The University of Chicago

On r- and K-Selection
Author(s): Eric R. Pianka
Source: The American Naturalist, Vol. 104, No. 940 (Nov. - Dec., 1970), pp. 592-597
Published by: The University of Chicago Press for The American Society of Naturalists
Stable URL: http://www.jstor.org/stable/2459020 .
Accessed: 25/11/2013 10:21

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

The University of Chicago Press, The American Society of Naturalists, The University of Chicago are
collaborating with JSTOR to digitize, preserve and extend access to The American Naturalist.

http://www.jstor.org

This content downloaded from 161.116.100.92 on Mon, 25 Nov 2013 10:21:24 AM

All use subject to JSTOR Terms and Conditions
592 THE AMERICAN NATURALIST

ON r AND K-SELECTION

Dobzhansky(1950) proposedthatnatural selectionin the tropicsoperates

in a fundamentallydifferent way than it does in temperatezones. He argued
that muchof the mortalityin the temperatezones is relativelyindependent
of the genotype(and phenotype) of the organismconcerned,and has little
to do with the size of the population. Traditional examples of mass winter
kills of fishand sparrows are extremesof this sort. Dobzhansky reasoned
that in the relatively constant tropics, most mortalityis more directed,
generallyfavoringthoseindividualswithbettercompetitiveabilities.Thus,
in the temperatezones selection often favors high fecundityand rapid
development,wherasin the tropicslowerfecundityand slowerdevelopment
could act to increasecompetitiveability.By puttingmore energyinto each
offspringand producingfewer total offspring, overall individual fitnessis
increased. The small clutch sizes characteristicof many tropical birds are
consistentwith Dobzhansky's hypothesis.Dobzhansky's ideas were framed
in termstoo specificto reach the general ecologicalaudience and have gone
moreor less unnoticeduntil fairlyrecently.
MacArthurand Wilson (1967) coined the terms"K-selection" and "r-
selection" for thesetwo kinds of selection,whichare clearly not restricted
to the tropicsand the temperatezones (K refersto carryingcapacity and
r to the maximal intrinsicrate of natural increase [rmax]). To the extent
that these termsinvokethe much overused logistic equation, they are per-
haps unfortunate.However,it is clear that thereare two opposingkinds of
selection,whichusually have to be compromised.Certainly,no organismis
completely"r-selected" or completely"K-selected," but all must reach
some compromisebetween the two extremes.Fisher (1930) stated the
problem as follows: "It would be instructiveto know not only by what
physiologicalmechanisma just apportionmentis made betweenthe nutri-
ment devotedto the gonads and that devoted to the rest of the parental
organism,but also what circumstancesin the life-historyand environment
would render profitablethe diversionof a greater or lesser share of the
available resourcestowardsreproduction." Fisher's early statementis one
of the cleareston the idea of the budgetingof time,matter,and energyinto
these components (see also, Williams 1966; Gadgil and Bossert 1970).
Presumably,natural selectionwill usually act to maximizethe amountsof
matterand energygatheredper unit time; the problem is to understand
how thismatterand energyare partitionedamongsomaticand reproductive
tissuesand activities.
We can visualize an r-K continuum,and a particularorganism'sposition
along it. The r-endpointrepresentsthe quantitative extreme-a perfect
ecologic vacuum, with no density effectsand no competition.Under this
situation,the optimalstrategyis to put all possible matterand energyinto
reproduction,with the smallest practicable amount into each individual
offspring, and to produce as many total progenyas possible. Hence r-selec-
tion leads to high productivity.The K-endpointrepresentsthe qualitative

This content downloaded from 161.116.100.92 on Mon, 25 Nov 2013 10:21:24 AM

All use subject to JSTOR Terms and Conditions
 LETTERS TO THE EDITOR 593

TABLE 1
SOME OF THE CORRELATES OF r-AND K-SELECTION

Climate .Variable
Mortality .Often
Survivorship .Often
Population size .Variable
Intra- and interspecific
competition.Variable,
Relative abundance .....
Selection favors .1.
Length of life .Short,
Leads to .Productivity
r-Selection K-Selection
and/or unpredict- Fairly constant and/or pre-
able: uncertain dictable: more certain
catastrophic, nondi- More directed, delisity-de-
rected, density-independent pendent
Type III (Deevey Usually Type I and II
1947) (Deevey 1947)
in time, ionequilib- Fairly constantin time,equi-
rium; usually well below librium; at or near carry-
carrying capacity of en- ing capacity of the
vironment; unsaturated enviroilment; saturated
communitiesor portions communities; no recoloin-
thereof; ecologic vacuums; ization necessary
recoloniizationeach year
often lax Usually keen
Often does not fitMacArthur's Frequently fits the Mac-
broken stick model (King Arthur model (King
1964) 1964)
Rapid development 1. Slower development,
2. High ritax greater competitive abil-
3. Early reproduction ity
4. Small body size .. Lower resourcethresholds
5. Semelparity: single repro- 3. Delayed reproduction
duction 4. Larger body size
5. Iteroparity: repeated re-
productions
usually less than 1 Longer, usually more than
year 1 year
Efficiency,

extreme-densityeffectsare maximaland the environment is saturatedwith

organisms.Competitionis keen and the optimal strategyis to channel all
available matterand energyinto maintenanceand the productionof a few
extremelyfitoffspring.Replacementis the keynotehere. K-selectionleads
to increasing efficiency of utilization of environmentalresources.Table 1
summarizessome of the correlatesof the r- and K-selectedextremes.How-
ever,even in a perfectecologicvacuum, as soon as the firstorganismrepli-
cates itself,thereis thepossibilityof somecompetition,and natural selection
should favor compromisinga little more toward the K-endpoint.Hence, ax
an ecologic vacuum is filled,selectionwill shift a population fromthe r-
toward the K-endpoint (MacArthur and Wilson 1967).
One whole class of terrestrialorganisms (vertebrates) seems to be rela-
tively K-selected, while another large group (most insects, and perhaps
terrestrialinvertebratesin general) apparently is relatively r-selected.
There are, of course,a few exceptionsamongboth the insects (e.g., 17-year
cicada) and the vertebrates(some amphibians). Nevertheless,many of the
correlatesof the two kinds of selectionlisted in table 1 are characteristic
of thesetwo natural groups of terrestrialorganisms.Presumablyperennial
and annual plants differin a similarway. Aquatic organismsdo not appear
to obey this generalization;fish,,in particular,span the ravageof the r-K
continuum.

This content downloaded from 161.116.100.92 on Mon, 25 Nov 2013 10:21:24 AM

All use subject to JSTOR Terms and Conditions
594 THE AMERICAN' INIATURALIST

100

90

80-

70-

60-

150 50 - VERTEBRATES
140- 40 -

130 -30-

120 -20-

0,110 0

FIG. 10 Fqn drt Io any s s -

80
m 0

D 0-
z
5O0 INET
42

30-

20

10

0.1 1 10 100 1000

LOG, BODY LENGTH (cm)

FIG. 1.-Frequency distributionsof body lengths for many species of terres-

trial insects and vertebrates from eastern North America (data taken from
numerousfield guides and taxonomic accounts).

While the existenceof this dichotomycan (and doubtlesswill) be chal-

lenged, there are a number of reasons to believe it is real. For instance,
figure1 showsthe distributionof body lengthsfor a wide varietyof terres-
trial insects and vertebratesfrom eastern North America. Body length is
far fromthe most desirable measurementto demonstratethe polarity,but
the stronginverse correlationof rmaxwith generationtime and body size
(below) suggests that, when frequency distributionsfor the formertwo
parametersbecomeavailable, a similarbimodalitywill emerge.
Some interestingand importantgeneralizationshave been made concern-
ing the relationshipsbetween body size and rmaxand generation time.
Bonner (1965) plotted the logarithmof body lengthagainst the logarithm
of generationtime for a wide variety of organismsand demonstrateda
strong,nearly linear, positive correlation.Smith (1954) demonstrateda
similar,but inverse,correlationon a log-logplot of rmaxversus generation
time. Smith pointed out that rmaxmeasures the rate at which an organism
can fill an ecologic vacuum (at zero density); it is thereforeone of the
better indices of an organism's position on the r-K continuum.He also
notedthatrmax was inverselyrelatedto body size (i.e., that largerorganisms
are usually more K-selected than smaller ones). Now, rmaxis inversely
related to generationtime, T, by the followingformula: rmax= logeRo/T,
where Ro is the net reproductiverate. From this equation it can be seen
that variations in R? alter i'Ma.,only slightlycompared to changes in T.

This content downloaded from 161.116.100.92 on Mon, 25 Nov 2013 10:21:24 AM

All use subject to JSTOR Terms and Conditions
 LETTERS TO THE EDITOR 595

-
A.w-7-7- -7I I - x
Sectus
Minnus

A INSECTS
E 0 VERTEBRATES
0.3_

Z Gibbum
O 7rzgonogenius

- 0.2 _
X ASlethomaz/lm
u AMez'urn

Z Co/ondro
- 7r/bo/ium
Pediculus
0.1
4 Pfinus fur
A Furostus

:E P/inus sexpinc/ohls

NipfgMcov
Sclitary Bee
4{; cow *Cod ACicoda MCI
0 365 1000 2000 3000 4000 5000 6000
MEAN GENERATION, in DAYS

FIG. 2.-Arithmetic plot of rmax:against generation tine, using data of

Howe (1953) and Smith (1954).

Hence a hyperbola is expected. When Smith's data and that of Howe

(1953) are plottedon arithmeticaxes (fig.2), the data do fita stronghyper-
bola, the arms of which appear to representthe two "natural" groups of
organismsreferredto earlier. Cole (1954) presentsa similar plot showing
the reductionin rmax associated with delayed reproduction(or increased
generationtime).
Over a long period of time,the average r of any stable population must
equal zero. Smith (1954) pointed out that organismslike Escherichia coli
withveryhighrmsxvalues are muchfurtherfromrealizingtheirfull "biotic
potential" than organismswithlow rmgsvalues, such as man. He suggested
thatrmaxrepresentsa rate of increasenecessaryforthe populationto persist
in the face of its inevitable "environmentalresistance." Furthermore,he
argued that the resistance of the environmentmust exactly balance the
biotic potentialfor an actual r equal to zero. Hence rmax also measures en-
vironmentalresistance. Increased body size no doubt reduces environ-
mental resistancein many ways; an obvious one is that a larger organism
has fewer potential predators. Larger organismsare also better buffered
fromchanges in theirphysical environment.But the dividends of reduced
environmental resistanceattainedby increasedbody size are offsetby loss of
biotic potential. Nonetheless,the fossil record does show frequent evolu-
tionarytrendstowardlarger size (Newell 1949).
There are three major cycles in nature-daily, lunar, and annual. Most
organismslive longer than a day. Lunar cycles are probably of relatively
littleimportanceto the majorityof terrestrialorganisms.Thus it is unlikely
that either daily or lunar rhythmsunderlie the apparent bimodality of
theseorganisms.However,the annual cycle is anothercase. To survive,any
resident organism with a generationtime greater than a year must be

This content downloaded from 161.116.100.92 on Mon, 25 Nov 2013 10:21:24 AM

All use subject to JSTOR Terms and Conditions
596 THE AMERICAN NATURALIST

adapted to cope withthe full range of physical and biotic conditionswhich

prevail at a given locality. An organismwhich lives less than a year en-
countersonly a portionof the total annual range of conditions.The latter
usually survivethe harshestperiodsby formingrestingeggs or pupae. Their
population sizes vary with the particular local climatic conditionsand the
lengthof time duringwhich r is positive (for descriptionsand discussions
of such organisms,see Andewarthaand Birch 1954). Because longer-lived
larger organismsare betterbufferedfromenvironmentalvicissitudes,their
populationsizes do notvary as muchas thoseof smaller,shorter-lived organ-
isms. Furthermore,presumably their competitiverelationships are also
morepredictableand constant.
The attainmentof a generationtime exceeding a year may well be a
thresholdevent in the evolutionaryhistoryof a population. When peren-
niality is reached,thereare substantiallyfewerenvironmental"surprises"
and a ratherdrasticshiftfromr- to K-selection.I concludethat theremay
well be a natural bimodalityin environmentalresistance.
There are, of course,otherpossible reasons for the apparent bimodality
of relativelyr- and relativelyK-selected organismsin nature. It could be
simplyhistoricaccidentthat body sizes and generationtimesof insectsand
vertebratesare largely non-overlapping.It is also quite possible that an
either/orstrategyis usually superiorto some compromise.This mightbe a
simple consequence of the hyperbolicinverse relationship between rman
and T.

ACKNOWLEDGMENTS

A numberof the ideas presentedhere were developedin discussionswith

studentsat the Universityof Texas. V. K. Johnsonperformedthe tedious
task of extractingand convertingthe body lengthsfor figure1; her assis-
tance is gratefullyacknowledged.P. West prepared the figures.My wife
Helen read and commentedon the manuscript.It is a pleasure to acknowl-
edge the financialsupport of the National Science Foundation (grant no.
GB-8727).

LITERATURE CITED

Andewartha, H. G., and L. C. Birch. 1954. The distributionand abundance of animals.

Univ. Chicago Press, Chicago. 782 p.
Bonner, J. T. 1965. Size and cycle: an essay on the structureof biology. Princeton Univ.
Press, Princeton, N.J. 219 p.
Cole, L. C. 1954. The population consequences of life history phenomena. Quart. Rev.
Biol. 29:103-137.
Deevey, E. S. 1947. Life tables for natural populations of animals. Quart. Rev. Biol.
22: 283-314.
Dobzhansky, T. 1950. Evolution in the tropics. Amer. Sci. 38:209-221.
Fisher, R. A. 1930. The genetical theoryof natural selection. 2d revised ed. Dover, N.Y.,
1958. 287 p.
Gadgil, M., and W. H. Bossert. 1970. Life historical consequences of natural selection.
Amer. Natur. 104:1-24.

This content downloaded from 161.116.100.92 on Mon, 25 Nov 2013 10:21:24 AM

All use subject to JSTOR Terms and Conditions
 LETTERS TO THE EDITOR 597

Howe, R. W. 1953. Studies on beetles of the family Ptinidae. VIII. The intrinsicrate of
increase of some ptinid beetles. Ann. Apple.Biol. 40:121-134.
King, C. E. 1964. Relative abundance of species and MacArthur's model. Ecology 45:716-
727.
MacArthur,R. H., and E. 0. Wilson. 1967. The theoryof island biogeography. Princeton
Univ. Press, Princeton, N.J. 203 p.
Newell, N. D. 1949. Phyletic size increase-an important trend illustrated by fossil
invertebrates.Evolution 3:103-124.
Smith, F. E. 1954. Quantitative aspects of population growth, 277-294 p. In E. Boell
[ed.], Dynamics of growthprocesses. Princeton Univ. Press, Princeton, N.J.
Williams, G. 0. 1966. Adaptation and natural selection. Princeton Univ. Press, Princeton,
N.J. 307 p.
ERIC R. PIANKA
DEPARTMENT OF ZOOLOGY
UNIVERSITY OF TEXAS
AUSTIN, TEXAS 78712
June 9, 1970

This content downloaded from 161.116.100.92 on Mon, 25 Nov 2013 10:21:24 AM

All use subject to JSTOR Terms and Conditions