Art 4 - Clima

7/29/2019 Art 4 - Clima

1/12

Evolutionary Responses to Changing Climate

Author(s): Margaret B. Davis, Ruth G. Shaw and Julie R. EttersonReviewed work(s):Source: Ecology, Vol. 86, No. 7 (Jul., 2005), pp. 1704-1714Published by: Ecological Society of AmericaStable URL: http://www.jstor.org/stable/3450613 .

Accessed: 28/01/2013 20:47

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 [email protected].

.

Ecological Society of America is collaborating with JSTOR to digitize, preserve and extend access toEcology.

http://www.jstor.org
http://www.jstor.org/action/showPublisher?publisherCode=esahttp://www.jstor.org/stable/3450613?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/3450613?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=esa

7/29/2019 Art 4 - Clima

2/12

Ecology, 86(7), 2005, pp. 1704-1714C2005 by the Ecological Society of America

EVOLUTIONARY RESPONSES TO CHANGING CLIMATEMARGARET B. DAVIS,,'3 RUTH G. SHAW,' AND JULIE R. ETTERSON2

'Department f Ecology,Evolutionand Behavior,University f Minnesota,St. Paul, Minnesota55108 USA2Department f Biology, Universityof Minnesota,Duluth,Minnesota55812 USA

Abstract. Until now, Quaternary paleoecologists have regarded evolution as a slowprocess relative to climate change, predicting that the primary biotic response to changingclimate is not adaptation, but instead (1) persistence in situ if changing climate remainswithin the species' tolerance limits, (2) range shifts (migration) to regions where climateis currently within the species' tolerance limits, or (3) extinction. We argue here that allthree of these outcomes involve evolutionary processes. Genetic differentiation withinspecies is ubiquitous, commonly via adaptation of populations to differing environmentalconditions. Detectable adaptive divergence evolves on a time scale comparable to changein climate, within decades for herbaceous plant species, and within centuries or millenniafor longer-lived trees, implying that biologically significant evolutionary response can ac-company temporal change in climate. Models and empirical studies suggest that the speedwith which a population adapts to a changing environment affects invasion rate of newhabitat and thus migration rate, population growth rate and thus probability of extinction,and growth and mortality of individual plants and thus productivity of regional vegetation.Recent models and experiments investigate the stability of species tolerance limits, theinfluence of environmental gradients on marginal populations, and the interplay of de-mography, gene flow, mutation rate, and other genetic processes on the rate of adaptationto changed environments. New techniques enable ecologists to document adaptation tochanging conditions directly by resurrecting ancient populations from propagules buriedin decades-old sediment. Improved taxonomic resolution from morphological studies ofmacrofossils and DNA recovered from pollen grains and macroremains provides additionalinformation on range shifts, changes in population sizes, and extinctions. Collaborationbetween paleoecologists and evolutionary biologists can refine interpretations of paleo-records, and improve predictions of biotic response to anticipated climate change.

Key words: adaptation;climate change; evolutionaryconstraints;Quaternarypaleoecology;range shifts;tolerance imits.

INTRODUCTIONAdaptation to changes in the environment is an im-portant and nearly universal aspect of biotic responseto climate change. We argue here that adaptive re-sponses affect species persistence, migration rate, andforest productivity. Thus more complete understandingof adaptive responses to climate must be incorporatedinto interpretations of Quaternary paleorecords. Fur-thermore, fossil records can provide tests of evolu-tionary hypotheses, enhancing understanding of theevolutionary consequences of climate change.It is remarkable that paleoecologists reconstructinglate-Quaternary environments so seldom discuss thepossibility of adaptation to changing climate. Theyhave tended to adopt the perspective of Darwin, whoemphasized that vast periods of geologic time allowgreat opportunity for the accumulation of many slightdifferences among organisms that, collectively, can re-

sult in major differences in biological form. Accord-ingly, a fundamental assumption of paleoecology hasbeen that the rate of evolution is far slower than therate of climate change (Bennett 1997, Webb 1997, Jack-son 2000), with meaningful evolutionary adaptations,such as the origination of taxa, occurring on the time-scale of millions of years. Climate changes much morerapidly, because superimposed on glacial cycles100000 years in length are variations on' millennial,centennial, and even decadal time scales. For example,during the most recent glacial-interglacial transition, atime of rapid warming, there was a brief reversal lastingseveral centuries. At this time paleorecords from thenorth Atlantic region record a temperature drop of sev-eral degrees C within decades (Huntley et al. 1997).Paleoecologists have argued that such rapid environ-mental changes overwhelm evolutionary processes,causing extinction except where climate remains withinpreexisting tolerance limits for a species (Bennett1997). Contributing to this view has been the sparse-ness of the record of new plant species during the Qua-ternary and the apparent limitations of fossils, espe-cially pollen, to record traits involved in adaptation toclimate, such as phenology and physiology. Further-

Manuscripteceived24 November 003; accepted16 June2004;finalversion eceived October 004.Correspondingd-itor adhoc):L.J.Graumlich. orreprintsf thisSpecialFeature,see footnote1, p. 1667.3E-mail:[email protected]

This content downloaded on Mon, 28 Jan 2013 20:47:36 PMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp

7/29/2019 Art 4 - Clima

3/12

July2005 PALEOPERSPECTIVESN ECOLOGY 1705more, latitudinal range shifts of tree taxa, well docu-mented in Quaternary sediments, are compatible withthe view that species persist only within the environ-ments to which they are already adapted, even as thoseenvironments shift in space (Prentice et al. 1991, Jack-son 2000).Moreover, paleoclimate reconstructions rely on thepremise that each species has a unique set of limits inits tolerance to various aspects of the environment, de-fining a multidimensional ecological niche (Jackson2000). Paleoecologists have focused on physical nichedimensions, determining for particular species the re-alized limits of physiological tolerances to factors suchas minimum temperature or growing degree days.When species-specific tolerances are plotted on a mapof modern climate, they quite accurately delimit thepresent geographical distribution of the species in ques-tion (Peterson et al. 1999). Paleoclimate reconstructionemploys these environmental attributes of species toinfer the distribution of climate at an ancient time fromthe distribution and abundance of fossils. Thus, paleo-climate reconstruction depends on the assumption thatspecies tolerance limits remain stable in time, that is,evolutionarily inert. In practice, a certain amount ofchange might not be noticed because reconstructionsare often at a coarse geographical scale (Prentice et al.1991), with broad confidence intervals surroundingtemperature and precipitation estimates. Fossil recordsused for validation are often sparse, especially for oldertime horizons. Nevertheless, recent reviews regard rap-id evolution as the exception, affecting only "somelocal populations of a species . .. while others undergoextinction or disperse to newly suitable sites" (Jackson2000:294). Given this context, discussions of bioticresponse to future climate emphasize predicted rangeshifts, focusing on seed dispersal and establishment aspotentially limiting processes (e.g., Clark et al. 1998).Here we consider evolution in relation to climate anddiscuss its relevance to processes paleoecologists con-sider important, including population persistence insitu, range shifts, and extinction. We review the evi-dence for adaptation of plant populations to spatial var-iation in environment and discuss the potential for evo-lutionary response to temporal variation. We emphasizethat climate change on various time scales imposesselection regimes that may lead to adaptive changes inplants and animals, whether or not their ranges shift(Davis and Shaw 2001). The degree of adaptation de-pends on the interplay of natural selection with otherevolutionary processes, such as gene flow, genetic drift,and mutation, and also with demography. A lag in ad-aptation implies reduction in growth and survival ofindividual plants as well as in the overall productivityof regional vegetation (Rehfeldt et al. 1999, 2002). Therate of adaptation influences the rate at which popu-lations invade newly available habitat (Garcia-Ramosand Rodriguez 2002), and also the probability that pop-ulations will die out within their present ranges (Pease

et al. 1989, Burger and Lynch 1995). Thus, adaptiveevolution affects all of the primary responses to chang-ing climate predicted by paleoecologists-"tolerance,migration, or extinction" (Jackson 2000:294). We re-interpret examples from the Quaternary record thatdemonstrate how evolutionary models can expand un-derstanding. Last, we review recent contributions byQuaternary paleoecologists that utilize paleorecords todemonstrate adaptations to environmental change.

GENETICCHANGE N PLANTSDURINGTHEQUATERNARY

For animal phyla, particularly mammals, both spe-ciation events and extinctions are well documented dur-ing the past two million years. In contrast, the Qua-ternary fossil record documents few examples of newspecies of vascular plants (Comes and Kadereit 1998),reinforcing the impression that, particularly in theseorganisms, evolution is a slow process. Speciation is,however, only one aspect of evolutionary change. With-in species, genetic differentiation among populationsattests to the ubiquity and rapidity of evolutionarychange.

Spatial substructuring of populations is evidenced byvariation at putatively neutral genetic marker loci, re-flecting current mating patterns and/or previous iso-lation of populations. Phylogeographers use such mark-ers as a basis for inferring locations of refuges duringthe last glaciation, and pathways of migration sincethen (e.g., Cruzan and Templeton 2000, Hewitt 2000).Fossil evidence has valuably informed such studies.For example, allozyme and mitochondrial DNA(mtDNA) divergence has been calibrated against theactual time since isolation of tree populations in theAmerican Southwest. Time of isolation was determinedfrom migration histories documented by macroremainspreserved in packrat middens (Hamrick et al. 1994).Allozymes in a series of populations of lodgepole pine(Pinus contorta), the ages of which were determinedfrom radiocarbon-dated fossil-pollen records, suggesta progressive loss of alleles in the course of northwardmigration in western Canada during the Holocene(Cwynar and MacDonald 1987). Reduced allelic di-versity away from putative glacial refugia has beendocumented for several plant species, while Pinuspum-ila shows reductions in diversity at some loci and in-creased diversity at others (Tani et al. 1996).

Adaptative differentiation among populations withina species is documented by clinal variation in physi-ological, phenological, and fitness traits in relation tolatitudinal or elevational gradients in climate. Such var-iation has been shown for many species, beginning withthe classic studies of Turesson (1922) and Clausen etal. (1940) and continuing with recent papers too nu-merous to list here. The distribution of climate-sensi-tive traits within species is studied in common-garden(provenance) experiments, in which plants from dif-ferent geographical locations are grown together at a

-U

r-r

P1


7/29/2019 Art 4 - Clima

4/12

1706 MARGARETB. DAVIS ET AL. Ecology,Vol. 86, No. 7A BS. Stable climate pollen/seeddispersalS., pollen/seedispersal

B -Fe d c b a meanphenotypeo . E D C B A fitnessoptimumA /climate gradient

? /' Changing climate

C E F. t d c b a meanphenotype'

E D C B A newoptimum0 /changed climategradient

Annual temperatureFIG.1. (A) Diagrammatical epresentation f the growthandmortality esponseof trees to gradients n the mean annualtemperaturefundamental iche) (top), and the frequencyof occurrenceof the samepopulations realized niche) across thesame temperature radient(bottom).Diagramsare generalized rom common-garden tudies of Pinus contortain westernCanada. Panel A is reproduced rom Rehfeldt et al. [1999].) (B) Diagram depictinghypotheticalphenotypicmeans andoptimalfitness means of populationsof a plant species along a climategradientundera stable climate (top), and under achangingclimate(bottom). n time,thechangedselectionregimewill bring hephenotypicmeansof local populations loserto the changed itnessoptima.Meanwhile, he range s shiftingin space as populations pread ntonewly suitablehabitatatone end of the gradientand retreat rom unsuitablehabitatat the otherend of the gradient. PanelB is modified rom Davisand Shaw [2001].)

series of locations with differing climate, or as a specialcase of this approach, are transplantedreciprocally intothe different sites from which populations are sampled.These studies show that most tree and herbaceous plantspecies comprise a series of populations, each of whichis relatively well adapted to its local environment (Fig.1; see e.g., Reinartz 1984, Lacey 1988, Rehfeldt et al.1999, 2002, Etterson 2004a; also see Plate 1). For manytree populations, growth rates are highest at test siteswith temperatures similar to the source location;growth tends to be less and mortality greater at bothwarmer and colder locations (Carter 1996) (Fig. 2).Thus, a novel climate reduces individuals' growth, aswell as survival and fecundity, i.e., fitness. The reduc-tion can be large: for example climate differences pre-dicted for doubled concentrations of greenhouse gasescould reduce biomass production by 20% in Canadianpopulations of lodgepole pine (Rehfeldt et al. 1999)and reduce seed production by 30% in some herbaceousspecies (Etterson 2004a).Numerous examples document the speed of adaptivedifferentiation. For example, adaptation of Scots pine(Pinus sylvestris) to the diverse elevations and latitudesit currently occupies in Finland (Muona 1989, Hurmeet al. 1997) must have happened within the past severalthousand years at most, because the region was coveredby continental ice until the mid-Holocene. Similarly,studies of introduced herbaceous species in NorthAmerica and Europe show that genetic differentiationand clinal variation in response to climate has accrued

since the time of establishment on the continent, i.e.,within decades or a century (Verbascum thapsus, Rei-nartz 1984; Daucus carota, Lacey 1988; Solidago sp.,Weber and Schmid 1998).To the extent that spatial differences in climate elicitthe expression of genetic variation in fitness, a temporalchange in climate will also impose a new selectionregime, under which traits may evolve adaptively.When climate changes, fitness may be reduced initially,but after a number of generations populations through-out a species range could, via selection response, re-cover the fitness and biomass they exhibited before theonset of climate change. Is adaptationof all populationsto a new climate gradient equally likely? And howmany generations would it require for fitness and pro-ductivity to match those prior to climate change? Ofcourse the magnitude and rate of climate change andits duration are critical. But answers will also dependupon the ecological breadth of individual genotypes,the distribution and nature of genetic variation for rel-evant traits, the extent of gene flow among populationsthrough dispersal of both pollen and seeds, and thedemographic flux of populations. The potential foradaptive evolution may also be influenced by changingcompetition among species that differ in their responseto climate change. To date, stand-simulators examiningforest response to changing climate have not consid-ered these issues because they treat species as collec-tions of identical individuals rather than taking geneticvariation into account.


7/29/2019 Art 4 - Clima

5/12

7/29/2019 Art 4 - Clima

6/12

1708 MARGARETB. DAVIS ET AL. Ecology,Vol. 86, No. 7

-d

TABLE 1. Demographyof sugar maple (Acer saccharum),1991-1998, in plot A, SylvaniaWildernessArea, Gogebic County,Michigan,USA.Parameter Value

Seeds (rangeof annualvalues) 3000-10 000 000 seeds-ha-'-yr-'Seedlings 725 000 seedlings/haSaplingsandunderstory rees 500 individuals/haCanopytrees 145 trees/haTreesreachingcanopy (average,1991-1995) 1 tree-ha-1-yr-'

tablish each year (Table 1). This prolificacy from sexualreproduction facilitates exposure of the available ge-netic variation to selection. A miniscule fraction ofeach cohort of seeds and seedlings survives to reachthe forest canopy and reproduce (Table 1). Over theseveral decades thatjuveniles are thinned from a foreststand, selection could increase the frequency of ge-notypes that tolerate novel climate conditions. Rehfeldtet al. (1999, 2002) use data from generational changesin growth rate in selection experiments to estimate thatrecovery of adaptation following anticipated warmingin coming decades would require one to 13 generationsfor particular populations of lodgepole pine and Scotspine. Thus, adaptation of these long-lived trees couldrequire as little as a century for the northern popula-tions that grow well under warmer conditions, but wellover a millennium for central and southern populationsthat grow poorly under warmer conditions. These es-timates assume sufficient genetic variation, but 10 ormore generations of strong directional selection couldseverely reduce genetic variability available to supportselection response.Depletion of variability is only one of the geneticconstraints that could impede adaptation to changingclimate. A recent study of the prairie annual, Cha-maecrista fasciculata, demonstrates that genetic cor-relations among traits important for fitness may impedeadaptive evolution under a warmer and drier climate(Etterson and Shaw 2001). Pedigreed seedlings werereciprocally planted in common gardens along an arid-ity gradient in the Great Plains. Traits that influencefitness were measured, including the rate of phenolog-ical development, leaf number, and leaf thickness.Clines in selection corresponding to latitude were ob-served and are predicted to move northward in the fu-ture (Etterson 2004a). Populations harbored geneticvariation for most traits under selection, although theamount was lower at the periphery of the species range(Etterson 2004b). Despite significant selection and ge-netic variance, the rate of adaptive evolution in thesepopulations is predicted to be slower than the antici-pated rate of climate change. Adaptive evolution maybe substantially slowed because of genetic correlationsamong traits that are antagonistic to the direction ofselection under a changed climate (Fig. 3, Etterson andShaw 2001). The genetic correlations are due either topleiotropy, in which the effect of an allele on one trait,e.g., leaf thickness, enhances fitness, whereas its effect

on another trait, e.g., rate of phenological development,reduces fitness, or to linkage disequilibrium, the as-sociation of alleles at different loci. Adverse geneticcorrelations may reduce the likelihood that combina-tions of traits that result in highest fitness in hotter,drier climates will evolve. The lability of genetic cor-relations among traits and thus the extent of geneticconstraint, depends on the genetic nature of the cor-relation (pleiotropy, linkage disequilibrium), which isnot readily determined. The extent of genetic impedi-ments to adaptation to climate change is generally poor-ly known, but impediments to adaptation are particu-larly important today, as the future rate of change inclimate is predicted to exceed past changes by at leastone order of magnitude.

Despite these potential impediments, evolutionaryresponse to change in physical conditions has beenshown in a number of different species. Experimentsin bacteria demonstrate both adaptation to changedtemperature regime and lability of species tolerancelimits. Replicate lines of the bacterium E. coli, all de-rived from a single cell, were subjected to directionalselection under distinct temperatures throughout therange typically tolerated, and one set of lines was sub-jected to variable temperatures (10'C diurnal range;Lenski 2001). Over 2000 generations, each populationadapted to the temperature it experienced, with the rateof increase in fitness, relative to the ancestor, greatestat the highest temperature used, 42?C, within a degreeof the lethal limit (Bennett et al. 1992). Lines subjectedto variable temperature evolved increased fitnessthroughout a 20'C range of temperatures, broader thanthe range over which fitness increased for any of thelines held at constant temperatures (Bennett et al. 1992,Lenski 2001). This argues against the view that variableclimate, as experienced during the Quaternary Period,generally impedes adaptation. Further investigation ofthe evolutionary potential to extend the range above42?C revealed that mutations conferring tolerance tohigher temperatures consistently reduce competitiveperformance (Mongold et al. 1999). Consequently, inthese E. coli cultures, adaptation to temperature moreextreme than that tolerated by the ancestral populationevolves only in declining populations. However, oncepopulations reached a critical low density, new muta-tions conferring tolerance of higher temperatures didincrease in frequency. These studies demonstrate therole of genetic constraints in evolution under changing


7/29/2019 Art 4 - Clima

7/12

July2005 PALEOPERSPECTIVESN ECOLOGY 17095.0 -0.72

RA B c, CD *n -0.74*- *# 00

> * . )-0.78*F_2-0.7680S-082 *** .3.0 '00a) o0-0.80-0rr = -0.82 *** ? rA=0.47 *2.0 -0.82

Trait 2 1.2 1.3 1.4 1.5 1.6 1.7 1.2 1.3 1.4 1.5 1.6 1.7log(Leafnumber) log(Leafnumber)FIG.3. Illustration f the influenceof genetic correlationsamongtraitson selectionresponse,reproducedrom EttersonandShaw(2001). (A) Hypotheticalpositive geneticcorrelationbetweentwo traits(eachpointrepresents hebreedingvalue,similar o a familymean,foreachof two traits).Thereare two selectionscenarios: orR (reinforcing election),thedepictedrA correlation etween raits) s in accordwiththedirectionof selection,enhancing volutionary esponse; orA (antagonisticselection), the depicted rA (correlationbetween traits) is in the opposite directionfrom selection, inhibitingevolutionaryresponse.(B) Scatterplot of reproductive tageand leaf numberbreedingvaluesfora Minnesotapopulationof Chamaecristafasciculata when grown in the hotter,drier climate of Oklahoma,showing significantnegative genetic correlation hat isantagonistic o thepositivevectorof joint selectionon these traits. C) Scatterplotof the Minnesotapopulationeaf thicknessand leaf numberbreedingvalues, showing a significantpositive genetic correlation hat s also antagonistic o the negativevector of joint selection. Both (B) and (C) are examplesof adversecorrelationsbetween traits that are predicted o limitadaptiveevolutionaryresponseof Minnesotapopulationsof this species underconditions of higher temperature nd soilmoisturestress.

climate. Most importantly, they demonstrate the evo-lutionary lability of limits of environmental tolerance,even when genetic variation is restricted to variationgenerated by spontaneous mutation. Clearly, when in-flux of variation due to mutation is added to the stand-ing genetic variation already present in populations ofdiploid organisms, tolerance limits for a species or pop-ulation should not be regarded as fixed.

EVOLUTIONARY MODELS INFORM INTERPRETATIONOF THE PALEORECORD

Interpretations of fossils often assume stability ofspecies tolerance limits. This issue is examined in re-cently developed evolutionary theory, which considersthe interplay of evolutionary and demographic pro-cesses in the evolution of tolerance limits. The differ-ential population growth rates of central and peripheralpopulations appear to produce a bias toward stabilityof tolerance limits. Gene flow from more central pop-ulations through pollen or seed dispersal overwhelmsgenetic adaptation in the much smaller peripheral pop-ulations, reducing fitness (Antonovics 1976, Garcia-Ramos and Kirkpatrick 1997).Although paleoecologists generally assume constan-cy of environmental tolerance limits, paleoecologicaldata can sometimes be used to test the hypothesis thatlimits have changed over time. An example is a studythat predicts fossil pollen distributions at various timehorizons from global climate model output, using trans-fer functions derived from patterns of modern pollenand modern climate (Prentice et al. 1991). Predictionsof oak pollen distributions in North America during thelast glacial maximum and at various time horizons

within the Holocene appear to correspond well withthe distributions of fossils (Fig. 4a), suggesting that therelationship between climate and pollen abundance hasremained stable over the past 24 000 years for the morethan two dozen eastern North American oak speciesconsidered collectively. However, spruce pollen distri-bution is underpredicted by the model 24 000, 21 000,and 16000 calendar years ago (Fig. 4b). In this case aprobable explanation is greater breadth of tolerancelimits in the past for the genus Picea because the extinctspruce species, Picea Critchfieldii, was distributedwidely in southeastern United States. This species,known only from fossil remains, was apparently adapt-ed to warmer climate than other eastern Americanspruce species, given its co-occurrence with temperatedeciduous trees in the southern Mississippi valley. Itdied out around 15 000 years ago (Jackson and Weng1999). Ackerly (2003) suggests that diminished com-petition from congeners might promote adaptation oftrailing-edge species to changing climate, but in thiscase extinction occurred instead, even as other sprucespecies migrated northward. Climate changes were ex-ceptionally rapid at the time Picea Critchfieldii becameextinct, in some regions rivaling in speed future chang-es predicted to result from greenhouse warming (Hunt-ley et al. 1997).A third illustration is provided by eastern hemlock(Tsuga canadensis). The distribution of pollen fromthis tree species in the three oldest time horizons isoverpredicted by the climate model (Fig. 4c). Over-prediction using parameters derived from modern pop-ulations suggests that tolerance limits for this speciesmight have been narrower in the past than they are at

-Ur-rP1


7/29/2019 Art 4 - Clima

8/12

1710 MARGARET B. DAVIS ET AL. Ecology, Vol. 86, No. 7

L

A)Oak,observed

Oak, imulatedbyinferredlimate

BSpruce, observed

ruce, imulatedyinferredlimate

C)Hemlock,bserved

Hemlock,imulatedbyinferredlimate

24000 21 000 16000 11 000 6000 3000 0calyrBPFIG. 4. (A) Fossil oak (Quercus spp.) pollen abundance in eastern North America during the full-glacial (24 000 yearsago) and six younger time intervals, compared to pollen abundance predicted from climate model output, using transferfunctions derived from modern pollen and climate. Time is expressed in calibrated years before 1950. Observed and predictedabundance correspond well. (B) Similar comparisons for spruce (Picea spp.). In this case, pollen abundance is underpredictedfor the three oldest time horizons, suggesting broader environmental tolerances at that time than at present, perhaps becausethe extinct species Picea Critchfieldii was still present. (C) Similar comparisons for eastern hemlock (Tsuga canadensis).The location of the hemlock refuge (judged from locations where hemlock pollen first appears) is correctly predicted, butpopulation density is overpredicted at the earlier time horizons, when latitudinal temperature gradients were steeper thantoday. One explanation is a narrowing of the tolerance limits of the species at that time. Predictions correspond better tofossil records deposited during the last 6000 yr. (The figure is modified from Prentice et al. [1991].)


7/29/2019 Art 4 - Clima

9/12

July2005 PALEOPERSPECTIVESN ECOLOGY 1711

K*

> % - 0 . 2 5

00 15 30Location X)

FIG.5. Theoreticaldepictionof populationdensity alongan environmentalradient.Xindicates ocationalongthegra-dient,only half of whichis shownin the figure.X = 0 wherethe environment s optimalfor the species. B is an index ofthe steepness of the gradient.WhenB is small, populationdensity is at carrying capacityK* whereverenvironmentalconditions are close to optimal,and the environmental m-plitudefor the species is large. When B is large, indicatinga steep gradient,populationdensityfalls to low levels, genemigration verwhelmsmarginalpopulations, ndrange imitscontract. Thefigure s reproducedromKirkpatrickndBar-ton [1997].)

present. Of course, errors may explain the overpred-iction, for example, problems in the climate predictionsor errors in evaluating the climate-pollen transfer func-tion. Background vegetation with higher pollen pro-ductivity suppressing hemlock pollen percentages is athird possibility.Kirkpatrick and Barton (1997) develop a model thatoffers a possible evolutionary explanation for this dis-crepancy, considering that latitudinal temperature gra-dients were steeper than today during the last glacialmaximum. Their model investigates adaptation in a se-ries of populations along an environmental gradient(Fig. 5). Parameters describe gene flow and steepnessof the environmental gradient (B in Fig. 5). When thegradient is gradual, pollen and seed dispersal bring toeach adjacent population genes selected in similar en-vironments. Under this circumstance, marginal popu-lations in the model may adapt to extreme conditions,and the species tolerance limits and geographic rangeexpand. When the environmental gradient is steep,however, populations subject to drastically different en-vironmental conditions are growing in close juxtapo-sition, and gene flow from more central populationsmay overwhelm adaptation in marginal populations,such that they become demographic sinks or go extinct.In this way, both the species boundary and physiolog-ical tolerance range can shrink. We note, however, thatmore recent work of Barton (2001) demonstrates thatthe outcome, in which limits to a species range evolve,is highly sensitive to Kirkpatrick and Barton's (1997)assumption of constant genetic variance. The plausiblealternative, in which genetic variance evolves alongwith traitmeans, can lead to indefinite range expansion.Because genetic details that would influence the evo-lution of variance of quantitative traits are elusive, it

is not possible at present to assess how applicable thesedifferent cases are to particular species.Latitudinal temperature gradients are anticipated tobecome less steep in future as high latitudes warm morerapidly than low latitudes under the influence of green-house gases. This scenario was examined by Garcia-Ramos and Kirkpatrick (1997). They modified theKirkpatrick-Barton model to examine population dy-namics and genetic changes within a population alonga climate gradient that varied in steepness. Lesseningof the gradient in the model increased the speed ofadaptation and thus the expansion rate for marginalpopulations invading new habitat.Case and Taper(2000) extended the Kirkpatrick-Bar-ton model to explore the interaction between environ-mental gradients and interspecific competition. Alongenvironmental gradients, competition from neighbor-ing species plays a critical role by reducing densitiesof marginal populations, augmenting the effect of geneflow from populations nearer the center in swampingout adaptations to local environments. Case and Taperexamined different scenarios in which interspecificcompetition interacts with environmental gradients ofvarying steepness to limit the sizes of species ranges.Gomulkiewicz and Holt (1995) have modeled the prob-lem of adaptation in populations that are sufficientlymaladapted that they persist only as a result of im-migration, "black-hole sinks." They investigated con-ditions under which the population evolves such thatits growth becomes positive before its numbers declineto zero. Their model shows that, the smaller the sinkpopulation and the more maladapted it is, the morelikely it is that it will become extinct before it achievesa positive growth rate. The consequence, in this in-stance, is that the range contracts.The effect on evolutionary response of temporal con-tinuity in the trajectory of environmental change isaddressed by Pease et al. (1989). These authors useda model to examine the relative importance of evolu-tion, migration and habitat selection to population fit-ness in the course of a range shift induced by contin-uously changing climate. They modeled a populationwith maximum fitness at a point along a climate gra-dient. Environmental parameters affecting fitnesschanged steadily through time at all points along thegradient. If climate changed sufficiently slowly, thespecies persisted, shifting its range along the climategradient. Persistence of the species also depended ongenetic variability and the pace of evolutionary ad-aptation to the changing climate.Evolution in an environment undergoing constanttemporal change has also been modeled by Btirger andLynch (1995). With effects of mutations on phenotypessymmetrically distributed around zero, they found thatthe mean phenotype evolves in a lag behind the phe-notypic optimum as the optimum changes with chang-ing environment. The chance of extinction depends onthe magnitude of this lag, which, in turn, depends both

"0Of-DP1C)--m

SjlP17


7/29/2019 Art 4 - Clima

10/12

1712 MARGARETB. DAVIS ET AL. Ecology,Vol. 86, No. 7

3

on the rate of environmental change and on the influxof variation due to mutation. It will complicate thisscenario to consider climate variability on the manydifferent time scales that paleoecologists find super-imposed on long-term climate trajectories.Depictions of climate in genetic models are increas-ingly sophisticated. Older models imposed a step-change in climate, preceded and followed by climatestasis, whereas these more recent papers incorporatepaleoclimate reconstructions and the predictions ofglobal climate models.

USINGTHEFOSSILRECORDTOTESTEVOLUTIONARYYPOTHESES

Paleorecords reveal changes in past distribution andabundance of plant and animal species. This infor-mation provides a historical context that can be usedin conjunction with studies of contemporary geneticpatterns to test evolutionary hypotheses (e.g., Hamricket al. 1994). Well-preserved fossils are now providingdirect evidence of evolutionary changes coincidentwith environmental change, supplementing the infor-mation derived from neutral markers in contemporarypopulations. Recently, DNA recovered from Daphniaeggs in lake sediments several decades in age, and fromfossil pollen grains 150000 years old has providedidentifications at the species level, refining our under-standing of the historical distributions of species (Suy-ama et al. 1996, Hairston et al. 2005). Like the dis-covery of Picea Critchfieldii, such identifications il-luminate the role of individual species in ecosystemprocesses, and bear on the lability of species' tolerancelimits under changed climates of the past.The new field of "Resurrection Ecology" goes a stepfurther, reviving propagules of plants and animals pre-served in sediment deposited at different time horizons,and directly comparing the environmental responses ofthe resulting "resurrected" populations. Seeds of twograminoid species, buried in arctic soils by solifluctionlobes over a century earlier, were germinated andgrown in common conditions with seeds from muchyounger soil horizons. Significant differences in phys-iological response to temperature were found for theseperennial plants (Fig. 6A) (McGraw and Fetcher 1992).Similarly, evolutionary changes in zooplankton havebeen observed in populations hatched from resting eggsburied in lake sediments (see references cited in Hair-ston et al. 2005). For example, comparison of Daphniasubpopulations dating from successive time horizonsranging from 0 to 35 years in age permitted directobservation of a time series of traits associated withfitness. In this study, the physiological responses ofDaphnia demonstrated evolutionary adaptation on thedecadal time scale: clones of Daphnia from modern,eutrophic Lake Constance are significantly more resis-tant to toxin-containing cyanobacteria than clones orig-inating from 35-year-old resting eggs deposited beforecyanobacteria became abundant (Hairston et al. 1999)

A6 -S5= 4F-_.o3S2a Youngeeds1 --- Intermediateeeds----- Old eeds05 a10 20 30

Temperature?C)B

1995-1997and 0** * .a 1992-19940)Q) 1978-1980 * *.0 1969-1971and . * * *1962-1964

0 0.1 0.2 0.3 0.4 0.5highresistance lowresistanceGrowth ate reduction

FIG. 6. (A) Differentialresponse to temperatureundergrowth-chamber onditions of Carex bigelowii subpopula-tions resurrected rom seeds of differentages buriedundersolifluction obes. The oldest seedsare 175yrold;theyoun-gest seeds are 0-20 yr old. (Panel A is reproduced romMcGrawandFetcher 1992], withpermission romElsevier.)(B) Differentialsensitivityto cyanobacteriaoxin of subpop-ulationsof Daphniahatched romrestingeggs recovered romLake Constancesediment. The oldest populationsdatefrom1962-1971, and the youngest from 1992-1997. Cyanobac-teria began to increaseduringthe late 1960s, reachedhighlevels by 1970, and have remainedabundant ubsequently.At the sametime,Daphniapopulationsdeveloped ncreasedresistance to cyanobacteria oxins. (Panel B is reproducedfrom Hairstonet al. [1999], with permission romNature.)Both examples(A) and(B) demonstrate hanges n adaptivetraits on the decadaltime scale.

(Fig. 6B). In Lake Superior, Daphnia resurrected fromsediments dating from early in the 20th century up tothe present document rapid morphological change andspecies replacement, as dredging and enrichmentchanged the near-shore environment (Kerfoot et al.1999). Using similar methods, Hairston et al. (2005)document the effects on a lake ecosystem of speciesreplacements induced by industrial contamination dur-ing the past half-century.Resurrected populations include only the last 150years (so far). Nevertheless, they demonstrate evolu-tionary change in adaptive traits, even in organismswith relatively long generation times. They providestrong evidence against the idea that rapid environ-mental change overwhelms evolutionary processes,


7/29/2019 Art 4 - Clima

11/12

July2005 PALEOPERSPECTIVESN ECOLOGY 1713

preventing adaptation to local environments as theyvary in time and space (Webb 1997, Jackson 2000).CONCLUSIONS

Adaptive responses to climate, while clearly impor-tant, are not yet well understood. Future researchshould have the goal of determining the pace of adap-tive changes in different species and groups of organ-isms. How closely can adaptation be expected to trackclimate change, particularly changes as rapid as en-visioned for the future? How will adaptationbe affectedby changed environmental gradients? We have re-viewed several processes that might limit rates of ad-aptation to changing climate. But how prevalent andhow strong are evolutionary constraints due to adversegenetic correlations among traits, as encountered inChamaecrista fasciculata (Etterson and Shaw 2001)and E. coli (Mongold et al. 1999), and how persistentwill they be? Do differences in rates of adaptationamong organisms differing in breeding systems, spatialgenetic structure, or demographic properties accountfor phenomena observed in the fossil record, such asdifferent migration rates? Does adaptation explain whysome species survived as tiny populations in multiplerefuges, while others maintained continuous popula-tions, shifting to different latitudes? Can differencesamong taxa in rates or degrees of adaptation explainwhy some ancient forest communities have no analogueat present? Do differences in rates of adaptation tochanging environments determine which taxa persistand which become extinct?

Answers to these questions will help in predictinghow populations and ecosystems are likely to respondto the changes in climate occurring as greenhouse gasesaccumulate in the atmosphere. But they will also helpto provide more complete explanations of the past. Un-til now evolutionary responses to changing environ-ments have hardly been considered in the vast literatureabout the late Quaternary. In contrast, genetic theory,models and experiments that deal with adaptation to acontinuously changing environment are progressingrapidly. By using knowledge from the past we may beable to accelerate research into the mechanisms of ad-aptation to changing environments, obtaining a morecomplete understanding of the past as well as morerealistic predictions of the future.

ACKNOWLEDGMENTSOurparticularhanksgo to the editors of this SpecialFea-ture,LindaBrubaker, isaGraumlich, ndShinyaSugita.Wereceivedstimulating ommentsand criticismfrom reviewersDavidAckerlyand BrianHuntley,and fromGeraldRehfeldt,Linda Brubaker,Eric Lonsdorf, Stacy Halpern,and BryanHoltz. We also express our gratitude o ChristineDouglas,James Ferrari and numerousundergraduate ssistants forcompilingthe datapresented n Table1. Weapologizeto themanyauthorswhose workon this subjectcould not be citedbecauseof spacelimitations.Readerswho wish a morecom-plete list of references should consult the Appendix. Thiswork was supportedby the National Science Foundation

(grantsDMS 0083468 and DEB9221371), the GuggenheimFoundation,andby the MellonFoundation.LITERATURE CITED

Ackerly, D. D. 2003. Communityassembly,niche conser-vatism,and adaptiveevolutionin changingenvironments.International ournalof Plant Science 164:S165-S184.Antonovics,J. 1976. Thenatureof limitsto natural election.Annals of the MissouriBotanicalGarden63:224-247.Barton,N. H. 2001. Adaptationat the edge of a species'range. Pages 365-392 in J. Silvertownand J. Antonovics,editors.Integrating cology and evolution n a spatialcon-text. Special Symposiaof the BritishEcological Society,BlackwellScience, Oxford,UK.Bennett,A. E, R. E. Lenski, and J. E. Mittler.1992. Evo-lutionaryadaptationo temperature.. Fitnessresponses oEscherichiacoli to changes in its thermalenvironment.Evolution 46:16-30.Bennett,K. D. 1997. Evolutionandecology: thepaceof life.CambridgeUniversityPress, Cambridge,UK.Birger, R., and M. Lynch.1995. Evolutionandextinction na changing environment:a quantitativegenetic analysis.Evolution 49:151-163.Carter,K. K. 1996. Provenance ests as indicatorsof growthresponse o climatechangein 10 north emperatereespe-cies. Canadian Journal of Forestry Research 26:1089-1095.Case,T.J., andM. L. Taper. 000. Interspecific ompetition,environmental radients, ene flow and coevolutionof spe-cies' borders.AmericanNaturalist155:583-605.Clark,J. S., et al. 1998. Reid's paradoxof rapidplantmi-gration: dispersaltheory and interpretation f paleoeco-logical records.BioScience 48:13-24.Clausen, J., D. D. Keck, and W. M. Hiesey. 1940. Experi-mental studies on the natureof species. I. The effect ofvaried environmenton western North American Plants.Publication 20. Carnegie nstitution f Washington,Wash-ington,D.C., USA.Comes, H. P.,and J. W. Kadereit.1998. The effect of Qua-ternaryclimaticchanges on plantdistributionand evolu-tion. Trends n Plant Science 3:432-438.Cruzan,M.B., and A. R. Templeton. 000. Paleoecologyandcoalescence:phylogeographic nalysisof hypotheses romthe fossil record.Trendsn EcologyandEvolution15:491-496.Cwynar,L. C., and G. M. MacDonald. 1987. Geographicvariation f lodgepolepineinrelation opopulation istory.AmericanNaturalist129:463-469.Davis, M. B., and R. G. Shaw. 2001. Rangeshiftsandadap-tive responses o Quaternary limatechange.Science 292:673-679.Etterson,J. R. 2004a. Evolutionarypotentialof Chamae-cristafasciculata in relationto climate change:I. Clinalpatterns f selectionalonganenvironmental radientntheGreatPlains. Evolution58:1446-1458.Etterson,J. R. 2004b. Evolutionarypotential of Chamae-crista asciculata in relation o climatechange:II. Geneticarchitecture f threepopulations eciprocallyplantedalongan environmental radient n the GreatPlains. Evolution58:1459-1471.Etterson,J. R., andR. G. Shaw.2001. Constrainto adaptiveevolution in response o global warming.Science294:151-154.Garcia-Ramos,G.,andM. Kirkpatrick.997. Geneticmodelsof adaptation ndgene flowinperipheral opulations.Evo-lution 51:21-18.Garcia-Ramos,G., and D. Rodriguez.2002. Evolutionaryspeed of species invasions. Evolution56:661-668.Gomulkiewicz,R., and R. D. Holt. 1995. Whendoes evo-lution by natural selection preventextinction.Evolution49:201-207.

-UrP1?


7/29/2019 Art 4 - Clima

12/12

1714 MARGARET B. DAVIS ET AL. Ecology, Vol. 86, No. 7

Hairston, N. G., C. M. Kearns, L. P. Demma, and S. W. Effler.2005. Species-specific Daphnia phenotypes: a history ofindustrial pollution and pelagic ecosystem response. Ecol-ogy 86:1669-1678.Hairston, N. G., Jr., W. Lampert, C. E. Caceres, C. L. Holt-meier, L. J. Weider, U. Gaedke, and D. M. Post. 1999.Rapid evolution revealed by dormanteggs. Nature401:466.Hamrick, J. L., A. E Schnabel, and P. V. Wells. 1994. Dis-tribution of genetic diversity within and among populationsof Great Basin conifers. Pages 147-161 in K. T. Hager, L.L. St. Clair, K. H. Thome, and W. W. Witters, editors.Natural history of the Colorado Plateau and Great Basin.University of Colorado Press, Niwot, Colorado, USA.Hewitt, G. 2000. The genetic legacy of the Quaternary iceages. Nature 405:907-913.Huntley, B., W. Cramer, A. V. Morgan, H. C. Prentice, andJ. R. M. Allen. 1997. Predicting the response of terrestrialbiota to future environmental changes. Pages 487-504 inB. Huntley, W. Cramer, A. V. Morgan, H. C. Prentice, andJ. R. M. Allen, editors. Past and future rapid environmentalchanges: the spatial and evolutionary responses of terres-trial biota. NATO ASI Series. Volume I 47. Springer-Ver-lag, Berlin, Germany.Hurme, P., R. Repo, O. Savolainen, and T. Paakkonen. 1997.Climatic adaptation of bud set and frost hardiness in Scotspine (Pinus sylvestris). Canadian Journal of Forest Re-search 27:716-723.Jackson, S. T. 2000. Out of the garden and into the cooler?A Quaternaryperspective on deep- time paleoecology. Pag-es 287-308 in A. Gastaldo and W. A. DiMichele, editors.Evolution of phanerozoic terrestrial ecosystems. Paleon-tological Society Papers. Volume 6. New Haven, Con-necticut, USA.Jackson, S. T., and C. Weng. 1999. Late Quaternary extinc-tion of a tree species in eastern North America. Proceedingsof the National Academy of Science (USA) 96:13847-13852.

Kerfoot, W. C., J. A. Robbins, and L. J. Weider. 1999. A newapproach to historical reconstruction: combining descrip-tive and experimental paleolimnology. Limnology andOceanography 44:1232-1247.Kirkpatrick, M., and N. H. Barton. 1997. Evolution of aspecies range. American Naturalist 150:1-23.Lacey, E. P. 1988. Latitudinal variation in reproductive tim-ing of a short-lived monocarp, Daucus carota (Apiaceae).Ecology 69:220-232.Lenski, R. E. 2001. Testing Antonovics' five tenets of eco-logical genetics: experiments with bacteria at the interfaceof ecology and genetics. Pages 25-45 in M. C. Press, N.J. Huntly, and S. Levin, editors. Ecology: achievement andchallenge. Blackwell Science, Oxford, UK.McGraw, J. B., and N. Fetcher. 1992. Response of tundraplant populations to climatic change. Pages 359-376 in EChapin, editor. Arctic ecosystems in a changing climate.Academic Press, New York, New York, USA.

Mongold, J. A., A. E Bennett, and R. E. Lenski. 1999. Evo-lutionary adaptation to temperature. VII. Extension of theupper thermal limit of Escherichia coli. Evolution 53:386-394.Muona, O. 1990. Population genetics in forest tree improve-ment. Pages 282-298 in A. H. Brown, M. T. Clegg, A. L.Kahler, and B. S. Weir, editors. Plant population genetics,breeding, and genetic resources. Sinauer, Sunderland, Mas-

sachusetts, USA.Pease, C. M., R. Lande, and J. J. Bull. 1989. A model ofpopulation growth, dispersal, and evolution in a changingenvironment. Ecology 70:1657-1664.Peterson, A. T, J. Soberon, and V. Sanchez-Cordero. 1999.Conservatism of ecological niches in evolutionary time.Science 285:1265-1267.

Prentice, C., P.J. Bartlein, and T.Webb, III. 1991. Vegetationand climate change in eastern North America since the lastglacial maximum. Ecology 72:2038-2056.Rehfeldt, G. E., N. M. Tchebakova, and E. I. Parfenova. 2004.Genetic responses to climate and climate-change in conifersof the temperate and boreal forests. Recent Research andDevelopments in Genetics and Breeding 1:113-130.Rehfeldt, G. E., J. M. Tchebakova, Y. I. Parfenova, W. R.Wykoff, N. A. Kuzmina, and L. I. Milyiutin. 2002. Intra-specific response to climate in Pinus sylvestris. GlobalChange Biology 8:1-18.Rehfeldt, G. E., C. C. Ying, D. L. Spittlehouse, and D. A.Hamilton, Jr. 1999. Genetic responses to climate in Pinuscontorta: niche breadth, climate change, and reforestation.Ecological Monographs 69:375-407Reinartz, J. A. 1984. Life history variation of common mul-lein (Verbascum thapsus). I. Latitudinal differences in pop-ulation dynamics and timing of reproduction. Journal ofEcology 72:897-912.Suyama, Y., K. Kawamuro, I. Kinoshita, K. Yoshimura, Y.Tsumura, and H. Takahara. 1996. DNA sequence from fos-sil pollen of Abies spp. from Pleistocene peat. Genes andGenetic Systems 71:145-149.Tani, N., N. Tomaru, M. Araki, and K. Ohba. 1996. Geneticdiversity and differentiation in populations of Japanesestone pine (Pinus pumila) in Japan. Canadian Journal ofForest Research 26:1454-1462.

Turreson, G. 1922. The genotypic response of the plant spe-cies to the habitat. Hereditas 3:211-350.Webb, T., III. 1997. Spatial response of plant taxa to climatechange: a palaeoecological perspective. Pages 57-72 in B.Huntley, W. Cramer, A. V. Morgan, H. C. Prentice, and J.Allen, editors. Past and future rapid environmental chang-es: the spatial and evolutionary responses of terrestrial bi-ota. Springer-Verlag, Berlin, Germany.Weber, E., and B. Schmid. 1998. Latitudinal population dif-ferentiation in two species of Solidago (Asteraceae) intro-duced into Europe. American Journal of Botany 85:1110-1121.

APPENDIXA complete bibliography of references used in preparing the review is available in ESA's Electronic Data Archive: EcologicalArchives E086-092-A1.

Art 4 - Clima

Documents

Transcript of Art 4 - Clima