66 Opinion TRENDS in Ecology & Evolution Vol.17 No.
2 February 2002
Testing the beneficial
acclimation
hypothesis
Robbie S. Wilson and Craig E. Franklin
Recent developments in evolutionary physiology have seen many of the
long-held assumptions within comparative physiology receive rigorous
experimental analysis. Studies of the adaptive significance of physiological
acclimation exemplify this new evolutionary approach. The beneficial
acclimation hypothesis (BAH) was proposed to describe the assumption that
all acclimation changes enhance the physiological performance or fitness of
an individual organism. To the surprise of most physiologists, all empirical
examinations of the BAH have rejected its generality. However, we suggest
that these examinations are neither direct nor complete tests of the functional
benefit of acclimation. We consider them to be elegant analyses of the adaptive
significance of developmental plasticity, a type of phenotypic plasticity that
is very different from the traditional concept of acclimation that is used by
comparative physiologists.
To bring t r adition al compa r a tive physiology in to
line wit h con tempor a ry evolu tion a ry biology,
physiologists over t he past decade or so h ave been
using a more t heoretical a nd hypot hesis-driven
approach to evolu tion a ry questions in physiological
resea rch. H istorically, m a ny st udies in compa r a tive
physiology proposed post-hoc adaptive stories to
explain t he fu nction al significa nce of a physiological
t r ait after elucida ting its mech a nistic basis.
H owever, m a ny critics of t he adapt a tionist
progr a m me h ave highligh ted t h a t t here a re m a ny
alter n a tives to adaptive scen a rios [1,2], including
genetic drift, past selection, genetic cor rela tions a nd
historical a t t ribu tes [3]. T he st rengt h a nd success
of t his new evolu tion a ry approach to compa r a tive
physiology is reflected by t he diversity of st udies
t h a t a re producing a deeper u nderst a nding of t he
evolu tion of physiological systems (e.g. Refs [4–6],
reviewed in Ref. [7]).
The beneficial acclimation hypothesis
O ne of the best examples of this new approach to
Robbie S. Wilson* physiological research has been the experimental
Dept of Biology, analysis of the adaptive significance of physiological
University of Antwerp,
A C C L I M A T I O N [8–12] (see G lossary). Traditionally,
Universiteitsplein 1,
B-2610 Wilrijk, Belgium. acclimation has been defined as the adjustment of
*e-mail: physiological traits in response to changes in a single
rwilson @ uia.ua.ac.be environmental variable in the lab [13], whereas
Craig E. Franklin A C C L I M A T IZ A T I O N refers to physiological responses to
Physiological Ecology environmental variables in the field [13]. Physiologists
Laboratory, Dept of
often assumed that all acclimation changes to the
Zoology, The University
of Queensland, St Lucia, phenotype enhanced the physiological performance or
QLD 4072, Australia. fitness of an individual organism in the environment
http://tree.trends.com 0169-5347/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S0169-5347(01)02384-9
in which those changes were induced. A daptive
arguments were often formulated after identifying
the functional role of the phenotypic modification and
usually involved logical arguments that showed how
the phenotype enhanced reproductive success, growth
or survival. T his long-held assumption, now referred
to as the beneficial acclimation hypothesis (B A H) [8],
has recently received a significant amount of
experimental interrogation.
T he B A H h as been tested predomin a ntly by
exa mining the acclim a tory responses of ectotherms to
temper a ture. T he hypothesis predicts th a t a nim als
acclim a ted to a pa rticula r temper a ture h ave
enh a nced perform a nce or fitness a t th a t temper a ture
in compa rison with a nim als acclim a ted to other
temper a tures. However, to the surprise of m a ny
compa r a tive physiologists, all empirical
exa min a tions of the B A H so fa r h ave rejected its
gener ality [8,10–12]. T hese studies h ave
demonstr a ted th a t the phenotypic ch a nges
(P H E N O T Y P I C P L A S T I C I T Y ) th a t occur in orga nisms during
development in different therm al environments do
not alw ays lead to a n increased fitness in th a t
environment when compa red with the fitness of
orga nisms r aised a t other temper a tures.
I n the first test of the B A H , L eroi et a l. [8] exposed
genetically identical lines of the bacteria Escherichi a
coli to either 32°C or 41.5°C for 24 h (~6.7 cell
gener a tions d−1 a t 37°C) a nd then competed the two
groups a t both exposure temper a tures ( F ig. 1). T he
B A H w as used to predict th a t the 32°C group would
outcompete the 41.5°C group a t 32°C, a nd vice versa
a t 41.5°C. However, bacteria grown up a t 32°C
outcompeted the 41.5°C-group a t both temper a tures,
a nd so the B A H w as rejected ( F ig. 1). I n a more
extensive test of the B A H , Bennett a nd L ensk i [9]
r aised E . coli a t 22, 27, 32, 37 or 40°C a nd then
competed the different acclim a tion groups against
each other a t each temper a ture. As in the previous
study, m a ny groups were outcompeted a t their
‘acclim a tion’ temper a ture by bacteria r aised a t other
temper a tures (benefit for acclim a tion w as found in
only seven out of the 12 compa risons). Again, these
results were used to reject the gener ality of the B A H .
G ibert a nd co-wor kers [14] recently outlined
a nother experiment al test of the B A H . T hey r aised
D rosophil a mel a nogaster from two different
popula tions a t 18, 25 or 29°C a nd then tested the
w al k ing speed of each development al group a t each
temper a ture. T he B A H w as used to predict th a t flies
would w al k faster a t their actu al development al
temper a ture th a n would flies developed a t other
temper a tures. However, in contr ast with their
predictions, flies rea red a t 25°C w al ked faster a t all
other temper a tures th a n did those r aised a t 18 or
29°C, a nd the B A H w as again rejected.
Acclimation or developmental plasticity?
We suggest th a t the empirical studies discussed here
a re neither direct nor complete tests of the function al
Opinion TRENDS in Ecology & Evolution Vol.17 No.2 February 2002
benefit of therm al acclim a tion, as defined from
tr adition al physiological studies of acclim a tion.
R a ther, we suggest these studies a re elega nt a n alyses
of the A D A P T I V E SI G N I F I C A N C E of D E V E L O P M E N T A L
P L A S T I C I T Y . Acclim a tion responses studied by
tr adition al compa r a tive physiologists differ
subst a ntially to the development al plasticity
exa mined by L eroi et a l. [8], Bennett a nd L ensk i [9]
a nd G ibert et a l. [14]. H istorically, compa r a tive
physiologists considered acclim a tion as a reversible
response by a n orga nism to ch a nges (often season al)
in a single environment al va riable [13]. B y contr ast,
development al plasticity deals with the entire suite
of phenotypic ch a nges th a t occur as a result of
differences in the development al environment, not
just the facult a tive physiological responses of a n
orga nism (acclim a tion).
Beca use of the highly sensitive n a ture of
orga nisms during the ea rly st ages of development,
sm all va ria tions in the development al environment
ca n lead to a cascade of phenotypic ch a nges [15–17].
Besides acclim a tion responses, development al
plasticity ca n a rise from the direct biophysical effects
of the environment, a nd ca n be detriment al, neutr al
or beneficial. E nvironment al factors th a t lead to these
un avoidable, a nd often nonreversible phenotypic
ch a nges include temper a ture [18], oxygen tension
[19–21], nutrition [22,23] a nd density of conspecifics
[24]. F or exa mple, M a tsch a k et a l. [19] found th a t
temper a ture-induced ch a nges in muscle cellula rity
during embryonic development of the A tla ntic salmon
S a lm a sa l a r were pa rtly due to restricted oxygen
availability a t higher temper a tures r a ther th a n to
facult a tive responses to temper a ture. T he egg capsule
of embryonic salmon ca n act as a n oxygen ba rrier,
pa rticula rly a t higher temper a tures when there is a n
increased oxygen dem a nd. I rreversible ch a nges in
the size a nd number of muscle fibres occur a t high
development al temper a ture as a direct consequence
of a constr aint in oxygen availability. T hese high-
temper a ture-induced development al ch a nges in
muscle cellula rity a re clea rly not facult a tive
acclim a tion responses.
O bliga tory developmen t al ch a nges a re
pa r ticula rly prevalen t following exposu re to st ressful
conditions, bu t t heir effects a re often subtle.
H offm a n n a nd H ew a- K apuge [18] distinguished t he
rela tive con t ribu tions of differen t types of phenotypic
ch a nge following exposu re to high temper a t u res in
t he pa r asitic w asp Tr ichogr a m m a n r. br assicae .
I mpor t a n tly, t hey fou nd t h a t some bu t not all
phenotypic ch a nges du ring developmen t were
t he result of facult a tive acclim a tion responses.
H offm a n n a nd H ew a- K apuge [18] initially observed
t h a t adults of T. n r. br assicae exhibited a n increased
resist a nce to st ressful temper a t u res following
exposu re to 33°C as pupae, bu t t h a t t hese ch a nges
were accompa nied by deleterious fit ness effects.
T hey suggested t h a t t hese fit ness decreases eit her
reflected a gener al cost of increasing resist a nce to
http://tree.trends.com
67
Exposure of 32oC 41.5oC
bacteria for~7
generations
Compete both 32oC-group 41.5oC-group
groups against versus versus
each other 41.5oC-group 32oC-group
32oC-group 32oC-group
Results outcompeted outcompeted
41.5oC-group 41.5oC-group
at 32oC at 41.5oC
TRENDS in Ecology & Evolution
Fig. 1. Experimental investigation of the Beneficial Acclimation
Hypothesis (BAH) by Leroi et al. [8]. The BAH was rejected in this case.
st ressful temper a t u res (acclim a tion response) or
were associa ted wit h direct phenotypic effects
a rising from da m age or developmen t al const r ain ts
ca used by t he high temper a t u res.
To test this idea, the a uthors exa mined whether
the increased resist a nce to high temper a tures in
T. nr. br assicae could occur without a ny of the
observed decreases in fitness [18]. Pupae of
T. nr. br assicae were exposed to 33°C for 2, 3 or 4 h d−1
for four days [18]. F or each trea tment group, there
w as a n increase in adult resist a nce to higher
temper a tures. However, fitness decreased only in
the groups exposed to 33°C for 3 or 4 h d−1 . T hus,
decreases in fitness following exposure to high
temper a tures were clea rly not ca used by the
increased resist a nce to stressful temper a ture
(acclim a tion) but r a ther to either gener al da m age to
the phenotype or development al constr aints imposed
by the high temper a tures. I ncreased resist a nce to
temper a ture without fitness costs h as also been
observed in the egg pa r asitoid T. ca rver ae in both
labor a tory a nd field experiments [25].
E xposing orga nisms to stressful conditions
confounds a ny a n alysis of the B A H (Box 1). Besides
acclim a tion responses possibly aimed a t minimizing
the stress of the environment, pa thological da m age
to the phenotype also occurs. I n spite of these
confounding effects, sever al a n alyses of the B A H
h ave incorpor a ted stressful conditions [8,9,26,27].
Met abolic costs a nd gener al phenotypic da m age
could overwhelm a ny positive acclim a tion responses
in a stressful environment a nd the B A H might be
incorrectly rejected. R a ther th a n compa risons a mong
orga nisms r aised under stressful conditions for
a n alyses of the B A H , Woods a nd H a rrison [27]
advoca te exa mining the costs a nd benefits of specific
acclim a tion responses.
N ot a ll developmen t a lly i nevi t able ch a nges to t he
phenot ype a re ca used by st ressfu l condi t ions. I n t he
specific case of temper a t u re, i t is doubtfu l t h a t ever y
ph ysiologica l process t h a t is affected by temper a t u re
68 Opinion TRENDS in Ecology & Evolution Vol.17 No.2 February 2002

Box 1. Experimental analysis of the BAH under stressful conditions

Woods and Harrison [a] addressed whether the acclimation of • Conclusions

Manduca sexta caterpillars to water stress was beneficial.
• Treatments
Larvae were raised from the first instar on low- (69%) or high-water
(80%) artificial diets. After reaching the fifth instar, growth rates for
both groups were measured on the same or opposite diet for 36 h.
Several determinants of larval water budget were also recorded.
• Results
Caterpillars raised on high-water diet grew faster than did those
raised on low-water diet when tested on both diets. Thus, no
benefit for previous exposure to low-water diet was observed.
However, larvae responded to short-term hydric stress (low-water
diet) by minimizing water excretion by increasing rectal water
absorption, and to long-term hydric stress by significantly
reducing faecal water excretion.
The authors concluded that, under current usage, the
beneficial acclimation hypothesis (BAH) had to be rejected in
this case. However, we suggest that phenotypic damage owing
to hydric stress confounds any analysis of the benefits of
acclimation.
Do the lower growth rates of larvae raised on a low-water
diet (when tested on both diets) reflect pathological
phenotypic changes owing to stress OR a lack of beneficial
acclimation? We suggest that the pathological effects of chronic
stress might overwhelm any possible beneficial acclimation
responses.
Reference
a Woods, H . A. a nd H a rrison, J. F. (2001) T he beneficial acclim a tion hypothesis
versus acclim a tion of specific tr aits: physiological ch a nges in w a ter-stressed
M a nduca sexta ca terpilla rs. Physiol. Zool. 74, 32–44

ca n ‘accli m a te’ to differen t temper a t u res. T h is
w ill i nevi t ably lead to phenot ypic differences
bet ween orga n isms from differen t developmen t a l
temper a t u res t h a t a re si mply due to t he di rect
effects of temper a t u re on t hese developmen t a l
pa t h w a ys [17,19–21]. T hese phenot ypic ch a nges
a re cer t a i n ly not t hose t h a t were t r adi t ion a lly
descr ibed as accli m a t ion responses by compa r a t ive
ph ysiologists, bu t a re u ndoubtedly i ncl uded i n t he
st udies of L eroi et a l. [8] a nd B en net t a nd L ensk i [9].
T h us, previous a n a lyses of t he B A H usi ng
developmen t a l pl ast ici t y a re confou nded by
i ncl udi ng sever a l t ypes of phenot ypic pl ast ici t y.
A more compelli ng exper i men t a l a n a lysis of t he
benefi t of accli m a t ion wou ld be based a rou nd t he
concept of accli m a t ion t h a t t r adi t ion a l compa r a t ive
ph ysiologists were cr i t icized for assu m i ng w as
a l w a ys benefici a l.
Exploring the BAH using competing hypotheses
I n t wo addi t ion a l st udies explor i ng t he B A H , bot h
H uey a nd B er r iga n [11] a nd H uey et a l. [12]
advoca ted a st rong i nference approach to
ex a m i n i ng quest ions rel a t i ng to t he t her m a l
accli m a t ion of ectot her ms. T hei r approach i n volved
test i ng a mong compet i ng h ypot heses t h a t m a k e
differen t predict ions as to how developmen t a l
temper a t u re i nfl uences t he t her m a l sensi t ivi t y of
perfor m a nce (Box 2).
H uey a nd B er r iga n [11] a nd H uey et a l. [12]
t hen used t he da t asets of sever a l previous st udies,
such as t h a t by Z w a a n et a l. [28], to compa re t he
h ypot heses. Z w a a n et a l. [28] a n a lysed t he effect
of developmen t a l temper a t u re on adu l t longevi t y
i n D . mel a nogaster a nd fou nd t h a t flies r a ised a t
i n ter medi a te temper a t u res su r vived longer as
adu l ts t h a n did t hose flies r a ised a t cool or h igh

Box 2. Set of competing hypotheses

This set of competing hypotheses is as suggested by Huey and
Berrigan [a], and Huey et al. [b].
Beneficial Acclimation Hypothesis (BAH): organisms acclimated
to a particular environment have enhanced performance or
fitness in that environment relative to organisms acclimated to
other environments [c].
Optimal developmental Temperature Hypothesis (OTH):
organisms raised at intermediate temperatures have higher
relative fitness across all temperatures than do organisms raised
at high or low temperatures. The OTH was suggested as an
alternative to the BAH by Zamudio et al. [d], Huey and Berrigan [a]
and Huey et al. [b].
Cooler is Better Hypothesis (CBH): organisms raised at cool
temperatures have higher relative fitness across all
temperatures than do organisms raised at intermediate or
high temperatures. The CBH is based on the assumption
that the larger size of cool-developed organisms is
sufficiently advantageous to outweigh any benefits of
acclimation [b].
Warmer is Better Hypothesis (WBH): organisms raised at high
temperatures have higher relative fitness across all temperatures
than do those raised at intermediate or cool temperatures. The
WBH is the reciprocal of the CBH of Huey et al. [b].
References
a H uey, R. B. a nd Berriga n, D . A. (1996) Testing evolution a ry hypotheses of
acclim a tion. I n A nim a ls a nd Temper ature: Phenotypic a nd E volution a ry
A d aptation. Society for E xperimenta l B iology Semin a r Series (Johnston, I. A.
a nd Bennett, A. F., eds), pp. 205–237, C a mbridge U niversity Press
b H uey, R. B. et a l. (1999) Testing the adaptive significa nce of acclim a tion: a
strong inference approach. A m. Zool. 39, 323–336
c L eroi, A. M. et a l. (1994) Temper a ture acclim a tion a nd competitive fitness: a n
experiment al test of the beneficial acclim a tion assumption. Proc. N atl. Aca d.
Sci. U . S. A. 91, 1917–1921
d Z a mudio, K .R. et a l. (1995) B igger isn’t alw ays better: body size, temper a ture
a nd m ale territorial success in D rosophil a mel a nogaster . A nim. Beh av.
49, 671–677

http://tree.trends.com
 Opinion TRENDS in Ecology & Evolution Vol.17 No.2 February 2002 69

Conclusions and future directions

Glossary
Previous empirical tests of the B A H h ave elega ntly
Acclimation: any facultative modification in a physiological trait in response to changes in demonstr a ted the evolution a ry significa nce of
an environmental variable in the lab. Changes can be in response to the developmental therm ally induced development al plasticity [8,9],
environment or long-term environmental shifts during the later stages of the life history of an
especially with the advent of a rigorous experiment al
organism (more traditionally studied). Responses can be beneficial, neutral or negative.
Acclimatization: facultative modifications in a physiological trait in response to changes in design testing sever al competing hypotheses [11,12].
one or more environmental variables in the field. I mport a ntly, these studies of the B A H h ave forcefully
Adaptive significance: (in context of plasticity) fitness advantages and disadvantages associated m ade the point th a t acclim a tion ch a nges ca nnot just
with the expression of phenotypic plasticity across a range of environments. Fitness
consequences can be positive, negative or neutral.
be assumed to be beneficial, but this is a hypothesis
Developmental plasticity: phenotypic changes induced by variation in the developmental th a t must be rigorously tested. However, we believe
environment. Changes can include facultative responses by the organism (e.g. acclimation) th a t a det ailed empirical exa min a tion of the adaptive
or unavoidable biophysical consequences of the environment (obligatory and/or
significa nce of more tr adition al measures of
pathological).
Phenotypic plasticity: environmentally induced variation in morphology, physiology and/or physiological acclim a tion is now required to test the
behaviour of an organism. B A H . We ch allenge compa r a tive physiologists to
develop new inventive experiment al designs to
explore the benefits a nd costs of the more tr adition al
temper a t u res, rega rdless of t he temper a t u re a t acclim a tion responses. However, this will not be
w h ich t he adu l t flies were k ept. T hese da t a were easy a nd using a nything less th a t a close correla te
used by H uey a nd B er r iga n [11] a nd H uey et a l. [12] of fitness to test the hypothesis, such as survival,
to aga i n reject t he gener a li t y of t he B A H i n fa vou r reproductive success or competitive ability, would be
of t he opt i m a l developmen t a l temper a t u re less th a n desir able. Previous empirical tests of the
h ypot hesis (O T H ). B A H h ave cert ainly set a benchm a r k for exa mining
However, as with the previous a n alyses of the the adaptive significa nce of phenotypic plasticity,
B A H , we suggest this experiment al design is also rega rdless of the source of phenotypic va ria tion.
confounded by sever al different types of plasticity O ne entertaining possibility for future tests of the
underlying the phenotypic ch a nges, not just B A H would be to examine the thermal acclimation
acclim a tion responses. We consider th a t, in this of reproductive performance, especially in a system
context, the O T H , cooler is better hypothesis (C B H) where females discriminate between displaying
a nd w a rmer is better hypothesis (W B H) all deal males. For example, in a species where females are
specifically with the adaptive consequences of the choosy about their mates, females could be given the
development al environment, rega rdless of the source opportunity to discriminate between cool- and warm-
of phenotypic ch a nges. However, the B A H refers only acclimated males at various temperatures. T he ability
to the facult a tive physiological responses of the to attract and procure a female might depend on
orga nisms a nd is thus only one specific type of characters such as swimming performance, aerobic
phenotypic plasticity. We suggest th a t the O T H , C B H capabilities and general activity, all of which have
Acknowledgements a nd W B H a re not mutu ally exclusive to the B A H . been shown to acclimate to temperature in a variety of
We thank Ian A. Johnston, F or exa mple, it is possible th a t the development al taxa [13]. I n this specific case, the B A H would predict
Helga Guderley, constr aints imposed on the phenotype by some that, at high temperatures, females would find the
Andy Clarke, Craig Moritz
and Raoul Van Damme for
temper a tures a re so grea t th a t the over all warm-acclimated males more attractive than they
stimulating discussions perform a nce is domin a ted not by the acclim a tion would the cool-acclimated males and vice versa at cool
and/or reading the article. responses (if they occur), but by these phenotypic temperatures. E ven more compelling would be the
This article was improved
inevit abilities. I n other words, there might be a n inclusion of males that had been raised at different
substantially by
comments from several optim al temper a ture for development th a t is temperatures, so the relative merits of developmental
anonymous referees. This determined solely by the un avoidable ch a nges to plasticity could be compared with the more reversible-
work was supported by a the phenotype th a t occur in the different therm al type acclimation responses. We suggest that future
small ARC grant awarded
environments. T his, of course, says nothing about tests of the B A H should investigate traditional types
to C.E.F. and R.S.W. and a
ARC Large Grant to C.E.F. the rela tive merit of the ‘acclim a tion’ ch a nges in of acclimation using the protocols developed for
and Craig Moritz. each environment. analysing developmental plasticity.
References 4 K ingsolver, J. G . a nd H uey, R. B. (1998) the beneficial acclim a tion assumption. Proc. N atl.
1 F eder, M. E . (1987) T he a n alysis of physiological E volution a ry a n alyses of morphological a nd Aca d. Sci. U . S. A. 91, 1917–1921
diversity: the prospects for pa ttern physiological plasticity in therm ally va riable 9 Bennett, A. F. a nd L ensk i, R. E . (1997)
document a tion a nd gener al questions in environments. A m. Zool. 38, 545–560 E volution a ry adapt a tion to temper a ture:
ecological physiology. I n N ew D irections in 5 Bennett, A. F. a nd L ensk i, R. E . (1999) V I. Phenotypic acclim a tion a nd its evolution in
E cologica l Physiology ( F eder, M. E . et a l., eds), E xperiment al evolution a nd its role in Escherichi a coli. E volution 51, 36–44
pp. 38–75, C a mbridge U niversity Press evolution a ry physiology. A m. Zool. 39, 346–362 10 Hoffm a nn, A. A. (1995) Acclim a tion: increasing
2 G arland, T., Jr and C arter, P.A. (1994) E volutionary 6 B r adley, T.J. et a l. (1999) Physiological responses survival a t a cost. Trends E col. E vol. 10, 1–2
physiology. Annu. Rev. E col. Syst. 56, 579–621 to selection for desicca tion resist a nce in 11 H uey, R.B. and Berrigan, D.A. (1996) Testing
3 Gould, S.J. a nd L ewontin, R. C. (1979) T he D rosophil a mel a nogaster . A m. Zool. 39, 337–345 evolutionary hypotheses of acclimation. In A nimals
spa ndrels of Sa n M a rco a nd the P a nglossia n 7 F eder, M. E . et a l. (2000) E volution a ry physiology. and Temperature: Phenotypic and E volutionary
P a r adigm: a critique of the adapt a tionist A nnu. Rev. E col. Syst. 31, 315–341 A daptation. Society for E xperimental B iology
progr a mme. Proc. R. Soc. London B B iol. Sci. 8 L eroi, A. M. et a l. (1994) Temper a ture acclim a tion Seminar Series (Johnston, I.A. and Bennett, A. F.,
205, 581–598 a nd competitive fitness: a n experiment al test of eds), pp. 205–237, C ambridge U niversity Press

http://tree.trends.com
70 Opinion TRENDS in Ecology & Evolution Vol.17 No.2 February 2002

12 H uey, R. B. et a l. (1999) Testing the adaptive
significa nce of acclim a tion: a strong inference
approach. A m. Zool. 39, 323–336
13 Prosser, C. L . (1991) E nvironmenta l a nd Metabolic
A nim a l Physiology: Compa r ative A nim a l
Physiology , 4th edn, Wiley– L iss
14 G ibert, P. et a l. (2001) Locomotor perform a nce of
D rosophil a mel a nogaster : inter actions a mong
development al a nd adult temper a tures, age, a nd
geogr aphy. E volution 55, 205–209
15 U sher, M. L . et a l. (1994) M uscle development in
A tla ntic salmon ( S a lmo sa l a r ) embryos a nd the
effect of temper a ture on muscle cellula rity. J. F ish
B iol. 44, 953–964
16 Vieir a, V. L . A. a nd Johnston, I. A. (1992) I nfluence
of temper a ture on muscle-fibre development in
la rvae of the herring C lupea h a rengus. M a r. B iol.
112, 333–341
17 Johnston, I. A. et a l. (1998) E mbryonic
temper a ture modula tes muscle growth
ch a r acteristics in la rval a nd juvenile herring.
J. E xp. B iol. 201, 623–646
18 Hoffm a nn, A. A. a nd H ew a-K apuge, S. (2000)
Acclim a tion for hea t resist a nce in Trichogr a mm a
nr. br assicae: ca n it occur without costs? F unct.
E col. 14, 55–60
19 M atscha k, T.W. et al. (1995) Is physiological hypoxia
the driving force behind temperature effects on
muscle development in embryonic A tlantic salmon
(Salmo salar)? D ifferentiation 59, 71–77
20 M a tsch a k, T.W. et a l. (1998) Met abolic enzyme
activities in A tla ntic salmon ( S a lmo sa l a r L .)
embryos respond more to chronic ch a nges in
oxygen availability th a n to environment al
temper a ture. F ish Physiol. B iochem. 18, 115–123
21 M a tsch a k, T.W. et a l. (1998) Temper a ture a nd
oxygen tension influence the development of
muscle cellula rity in embryonic r ainbow trout.
J. F ish B iol. 53, 581–590
22 Shine, R. a nd Downes, S.J. (1999) C a n pregn a nt
liza rds adjust their offspring phenotypes to
environment al conditions? Oecologi a 119, 1–8
23 Rasmussen, K .M. and F ischbeck, K .L. (1987) E ffect
of repeated reproductive cycles on pregnancy
outcome in ad libitum -fed and chronically food-
restricted rats. J. N utr. 117, 1959–1966
24 N ewm a n, R. A. (1998) E cological constr aints on
a mphibia n met a morphosis: inter actions of
temper a ture a nd la rval density with responses to
ch a nging food level. Oecologi a 115, 9–16
25 T homson, L .J. et a l. (2001) F ield a nd labor a tory
evidence for acclim a tion without costs in a n egg
pa r asitoid. F unct. E col. 15, 217–221
26 G ibbs, A. G . et a l. (1998) E ffects of cuticula r lipids
a nd w a ter bala nce in a desert D rosophil a : is
therm al acclim a tion beneficial? J. E xp. B iol.
201, 71–80
27 Woods, H . A. a nd H a rrison, J. F. (2001) T he
beneficial acclim a tion hypothesis versus
acclim a tion of specific tr aits: physiological
ch a nges in w a ter-stressed M a nduca sexta
ca terpilla rs. Physiol. Zool. 74, 32–44
28 Zw a a n, B. et a l. (1995) A rtificial selection for
development al time in D rosophil a mel a nogaster
in rela tion to the evolution of aging: direct a nd
correla ted responses. E volution 49, 635–648

Putting predators in general [4–8]. Wor k on behavioral predator–prey

interactions therefore provides an important bridge
between the studies of behavior and ecology.

back into behavioral I n spite of t hese m a ny adva nces, ou r

u nderst a nding of beh avior al preda tor–prey
in ter actions is limited by a simple oversigh t: we

predator–prey vir t u ally forgot abou t t he beh avior of preda tors!

H istorically, we h ave been so focused on prey
beh avior t h a t we (myself included) beca me

interactions comfor t able wit h t rea ting preda tors as u n responsive

‘black boxes’ r a t her t h a n pa r ticipa n ts in a beh avior al
in ter action. T his oversigh t h as not only led to a n
incomplete view of beh avior al in ter actions bet ween
Steven L. Lima preda tors a nd prey, bu t h as also obscu red a n en tire
class of such in ter actions t h a t occu rs a t la rge spa tial
scales. M y goal is to explore some of t he insigh ts
In the study of behavioral predator–prey interactions, predators have been gained from pu t ting preda tors back in to beh avior al
treated as abstract sources of risk to which prey respond, rather than preda tor–prey in ter actions.
participants in a larger behavioral interaction. When predators are put back
into the picture by allowing them to respond strategically to prey behavior, How were predators removed from the interaction?
expectations about prey behavior can change. Something as simple as allowing T he removal of preda tors from the beh avior al
predators to move in response to prey movements might not only (radically) preda tor–prey inter action is appa rent in the
alter standard expectations of prey behavior, but might also reveal new classes ubiquitous ‘fixed-risk’ assumptions of const a nt a tt ack
of behavioral phenomena that occur at large spatial scales. Similar revelations r a tes over time a nd pa tch-specific risks of preda tion
undoubtedly await many well-studied aspects of the behavioral interaction (e.g. Ref. [9]); such assumptions imply th a t preda tors
between predator and prey. Most examples studied to date, both theoretical and a re not influenced by prey beh avior. As few would
empirical, require attention from this ‘predatory’ perspective. Putting predators a rgue for the strict validity of this assumption,
back into the picture will be challenging, but doing so might change the way in why were preda tors relega ted to the st a tus of
which biologists think about predator–prey interactions in general. unresponsive entities? I n m a ny w ays, the fixed-risk
approach (i.e. the assumption of unresponsive
O ver the past 20 years, the study of behavioral preda tors) w as a sensible st a rting point.
interactions between predator and prey has shed much C h a r acterizing preda tion risk as a n environment al
light on prey behavior, and it is now clear that almost const a nt seemed reason able given th a t preda tors ca n
Steven L. Lima
any aspect of prey decision-ma king (from foraging stri ke opportunistically a nd could be a nywhere a t a
Dept of Life Sciences, behavior to mate choice) can be influenced by the risk given time. M a them a tical convenience might h ave
Indiana State University, of predation [1–3]. A growing literature also suggests also played a role: models of a ntipreda tor decision-
Terre Haute, IN 47809,
that nonlethal interactions between predator and prey m a k ing a re much simpler under a n assumption of
USA.
e-mail: (those driven by prey avoidance of predation) might be fixed risk th a n they a re when both preda tor a nd prey
S-Lima @ indstate.edu an important component of predator–prey interactions a re allowed to respond to one a nother. F urthermore,

http://tree.trends.com 0169-5347/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S0169-5347(01)02393-X