
Sarukhan J. (1980) Demographic problems in a tropical Simberloff D. S. & Abele L. G. (1976)Island biogeography and
system. In Demography and Evolution in Plant Populations conservation practice. Science, 191, 285-6
(Solbrig O.T. ed.), pp. 161-88. Blackwell ScientificPublica-
tions, Oxford Sinclair A. R. E. (1975) The resource limitation of trophic
Sarukhan J. & GadgilM. (1974)Studies on plant demography: levels in tropical grassland ecosystems. Journal of Animal
Ranunculus repens L., R. bulbosus L., and R. acris L. 111. A Ecology, 44, 497-520
mathematical model incorporating multiple modes or re-
production. Journal of Ecology, 62, 921-36 Siclair A. R. E. (1989) Population regulation in animals. In
Sarukhan J. & Harper J. L. (1973) Studies on plant Ecological Concepts. (Cherrett,J. M. ed.), Symposium of the
demography: Ranunculus repens L., R. bulbosus L., and R
acris L. I. Population flux and survivorship. Journal of British Ecological Society, No. 29, pp. 197-241. Blackwell
Ecology, 61, 675-716
Schaefer M. B. (1957)A study of the dynamics of the fishery for Scientific Publiations, Oxford
yellowfin tuna in the eastern tropical Pacific ocean. Inter-
Sjogren P. (1991) Extinction and isolation gradients in
American Tropical Tuna Commission Bulletin 2, No. 6 metapopulations: the case of the pool frog (Rana lessonae).
1957 Biological Journal ofthe Linnean Society, 42, 135-47
Schoener T. W. (1974) Resource partitioning in ecological
communities. Science, 185, 27-39 SkellamJ. G. (1972)Some philosophical aspects of mathemat-
Searle S. R. (1966) Matrix Algebra for the Biological Sciences. ical modelling in empirical sciencewith special reference to
John Wiley & Sons, New York ecology. In Mathematical Models in Ecology. (Jeffers J . N . R.
Sharitz R. R. & McCormick J. F. (1973)Population dynamics
ed.), Symposium of the British Ecological Society, No. 12,

pp. 13-29. Blackwell Scientific Publiations, Oxford
SladeN. A. (1977)Statisticaldetection of density dependence

from a series of sequential censuses. Ecology, 58, 1094-

1102
Slobodkin L. B. & Richman S. (1956) The effect of removal of

fixed percentages of the newborn on size and variabilityin
populations of Daphnia pulicaria (Forbes). Limnology and
Oceanography, 1,209-37

Smith F. E. (1961) Density dependence in the Australian
thrips. Ecology, 42, 403-7

REFERENCES 235

Snaydon R. W. (1991) Replacement or additive designs for Taylor A. D. (1993) Heterogeneity in host-parasitoid
competition studies?Journal of Applied Ecology, 28, 930-46 interactions: the 'aggregation of risk' and the 'CV2 > 1
rule'. Trends in Ecology and Evolution, 8, 400-5
Snyman A. (1949) The influence of population densities on
the development and oviposition of Plodia interpunctella Taylor L. R. (1986) Synoptic dynamics, migration and the
Hubn. (Lepidoptera). Journal of the Entomological Society of Rotharnstead Insect Survey. Journal of Animal Ecology, 55,
South Africa, 12, 137-71 1-38

Solomon M. E. (1949) The natural control of animal popula- Thomas C. D. & Harrison S. (1992) Spatial dynamics of a
tion~J.ournal of Animal Ecology, 18, 1-35 patchily distributed butterfly species. Journal of Animal
E C O ~ O6~1,Y4,37-46
Solomon M. E. (1964) Analysis of processes involved in the
natural control of insects. Advances in EcologicalResearch, 2, Thomas W. R., Pomerantz M. J. & Gilpin M. E. (1980)Chaos,
1-58 asymmetric growth and group selection for dynamical
stability. Ecology, 61, l 312-20
Solomon M. E. (1969) Population Dynamics. Edward Arnold,
London Thompson D. J. (1975) Towards a predator-prey model
incorporating age-structure: the effects of predator and
Southern H. N. (1970) The natural control of a population of prey size on the predation of Daphnia magna by Ischnura
tawny owls (Strix aluco). Journal of Zoology, 162, 197-285 elegans. Journal of Animal Ecology, 44, 907-16

Southwood T. R. E. & Reader P. M. (1976)Population census Thompson K. & Grime J. P. (1979) Seasonal variation in the
data and key factor analysis for the viburnum whitefly, seed banks of herbaceous species in ten contrasting habi-
Aleurotrachelusjelinekii (Fraunenf.),on three bushes. Journal tats. Journal of Ecology, 67, 893-921
ojAnimal Ecology, 45, 313-25
Tilman D. (1976) Ecological competition between algae.
Spitters C. J. T. (1983)An alternativeapproach to the analysis Experimental confirmation of resource-based competition
of mixed cropping experiments. 1.Estimation of competi- theory. Science, 192, 463-5
tion coefficients. NetherlandsJournal of Agricultural Science,
31,l-11 T h a n D. (1977)Resource competition between planktonic
algae: an exprimental and theoretical approach. Ecology,
Spradbery J. P. (1970)Host finding by Rhyssa persuasoria (L.), 58,338-48
an ichneumonid parasite of siricid woodwasps. Animal
Behaviour, 18, 103-14 Tilman D. (1980) Resources: a graphical-mechanistic ap-
proach to competition and predation. American Naturalist,
Stephens D. W. & Krebs J. R. (1986) Foraging Theory. 116,362-93
Princeton Unversity Press, Princeton
T h a n D. (1982) Resource Competition and Community Struc-
StilingP. (1988)Density-dependentprocessesand key factors ture. Princeton University Press, Princeton
in insect populations.Journal of Animal Ecology, 57, 581-93
Tilman D., Kilham S. S. & Kilham P. (1982) Phytoplankton
Stinnson J. S. (1973)The role of territory in the ecology of the community ecology: the role of limiting nutrients. Annual
intertidal limpet Lottia gigantea (Gray). Ecology, 54, Reviews of Ecology and Systematics 13, 349-72
1020-30
Tinbergen L. (1960) The natural control of insects in pine-
Stubbs M. (1977) Density dependence in the life-cycles of woods. 1:Factors influencing the intensity of predation by
animals and its importance in K- and r-strategies.Journal of songbirds. Archives Neerlandaises de Zoologie 13, 266-336
Animal Ecology, 46, 677-88
Turchin P. (1990)Rarity of density dependenceor population
Sugihara G. & May R. M. (1990) Nonlinear forecasting as a regulation with lags? Nature, 344, 660-3
way of distinguishing chaos from measurement error in a
time-series. Nature, 344, 7 3 4 4 1 Turchin P. & Taylor A. D. (1992) Complex dynamics in
ecological time series. Ecology, 73, 289-305.
Symonides E. (1979)The structure and population dynamics
of psarnmophytes on inland dunes. 111. Populations of Turnbull A. L. (1962) Quantitative studies of the food of
compact psammophyte communities. Ekologia Polska, 27, Linyphia triangularis Clerck (Aranaea: Linyphiidae). Cana-
235-57 dian Entomologist, 94, 1233-49

SynnonidesE. (1983)Population size regulation as a result of Turnbull A. L. (1964) The searching for prey by a web-
intra-population interactions. I. Effect of density on sur- building spider Archaeranea tepidariorum (C. L. Kock). Can-
vival and development of individuals of Erophila verna (L.) daian Entomologist, 96, 568-79
C.A.M. Ekologia Polska, 31, 839-82
Tybjerg H. & Vestergaard P. (1992) Growth dynamics in the
Takahashi F. (1968)Functional response to host density in a rhizomatousherb Polygonatum verticillatum Oikos, 65, 395-
parasitic wasp, with reference to population regulation. 408
Researches in Population Ecology, 10, 54-68
Ullyett G. C. (1949a) Distribution of progeny by Chelonus
Tamm C. 0. (1956)Further observations on the survival and texanus Cress. (Hymenoptera: Braconidae). Canadian Ento-
flowering of some perennial herbs. Oikos, 7, 274-92 mologist, 81, 25-44

236 REFERENCES

Ullyett G. C. (1949b) Distribution of progeny by Cryptus dependent effects in annual duck censuses. Ecology, 65,
inornatus Pratt (Hymenoptera: Ichneumonidae). Canadian 96-104
Entomologist, 81, 285-99 Vickery W. L. & Nudds T. D. (1991) Testing for density-
dependent effects in sequential censuses. Oecologia, 85,
Uranov A. A. (1975) Age spectrum of the phytocoenopopu- 419-23
lation as a function of time and energetic wave processes.
Biologicheskie Nauki, 2, 7-34 Volterra V. (1926)Variations and fluctuations of the numbers
of individuals in animal species living together. (Reprinted,
Usher M. B. (1972)Developments in the Leslie matrix model. 1931, in R. N. Chapman. Animal Ecology, McGraw-Hill,
In: Mathematical Models in Ecology (Jeffers J. N. R. ed.), New York)

Symposium of the British Ecological Society, No. 12, Watkinson A. R. (1980)Density-dependencein single species
pp. 29-60. Blackwell Scientific Publications, Oxford populations of plants. Journal of Theoretical Biology, 83,
Usher M. B. (1974) Biological Management and Conservation.
Chapman & Hall, London 345-57
Usher M. B., Longstaff B. C. & Southall D. R. (1971) Studies Watkinson A. R. (1981) Interference in pure and mixed
on populations of Folsomia candida (Insecta: Collembola):
the productivity of population in relation to food and populations of Agrostemma githago. Journal of Applied Ecol-
exploitation. Oecologia, 7, 68-79 ogy, 18, 967-76.
Utida S. (1950) On the equilibrium state of the interacting Watkinson A. R. (1983)Factors affectingthe density response
population of an insect and its parasite. Ecology, 31, of Vulpia fasiculata. Journal of Ecology, 70, 149-61
165-75 Watkinson A. R. (1984) Yield-density relationships: the
Utida S. (1957) Cyclic fluctuations of population density influence of resource availability on growth and self-
intrinsic to the host parasite system. Ecology, 38, 442-9 thinning in populations of Vulpia fasiculata. Annals of
Utida S. (1967) Damped oscillation of population density at Botany, 53, 469-82
equilibrium. Researches in Population Ecology, 9, 1-9 Watkinson A. R. (1985)On the abundance of plants along an
environmentalgradient,Journal of Ecology, 73, 569-78
Vance R. R. (1972a)Competition and mechanism of coexist- Watkinson A. R. & Harper J. L. (1978)The demography of a
sand dune annual Vulpia fasiculata: I. The natural regula-
ence in three syrnpatric species of intertidal hermit crabs. tion of populations. Journal of Ecology, 66, 15-33
Ecology, 53, 1062-74 Watkinson A. R. & White J. (1985). Some life history
Vance R. R. (1972b)The role of shell adequacyin behavioural consequences of modular construction in plants. Philo-
interactions involving hermit crabs. Ecology, 53, 1075-83 sophical Transactions of the Royal Society of London B, 313,
Vandermeer J. H. (1972) Niche theory. Annual Review of 31-51
Ecology and Systematics, 3, 107-32 Watson A. (1967) Territory and population regulation in the
Vandermeer J. H. (1989) Ecology of Intercropping. Cambridge red grouse. Nature, 215, 1274-5
University Press, Cambridge Watson D. (1987)Aspects of the population ecology of Senecio
van der Meijden E. (1971)Senecio and Tyrio (Callimorpha) in a vulgaris L. Unpublished PhD thesis, University of Liverpool
Dutch dune area. A study on an interaction between a Watt K. E. F. (1968) Ecology and Resource Management.
monophagous consumer and its host plant. Proceedings of McGraw-Hill, New York
the Advanced Study Institute on 'Dynamics of Numbers in Way M. J. & Cammell M. (1970) Aggregation behaviour in
Populations' (den Boer P. J. & Gradwell G. R. eds). Pudoc, relation to food utilization by aphids. In Animal Populaltions
Wageningen in Relation to their Food Resources (Watson A. ed.), pp. 2!29-
Varley G. C. (1947)The natural control of population balance 47. Blackwell Scientific Publications, Oxford
in the knapweed gall-fly (Urophora jaceana). Journal of Weiner J. (1993) Competition, herbivory and plant size
Animal Ecology, 16, 139-87 variability: Hypochaeris radiacata grazed by snails (Helix
Varley G. C. & Gradwell G. R. (1960)Key factors in population aspersa). Functional Ecoloy, 7, 47-53
studies.Journal of Animal Ecology, 29, 399-401 Weir D. A. (1985)The population ecology and clonal struc-
Varley G. C. & Gradwell G. R. (1963) The interpretation of ture of two grasses. Unpublished PhD thesis, University of
insect population changes. Proceedings of the Ceylon Associ- Liverpool.
ation for the Advancement of Science, 18, 142-56 Weller D. E. (1990) Will the real self-thinning rule please
Varley G. C. & Gradwell G. R. (1970) Recent advances in stand up? A reply to Osawa & Sugita. Ecology, 71, 1204-7
insect population dynamics. Annual Review of Entomology, Werner P. A. (1979)Competition and coexistence of similar
15,l-24 species. In Topics in Plant Population Biology. (Solbrig 0. T.,
Varley G. C., Gradwell G. R. & Hassell M. P. (1975) Insect Jain S., Johnson G. B. & Raven P. H. eds), pp. 287-310.
Population Ecology. Blackwell ScientificPublications,Oxford Macmillan, London

Vickery W. L. & Nudds T. D. (1984) Detection of density

REFERENCES 237

Werrner P. A. & Platt W. W. (1976)Ecological relationships of Related Problems in Quantitative Ecology. Academic Press,
New York
CO-occurringgolden rods (Solidago: Compositae). American Williamson M. (1972) Analysis of Biological Populations. Ed-
Naturalist, 110, 959-71 ward Arnold, London
Williamson M. (1981) Island Populations. Oxford University
Westley L. C. (1993)The effect of inflorescencebud removal Press, Oxford
on tuber production in Helianthus tuberosus L. (Asteraceae). Wilson D. E. & Janzen D. H. (1972) Predation on Scheelea
Ecology, 74, 2136-44 palm seeds by bruchid beetles: seed density and distance
from the parent palm. Ecology, 53, 954-9
Westoby M. (1984)The self-thinningrule. Advances in Ecolog- Woiwod I. P. & Hanski, I. (1992)Patterns of density depen-
ical Research, 14, 167-225 dence in moths and aphids. Journal of Animal Ecology, 61,
619-29
White J. (1980)Demographic factors in populations of plants. Wolf L. L. (1969)Female territoriality in a tropical humming-
In Demography and Evolution in Plant Populations, (Solbrig bird. Auk, 86, 490-504.
Wynne-EdwardsV. C. (1962)Animal Dispersion in Relation to
0. T. ed.), pp. 21-48. Blackwell Scientific Publications, Social Behaviour. Oliver & Boyd, Edinburgh
Wynne-Edwards V. C. (1977)Intrinsicpopulation control and
Oxford introduction. In Population Control by Social Behaviour
(Ebling F . J . & Stoddart D. M. eds), pp. 1-22. Institute of
White J. (1985) The thinning rule and its application to Biology, London
mixtures of plant populations. In Studies on Plant Demogra-
phy (White J. ed), pp. 291-309. Academic Press, London Yoda K., Kira T., Ogawa H. & Hozumi K. (1963)Self thinning
in overcrowded pure stands under cultivated and natural
White T. C. R (1978)The importance of a relative shortageof conditions. Journal of Biology, Osaka City University, 14,
food in animal ecology. Oecologia, 33, 71-86. 107-29

Whitham T. G. & Mopper, S. (1985) Chronic herbivory: im- Yodzis P. (1989)Introduction to Theoretical Ecology. Harper &
Row, New York
pacts on architecture and sex expression on pinyon pine.
Science, 228, 1089-91 Zeide B. (1987) Analysis of the - 3/2 power law of self-
Whittaker J. B. (1979)Invertebrate grazing, competition and thinning. Forest Science, 33, 517-3 7
plant dynamics. In Population Dynamics. (Anderson R. M.,

Turner B. D. & Taylor L. R. eds), pp. 207-22. Blackwell

ScientificPublications, Oxford

Willey R. W. & Heath S. B. (1969)The quantitative relation-
ships between plant population and crop yield. Advances in
Agronomy, 21, 281-321

Williams A. G. & Whitham T. G. (1986) Premature leaf

abscission: an induced plant defense against gall aphids.
Ecology, 67, 1619-27
Williams C. B. (1964) Patterns in the Balance of Nature and

This page intentionally left blank

Page numbers in italic are in the Carnrnell, M., 139,236 Finlayson, L.H., 118, 228
references Carl, E., 202, 226 Firbank, L.G., 48, 85, 89, 110, 205,
Abele, L.G., 215, 234 Caswell, H., 72, 226
Abrarns, P,, 113, 114, 225 Caughley, G., 150, 226 228
Ackert, J.E., 146, 225 Charles-Edwards,P.A., 48, 226 Ford, E.D., 42, 45, 228
Akinlosotu, T.A., 140, 225 Chesson, P.L., 156, 226 Free, C.A., 146, 147, 228
AUee, W.C., 51, 225 CIark, L.R., 118, 142, 177, 226 Fujii, K., 64, 228
Anderson, R.M., 137, 138, 146, 150, Clarke, T.A., 202, 226 Gadgl, M., 12-13,234
Cockrell, B.J., 148-9, 226 Gagnon, D., 23, 232
164,225,232 Comins, H.N., 110, 112, 149, 153, Garwood, N.C., 27, 228
Andrewartha, H.G., 178, 179, 180, Gaston, K.J., 212, 228
154,226,229 Gatsuk, L.E., 11, 228
210,225,227 Connell,J.H., 11,92, 94, 138, Gause, G.F., 108, 228
Archer, S., 130, 225 Gill, F.B., 203, 228
Atkinson, W.D., 114, 225 219-20,221,226 Gilpin, M., 210, 212, 214, 228, 229
Auer, C., 65, 225 Connolly, J., 88-9, 226 Godfray, H.C.J., 209, 228
Ayala, F.J., 109-10, 207, 225, 232 Cook, R.M., 148-9,226 Goh, B.S., 222, 228
Bailey, V.A., 150, 232 Cousens, R., 110, 226 Goldberg, D.E., 84, 89, 228
Bakker, K., 50, 177, 225 Crawley, M.J., 128, 129, 130, 131, Gold. W.G., 130. 228
Banks, C.J., 142, 225 Gorman, M.L., 222, 223,228
Baskin, C.C., 26, 225 205,226-7,233 Gotelli, N.J., 214, 228
Baskin, J.M., 26, 225 Crisp, D.J., 202, 227 Gottleib, L.D., 21,-228
Beddington,J.R., 151, 152, 165, 169, Crofton, H.D., 150, 227 Gradwell, G.R., 15, 181, 183, 184,
Crombie, A.C., 92, 93, 227
174,222,225 Curio, E., 142, 227 236
Begon, M., 42, 90, 225 Davidson, D.W., 78-83, 98, 128, 226, Griffiths, K.J., 134, 136, 228
Bell, A.D., 21, 23, 24, 225, 229 Grime, J.P., 26, 235
Bellows, T.S., 208, 225 227 Gmbb, P.J., 102, 104, 228
Belsky, A.J., 130, 226 Davidson, J., 62, 63, 178, 179, 180, Gulland, J.A., 167, 228
Bengtsson,J., 212, 226 Gutsell,J.S., 172, 173, 234
Beverton, R.J.H., 167, 226 227 Haines, B., 129, 228
Birch, L.C., 178, 179, 210, 225 Davies, N.B., 120, 201, 202, 227 Haldane, J.B.S., 15, 228
Birkhead, T.R., 50-1, 226 Davy, A.J., 3, 227 Hancock, D.A., 169, 228
Blackman, G.E., 40, 160, 226, 230 Dawkins, P.A., 27, 233 Hanski, I., 114, 115, 189, 190, 209,
Boyce, S.G., 11, 232 DeBach, P., 152, 227
Boycott, A.E., 211, 226 Dempster, J.P., 186, 188, 227 210, 211, 212, 214, 215, 228, 229,
Bradshaw, A.D., 40, 226 de Wit, C.T., 84, 88, 227 237
Branch, G.M., 48-9,226 Diamond, J.M., 95-6, 214, 227 Hara, T., 48, 229
Brougham, R.W., 161, 226 Dixon, A.F.G., 130, 227 Harberd, D.J., 200, 229
Brown, J.H., 78-80, 128, 226 Donald, C.M., 40, 42, 160, 233 Harcourt, D.G., 181-6, 229
Brown, M.W., 158,226 Dunn, E., 203, 227 Harper, J.L., 13, 21, 23, 24, 26, 39,
Bryant, J.P., 128, 226 Eberhardt, L., 20, 227 42, 44, 45, 46, 68, 77, 101, 131,
Bulmer, M.G., 186, 188, 190, 226 Ehler, L.E., 158, 227 133,192-3,199, 200, 201, 203,
Burnett, T., 145, 151, 226 Elner, R.W., 120, 227 203-4, 205,206,216-17,229,
Burrows, F.M., 47, 226 Ennos, R., 102, 227 230,233,234,236
Caldwell, MM., 130, 228 Errington, P.L., 125, 227 Harrison, S., 212, 213, 214, 215,
Callaghan, T.V., 23, 230 Erwin, T.L., 215, 227 229,235
Cameron. E.A., 158. 226 Esau, K., 21,227 Hassell, M.P., 64, 65, 110, 112, 119,
Fenchel, T., 98, 99, 227 131, 134, 136, 138, 140, 142, 143,
Fernando, M.H.J.P., 145, 227 145, 146, 147, 149, 150, 151,
152-8,187,188,206, 208, 209,
226,229

240 AUTHOR INDEX

Hastings, A., 209, 229 Lonsdale, W.M., 47-8,231 Pearl, R., 18-20, 62, 233
Hatto, J., 133, 229 Lotka, A.J., 105, 149, 231 Perrins, C.M., 9, 233
Heads, P.A., 188, 230 Lovett Doust, J., 85, 205, 234 Perry, J.N., 209, 233
Healey, M.C., 202, 230 Lowe, V.P.W., 17-19, 231 Pianka, E.R., 121-2, 231
Heath, S.B., 38, 56, 237 MacArthur, R.H., 95, 98, 112-13, Pibeam, C.J., 40, 41, 233
Heed, W.B., 121, 230 P i e n t e l , D., 201. 233
Hodgson, G.L., 40, 230 121-2,149, 210,214,223,231, P i , S.L., 222, 231, 233
Holling, C.S., 133, 136, 230 234 Piper, E.L., 129, 233
Holt, S.J., 167, 226 MacLulick, D.A., 4, 118, 231 Platt, W.W., 102, 236
Hoppensteadt,F.C., 65, 230 McClure, MS., 158, 232 Podoler, H., 202, 233
Horn, H.S., 22, 230 McConnick, J.F., 102-3, 234 Pollard, E., 186, 188, 189, 190, 233
Horton, K.W., 122,123, 230 McNaughton, S.J., 128, 232 Poole. R.W., 178, 233
Hubbard, S.F., 141, 230 McNeil, S., 121, 132-3, 164, 231, Porter, J.R., 24, 43, 45, 233
Huffaker, C.B., 118, 143, 230 232 Pratt, D.M., 63, 233
Hughes, R.N., 120, 227 Manlove, R.J., 198, 199, 231 Price, P.W., 121, 233
Hughes, T.P., 11,12, 230 Marshall, D.R., 85-7, 110, 231 Puckridge, D.W., 40, 42, 233
Hutchinson, G.E., 91, 230 May, R.M., 60-1,112-14,138, 146, Putwain, P.D., 216-17, 233
Iles, T.D., 170, 230 150, 154, 155, 160, 165, 205, 208, Radovich, J., 168,169,233
Inouye. D.W., 94, 230 209, 214, 215, 221, 222, 225, 227, Randolph, S.E., 137, 233
Ives, A.R., 115. 159, 160, 230 231,232,235 Raup, M.J., 128, 233
Jain, S.K., 85-7, 110, 231 Maynard Smith, J., 202, 232 Reader, P.M., 187, 235
Janzen, D.H., 77-8, 220, 230 Mead-Briggs,A.R., 124, 232 Redfern, M., 157, 233
Jeffries, R.L., 3, 227 Mertz, R.W.. 11, 232 Rediske, J.H., 127, 231
Jonsdottir,I.S., 23, 230 Michelakis, S., 145, 232 Rees, M., 130, 233
Jordano, D.. 212, 230 Monro, J., 140, 141, 232 Rice,B., 128. 233
Kaban, R., 128, 230 Moore, N.W., 202, 232 Richards, O.W., 6,233
Kays, S., 13, 45, 46, 230 Mopper, S., 128, 237 Richman, S., 131,170,171,233,234
Keddy,P.A., 193, 195, 196, 230 Morris, R.F., 181, 183, 232 Ricker, W.E., 189, 190, 233
Keith, L.B., 164, 230 Mortimer, A.M., 5, 26, 112, 234 Rigler, F.H., 135, 233
Kennett, C.E., 118, 230 Mueller, L.D., 207, 232 Roberts, E.H., 26, 233
Kira, T., 36, 39, 46, 230, 234 Murdoch, W.W., 122-4, 144, 153, Roberts, H.A., 27,233
Klomp, H., 177, 230 155,156,160, 226, 232 Rogers, D., 202, 233
Kluyver, H.N., 63, 231 Murphy, G.I., 168, 232 Root, R.B., 79, 233
Kontkanen, P,, 211, 231 Murton, R.K., 124, 232 Rosen, B.R, 21, 234
Krebs, J.R., 120, 140, 148, 201, 203, Myers, J.H., 128, 230 Rosenzweig, M.L., 149, 234
Nault, A., 23, 232 Rosewell, J., 115, 234
231 Newsome, A.E., 118, 232 Ross, M.A., 203-4, 205, 234
Kuchlein, J.H., 144, 231 Nicholson, A.J., 31, 64, 150, 165-6, Rotheray, G.E., 142, 234
Kuusela, S., 115, 229 177,180,232 Rudge, A.J.B., 124, 232
Lack, D., 99-100, 231 Noy-Meir, I., 150, 160, 161, 163, 232 Sagar, G.R., 5, 26, 234
Lanciani, C.A., 126, 231 Noyes, J.S., 141, 145, 232 Sar~khanJ,., 12-13,199,200,201,
Langer, R.H.M., 25, 231 Nudds, T.D., 184, 186, 236
Law, R., 14-16, 72-4, 87-8.111, Oaten, A., 122-4, 144, 153, 155, 232 234
Obeid, M,, 42, 44, 232 Schaefer, M.B., 169, 234
174,231 Oinonen, E., 200, 232 Schoener,T.W., 98, 234
Lawrence, W.H., 127, 231 Osawa, A., 48, 232 Scott, J.C., 215, 227
Lawton, J.H., 121, 123, 132-3, 150, Oster, G.F., 205, 232 Searle, S.R., 70, 234
Pacala, S.W., 131, 156, 157, 205, Sharitz, R.R, 102-3, 234
188, 212, 222, 225, 226, 228, 230, 227,229,233 Shinozaki, K., 39, 234
231,233,237 Paine, R.T., 217, 218, 233 Shorrocks, B., 114, 115, 201, 225,
Leslie, P.H., 70, 231 Palmblad, I.G., 33-6, 64, 233
Levin. D.A.. 231 Park, T., 92, 95, 233 234
Levins, R., 52, 98, 112, 210, 211, Silander, J.A., 205, 233
231 SiIliman, R.P., 172, 173, 234
Silvertown,J.W., 11,85, 191, 205,

234
SimberlofT,D.S., 211, 215, 234