ARTICLES: The Evolution of Plant Ecophysiological Traits ... · repeated occurrence of plants with CAM ... fessor in the Department of Ecology and Evolutionary Biology, ... The Evolution

Plants exhibit enormous ecophysiological andfunctional diversity which underlies variation in

growth rates productivity population and communitydynamics and ecosystem function The broad congruenceof these variations with climatic and environmentalconditions on local regional and global scales has fosteredthe concept that plant ecophysiological characteristics arewell adapted to their local circumstances For example therepeated occurrence of plants with CAM (CrassulaceanAcid Metabolism) photosynthesis and succulent leaves orstems in severely water-limited environments and theindependent evolution of these traits in numerous plantlineages provides compelling evidence of the physiologicalevolution of these water-conserving traits under theinfluence of natural selection (Ehleringer and Monson1993) Similarly studies of the evolution of heavy metaltolerance confirm that natural selection may cause rapidecophysiological evolution in just a few generations leadingto local adaptation in populations just a few meters apart(Antonovics et al 1971)

Many ecophysiological traitsmdashconsidered here as allaspects of resource uptake and utilization including bio-chemistry metabolism gas exchange leaf structure andfunction nutrient and biomass allocation canopy struc-ture and growthmdashare likely to influence fitness andundergo adaptive evolution Traits affecting the assimila-tion and use of resources such as carbon water and nutri-ents directly influence plant growth Patterns of resourceallocation to growth reproduction defense and stress tol-erance are also likely to be under strong selection Pheno-typic plasticity the expression of different phenotypes by

November 2000 Vol 50 No 11 BioScience 979

Articles

David D Ackerly (e-mail dackerlystanfordedu) is an assistant professor in the Department of Biological Sciences Stanford University StanfordCA 94305 Susan A Dudley is an assistant professor in the Department of Biology McMaster University Hamilton Ontario Canada L8S 4K1

Sonia E Sultan is an associate professor in the Department of Biology Wesleyan University Middletown CT 06459 Johanna Schmitt is a pro-

fessor in the Department of Ecology and Evolutionary Biology Brown University Providence RI 02912 James S Coleman (e-mail

jcolemandriedu) is vice president for Research Business Development Desert Research Institute Reno NV 89506 C Randall Linder (e-mail

rlindermailutexasedu) is an assistant professor in the Section of Integrative Biology University of Texas Austin TX 78712 Darren R Sandquist

(e-mail dsandquistfullertonedu) is an assistant professor in the Department of Biological Science California State University Fullerton CA

92834 Monica A Geber is an associate professor in the Department of Ecology and Evolution Cornell University Ithaca NY 14853 Ann S Evans

is an associate professor in the Department of Biology University of New Mexico Albuquerque NM 87131 Todd E Dawson is an associate pro-

fessor in the Department of Integrative Biology University of California Berkeley CA 94270 Martin J Lechowicz (e-mail

martinbio1lanmcgillca) is a professor in the Department of Biology McGill University Montreal Quebec Canada H3A 1B1 copy 2000 American

Institute of Biological Sciences

The Evolution of PlantEcophysiological TraitsRecent Advances and FutureDirectionsDAVID D ACKERLY SUSAN A DUDLEY SONIA E SULTAN JOHANNA SCHMITT JAMES S COLEMANC RANDALL L INDER DARREN R SANDQUIST MONICA A GEBER ANN S EVANS TODD E DAWSON ANDMARTIN J LECHOWICZ

NEW RESEARCH ADDRESSES NATURAL

SELECTION GENETIC CONSTRAINTS AND

THE ADAPTIVE EVOLUTION OF PLANT

ECOPHYSIOLOGICAL TRAITS

the same genotype in response to environmental varia-tion also affects plant function and hence fitness indiverse environments However only recently have plantbiologists directly studied selection in natural popula-tions Moreover until the last few years little was knownabout the genetic basis for evolutionary change in eco-physiological traits or about genetic developmental orphylogenetic constraints on the evolutionary response tonatural selection

The authors of this article presented research findingsand discussed future directions in evolutionary plant eco-physiology in a symposium at the 1998 annual meeting ofthe Ecological Society of America This article presents anoverview of advances in the field and addresses a numberof questions regarding the evolution of ecophysiologicaltraits

How are ecophysiological traits related to fitness Howdo patterns of natural selection for these traits vary withthe environment

What is the genetic basis for ecophysiological traitsHow much genetic variation for these traits exists innatural populations How do genetic and developmen-tal constraints influence the adaptive evolution of eco-physiological traits

What is the evidence for adaptive evolution of ecophysi-ological traits in natural populations Do closely relatedpopulations or species demonstrate ecophysiologicaldivergence and do their differences reflect patterns ofenvironmental variation as expected on functionalgrounds

What are the macroevolutionary patterns in ecophysio-logical traits viewed from a phylogenetic perspectiveDo these patterns suggest constraints on ecophysiologi-cal evolution or do they suggest high evolutionary labil-ity and frequent convergence

The adaptive value of ecophysiologicaltraitsAn ecophysiological trait can be considered adaptive if ithas a direct impact on fitness in natural environmentsHere we examine the successes and limitations of variedperspectives from which this problem has been addressedPhenotypic selection analysis The most straightfor-ward evidence for the adaptiveness of ecophysiologicaltraits is the observation of correlations between traits andfitness in natural populations but this approach hasproven problematic For example direct correlationsbetween photosynthetic rates and fitness are rarelyobserved in natural populations (eg Farris and Lechow-icz 1990 and references) Moreover even when correla-tions are observed it is difficult to determine whetherindividual traits contribute directly to variation in fitnessor whether these relationships reflect indirect selection via

correlated traits For example a study of Plantago lanceo-lata found a significant positive correlation between pho-tosynthetic capacity and reproductive dry weight but cor-relations were also observed for corm diameter number ofleaves leaf size specific leaf weight and transpiration rate(Tonsor and Goodnight 1997) Thus it is impossible to saythat variation in photosynthetic rate per se contributeddirectly to individual fitness

In these situations when many interacting traits mayinfluence fitness multivariate selection analysis (Landeand Arnold 1983) provides a powerful tool for detectingselection on ecophysiological traits This method involvesmeasuring a suite of traits as well as a measure of fitness(eg lifetime seed production) on individual plantsDirect selection on each trait is measured as a partialregression coefficient in a multiple regression of relativefitness on all measured traits If causal relationshipsamong traits are known path analysis can also be used todetermine the direct and indirect effects of each trait onfitness (Kingsolver and Schemske 1991)

The first selection study on plant ecophysiological traitswas conducted by Farris and Lechowicz (1990) who mea-sured selection on ecophysiological phenological mor-phological and growth traits in a Xanthium populationgrown under uniform experimental conditions They cre-ated an experimental population in the greenhouse thatmaximized genetic variation and then planted seedlingsinto a garden preventing any correlation between geno-type and microenvironment that could confound theresults Twelve traits on each plant were measured at dif-ferent times during the season and path analysis was usedto integrate the functional relationships among traits intothe selection analysis The researchers found that seedlingemergence time height and branch growth rates andwater-use efficiency all influenced total vegetative bio-mass which in turn was strongly correlated with fitnessillustrating that these traits were under selection

Similar methods have been applied successfully in sev-eral recent studies One of the most important results ofsuch research is the demonstration that patterns of phe-notypic variation its genetic basis and natural selectionvary in different environmental conditions For examplein the sand-dune annual Cakile edentula high water-useefficiency and intermediate leaf sizes were correlated withfitness in an environment with low water availabilitywhile larger leaf sizes were favored in an environment withhigh water availability (Figure 1 Dudley 1996a) In thelow-water environment the finding that genotypes withintermediate leaf size had the highest fitness illustrates theimportance of nonlinear approaches in selection studiesThis pattern indicates the occurrence of stabilizing selec-tion in which intermediate trait values are favored in con-trast with directional selection which favors lower orhigher values of a trait In addition genetic differentiationbetween populations derived from low- and high-water

980 BioScience November 2000 Vol 50 No 11

Articles

environments mirrored the selection results The low-water population had higher water-use efficiency andsmaller leaf size when lines from both populations wereraised in a common greenhouse environment (Dudley1996b)

In another example Evans (1998) grew field-collectedsibships of Townsendia annua in a factorial experimentwith high and low levels of both water and nutrients toexamine patterns of selection (ie the strength and thesign [positive versus negative] of correlations betweenecophysiological traits and fitness) in contrasting environ-ments She examined both nitrogen-use efficiency andwater-use efficiency measured using carbonnitrogenratios and carbon isotope discrimination respectively (seeEhleringer 1991 for a discussion of carbon isotope meth-ods) The adaptive value of water-use efficiency changedwith environment In the low-nutrient treatments fitness

(measured as flower production) was higher in individu-als with lower water-use efficiency whereas in the high-water and high-nutrient treatment this pattern wasreversed There was also a positive correlation betweennitrogen-use efficiency and flower production in all treat-ments Changes in patterns of selection in contrastingenvironments illustrated in these studies provide thebasis for differentiation and local adaptation of popula-tions

To understand the functional mechanisms underlyingchanging patterns of selection it is valuable to examinehow different traits interact to influence fitness in con-trasting environments For example populations of theannual Polygonum arenastrum are genetically variable inphysiology (photosynthetic rates water-use efficiency) indevelopment (node position of the first flowering meris-tem leaf production rate) and in morphology (leaf size


Articles

Figure 1 Phenotypic selection analysis (a) Phenotypicselection analysis for leaf size and water-use efficiency inCakile edentula measured in two contrastingenvironments In the wet environment plants with largeleaves and intermediate water-use efficiency exhibit thehighest fitness while in the dry environment small leavesand high water-use efficiency are favored These resultsillustrate the importance of using nonlinear analysis todetect patterns of stabilizing selection (in whichintermediate trait values are favored) (b) Scatterplot offamily means for leaf size and water-use efficiency forpopulations derived from a wet site (open circles) and a drysite (closed circles) and grown in a common environmentto evaluate genetic differences in ecophysiological traitsThe dry-site population had higher water-use efficiencyand smaller leaf size (squares with error bars showpopulation means) consistent with predictions of theselection analysis Reprinted from Dudley (1996a 1996b)with permission

a

b

Geber and Dawson 1990 1997) In addition the traitsare genetically correlated so that early-flowering geno-types initiate their first flower at a basal node and havesmall leaves high photosynthetic rates and low water-use efficiency while the converse is true of late-floweringgenotypes To determine whether selection might favordifferent suites of traits in different environments green-house studies examined the influence of these traits onplant fitness (seed number) in environments that dif-fered in water availability (Figure 2)

Results of path analysis suggest that selection on thefunctional traits (physiology development and mor-phology) can be accounted for by the effect of thesetraits on components of fitness such as life history (dateof flowering) growth (plant relative growth rate) andarchitecture (meristem number) Environmental differ-ences in patterns of selection on functional traits can beattributed to differences in the effects of these traits onfitness components in the relationships among fitnesscomponents and in the effects of fitness components onfitness (Figure 2 Monica Geber Cornell Universityunpublished data) For example selection favors a morebasal node of first flowering in drier environmentsbecause early-flowering plants have higher fitness underdrought while flowering time does not affect fitnessunder well-watered conditions Selection favors a higherleaf initiation rate in dry compared with wet environ-ments because in dry environments a high leaf initiationrate leads to a high relative growth rate and a high rela-tive growth rate results in turn in plants with manymeristems By contrast the links between leaf initiationrate relative growth rate and meristem number areweaker or absent under well-watered conditions Theresults of this experiment point to the complex relation-ships that link development and physiology to fitnessand the variable nature of these relationships across


Articles

Figure 2 Path analysis illustrating how functional traits affectfitness in Polygonum arenastrum Low-level traits include thoserelated to development (node of first flowering leaf initiationrate) morphology (leaf size) and leaf gas exchange(photosynthetic rate per unit leaf area) Fitness componentsinclude traits related to life history (flowering time) growth inbiomass (relative growth rate) and architecture (total meristemnumber) Double-headed arrows between traits indicate thatthey are correlated single-headed arrows indicate that the traitat the base of the arrow causally affects the trait at the tip of thearrow The path coefficient (not shown) which measures theeffect of one trait on another may be positive or negative Forexample the node of first flowering has a positive effect onflowering time because plants that begin flowering at a distal(high) node flower late Conversely leaf initiation rate has anegative effect on flowering time because plants that initiateleaves (and nodes) rapidly attain a reproductive size earlier thanplants with slow rates of leaf initiation The magnitude and signof paths can also change among environments

To understand the pattern of selection on traits and how thepatterns differ among environments 20 genotypes of Parenastrum were grown in replicate in three environmentsdiffering in water availability (high medium and low) andfunctional traits fitness components and fitness were measuredon each plant Path coefficients were estimated for the pathdiagram for each environment The thick colored arrows in thepath diagram are paths whose coefficients differ significantly inmagnitude or sign among environments (see text) For exampleleaf initiation rate has a strong positive effect on relative growthrate in drier (low medium) environments but no effect in a wetenvironment Differences in the pattern of selection amongenvironments are due to differences in the effect of functionaltraits on fitness components (red paths) in relationships amongfitness components (green path) and in the effects of fitnesscomponents on fitness (blue paths) From Monica Geber CornellUniversity (unpublished data)

environments Understanding the interactions amongmultiple traits and their consequences for plant perfor-mance remains one of the outstanding challenges in evo-lutionary ecophysiologyManipulating genotypes The ability to manipulateexpression of individual genes has provided new insightsinto the ecological and evolutionary consequences of vari-ation in ecophysiological traits Directed mutagenesis andtransgenic manipulation (both knockout and overexpres-sion mutants) allow researchers to generate variationbeyond the range observed in natural populations whilemaintaining a constant genetic background (Schmitt et al1999) For example in studies of tobacco plants an anti-sense version of the gene coding for the small subunit ofRubisco was inserted resulting in a range of genotypeswith variable amounts of the Rubisco protein which isresponsible for carbon fixation in photosynthesis Thesegenotypes which were otherwise identical exhibited sig-nificant differences in Rubisco enzyme levels photosyn-thetic rates responses to light levels specific leaf area bio-mass allocation and relative growth rates (Stitt andSchulze 1994) The cascade of effects resulting from thissimple genetic manipulation dramatically illustrates howchanges at the molecular level can influence ecophysiolog-ical characteristics at many levels of organization as wellas whole plant performance

Several studies have employed transgenic plants to testthe hypothesis that traits conferring physiological resis-tance to herbicides pathogens or herbivores also incurcosts in performance or fitness (Bergelson and Purring-ton 1996) For example Arntz et al (1998) tested the fit-ness effects of resistance to the herbicide atrazine usingtransgenically modified Amaranthus hybridus with aresistance allele inserted into the chloroplast genomeIndividuals with the atrazine resistance allele in an other-wise-identical genetic background had reduced photo-synthetic electron transport resulting in reductions inphotosynthetic rates of 15ndash25 In the absence of theherbicide resistant individuals were less fit than wild-type plants particularly in a high-density competitiveenvironment supporting the hypothesis that increasedphotosynthetic rate results in increased fitness In studiesof transgenically modified Arabidopsis resistance to theherbicide chlorsulfuron also conferred a cost in lifetimeseed production but there was no apparent cost of resis-tance to the antibiotic kanamyacin (Purrington andBergelson 1997) Analysis of the mechanisms of resistanceand their associated physiological effects (eg Purringtonand Bergelson 1999) may reveal why resistance confers afitness cost in some cases and not others An understand-ing of these variable costs will enrich studies of physio-logical tradeoffs and ecophysiological evolution

Transgenic plants hold enormous promise for testinghypotheses in evolutionary and ecological physiology buttheir use comes with some cautionary notes Transforma-tion technologies often introduce mutations in addition to

the intended insertion event moreover the effect of thetransgene may depend on where it is inserted and on theoverall genetic background reflecting epistatic interac-tions (De Block 1993 Purrington and Bergelson 1999)Furthermore constructs that alter the physiological phe-notype in ways that are not observed in natural popula-tions may have limited value in unraveling the ecologicaland evolutionary consequences of natural patterns ofgenetic or phenotypic variation Genetic manipulation ispotentially useful to study the adaptive role of individualtraits but in many cases a genetic change results in multi-ple phenotypic effects (pleiotropy) that may influence fit-ness in different ways (De Block 1993) Although sucheffects are realistic as mimics of the natural mutationprocess the functional role of individual traits will beinterpreted incorrectly if pleiotropic effects are not detect-ed or examined An understanding of the genetic anddevelopmental mechanisms underlying such pleiotropicand epigenetic effects will play an increasingly importantrole in understanding evolutionary change in functionallyimportant traitsManipulating phenotypes In some systems theadaptive value of individual ecophysiological traits can beassessed by direct experimental manipulation A classicexample is the study of leaf pubescence in the genusEncelia Physical removal of pubescence from the leaf sur-face demonstrated that pubescence plays a direct role inleaf energy and water balance by reducing both energyabsorption and water loss (Ehleringer et al 1976) Com-plementary studies of intraspecific and interspecific varia-tion in pubescence demonstrated correlations with cli-mate that were consistent with this functional role(Ehleringer et al 1981 Sandquist and Ehleringer 19971998 see below)

Direct manipulation of this sort is most feasible withexternal morphological traits and it has also been appliedto the study of inflorescence and flower morphology (egFishbein and Venable 1996) In studies of biochemical andphysiological processes at the cellular and intracellular lev-els experimental manipulation is a standard techniqueFor example Heckathorn et al (1998) studied isolatedchloroplasts with and without the presence of heat shockproteins (hsps) to elucidate the proteinsrsquo role in protectingthe photosystem from heat stress (see box page 984)However manipulation of physiological traits in intactplants is much more difficult and this type of experimen-tation may be possible only for a small subset of ecophys-iological and functional traitsManipulating virtual plants Computer simula-tions Computer models of plant function allow manipu-lation of individual traits assessment of the consequencesand tests of the interactions among multiple traits Suchmodels have played a prominent role in studies of canopyroot and clonal architecture For example Niklas (1988)modeled light interception on a single shoot in relation tophyllotaxy leaf shape leaf angles and internode lengths


Articles


Articles

One exciting area of advance in thepast 10 years is in the study of heatshock proteins and their functions atthe biochemical physiological andecological levels (Coleman et al1995) In plants low molecularweight heat shock proteins (lmwhsps) are rapidly induced in responseto heat shock and other environmen-tal stresses The lmw hsp genes arelocated in the nucleus and in re-sponse to stress the proteins are syn-thesized and transferred to the chlo-roplast and mitochondria In anelegant experiment Heckathorn etal (1998) isolated chloroplasts sothat the nuclear-encoded proteinscould not be synthesized and ex-posed them to heat shock In theabsence of the lmw hsps photosys-tem II experienced heat damage andelectron transport levels declinedThe subsequent addition of purifiedhsp rapidly restored photosyntheticactivity demonstrating that theseheat shock proteins protect the elec-tron transport chain from damageand may contribute to repair of dam-aged photosystem proteins

Variation in patterns of hsp pro-duction may help explain the evolu-tion of broad variation in thermo-tolerance among plant species Someevidence suggests that thermotoler-ant species such as desert cacti allo-cate a far greater proportion (several orders of magnitude) of leaf protein to chloroplast hsps than do thermosensitivespecies (Downs et al 1998) The relative allocation of protein to chloroplast hsps is also highly correlated with photosyn-thetic thermotolerance of horticultural varieties of tomato (Heckathorn et al 1999) populations of the annual plantspecies Chenopodium album (Heckathorn et al 1999) and species of the perennial shrub genus Ceanothus (Knight andAckerly in press see figure above) Given the strong correlation between chloroplast hsp production and thermotoleranceand given that all plants tested to date have the machinery to make chloroplast hsps why do all plants not produce largequantities of hsps conferring greater thermotolerance One possibility is that production of chloroplast heat shock pro-teins is physiologically costly in terms of their nitrogen requirement nitrogen limitation reduces the relative allocation ofprotein to chloroplast hsps in corn and tomato and reduces photosynthetic thermotolerance (Heckathorn et al 1996)Together these results suggest that although hsps may be required for physiological function especially in the presence ofheat shock and other stresses variation in their production may reflect tradeoffs in allocation of protein between hsps andphotosynthetic proteins and optimization of the benefits of thermotolerance with the physiological costs of hsp produc-tion Because of the detailed work on the molecular biology of hsps their conservation throughout plant evolution andtheir significance with respect to plant tolerance of environmental stresses hsps represent a model trait for the study ofoptimizing selection and ecophysiological evolution

Variation in expression of low molecular weight hsps chlorophyll fluorescence andspecific leaf area among eight species of the shrub genus Ceanothus (a) Hspexpression levels in response to a 4-hour 45degC heat shock are significantly higherin four closely related species in subgenus Cerastes compared with species ofsubgenus Ceanothus (b) Hsp expression levels are positively correlated with FvFmchlorophyll fluorescence consistent with the function of hsps in protectingphotosystem II from damage (see text) (c) Hsp expression is also negativelycorrelated with specific leaf area (SLA the ratio of leaf area to mass) Species withhigh hsp expression and low SLA (ie thick leaves) also show a suite ofphysiological traits related to greater drought tolerance Reprinted from Knight andAckerly (in press) with permission

Box 1 Heat shock proteins A model system for the study of ecophysiological evolution

By comparing different combinations of these parametershe was able to show that phyllotaxy has a very strong effecton efficiency of light interception in rosette plants but thatthe effect greatly diminishes when leaves are vertically dis-placed by elongated internodes This study illustrates thepower of simulations to explore many different trait com-binations and examine how variation in one trait in thiscase internode elongation can dramatically alter the func-tional significance of other characteristics

Pearcy and Yang (1996) have developed a more-detailedmodel (YPLANT) incorporating light interception pat-terns biomass investments in internodes petioles andleaves and leaf physiology to estimate both the costs ofcanopy construction and resulting carbon gain This mod-el helps reveal the consequences of allocational tradeoffssuch as biomass investment in leaves to increase canopyarea versus in petioles to reduce self-shading (Pearcy andValladares 1999) Models of this type make it possible toexamine the function of hypothetical plants with a widevariety of different trait combinations many of whichmay never occur in nature (or occur only in the fossilrecord see Niklas 1997) and thus provide critical com-parisons between the performance of observed pheno-types and possible alternatives that may have been elimi-nated by natural selectionTesting the adaptive plasticity hypothesis Muchof the phenotypic variation in plant populations reflectsthe direct effects of environment on plant growth anddevelopment that is phenotypic plasticity Plasticity isobserved at all levels of organization For example plantsgrown in low versus high light exhibit differences in bio-chemical and physiological aspects of photosynthesis leafanatomy and morphology allocation and canopy struc-ture and whole plant growth (Givnish 1988) Plasticitymay range from complex and apparently adaptive devel-opmental responses mediated by signal transductionpathways (eg phytochrome-mediated stem elongation indense stands Schmitt and Wulff 1993) to apparently pas-sive responses to biophysical properties of the environ-ment (eg reduced growth of rhizomes in low-tempera-ture soils MacDonald and Lieffers 1993) In some casesecophysiological plasticity may contribute to homeostasisof growth and fitness in variable environments Phenotyp-ic plasticity creates a challenge in the analysis of the adap-tive value of a trait because plants in different environ-ments will also be phenotypically distinct Thus it isdifficult to separate the contributions of trait variationand environmental variation to relative fitness

The manipulation of genotypes and phenotypes (eitherreal or simulated) provides one solution to this dilemmaThe adaptive plasticity hypothesis predicts that pheno-types produced in contrasting environments will be morefit in their respective environments when compared withalternative phenotypes (Schmitt et al 1999) Testing thishypothesis requires a factorial experiment in which the fit-ness of each phenotype is assessed in each environment


Articles

Figure 3 Test of the adaptive plasticity hypothesisin Impatiens capensis To test the adaptive valueof stem elongation elongated and suppressedplants were experimentally induced and thenplanted in mixture at different densities to testtheir relative fitness at low and high density (a)Because stem elongation is highly plastic withrespect to density the plants responded to theexperimental environments At high densitysuppressed plants grew rapidly in heightalthough they remained shorter than the high-density elongated plants throughout theexperiment Elongated plants at low densityexhibited reduced height growth and wereeventually the same height as the low-densitysuppressed plants (b) Despite these adjustmentsthe adaptive plasticity hypothesis was confirmedby a crossover in cumulative reproductive outputa measure of relative fitness Elongated plants hadhigher fitness at high density and suppressedplants had higher fitness at low density FromDudley and Schmitt (1996) with permission

a

b

For example Dudley and Schmitt (1996) combined phe-notype manipulation with selection analysis to test theadaptive value of phytochrome-mediated stem elongationin dense stands of Impatiens capensis

Based on functional considerations it is predicted thatat high density stem elongation will be favored becausethe tallest plants will receive the most light while at lowdensity excess investment in height growth would result inopportunity costs for reproduction as well as a moreunstable stem These predictions were tested by experi-mentally inducing plants to produce either elongated orsuppressed stems by manipulating the ratio of RFR wave-lengths experienced by seedlings during development(The RFR ratio of light is perceived by the light-sensitivephytochrome molecule and low RFR ratios which occurin vegetation shade or high-density stands promote stemelongation) These seedlings were then planted out in thefield in high and low densities As predicted the elongatedplants exhibited higher fitness at high density while sup-pressed plants were more fit at low density (Figure 3)Selection analysis (based on correlations between fitnessand functional traits) suggested that increased stem heightrelative to leaf size was favored in high density but not lowdensity supporting the hypothesis that the stem elonga-tion response contributes to these fitness differences andis adaptive (Dudley and Schmitt 1996)

Genetic manipulation can also be used to test the adap-tive plasticity hypothesis For example Schmitt et al (1995)tested the adaptive value of phytochrome-mediated stemelongation responses to neighbors using genetically modi-fied plants lacking these plastic responses Transgenictobacco plants in which plastic elongation responses wereblocked showed a reduction in fitness when grown at highdensity with normally elongating wild type Converselyphytochrome-deficient mutants of Brassica rapa constitu-tively expressing the elongated phenotype had lower fitnessrelative to the plastic wild type at low density These resultsconfirm that phytochrome-mediated stem elongation isadvantageous for crowded plants but maladaptive at lowdensity they thus demonstrate that plasticity for stemelongation is adaptive across this range of environments

Computer models can also be used to assess the perfor-mance of different phenotypes across a range of environ-ments Ackerly and Bazzaz (1995) conducted a ldquovirtualrdquoexperiment evaluating plants with different crown struc-tures across a range of light environments Using theYPLANT model (Pearcy and Yang 1996 see above) theyreconstructed the crown structure and leaf display of trop-ical tree seedlings growing in different treefall gap envi-ronments and evaluated their light interception capacityusing hemispherical canopy photographs taken directlyabove each plant Light interception capacity was thenevaluated in alternative microsites by repeating this analy-sis for each plant using canopy photographs from othertreefall gaps of different size and orientation The resultsshowed that for most individuals the adjustments in leaf

display (principally leaf angle) in response to the asymme-try of their light environment enhanced light interceptioncompared with what it would have been in other localitiesin the forest

These studies of the performance of alternative pheno-types in contrasting environments are analogous to recip-rocal transplant studies of genetically divergent popula-tions (see below) and provide valuable insights into theadaptive value of the responses of individual plants totheir environments

The genetic basis of ecophysiologicalevolutionPhenotypic analyses as discussed above highlight thepotential for natural selection on ecophysiological traitsStudies of the genetic basis of ecophysiological variationare critical to understanding the potentials and limitationsfor adaptive evolutionary change and the nature of thegenetic mechanisms underlying contemporary ecophysio-logical diversityGenetic variation for ecophysiological traits Fornatural selection to cause evolutionary change in ecophys-iological traits genetic variation for those traits must existwithin natural populations Adaptive differentiationamong populations (see below) demonstrates that suffi-cient genetic variation was present in the past to permitdivergent responses to selection However measuring suchvariation in natural populations has proved to be a sub-stantial challenge because of the large sample sizesrequired for quantitative genetics and the sensitivity ofecophysiological traits to fine-grained environmental vari-ability Experiments in controlled conditions are criticalfor revealing physiological differences among genotypesUsing quantitative genetic techniques heritability esti-mates can be calculated to describe the proportion of totalvariation associated with either additive or overall geneticdifferences among individuals (narrow- and broad-senseheritability respectively) Heritability estimates rangefrom 0 to 1 with higher levels indicating that a trait hasgreater potential to respond rapidly to selection Heritabil-ity estimates for ecophysiological traits vary widely rang-ing from less than 01 for instantaneous photosyntheticrate and water-use efficiency (Dudley 1996b Tonsor andGoodnight 1997) to 020ndash081 for carbon isotope dis-crimination an integrated index of water-use efficiency(Schuster et al 1992 Donovan and Ehleringer 1994) andto 066ndash094 for measures of biochemical and stomatallimitations on photosynthesis measured under carefullycontrolled conditions (Geber and Dawson 1997)

Although heritability has been addressed in only ahandful of studies evidence suggests that instantaneouslymeasured physiological traits have lower heritabilitiesthan traits that reflect integrated ecophysiological process-es over time such as carbon isotope discrimination Thisdifference probably reflects the increased variability ofinstantaneous traits caused by environmental fluctuations


Articles

and developmental effects (eg leaf age) whereas thesefactors are integrated over longer time scales in traits suchas carbon isotope discrimination If we could measureintegrated photosynthetic rates over the lifetime of eachleaf and each individual under controlled conditionsgenetic differences among individuals might become moreapparent Moreover it is important to note that heritabil-ity does not necessarily relate to the evolutionary signifi-cance of a trait since low heritability may reflect strongprior selection that has eliminated genetic variation in atrait of adaptive importanceGenetic constraints on adaptive evolution Whileheritable variation provides the raw material for adaptiveevolution a variety of genetic factors may limit evolution-ary responses to selection The most direct constraint isthe absence of appropriate genetic variation For examplethe evolution of heavy metal tolerance in contaminatedsoils has been observed in populations of a variety of plantspecies (Antonovics et al 1971) However in a series offive Agrostis capillaris populations growing along a utilityline where the soils beneath each pylon were heavily con-taminated by zinc variable levels of zinc tolerance wereobserved In one population tolerance had failed toevolve Greenhouse experiments showed no genetic varia-tion for zinc tolerance in this population even though itwas located less than 1 km from other populations inwhich such variation did occur and tolerance had evolved(Alhiyaly et al 1993) The lack of appropriate genetic vari-ation for adaptive response to selective pressures may befrequent in natural populations and its role in under-standing patterns of adaptive evolution deserves greaterattention (Bradshaw 1991)

Plant mating systems and life histories will influence thetransmission of heritable variation and may also imposesignificant constraints on adaptive evolution In particu-lar asexual reproduction which prevents genetic recombi-nation is relatively common in flowering plants (Briggsand Walters 1997) While asexuality does not diminish theefficacy of selection (it may actually promote rapidresponses by selective ldquosweepsrdquo of favorable genotypes) itwill limit the generation of new and potentially adaptivegenotypes through recombination The very long life spanof some woody plants and the prolongation of geneticlongevity in clonal plants through vegetative reproduc-tion will also reduce the potential for rapid response toselection in many plant species

When natural selection acts on two or more traitssimultaneously evolutionary responses may be limited bygenetic correlations that reflect genetic physiologicalandor developmental constraints For example Dudley(1996b) observed positive genetic correlations (iebetween the mean values for different genotypes)between stomatal conductance and photosynthetic rateand between leaf size and water-use efficiency in Cakileedentula She also detected a genetic tradeoff betweenstomatal conductance and water-use efficiency indicating

that selection to enhance one trait will lead to a decline inthe other In populations of the annual plant Polygonumarenastrum there is genetic variation in and covariationbetween developmental (node position on the main stemof the first flowering meristem) morphological (leafsize) and ecophysiological traits (photosynthetic rateswater-use efficiency electron transport activity Geberand Dawson 1990 1997) Genotypes that flower first at ahigher node on the main stem have larger leaves and low-er rates of gas exchange than those that start flowering atlower nodes

The expression of genetic variation and covariation alsochanges with environmental conditions For example inTownsendia annua Evans (1998) observed significantfamilyndashnutrient treatment interactions for nitrogen- andwater-use efficiency suggesting genetic variation in theresponse to nutrient availability Genetic correlationsbetween water- and nitrogen-use efficiencies also changedwith environment At higher nitrogen availability water-use efficiency was positively correlated with nitrogen-useefficiency but at lower nitrogen levels there was a negativecorrelation between these traits This change in the trade-off in resource-use efficiency suggests that the action ofgenes governing pleiotropic effects changes with environ-ment Together these three examples of genetic correla-tions demonstrate that ecophysiological traits cannot nec-essarily respond independently to selection Strongselection on one trait may result in correlated evolutionaryresponses in others and genetic correlations betweentraits with opposing effects on fitness may constrain theirresponses to selection

Pleiotropic linkages between traits may also constrainresponses to selection For example one of the first report-ed etiolation mutants of Arabidopsis which lacked thenormal stem elongation response in darkness was alsofound to have green roots (Chory et al 1989) Apparentlythe absence of chlorophyll in roots is itself a developmen-tal response to darkness that is regulated in part by thesame gene as the etiolation response If in a particularenvironment selection favored a mutation at this locusreducing the etiolation response the linkage between eti-olation and root development might limit the response toselection because of the energy costs of unnecessarychlorophyll production in rootsGenotype by environment interactions Pheno-typic plasticity Several recent studies have shown theexistence of genetic diversity for the responses of ecophys-iological traits across diverse environments that is genet-ic variation in patterns of phenotypic plasticity In statisti-cal terms variation in plasticity appears as a significantgenotypendashenvironment interaction in analysis of varianceor as significant genetic variation in measures of plasticresponses Variation in ecophysiological plasticity has beenobserved among both populations (eg Emery et al 1994Dudley and Schmitt 1995) and genotypes within popula-tions (eg Schmitt 1993 Sultan and Bazzaz 1993a 1993b


Articles

1993c) The observation of widespread genotypendashenvi-ronment interaction for ecophysiological traits in naturalplant populations suggests that the genetic potential oftenexists for the pattern or amount of adaptive plasticity toevolve in response to variable selection in heterogeneousenvironments

Variation in plasticity among genotypes may have fur-ther effects on evolutionary responses to selection in nat-ural populations If reaction norms (ie patterns of phe-notypic expression) converge in one environment thenheritability and thus the potential for response to selectionwill be lower in that environment than in other environ-ments where greater genetic variation is expressed If reac-tion norms for performance traits cross so that differentgenotypes are favored in different environments then envi-ronmental heterogeneity may contribute to the mainte-nance of genetic variation Sultan and Bazzaz (1993a1993b 1993c) compared norms of reaction for naturallyoccurring genotypes of Polygonum persicaria alongresource gradients (Figure 4) Although the genotypes dif-fered significantly for many traits these differences variedfrom one environment to another and most genotypesexhibited appropriate phenotypic responses that reducedthe deleterious effects of stressful environments Such com-plex patterns of plasticity variation may permit diverse

genotypes to be maintained in populations inhabiting vari-able environments

Phenotypic plasticity may also evolve as an individualcharacteristic in response to fine-grained environmentalheterogeneity (Lloyd 1984) For example leaves develop-ing in sun and shade within the canopy of a single plantor below and above the water level in the case of hetero-phyllous aquatic plants often exhibit physiological andstructural differences Plants that produce these divergentleaf forms may have greater fitness than an individual ableto produce only one type or the other However althoughWinn (1996) documented the existence of genetic varia-tion for within-individual variation in leaf traits no fit-ness differences were detected in association with this vari-ation in fine-grained plasticity (Winn 1999b) Theexpression of phenotypic plasticity within individuals(eg in different leaves or at different developmentalstages) is extremely important in plants because of theirmodular growth and indeterminate development andstudies of the evolution of plasticity at this level meritgreater attention in the future (Winn 1999a)

These studies of genetic variation reflect a fruitful syn-thesis of the technical approaches of ecophysiology withthe statistical methods of population biology and evolu-tionary genetics The strength of this quantitative


Articles

Figure 4 Genetic variation in phenotypic plasticity Patterns of individual plasticity (reaction norms) for whole-plant leafarea in the widespread annual Polygonum persicaria Each line depicts the response of a single genotype based on four tosix clonal replicates grown in each environment Panels A and B show the responses of the same ten genotypes grown alonggradients of moisture or light availability These patterns exemplify how the amount of genetic variation in a trait dependsupon environmental conditions for example on the moisture gradient there is greater variation among genotypes underfield capacity conditions than in the dry or moist environments In addition the relative ranking of different genotypeschanges with environment and no single genotype exhibits highest performance in all environments within or acrossgradients As a result selection in variable environments may result in the maintenance of genetic variability From Sultanand Bazzaz (1993a 1993b) with permission

100

approach is that it permits the study of genetic variationand selective dynamics based on phenotypic data in exper-imental and natural populations However this is also itslimitation because the specific genes responsible for heri-table variation are not examinedQTL mapping and candidate genes One of themost exciting advances in recent years which addressesthis limitation is the use of tools from molecular anddevelopmental genetics to examine the genetic basis ofphenotypes at a mechanistic level Two of the most impor-tant approaches in this research are the study of candidategenes controlling a trait of interest and mapping of quan-titative trait loci (QTL) to search for unknown or multipleloci affecting trait expression

Techniques for mapping QTL have recently been usedto examine genes involved in plant responses to differentlight environments For example Dorn and Mitchell-Olds(unpublished data) used hybrid segregates of a crossbetween two natural Arabidopsis thaliana ecotypes to mapseveral loci involved in plasticity of flowering time androsette leaf number in response to differences in spectrallight quality Yanovsky et al (1997) used recombinantinbred lines of a cross between two Arabidopsis ecotypes tomap two QTL associated with naturally occurring varia-tion in signal transduction of very-low-fluence responsesThese responses which are mediated by phytochrome Ainclude seedling germination and de-etiolation inresponse to brief pulses of very dim light Once a quanti-tative trait locus has been mapped candidate gene studiesmay be used to determine whether it is linked with a par-ticular locus using mutant and transgenic approaches

Adaptive differentiation of populationsand speciesThe study of evolutionary processes within populations asdiscussed above is critical to understand the potentialsand limitations for adaptive evolution Variation in pat-terns of selection on genetically variable traits in differentenvironments provides the conditions for adaptive diver-gence of conspecific populations Comparative studies ofecophysiological variation between populations andspecies have long served as a powerful tool for examiningthese patterns of adaptive evolutionEcophysiological variation among populationsand species The study of genetic differentiation amongplant populations based on common garden experimentsand reciprocal transplants was pioneered by Turreson(1922) in Sweden and Clausen and colleagues in NorthAmerica (Clausen et al 1940) Since then numerous stud-ies have documented evolutionary divergence in ecophys-iological and life history traits including photoperiod andtemperature responses specific leaf area photosyntheticrate water-use efficiency leaf size allocation etc (Earlystudies were reviewed in Hiesey and Milner 1965 recentexamples include McGraw and Antonovics 1983 Gurevitch

1992 Monson et al 1992 Sawada et al 1994 Dudley andSchmitt 1995 Dudley 1996b)

Reciprocal transplant experiments test the adaptivenature of population divergence In these experimentsgenotypes from two contrasting environments are grownin their native site and in the site of the other population tocompare their function and performance Higher fitness ofgenotypes in their native environment when compared togenotypes transplanted from contrasting environmentsprovides evidence of local adaptation of populations totheir respective environments It is important to recognizethat reciprocal transplants do not address the adaptive sig-nificance of individual ecophysiological traits Instead theytest whether the entire suite of traits that has differentiatedbetween populations confers a performance advantage inthe local environment Combining reciprocal transplantswith selection analyses makes it possible to examine theunderlying causes of fitness variation across contrastingenvironments (eg Jordan 1992 Bennington and McGraw1995)

Divergence among populations provides the materialfor speciation and differentiation among closely relatedspecies For example over a broad rainfall gradient in thedeserts of southwestern North America leaf pubescence inpopulations of Encelia farinosa declined as mean wateravailability increased (Sandquist and Ehleringer 1997)Variation among populations in both pubescence and car-bon isotope discrimination persisted when they weregrown in common environments differing in water avail-ability indicating a genetic basis for variation in thesetraits (Figure 5 Sandquist and Ehleringer 1998) Thisgenetic variation in ecophysiological function within Efarinosa parallels similar variation found among specieswithin the genus Encelia (Ehleringer et al 1981) demon-strating both intraspecific and interspecific adaptation

The interface between intraspecific and interspecificvariation provides a critical link between microevolution-ary processes within populations and larger patterns ofdiversification and adaptation For example both withinand among species of Helianthus variation in the relativeproportions of saturated and unsaturated fatty acids inseed oils parallels climatic differences in the speciesrsquoranges Experimental studies suggest that within H annu-us seed oil composition influences germination perfor-mance producing a tradeoff between the timing of ger-mination at low temperature and the rate of growth athigh temperature (Linder 2000) This tradeoff suggeststhat the interspecific biogeographical patterns in seed oilcomposition of Helianthus species result from selectionwithin each lineage for optimal seed oil composition in itslocal environment By studying local adaptation ofspecies with broad geographic distributions it may bepossible to explain how selection has influenced themicroevolution of seed oil composition at the level ofindividual genes Patterns revealed at this level can thenbe compared with genetic differences among species with


Articles

different geographic ranges Such an approach wouldindicate whether physiologically important traits becomeadapted to different environments by the same geneticmechanisms in independently evolving lineages

Patterns of phenotypic plasticity also evolve in the courseof species differentiation and related species may expressdifferent characteristic plastic responses to environmentalconditions (Sultan 1995) Because adaptive plasticity influ-ences environmental tolerance such differences may con-tribute to differences in the range of environments that

species inhabit in the field For example four closely relat-ed species of Polygonum differed in the magnitude direc-tion and timing of plasticity for such key functional traitsas photosynthetic rate proportional biomass allocationleaf size and total area and root morphology and totallength (Sultan et al 1998 Bell and Sultan 1999) These dif-ferences in patterns of plasticity were broadly consistentwith differences in the speciesrsquo ecological distributions forexample genotypes of P persicaria (a species found inboth open and moderately shaded sites) more than dou-bled allocation to leaf tissue in low versus high light whileshaded individuals of P hydropiper (a species excludedfrom shaded sites) increased leaf biomass allocation lessthan half as much (Figure 6) In some cases the Polygon-um species differed in the timing as well as the amount ofplasticity For instance P persicaria plants showed signifi-cantly faster as well as more pronounced spatial redeploy-ment of roots to the soilair interface in response to soilflooding than plants of P cespitosum which does notoccupy flood-prone sites (Bell and Sultan 1999) Theseresults suggest that differences in plasticity may be animportant aspect of adaptive diversity among plantspeciesEcophysiological variation in a phylogeneticcontext Comparative studies of interspecific variationsuch as the examples just mentioned have long played acentral role in the study of adaptation and the examina-tion of evolutionary constraints and patterns ofmacroevolution In the context of speciesrsquo evolutionaryrelationships adaptive variation has been addressed fromtwo complementary perspectives focusing on divergentversus convergent evolution Differentiation among close-ly related species occupying different habitats providesevidence of adaptive divergence (eg Robichaux et al1990) The study of closely related species pairs (eg con-generic annual versus perennial grasses Garnier and Lau-rent 1994) is a particularly powerful method to detect sta-tistically repeatable patterns of divergence Alternativelyphenotypic similarity among distantly related speciesoccupying similar habitats demonstrates evolutionaryconvergence and provides powerful evidence for theaction of selection (as in the repeated independent evolu-tion of C4 and CAM photosynthesis)

Explicit consideration of phylogenetic relationshipsallows us to address a variety of novel questions regardingpatterns of ecophysiological evolution such as reconstruc-tion of the sequence and direction of evolutionary changesin complex traits For example the genus Flaveria hasthree species with the C

3 photosynthetic pathway eightspecies with C

4 photosynthesis and nine species withvarying levels of intermediate characteristics (C

3ndashC4)Based on the phylogenetic distribution of these traits itappears that even within this one genus the eight C

4

species result from three independent origins of this path-way at least two of which started with the transition toC

3ndashC4 intermediates (Figure 7 Monson 1996)


Articles

Figure 5 Variation in leaf traits among populations Leafabsorptance values (mean + 1 sd) from (a) field and (b)common garden studies of three Encelia farinosapopulations The variation in absorptance values amongpopulations is correlated with mean annualprecipitation in the region from which they werecollected Superior AZ 453 mm yndash1 Oatman AZ 111mm yndash1 Death Valley CA 52 mm yndash1 The consistency ofthe differences observed in the field and in a commongarden in which plants were grown under the sameconditions and then measured at periodic intervalsduring a drought treatment demonstrates that there is agenetic basis for the variation among populations(Different letters within each comparison indicatesignificantly different means based on arcsintransformations P lt 05) See Sandquist and Ehleringer(1997 1998) for experimental details

Dawson et al (1998) are studying ecophysiological vari-ation in Schiedea an endemic plant genus that has radiat-ed throughout the Hawaiian islands from a single colo-nization event As plants of this group colonized drierhabitats evolutionary transitions to the woody habit andsmaller leaves have been accompanied by changes in planthydraulic architecture (eg vulnerability to cavitationunder drought) water-use efficiency and plant carbonbalance As in studies of the Hawaiian silversword alliance(Robichaux et al 1990) the evolution of ecophysiologicaltraits in Schiedea illustrates that functional diversificationand adaptive radiation appear to have occurred simulta-neously Mapping of functional traits onto the phyloge-netic tree for these species will provide rigorous tests ofevolutionary hypotheses about the origin of the possibleecophysiological adaptations the underlying cause-and-effect relationships the trajectories modes and tempos offunctional changes and alternative explanations to ldquoadap-tive evolutionrdquo for the origin of ecophysiological charac-teristics and their maintenance in nature

Lechowicz and Wang (1998) have conducted one of thefew studies evaluating phenotypic plasticity in a phyloge-netic context (see also Pigliucci et al 1999) In a study of16 species of North American spruce growing in ambientand elevated CO2 and low and high water availability theyfound that interspecific variation in many morphologicaland ecophysiological traits was not associated with thespeciesrsquo phylogenetic relationships However relativegrowth rate which is the outcome of interactions amongmany ecophysiological traits showed consistent evolu-tionary trends across species Perhaps most interestingrelative growth rate was also less plastic across environ-ments than the many ecophysiological traits underlyingvariation in growth but the levels of plasticity in growthrate did not themselves show any pattern of phylogeneticconstraint Evolution of function in extant spruces hasapparently involved different patterns of diversification inthe mean value of traits affecting growth and in the plas-tic expression of these traits in differing environmentalregimes

Comparative analyses of interspecific variation are fre-quently invoked to test adaptive hypotheses based on cor-relations among different traits or between traits and envi-ronments (eg variation in pubescence in Encelia speciesalong rainfall gradients Ehleringer et al 1981) In the past15 years there has been widespread attention to the con-ceptual and statistical issues associated with comparativeanalyses in the context of phylogenetic relationships Inparticular it has been argued that phylogenetic relation-ships among species create a hierarchical data structurethat prevents closely related species from being used asindependent data points Nonindependence leads to inflat-ed type I errors in testing the significance of correlations(ie increased chance of rejecting a null hypothesis whenin fact there is no relationship) suggesting that claimsabout adaptation derived from interspecific correlations

between traits and environment should be reexamined(Kelly and Beerling 1995)

The method of phylogenetic independent contrastsbased on differences in traits between related taxa ratherthan on the trait values themselves provides a powerfultool for evaluating trait correlations in a phylogenetic con-text (Figure 8 see Ackerly 1999) For example Reich et al(1999) have demonstrated significant and consistent cor-relations among leaf functional traits including life spanspecific leaf area assimilation rates and nitrogen concen-tration across a broad range of conifers and angiospermsinhabiting diverse environments Ackerly and Reich(1999) found that these relationships were robust from aphylogenetic perspective that is correlations observed inthe species data were also significant using independentcontrasts (Figure 8) However they also examined the cor-relation between leaf size and leaf life span and found astrong negative correlation in the interspecific data Thissuggests the rather unexpected result that smaller-leavedspecies also have longer-lived leaves The use of indepen-dent contrasts showed that this correlation was due entire-ly to the single divergence in each of these traits betweenconifers and angiosperms with no correlation betweenthese traits within each group and no significant associa-tion overall

In a review of comparative studies of plant functionaltraits Ackerly (1999) found that there does not appear tobe any consistent bias in trait correlations or regressionslopes when comparing phylogenetic and nonphylogenetic


Articles

Figure 6 Plasticity in closely related species Plasticity forproportional biomass allocation to leaves in low (20)light versus high (100) light in Polygonum persicariaand P hydropiper P persicaria occurs in bothmoderately shaded and open habitats while Phydropiper is excluded from shaded sites Means plusmn 2standard errors are shown based on 48 plants per speciesper light treatment over a range of three moisture twonutrient levels within each light treatment (Sonia ESultan and Amity M Wilczek unpublished data)

analyses although the two methods at times lead to quitedivergent results (as in the above example of leaf size andlife span) Quantitative analysis of levels of convergent evo-lution (a high level of convergence for a trait indicates thatclosely related species are highly divergent while distantlyrelated species are most phenotypically similar) demon-strated that when traits are highly convergent analyses oftrait correlations based on interspecific patterns are mostsimilar to results based on phylogenetically structuredindependent contrasts However these methods producedifferent results when convergence is low as in the leaf sizeexample in which two major phylogenetic groups ofspecies (conifers versus angiosperms) were markedly diver-gent This result accords with the long-held view that stud-ies of convergent traits provide particularly valuable androbust insights in the study of adaptive evolution

Future directionsIn this article we have reviewed recent advances in evolu-tionary ecophysiology of plants focusing on the analysisof adaptive value of ecophysiological traits the geneticbasis of ecophysiological variation and the evidence foradaptive differentiation among populations and species

Enormous progress has been made in recent decadesreflecting advances in molecular biology phylogeneticsand related fields and the application of refined experi-mental and quantitative evolutionary approaches We havealso sought to highlight the technical and conceptual dif-ficulties that accompany these approaches No singleapproach will succeed in unraveling the complexities ofthe evolutionary process and a recognition of theirrespective limitations is critical to gaining complementaryperspectives on these problems A comprehensive andintegrated approach to ecophysiological evolution is par-ticularly critical as global environmental changes imposenew selective pressures on plant populations and species(eg Bazzaz et al 1995 Potvin and Tousignant 1996) Sev-eral areas are particularly promising for research in thenext decade

Analysis of the genetic architecture of ecophysiologicaltraits including identification of molecular markers fortrait variation and the study of candidate genes respon-sible for naturally occurring variation in ecophysiologi-cal traits and plasticity in these traits once identifiedthese genes would open up further opportunities for


Articles

Figure 7 Evolution of C4

photosynthesis in Flaveria (Asteraceae Monson 1996) The colors along each branch of the phylogenyrepresent a hypothesized reconstruction of the evolution of photosynthetic pathways based on phylogenetic parsimonymethods (ie the reconstruction that requires the fewest evolutionary transitions leading to the observed present-daydistribution of photosynthetic types) The hatched bar indicates an uncertain reconstruction If this branch is inferred to beC

3 there are three independent origins of C

3ndashC

4intermediate pathways (including F angustifolia) alternatively if this

branch is reconstructed as C3ndashC

4intermediate there is one origin of C

3ndashC

4and one subsequent reversal to C

3 The C

4 pathway

is inferred to have evolved independently three times and in at least two of these cases the C3ndashC

4type represented an

intermediate evolutionary stage A more recent molecular phylogeny of Flaveria for 12 of the 20 species shown here suggestsat least two independent origins of C

4photosynthesis (Kopriva et al 1996) Redrawn with permission from Monson (1996)

mutant and transgenic studies of ecophysiological varia-tion and covariation among multiple genes and traits

Elucidation of the developmental cascades that linkchanges at the molecular and biochemical level tochanges in physiology resource allocation and perfor-mance

Development of general models of plant functionaddressing the interactions among traits that influencesurvival growth and reproduction and how these leadto diversification of suites of traits in contrasting envi-ronments

Integrated studies of model systems for ecophysiologicalevolution across multiple levels of organization (egphytochrome-mediated responses heat shock proteinsor seed oil composition) linking genetic variationintrapopulation selection and intraspecific and inter-specific differentiation across contrasting environments

Long-term studies of selection in natural populationsin order to understand the role of temporal fluctuationsand rare events on long-term evolutionary dynamics

Comparative study of ecophysiological evolution inclosely related species examining variation at genetic


Articles

Figure 8 Analysis of correlations among leaf traits using independent contrasts Scatterplots of leaf life span versus specificleaf area (a c) and leaf life span versus leaf size (b d) Left-hand panels show correlations for 90 species of angiosperms(blue circles) and 22 conifers (red squares) based on the species values themselves (from Reich et al 1999 leaf size dataPeter B Reich University of Minnesota unpublished data) The right-hand panels show the corresponding patterns forphylogenetic independent contrasts (Ackerly and Reich 1999) Each contrast represents the difference in the trait valuesbetween two related taxa and reflects an independent evolutionary divergence The X indicates the contrast betweenangiosperms and conifers reflecting the basal divergence in each of the three traits between these two major lineagesCorrelations of independent contrasts provide an estimate of correlated changes in the evolution of the two traits For leaflife span versus specific leaf area the strong correlations among extant taxa (a) reflect a significant pattern of correlatedevolutionary change (c) In contrast for leaf life span versus leaf size the correlation among species (b) is due to thedivergence between angiosperms and conifers (d) whereas there is no correlation between these traits within each groupand no evidence of correlated evolutionary changes overall Modified from Ackerly and Reich (1999) with permission

and phenotypic levels and mapping evolutionarychanges onto phylogenies to understand the sequenceand direction of evolutionary transitions

Evolutionary responses to changing environmentsespecially the effects of global change on plant popula-tions and the influence of evolutionary changes oncommunity and ecosystem responses to climate change

AcknowledgmentsThis paper resulted from a symposium organized by D DA and J S C at the Ecological Society of America (ESA)1998 annual meeting in Baltimore MD We wish to thankthe ESA Physiological Ecology section for sponsorship ofthe symposium Research by the authors presented in thispaper was supported by the US National Science Founda-tion (D D A S A D S E S J S J S C C R L D R Sand T E D) the US Department of Agriculture (M A G)the Andrew W Mellon Foundation (J S C) and theNational Science and Engineering Research Council ofCanada (M L) Preparation of the manuscript was sup-ported by a Terman Fellowship from Stanford University toD D A

References citedAckerly DD 1999 Phylogeny and the comparative method in plant func-

tional ecology Pages 391ndash413 in Press M Scholes JD Barker MG edsPhysiological Plant Ecology Oxford Blackwell Scientific

Ackerly DD Bazzaz FA 1995 Seedling crown orientation and interceptionof diffuse radiation in tropical forest gaps Ecology 76 1134ndash1146

Ackerly DD Reich PB 1999 Convergence and correlations among leaf sizeand function in seed plants A comparative test using independent con-trasts American Journal of Botany 86 1272ndash1281

Alhiyaly S McNeilly T Bradshaw AD Mortimer AM 1993 The effect ofzinc contamination from electricity pylons Genetic constraints onselection for zinc tolerance Heredity 70 22ndash32

Antonovics J Bradshaw AD Turner RG 1971 Heavy metal tolerance inplants Advances in Ecological Research 71 1ndash85

Arntz AM DeLucia EH Jordan N 1998 Contribution of photosyntheticrate to growth and reproduction in Amaranthus hybridus Oecologia117 323ndash330

Bazzaz FA Jasienski M Thomas SC Wayne P 1995 Microevolutionaryresponses in experimental populations of plants to CO2-enriched envi-ronments Parallel results from two model systems Proceedings of theNational Academy of Sciences of the United States of America 928161ndash8165

Bell DL Sultan SE 1999 Dynamic phenotypic plasticity for root growth inPolygonum A comparative study American Journal of Botany 86807ndash819

Bennington CC McGraw JB 1995 Natural selection and ecotypic differ-entiation in Impatiens pallida Ecological Monographs 65 303ndash323

Bergelson J Purrington CB 1996 Surveying patterns in the cost of resis-tance in plants American Naturalist 148 536ndash558

Bradshaw AD 1991 Genostasis and the limits to evolution PhilosophicalTransactions of the Royal Society of London B Biological Sciences 333289ndash305

Briggs D Walters SM 1997 Plant Variation and Evolution 3rd ed Cam-bridge (UK) Cambridge University Press

Chory J Peto C Feinbaum R Pratt L Ausubel F 1989 Arabidopsis thalianamutant that develops as a light-grown plant in the absence of light Cell58 991ndash999

Clausen J Keck DD Hiesey WM 1940 Experimental Studies on theNature of Species I Effect of Varied Environments on Western North

American Plants Washington (DC) Carnegie Institute of WashingtonPublication No 520

Coleman JS Heckathorn SA Hallberg RL 1995 Heat-shock proteins andthermotolerance Linking molecular and ecological perspectivesTrends in Ecology and Evolution 10 305ndash306

Dawson T Sakai A Weller S 1998 Physiology phylogeny and adaptationin Hawaiian plants Page 8 Ecological Society of America 83rd Annu-al Meeting Abstracts

De Block M 1993 The cell biology of plant transformation Current stateproblems prospects and the implications for plant breeding Euphyti-ca 71 1ndash14

Donovan LA Ehleringer JR 1994 Potential for selection on plants forwater-use efficiency as estimated by carbon isotope discriminationAmerican Journal of Botany 81 927ndash935

Downs CA Heckathorn SA Bryan JK Coleman JS 1998 The methionine-rich low molecular weight chloroplast heat shock protein Evolutionaryconservation and accumulation in relation to thermotolerance Amer-ican Journal of Botany 85 175ndash183

Dudley SA 1996a Differing selection on plant physiological traits inresponse to environmental water availability A test of adaptivehypotheses Evolution 50 92ndash102

______ 1996b The response to differing selection on plant physiologicaltraits Evidence for local adaptation Evolution 50 103ndash110

Dudley SA Schmitt J 1995 Genetic differentiation in morphologicalresponses to simulated foliage shade between populations of Impatienscapensis from open and woodland sites Functional Ecology 9655ndash666

______ 1996 Testing the adaptive plasticity hypothesis Density-depen-dent selection on manipulated stem length in Impatiens capensisAmerican Naturalist 147 445ndash465

Ehleringer J 1991 13C12C fractionation and its utility in terrestrial plantstudies Pages 187ndash200 in Coleman DC Fry B eds Carbon IsotopeTechniques San Diego Academic Press

Ehleringer J Bjoumlrkman O Mooney HA 1976 Leaf pubescence Effects onabsorptance and photosynthesis in a desert shrub Science 192376ndash377

Ehleringer J Mooney HA Gulmon SL Rundel PW 1981 Parallel evolutionof leaf pubescence in Encelia in coastal deserts of North and SouthAmerica Oecologia 49 38ndash41

Ehleringer JR Monson RK 1993 Evolutionary and ecological aspects ofphotosynthetic pathway variation Annual Review of Ecology and Sys-tematics 24 411ndash439

Emery RJN Reid DM Chinnappa CC 1994 Phenotypic plasticity of stemelongation in 2 ecotypes of Stellaria longipes The role of ethylene andresponse to wind Plant Cell and Environment 17 691ndash700

Evans A 1998 How do tradeoffs in resource use efficiencies arise Page 9Ecological Society of America 83rd Annual Meeting Abstracts

Farris MA Lechowicz MJ 1990 Functional interactions among traits thatdetermine reproductive success in a native annual plant Ecology 71548ndash557

Fishbein MVenable L 1996 Evolution of inflorescence design Theory anddata Evolution 50 2165ndash2177

Garnier E Laurent G 1994 Leaf anatomy specific mass and water contentin congeneric annual and perennial grass species New Phytologist 128725ndash736

Geber MA Dawson TE 1990 Genetic variation in and covariationbetween leaf gas exchange morphology and development in Polygon-um arenastrum an annual plant Oecologia 85 153ndash158

______1997 Genetic variation in stomatal and biochemical limitations tophotosynthesis in the annual plant Polygonum arenastrum Oecologia109 535ndash546

Givnish TJ 1988 Adaptation to sun and shade A whole plant perspectiveAustralian Journal of Plant Physiology 15 63ndash92

Gurevitch J 1992 Differences in photosynthetic rate in populations ofAchillea lanulosa from two altitudes Functional Ecology 6 568ndash574

Heckathorn SA Poeller GJ Coleman JS Hallberg RL 1996 Nitrogen avail-ability alters patterns of accumulation of heat stress-induced proteinsin plants Oecologia 105 413ndash418


Articles

Heckathorn SA Downs CA Sharkey TD Coleman JS 1998 The smallmethionine-rich chloroplast heat shock protein protects photosystemII electron transport during heat stress Plant Physiology 116 439ndash444

Heckathorn SA Downs CA Coleman JS 1999 Small heat-shock proteinsprotect electron transport in chloroplasts and mitochondria duringstress American Zoologist 39 865ndash876

Hiesey WM Milner HW 1965 Physiology of ecological races and speciesAnnual Review of Plant Physiology 16 203ndash213

Jordan N 1992 Path analysis of local adaptation in two ecotypes of theannual plant Diodia teres Walt (Rubiaceae) American Naturalist 140149ndash165

Kelly CK Beerling DJ 1995 Plant life form stomatal density and taxo-nomic relatedness A reanalysis of Salisbury (1927) Functional Ecolo-gy 9 422ndash431

Kingsolver JG Schemske DW 1991 Path analyses of selection Trends inEcology and Evolution 6 276ndash280

Knight C Ackerly DD In press Correlated evolution of chloroplast heatshock protein expression in closely related plant species AmericanJournal of Botany

Kopriva S Chu C-C Bauwe H 1996 Molecular phylogeny of Flaveria asdeduced from the analysis of nucleotide sequences encoding the H-protein of the glycine cleavage system PlantCell and Environment 191028ndash1036

Lande R Arnold SJ 1983 The measurement of selection on correlatedcharacters Evolution 36 1210ndash1226

Lechowicz M Wang ZM 1998 Comparative ecology of seedling sprucesA phylogenetic perspective on adaptation Page 16 Ecological Societyof America 83rd Annual Meeting Abstracts

Linder CR 2000 Adaptive evolution of seed oil composition AmericanNaturalist 156 442ndash458

Lloyd DG 1984 Variation strategies of plants in heterogeneous environ-ments Biological Journal of the Linnean Society 21 357ndash385

MacDonald SE Lieffers VJ 1993 Rhizome plasticity and clonal foraging ofCalamagrostis canadensis in response to habitat heterogeneity Journalof Ecology 81 769ndash776

McGraw JB Antonovics J 1983 Experimental ecology of Dryas octopetalaecotypes I Ecotypic differentiation and life-cycle stages of selectionJournal of Ecology 71 879ndash897

Monson RK 1996 The use of phylogenetic perspective in comparativeplant physiology and developmental biology Annals of the MissouriBotanical Garden 83 3ndash16

Monson RK Smith SD Gehring JL Bowman WD Szarek SR 1992 Physi-ological differentiation within an Encelia farinosa population along ashort topographic gradient in the Sonoran Desert Functional Ecology6 751ndash759

Niklas KJ 1988 The role of phyllotactic pattern as a ldquodevelopmental con-straintrdquo on the interception of light by leaf surfaces Evolution 42 1ndash16

______1997 Adaptive walks through fitness landscapes for early vascularland plants American Journal of Botany 84 16ndash25

Pearcy RW Valladares F 1999 Resource acquisition by plants The role ofcrown architecture Pages 45ndash66 in Press M Scholes JD Barker MGeds Advances in Physiological Plant Ecology Oxford Blackwell Scien-tific

Pearcy RW Yang W 1996 A three-dimensional crown architecture modelfor assessment of light capture and carbon gain by understory plantsOecologia 108 1ndash12

Pigliucci M Cammell K Schmitt J 1999 Evolution of phenotypic plastic-ity A comparative approach in the phylogenetic neighborhood of Ara-bidopsis thaliana Journal of Evolutionary Biology 12 779ndash791

Potvin C Tousignant D 1996 Evolutionary consequences of simulatedglobal change Genetic adaptation or adaptive phenotypic plasticityOecologia 108 683ndash693

Purrington C Bergelson J 1997 Fitness consequences of genetically engi-neered herbicide and antibiotic-resistance in Arabidopsis thalianaGenetics 145 807ndash814

______ 1999 Exploring the physiological basis of costs of herbicide resis-tance in Arabidopsis thaliana American Naturalist 154 S82ndashS91

Reich PB Ellsworth DS Walters MB Vose JM Gresham C Volin JC Bow-man WD 1999 Generality of leaf trait relationships A test across sixbiomes Ecology 80 1955ndash1969

Robichaux RH Carr GD Liebman M Pearcy RW 1990 Adaptive radiationof the Hawaiian silversword alliance (Compositae-Madiinae) Ecologi-cal morphological and physiological diversity Annals of the MissouriBotanical Garden 77 64ndash72

Sandquist DR Ehleringer JR 1997 Intraspecific variation of leaf pubes-cence and drought adaptation in Encelia farinosa associated with con-trasting desert environments New Phytologist 135 635ndash644

______1998 Intraspecific variation of drought adaptation in brittlebushLeaf pubescence and timing of leaf loss vary with rainfall Oecologia113 162ndash169

Sawada S Nakajima Y Tsukuba M Sasaki K Hazama Y Futatsuya MWatanabe A 1994 Ecotypic differentiation of dry matter productionprocesses in relation to survivorship and reproductive potential inPlantago asiatica populations along climatic gradients FunctionalEcology 8 400ndash409

Schmitt J 1993 Reaction norms of morphological and life-history traits tolight availability in Impatiens capensis Evolution 47 1654ndash1668

Schmitt J Wulff R 1993 Light spectral quality phytochrome and plantcompetition Trends in Ecology and Evolution 8 47ndash51

Schmitt J McCormac AC Smith H 1995 A test of the adaptive plasticityhypothesis using transgenic and mutant plants disabled in phy-tochrome-mediated elongation responses to neighbors American Nat-uralist 146 937ndash953

Schmitt J Dudley SA Pigliucci M 1999 Manipulative approaches to test-ing adaptive plasticity Phytochrome-mediated shade avoidanceresponses in plants American Naturalist 154 S43ndashS54

Schuster WS Phillips SL Sandquist DR Ehleringer JR 1992 Heritability ofcarbon isotope discrimination in Gutierrezia microcephala (Aster-aceae) American Journal of Botany 79 216ndash221

Stitt M Schulze ED 1994 Does Rubisco control the rate of photosynthesisand plant growth An exercise in molecular ecophysiology Plant Celland Environment 17 465ndash487

Sultan SE 1995 Phenotypic plasticity and plant adaptation Acta BotanicaNeerlandica 44 363ndash383

Sultan SE Bazzaz FA 1993a Phenotypic plasticity in Polygonum persicariaI Diversity and uniformity in genotypic norms of reaction to lightEvolution 47 1009ndash1031

______1993b Phenotypic plasticity in Polygonum persicaria II Norms ofreaction to soil moisture and the maintenance of genetic diversity Evo-lution 47 1032ndash1049

______1993c Phenotypic plasticity in Polygonum persicaria III The evo-lution of ecological breadth for nutrient environment Evolution 471050ndash1071

Sultan SE Wilczek AM Bell DL Hand G 1998 Physiological response tocomplex environments in annual Polygonum species of contrastingecological breadth Oecologia 115 564ndash578

Tonsor SJ Goodnight CJ 1997 Evolutionary predictability in natural pop-ulation Do mating system and nonadditive genetic variance interact toaffect heritabilities in Plantago lanceolata Evolution 51 1773ndash1784

Turesson G 1922 The genotypical response of the plant species to thehabitat Hereditas 3 211ndash350

Winn AA 1996 Adaptation to fine-grained environmental variation Ananalysis of within-individual leaf variation in an annual plant Evolu-tion 50 1111ndash1118

______ 1999a The functional significance and fitness consequences ofheterophylly International Journal of Plant Sciences 160 S113ndashS121

______ 1999b Is seasonal variation in leaf traits adaptive for the annualplant Dicerandra linearifolia Journal of Evolutionary Biology 12306ndash313

Yanovsky MJ Casal JJ Luppi JP 1997 The VLF loci polymorphic betweenecotypes Landsberg erecta and Columbia dissect two branches of phy-tochrome A signal transduction that correspond to very-low-fluenceand high-irradiance responses Plant Journal 12 659ndash667


Articles

the same genotype in response to environmental varia-tion also affects plant function and hence fitness indiverse environments However only recently have plantbiologists directly studied selection in natural popula-tions Moreover until the last few years little was knownabout the genetic basis for evolutionary change in eco-physiological traits or about genetic developmental orphylogenetic constraints on the evolutionary response tonatural selection

The authors of this article presented research findingsand discussed future directions in evolutionary plant eco-physiology in a symposium at the 1998 annual meeting ofthe Ecological Society of America This article presents anoverview of advances in the field and addresses a numberof questions regarding the evolution of ecophysiologicaltraits

How are ecophysiological traits related to fitness Howdo patterns of natural selection for these traits vary withthe environment

What is the genetic basis for ecophysiological traitsHow much genetic variation for these traits exists innatural populations How do genetic and developmen-tal constraints influence the adaptive evolution of eco-physiological traits

What is the evidence for adaptive evolution of ecophysi-ological traits in natural populations Do closely relatedpopulations or species demonstrate ecophysiologicaldivergence and do their differences reflect patterns ofenvironmental variation as expected on functionalgrounds

What are the macroevolutionary patterns in ecophysio-logical traits viewed from a phylogenetic perspectiveDo these patterns suggest constraints on ecophysiologi-cal evolution or do they suggest high evolutionary labil-ity and frequent convergence

The adaptive value of ecophysiologicaltraitsAn ecophysiological trait can be considered adaptive if ithas a direct impact on fitness in natural environmentsHere we examine the successes and limitations of variedperspectives from which this problem has been addressedPhenotypic selection analysis The most straightfor-ward evidence for the adaptiveness of ecophysiologicaltraits is the observation of correlations between traits andfitness in natural populations but this approach hasproven problematic For example direct correlationsbetween photosynthetic rates and fitness are rarelyobserved in natural populations (eg Farris and Lechow-icz 1990 and references) Moreover even when correla-tions are observed it is difficult to determine whetherindividual traits contribute directly to variation in fitnessor whether these relationships reflect indirect selection via

correlated traits For example a study of Plantago lanceo-lata found a significant positive correlation between pho-tosynthetic capacity and reproductive dry weight but cor-relations were also observed for corm diameter number ofleaves leaf size specific leaf weight and transpiration rate(Tonsor and Goodnight 1997) Thus it is impossible to saythat variation in photosynthetic rate per se contributeddirectly to individual fitness

In these situations when many interacting traits mayinfluence fitness multivariate selection analysis (Landeand Arnold 1983) provides a powerful tool for detectingselection on ecophysiological traits This method involvesmeasuring a suite of traits as well as a measure of fitness(eg lifetime seed production) on individual plantsDirect selection on each trait is measured as a partialregression coefficient in a multiple regression of relativefitness on all measured traits If causal relationshipsamong traits are known path analysis can also be used todetermine the direct and indirect effects of each trait onfitness (Kingsolver and Schemske 1991)

The first selection study on plant ecophysiological traitswas conducted by Farris and Lechowicz (1990) who mea-sured selection on ecophysiological phenological mor-phological and growth traits in a Xanthium populationgrown under uniform experimental conditions They cre-ated an experimental population in the greenhouse thatmaximized genetic variation and then planted seedlingsinto a garden preventing any correlation between geno-type and microenvironment that could confound theresults Twelve traits on each plant were measured at dif-ferent times during the season and path analysis was usedto integrate the functional relationships among traits intothe selection analysis The researchers found that seedlingemergence time height and branch growth rates andwater-use efficiency all influenced total vegetative bio-mass which in turn was strongly correlated with fitnessillustrating that these traits were under selection

Similar methods have been applied successfully in sev-eral recent studies One of the most important results ofsuch research is the demonstration that patterns of phe-notypic variation its genetic basis and natural selectionvary in different environmental conditions For examplein the sand-dune annual Cakile edentula high water-useefficiency and intermediate leaf sizes were correlated withfitness in an environment with low water availabilitywhile larger leaf sizes were favored in an environment withhigh water availability (Figure 1 Dudley 1996a) In thelow-water environment the finding that genotypes withintermediate leaf size had the highest fitness illustrates theimportance of nonlinear approaches in selection studiesThis pattern indicates the occurrence of stabilizing selec-tion in which intermediate trait values are favored in con-trast with directional selection which favors lower orhigher values of a trait In addition genetic differentiationbetween populations derived from low- and high-water


Articles






Articles


a

b




Articles









Articles


Articles









Articles


a

b










Articles








Articles







Articles


100









Articles








4




Articles









Articles







Articles




3ndashC




3ndashC


3 The C

4 pathway












Articles













































Articles
















































Articles






Articles


a

b




Articles









Articles


Articles









Articles


a

b










Articles








Articles







Articles


100









Articles








4




Articles









Articles







Articles




3ndashC




3ndashC


3 The C

4 pathway












Articles













































Articles
















































Articles




Articles









Articles


Articles









Articles


a

b










Articles








Articles







Articles


100









Articles








4




Articles









Articles







Articles




3ndashC




3ndashC


3 The C

4 pathway












Articles













































Articles
















































Articles







Articles


Articles









Articles


a

b










Articles








Articles







Articles


100









Articles








4




Articles









Articles







Articles




3ndashC




3ndashC


3 The C

4 pathway












Articles













































Articles
















































Articles


Articles









Articles


a

b










Articles








Articles







Articles


100









Articles








4




Articles









Articles







Articles




3ndashC




3ndashC


3 The C

4 pathway












Articles













































Articles
















































Articles





Articles


a

b










Articles








Articles







Articles


100









Articles








4




Articles









Articles







Articles




3ndashC




3ndashC


3 The C

4 pathway












Articles













































Articles
















































Articles










Articles








Articles







Articles


100









Articles








4




Articles









Articles







Articles




3ndashC




3ndashC


3 The C

4 pathway












Articles













































Articles
















































Articles








Articles







Articles


100









Articles








4




Articles









Articles







Articles




3ndashC




3ndashC


3 The C

4 pathway












Articles













































Articles
















































Articles







Articles


100









Articles








4




Articles









Articles







Articles




3ndashC




3ndashC


3 The C

4 pathway












Articles













































Articles
















































Articles









Articles








4




Articles









Articles







Articles




3ndashC




3ndashC


3 The C

4 pathway












Articles













































Articles
















































Articles








4




Articles









Articles







Articles




3ndashC




3ndashC


3 The C

4 pathway












Articles













































Articles
















































Articles








Articles







Articles




3ndashC




3ndashC


3 The C

4 pathway












Articles













































Articles
















































Articles






Articles




3ndashC




3ndashC


3 The C

4 pathway












Articles













































Articles
















































Articles








Articles













































Articles
















































Articles












































Articles
















































Articles
















































Articles

Documents

ARTICLES: The Evolution of Plant Ecophysiological Traits ... · repeated occurrence of plants with CAM ... fessor in the Department of Ecology and Evolutionary Biology, ... The Evolution