a trait-based approach to comparative functional plant ecology ... › uploads › membres ›...

REVIEW ARTICLE

A trait-based approach to comparative functionalplant ecology concepts methods and applicationsfor agroecology A review

Eric Garnier amp Marie-Laure Navas

Accepted 25 November 2010 Published online 11 August 2011 INRA and Springer Science+Business Media BV 2011

Abstract Comparative functional ecology seeks to under-stand why and how ecological systems and their compo-nents operate differently across environments Althoughtraditionally used in (semi)-natural situations its conceptsand methods could certainly apply to address key issues inthe large variety of agricultural systems encountered acrossthe world In this review we present major advances incomparative plant functional ecology that were madepossible over the last two decades by the rapid developmentof a trait-based approach to plant functioning and prospectsto apply it in agricultural situations The strength of thisapproach is that it enables us to assess the interactionsbetween organisms and their environment simultaneouslyon a large number of species a prerequisite to addressquestions relative to species distribution communityassembly and ecosystem functioning The trait concept willbe first defined before presenting a conceptual frameworkto understand the effects of environmental factors on plantcommunity structure and ecosystem properties via planttraits We will then argue that leading dimensions ofvariation among species can be captured by some selectedtraits and show that a combination of three easily measuredtraitsmdashspecific leaf area (the ratio of leaf area to leaf drymass) plant height and seed massmdashenables us to assesshow different species use their resources interact with

neighbours and disperse in time and space The use of traitsto address central questions in community ecology will bereviewed next It will be shown that traits allow us to (1)understand how plant species are sorted according to thenature of environmental gradients (2) evaluate the relativeimportance of habitat filtering and limiting similarity in theprocess of community assembly and (3) quantify two maincomponents of community functional structure namelycommunity-weighted means of traits and communityfunctional divergence The relative impacts of these twocomponents on ecosystem properties will then be discussedin the case of several components of primary productivitylitter decomposition soil water content and carbon seques-tration There is strong support for the biomass ratiohypothesis which states that the extent to which the traitsof a species affect those ecosystem properties depends onthe abundance of this species in the community Assessingthe role of functional divergence among species onecosystem properties will require major methodologicalbreakthroughs both in terms of metrics and statisticalprocedures to be used In agricultural situations we showthat trait-based approaches have been successfully devel-oped to assess the impacts of management practices on (1)the agronomic value of grasslands and (2) the functionalcomposition and structure of crop weed communities andhow these could affect the functioning of the cropApplications in forestry are still poorly developed espe-cially in temperate regions where the number of species inmanaged forest remains relatively low The last decadesof research have led to the constitution of large data setsof plant traits which remain poorly compatible andaccessible Recent advances in the field of ecoinformaticssuggest that major progress could be achieved in thisarea by using improved metadata standards and advanc-ing trait domain ontologies Finally concluding remarks

E Garnier ()CNRS Centre drsquoEcologie Fonctionnelle et Evolutive (UMR 5175)1919 Route de Mende34293 Montpellier Cedex 5 Francee-mail ericgarniercefecnrsfr

M-L NavasMontpellier SupAgroCentre drsquoEcologie Fonctionnelle et Evolutive (UMR 5175)1919 Route de Mende34293 Montpellier Cedex 5 France

Agron Sustain Dev (2012) 32365ndash399DOI 101007s13593-011-0036-y

unanswered questions and directions for research usingthe functional approach to biodiversity made possible bythe use of traits will be discussed in the contexts ofecological and agronomical systems The latter indeedcover a wide range of environmental conditions andbiological diversity and the prospect for reducingenvironmental impacts in highly productive low-diversity systems will certainly imply improving ourskills for the management of more diverse systems proneto a trait-based approach as reviewed here

Keywords Agroecology Biodiversity Communitystructure Comparative ecology Ecoinformatics

Ecological strategy Ecosystem properties Environmentalconditions Functional diversity Plant functional trait

Contents

1 Introduction a trait-based approach to comparativefunctional ecology 2

2 Definitions and a framework 321 What is a trait 322 A response-and-effect framework 4

3 A functional characterization of plants 531 The resource use dimension 632 The height dimension 833 The seed size dimension 1034 Beyond a three dimensional scheme 11

4 A functional approach to plant communitystructure 11

41 Environmental gradients and response traits 1142 Assembly rules and environmental filters 1243 Assembly rules and trait distribution within

communities 1344 Phylogeny traits and community assembly 13

5 Traits and ecosystem properties 1551 The biomass ratio hypothesis community-level

weighed means of traits 15511 Components of primary productivity 16512 Decomposition and accumulation of

litter 17513 Soil water content 17514 Other components of biogeochemical

cycles 1852 Complementarity among species incorporating

functional divergence 1953 The role of plant traits on ecosystem properties in

controlled experiments 2054 A different perspective ecosystem allometry 21

6 Plant traits in agricultural and forestry contexts 2161 Grasslands 2162 Crop weed communities 22

63 Forests 237 Trait databases and ecoinformatics 24

71 Availability of trait data 2472 Towards a trait-based ecoinformatics 24

8 Conclusions and perspectives 25

1 Introduction a trait-based approach to comparativefunctional ecology

A prominent field of research in functional ecology seeks tounderstand why and how ecological systems and theircomponents operate differently across environments (egCalow 1987 Duarte et al 1995 Westoby 1999 Pugnaireand Valladares 2007) The comparative approach is thusconsubstantial to this field of ecology (eg Bradshaw 1987Keddy 1992b Westoby 1999) If we are to understandquestions such as why and how are functions coordinatedwithin organisms why and how do different species performdifferently along ecological gradients why and how dospecies interact within a community and why and how dospecies affect the functioning of ecosystems comparingspecies is an absolute necessity Keddy (1992b) and Westoby(1999) also argue that for at least two reasons suchcomparisons should be conducted on a large number ofspecies First in the struggle to find broad generalisations wehave to make sure that the patterns observed are not simplepeculiarities of the species examined Second the tests ofrelationships between species and environment andor eco-system functioning need to be statistically powerful androbust As stressed by Keddy (1992b) this approach isliterally orthogonal to that followed in population ecologywhere many aspects pertaining to population dynamics areassessed in relatively few species There is also no reasonwhy the comparative approach could not apply to the largevariety of agricultural systems encountered across the world(eg Swift and Anderson 1993 Maleacutezieux et al 2009) Thesesystems indeed cover a wide range of environmentalconditions and biological diversity and the prospect forreducing environmental impacts in highly productive low-diversity systems (Giller et al 1997 Wezel et al 2009Griffon 2010) will certainly imply improving our skills forthe management of more diverse systems

There is a growing consensus that a trait-based approachhas a strong potential to address several of the issuesintroduced above A non-exhaustive list includes (1) thefunctioning of organisms and how it relates to the environment(eg Grime 1979 Chapin et al 1993 Grime et al 1997Westoby et al 2002) (2) the understanding of unsolvedquestions in community ecology such as the identification ofrules governing the assembly of communities (McGill et al2006 Suding et al 2008 Shipley 2010) and (3) the

366 E Garnier M-L Navas

understanding of how the functioning of organisms scales upto that of ecosystems (Reich et al 1992 Chapin 1993Lavorel and Garnier 2002) and controls some of the servicesthey deliver to humans (Diacuteaz et al 2006 Diacuteaz et al 2007b)including those delivered by agriculture As measurableproperties of individuals related to their functioning andmodulating their fitness (McGill et al 2006 Violle et al2007b) (Fig 1 and see Section 2) traits enable us to capturethe interactions between organisms and their environment(both abiotic and biotic) and bring a functional perspective tothe study of controls on biodiversity and how it affectsprocesses at higher levels of organisation

The aim of this review is to present the rationale for the trait-based approach and how it can be used to address somepending questions in functional ecology It will be restricted tothe case of terrestrial plants to which a large amount ofresearch has been devoted in the last 20 years (Lavorel et al2007) We will first discuss what a trait is before showingwhich are some of the most relevant traits widely utilised tocharacterise plant functioning We will then present recentdevelopments in community and ecosystem ecology based onthe use of traits This will be largely but not exclusivelybased on the responsendasheffect framework proposed by Lavoreland Garnier (2002) and further refined by Suding et al(2008) Response traits are those which show a consistentresponse to a particular environmental factor while effecttraits are those which have a similar effect on one or severalecosystem functions (Lavorel and Garnier 2002 and refer-ences therein) We will then present examples of various trait-based approaches in the context of agronomic and forestryapplications in selected temperate systems The wealth of

research devoted to plant traits has generated large amounts ofdata leading to the imperious need of devising systemsallowing potential users to access and value these data Somekey steps that will make this possible in the near future willthen be presented before concluding remarks will beproposed to close this review

2 Definitions and a framework

21 What is a trait

A thorough analysis of the literature reveals that the termlsquotraitrsquo has been used in somewhat different meanings in thefield of ecology and population biology (Violle et al2007b) Here we will stick to the idea that it should bemeasured at the level of an individual organism (McGill etal 2006 Lavorel et al 2007) with the following definitionlsquoany morphological physiological or phenological featuremeasurable at the individual level from the cell to thewhole-organism level without reference to the environmentor any other level of organizationrsquo (Violle et al 2007b)This definition implies that no information external to theindividual (environmental factors) or at any other level oforganisation (population community or ecosystem) isrequired to define a trait Figure 1 shows examples oftraits which may be actual functions related to egreproduction (fecundity dispersal etc) or vegetativegrowth (light interception nutrient uptake etc) or traitsrelated to these functions but which are more easilymeasured (lsquofunctional markersrsquo sensu Garnier et al 2004)

Functions laquo Functional markers raquo

Seed massFecundityReproductive heightDispersalReproductive phenologyRecruitment

Vegetative heightLight interceptionCompetitive ability

Traits of living leavesResource acquisitiongrowthNIRS spectrumLitter decomposition

Root densityAbsorption (nutrients water)Root diameter lengthCarbon fluxes (exsudationhellip)Root specific areaUnderground competition

Fig 1 Examples of functions and related functional markers NIRS near-infrared spectroscopy Hand drawing of Euphorbia helioscopia byBaptiste Testi Recomposed from Lavorel et al (2007)

A trait-based approach to functional plant (agro)ecology 367

As discussed by Violle et al (2007b) the definition givenabove requires further precisions (1) the particular value ormodality taken by the trait at any place and time is called anlsquoattributersquo (Lavorel et al 2007) (2) within a species the traiteither continuous or categorical may show different attrib-utes along environmental gradients or through time (seasonaland ontogenic variation) (3) the attribute for a trait is usuallyassessed for one population (average of attributes of a set ofindividuals) in space and time The latter assertion has twoconsequences first there is not a single-trait attribute for anyparticular species and assessing intra- vs inter-specificvariability of traits is clearly of interest in this context(Garnier et al 2001a Cornelissen et al 2003a Roche et al2004 Mokany and Ash 2008 Albert et al 2010) secondinformation on the local environment where the trait hasbeen measured is essential to interpret the ecologicalevolutionary meaning of trait attributes (McGill et al 2006Bartholomeus et al 2008) even if this information is notcompulsory to define a trait

Applying Arnoldrsquos (1983) lsquomorphology performancefitnessrsquo paradigm to plants Violle et al (2007b) furtherdiscussed the value of introducing a hierarchical perspectiveon plant traits They suggested that the three components ofplant fitness (growth survival and reproduction) be charac-terised by the three lsquoperformance traitsrsquo vegetative biomassreproductive output and plant survival while the morpholog-ical phenological and physiological traits that influence thesethree performance traits be called lsquofunctional traitsrsquo Althoughinferred in many ecological studies the relationships betweenfunctional and performance traits and hence fitness is actu-ally seldom demonstrated (cf Ackerly and Monson 2003)

Although the approach presented above takes anecological perspective on traits it should be borne in mindthat since all species have a common ancestor at somepoint in their evolutionary history there will always besome degree of similarity between species traits related totheir common ancestry (cf Silvertown et al 1997 andreferences therein) Trait conservatism might arise atdifferent depths of the phylogeny but some ecologicallyrelevant traits may be labile towards the tip of thephylogeny (eg Cavender-Bares et al 2009) Severalstudies have recently addressed these issues which wasmade possible by the increasing availability of phylogeneticdata and the accumulation of broad comparative data setsAs an example Fig 2 presents the distribution of leaf massper area (LMA the ratio of leaf mass to area seeSection 31) with respect to the major clades of theTracheophytes showing that divergence between gymno-sperms and angiosperms is a major contributor to thepresent-day spread of this trait (Flores et al submittedmanuscript) Comparable studies have been conducted onseveral other traits including seed mass (Moles et al2005b) and leaf chemical elements (Watanabe et al 2007)

The future development of such analyses on a broad rangeof traits should help teasing out the phylogenetic from theecological signal on trait values of extant species

22 A response-and-effect framework

The response-and-effect framework around which most ofthis review articulates involves several components andsteps as shown in Fig 3 In this scheme a species isrepresented by two sets of connected traits (same symbolbut different colours) to acknowledge the fact that somesets of traits might vary independently from one anotheracross species (as is the case for eg vegetative andregenerative traits cf Grime 2001) The upper part ofFig 3 shows the range of values displayed for these twotypes of traits within a species pool found in a particulargeographical region and from which local communitiesassemble (cf Belyea and Lancaster 1999) From this poolthe habitat filter which involves both abiotic and bioticcomponents is assumed to sort species according to theirattributes (based on Woodward and Diament 1991 Keddy1992a Grime 2006 Suding et al 2008 Cornwell andAckerly 2009) Habitat filtering acts in reducing the rangeof attributes (double horizontal line in Fig 3) compared tothat present in the species pool thereby inducing conver-gence in trait values By contrast the process of limitingsimilarity affects the spacing of trait values (trait diver-gence simple horizontal line in Fig 3) with two possiblehypotheses Hypothesis 1 assumes that both habitat filteringdue to environmental factors and limiting similarity affectthe same set of traits this may happen when filtering is dueeg to resource limitation and when competitive interac-tions among individuals are strong Hypothesis 2 postulatesthat different sets of traits are involved in the two processesthis may happen when limiting similarity is induced bydisturbances (Grime 2006 Navas and Violle 2009 and seeSections 42 and 43 for further discussion) The relativestrengths of the habitat filter and limiting similarity isexpected to depend on the identity of the trait in combinationwith the particular abiotic conditions at a site Traitsassociated with the response of organisms to these abioticand biotic factors are called lsquoresponse traitsrsquo (cf Lavorel andGarnier 2002)

The bottom part of the scheme (from the two lateralbended arrows downwards) presents hypotheses pertain-ing to the relationships between community structure andecosystem properties (EP hereafter a list of abbrevia-tions used in the review is given in Table 1) (based onChapin et al 2000 Loreau et al 2001 Lavorel andGarnier 2002 Suding et al 2008) These EP pertain bothto compartments (represented by symbols in the box at thebottom of Fig 3) such as standing biomass or soil watercontent and fluxes within or between compartments such


Trait values

Species pool

Habitat filter

Limit to traitHypothesis 1 Hypothesis 2

similarity

Trait dominanceand divergence

Ecosystem properties

Fig 3 A response-and-effect framework for effects of environmentalfactors on plant community structure and ecosystem properties (EP)via plant traits In this scheme a species is represented by two sets ofconnected traits with the same symbol but different colours The upperpart of the scheme (down to the two lateral bended arrows) presentshypotheses for assembly effects on within-community trait distribu-tion (based on Woodward and Diament 1991 Keddy 1992a Grime2006 Suding et al 2008 Cornwell and Ackerly 2009) The bottompart of the scheme (from the two lateral bended arrows downwards)

presents hypotheses pertaining to the relationships between commu-nity structure and EP (box at the bottom of the figure) compartments(eg standing biomass or soil water content) represented by symbolsand fluxes (eg primary productivity or decomposition of litter) withinor between compartments represented by arrows (based on Chapin etal 2000 Loreau et al 2001 Lavorel and Garnier 2002 Suding et al2008) Direct effects of habitat filters on EP as well as potentialfeedbacks are symbolised by the double-dotted arrow on the left-handside of the scheme See Section 22 for further explanations

Fig 2 Phylogeny of LMA (the ratio of leaf mass to leaf area the inverse of SLA for further discussion see Section 21) for major clades ofTracheophytes (data from Flores et al submitted manuscript) Numbers between brackets show the number of species in each clade


as primary productivity or decomposition of litter (arrowson Fig 3) EP are assumed to depend on at least twocomponents of community functional diversity (FD)attribute dominance and attribute divergence which bothrelate to the shape of the trait value distribution within thecommunity (see Sections 4 and 5) Those particular traitsthat determine the effects of plants on ecosystem func-tions such as biogeochemical cycling or propensity todisturbance are called lsquoeffect traitsrsquo (cf Lavorel andGarnier 2002) The two bended arrows point to twoexamples of species abundance distributions for a singletrait likely to induce different trait distributions (notshown on the figure) The question marks next to thesearrows point to the lack of sound hypotheses andknowledge linking presenceabsence of species in acommunity and their abundance distribution Finallydirect effects of habitat filters on EP as well as potentialfeedbacks are symbolised by the double-dotted arrow onthe left-hand side of the scheme

The following sections will illustrate and discuss variousaspects of this scheme We will begin by a prominentresearch area for functional traits which can be thought ofas developing a short list of trait dimensions (Westoby et al

2002) allowing us to grasp a functional perspective on thediversity of organisms

3 A functional characterisation of plants

Among the wealth of traits measurable on an individualthose of interest to comparative functional ecology mustfill at least four conditions (Lavorel et al 2007 Fig 1)They should (1) bear some relationship to plant function(2) be relatively easy to observe and quick to quantify(lsquofunctional markersrsquo Garnier et al 2004) (3) bemeasurable using standardised protocols across a widerange of species and growing conditions (Hendry andGrime 1993 Cornelissen et al 2003b Knevel et al 2005)and (4) have a consistent ranking mdash not necessarilyconstant absolute values mdash across species when environ-mental conditions vary (Garnier et al 2001a Cornelissenet al 2003a Mokany and Ash 2008)

Identifying such measurable traits provide the means tocompare species worldwide However individual traitsshould not be considered in isolation because pairs orgroups of traits often co-vary (Chapin et al 1993 Grime et

Category Abbreviation Meaning

Trait LDMC Leaf dry matter content

LLS Leaf life span

LNC Leaf nitrogen concentration

LPC Leaf phosphorus concentration

Lroot Root length

LMA Leaf mass per area

PSm Mass-based maximum photosynthetic rate

RGR Relative growth rate

RGRmax Maximum relative growth rate

SLA Specific leaf area

SRA Specific root area

SRL Specific root length

Ecosystem properties AGBmax Maximum above-ground standing biomass

ANPP Above-ground net primary productivity

BNCS Below-ground net carbon storage

EP Ecosystem properties

NPP Net primary productivity

SNPP Specific net primary productivity

SOC Soil organic carbon content

Others BRH Biomass ratio hypothesis

CWM Community-level weighted mean of trait

CFP Community functional parameter

FD Functional diversity

FDvg Functional divergence

Glopnet Global Plant Trait Network

LES Leaf economics spectrum

Table 1 List of abbreviationsused in the text


al 1997 Reich et al 1997 see Section 31) As recognisedby early work (Theophrastus ca 300 BC Raunkiaer 1934)growth and life forms are one expression of such co-variations among traits although there is actually animportant range of variation in trait values and co-variation among traits within each growthlife form Moregenerally quantifying these co-variations and asking whysome sets of traits are sufficiently closely coordinated leadto the identification of single dimensions of strategyvariation (Grime 2001 Westoby et al 2002) Opinionsvary but three such dimensions are now widely accepted asimportant to plant functioning and ecological strategies(Westoby 1998 Westoby et al 2002) resource use heightand seed size These are detailed below

31 The resource use dimension

Persistence in a site depends to a large extent on howindividuals use (ie acquire retain and loose) theirresources The functioning of green leaves is of primeimportance since it sets the pace of carbon input in theplants (and ecosystems) and controls many aspects of plantmetabolism

A fundamental trade-off between fast acquisition rateand efficient conservation of resources has been dis-cussed in the ecological literature for the last fourdecades (Small 1972 Grime 1977 Chapin 1980Berendse and Aerts 1987 Reich et al 1992) Howeveronly in the last 15 years or so has the availability of largesets of data allowed us to quantify these trade-offs and

identify the trait syndrome that could be used to describe iton a broad scale (Grime et al 1997 Reich et al 1997Wright et al 2004) A worldwide survey conducted by theGlobal Plant Trait Network (Glopnet) on individuals ofmore than 2500 species growing under field conditionsfor six leaf traits mdash mass-based maximum photosyntheticrate (PSm) dark respiration rate leaf life span (LLS)specific leaf area (SLA the ratio of leaf area to leaf mass)leaf nitrogen and phosphorus concentrations (LNC andLPC respectively) further expanded to leaf potassiumconcentration (Wright et al 2005) mdash shows a negativecorrelation between PSm and LLS largely independent ofgrowth form plant functional type or biome (Wright et al2004 Fig 4) Rapid photosynthesis tends to be associatedwith high specific nutrient (at least nitrogen) uptake andhigh relative growth rate (RGR Poorter and Garnier 2007and references therein) while a long LLS relates to a longresidence time of nutrients (the average time a moleculeremains in the plant once it has been taken up Berendseand Aerts 1987) within the plant (Berendse and Aerts1987 Garnier and Aronson 1998 Eckstein et al 1999Aerts and Chapin 2000 Kazakou et al 2007) Leafphotosynthetic rate and life spans are therefore two traitsthat can be used to assess the position of species on theacquisitionconservation trade-off The Glopnet data set alsoshowed that PSm and LLS were respectively positively andnegatively related to SLA LNC and LPC and leaf respirationrate (Wright et al 2004) while leaf potassium concentrationwas only loosely related to these various traits (Wright et al2005) Hence the idea of a worldwide lsquoleaf economicsspectrumrsquo (LES) composed of these different leaf traitswhich runs from quick to slow return on investments ofnutrients and dry mass in leaves The Integrated ScreeningProgramme conducted under laboratory conditions on a morelimited number of species but a wider range of traits point tothe same conclusions (Grime et al 1997)

It is recognised that many of the physiological andprotective features of green leaves persist throughsenescence leading to the so-called afterlife effects(Grime and Anderson 1986) whereby certain features ofliving organs relate to those of dead leaves ie leaf litter(Cornelissen and Thompson 1997 Wardle et al 1998Cornelissen et al 1999 Peacuterez-Harguindeguy et al 2000)A recent meta-analysis involving 818 species from 66decomposition experiments on 6 continents indeedshowed that the degree to which inter-specific variationsin living leaf structure and composition affect thedecomposition rate of litter could be as large as the effectof global climatic variation (Cornwell et al 2008) Thismeta-analysis showed that both SLA and LNC two traitsinvolved in the LES were positively related to litterdecomposition rate Leaves with high SLA and LNCtherefore tend to produce litter with chemical andor

Fig 4 Relationship between LLS and maximum leaf photosyntheticrate in the Glopnet data set (for details see Section 31) Each pointrepresents a species and colours are for different biomes yellowtropical forests green temperate forests blue arctic tundra darkpurple Mediterranean garrigue Correlation coefficient minus082(Plt0001 n=512) Data from Wright et al (2004)


structural properties leading to rapid decomposition TheLES therefore not only corresponds to features of greenliving leaves but also extends to those of leaf litter

Among the traits involved in the LES SLA appears asthe easiest quickest and cheapest to measure and is thuswidely used to assess the position of species on theacquisitionconservation trade-off (Westoby 1998 Weiheret al 1999 Westoby et al 2002) Wilson et al (1999)challenged this conclusion and proposed that leaf drymatter content (LDMC the ratio of leaf dry mass to leafwater-saturated fresh mass) be used instead of SLA for thispurpose LDMC which is one of the SLA componentstogether with leaf thickness and density (Roderick et al1999 Garnier et al 2001b Vile et al 2005) is indeedeasier to measure and less prone to measurement errors thanSLA since only masses have to be measured is lessvariable than the latter among replicates (Wilson et al1999 Garnier et al 2001a) and might be used when speciesfrom shaded habitats are involved in the comparisons(Wilson et al 1999) Although these statements may becorrect we currently lack an evaluation comparable to thatwhich has led to the identification of the traits involved inthe LES for LDMC Limited evidence suggest that LDMCis less tightly related to PSm and RGR than SLA at thewhole-plant level (Poorter and Garnier 2007) while thecomparison of three data sets where SLA LDMC and LLSare available (Ryser and Urbas 2000 Wright 2001 Prior etal 2003 Navas et al 2010) shows a looser relationshipbetween LDMC and LLS than between SLA and LLS(Fig 5) By contrast in the few experiments where litterdecomposition has been related to SLA and LDMC therelationship has always been found to be tighter with thelatter (Fortunel et al 2009a) This is probably becauseLDMC which reflects the proportion of hard tissues in theleaves (Garnier and Laurent 1994) is the SLA componentwhich relates best to initial litter properties controllingdecomposition (cf Kazakou et al 2009) Therefore if wewant to capture a maximum of aspects of plant functioningfrom resource acquisition to resource cycling in theenvironment via litter decomposition both SLA andLDMC should be measured

Interestingly recent broad-scale inter-specific compar-isons of wood traits in trees suggest that an economicsspectrum might also apply to other organs of the plant thetrunk in these particular cases (Chave et al 2009 Weedonet al 2009 Freschet et al 2010)

32 The height dimension

As an overall assessment of plant stature height is aquantitative trait which has been adopted by virtuallyeveryone doing comparative plant ecology (Westoby et al2002 and references therein) The height dimension of an

individual should be considered in relation to the height ofneighbours being taller than neighbours confers competi-tive advantage through prior access to light and istherefore central to a speciesrsquo carbon gain strategy (King1990 Westoby et al 2002) The height at which flowersand seeds are produced might also influence reproductivebiology dispersal in particular (Greene and Johnson 1989Bazzaz et al 2000)

Height has been shown to be the primary driver of lightextinction down the canopy of plant species (Fig 6a) (King1990 Violle et al 2009) explaining most of the variation ingrowth reduction of competing individuals (Gaudet andKeddy 1988 Violle et al 2009) In spite of this very cleareffect and its easiness of measurement the interpretation ofheight attributes should be done with caution sinceheight is very dynamic depending strongly on plant

Fig 5 Comparisons of relationships between a LLS and SLA andb LLS and LDMC Correlation coefficient for the relationship ina r=minus088 (Plt0001 n=165) and b r=080 (Plt0001 n=165)Black points data from Ryser and Urbas (2000) grey points datafrom Wright (2001) green points data from Prior et al (2003)white points data from Navas et al (2010)


ontogeny and disturbance regime in particular Thisoriginally led Westoby (1998) to propose the use ofcanopy height at maturity (or potential canopy height) asthe trait describing this strategy dimension

This notion of potential canopy height is widely used fortrees for which the distinction between vegetative andreproductive heights can be ignored This is not the case forherbaceous species where the top of reproductive struc-

tures may be substantially higher than the top of vegetativeparts It has therefore been argued that in these speciesvegetative and reproductive plant heights be explicitlyconsidered as two different traits (McIntyre et al 1999Vile et al 2006b Garnier et al 2007) McIntyre et al(1999) even suggested that the prominence of the inflores-cence above the vegetative plant body be considered aspecific response trait to disturbance in grasses In additionvegetative plant height might sometimes be difficult todefine from a morphological perspective in herbaceousspecies For example some annual grasses produce onlyone rapidly heading tiller while some rosette speciesproduce a stalk at the end of an elusive vegetative stageand some legumes with indeterminate growth produceleaves and flowers at the same time In these casesreproductive height at maturity is the only operationalassessment of plant height How it relates to the relativecompetitive ability for light still remains to be assessedhowever (see discussion in Violle et al 2009) It followsthat for broad-scale comparisons of herbaceous speciesdiffering in growth forms and habits reproductive plantheight might be the most relevant trait to measure

Height also correlates with a number of other dimensionspertaining to plant stature In trees it is positively related totrunk basal area up to a certain height with species-dependent relationships (eg King 1990 Niklas 1995a bWirth et al 2004) A combination of height with basal areain equations of the form ([Basal Area]2timesHeight) hasextensively been used to assess tree above-ground biomasswith improved estimations when wood-specific gravity istaken into account (Chave et al 2005 and referencestherein) Similarly in herbaceous plants plant height scalespositively with stem diameter (Niklas 1995b) and averagecross-sectional area of roots (Wahl and Ryser 2000Hummel et al 2007) Such increases in several plantdimensions are likely to confer higher stiffness and betteranchorage abilities necessary for enlarged height growthTall plants also tend to be deep-rooted (eg Schenk andJackson 2002 Violle et al 2009) As soil humidityincreases with soil depth during dry periods (eg Meinzeret al 1999) tall plants are likely to have access to largerwater sources than would smaller shallow-rooted plants(Schenk and Jackson 2002) Whether this can improvewater status during drought will depend in particular onhow plant height (and rooting depth) relates to total plantleaf area and stomatal regulation of transpiration

Finally plant height is also an important part of acoordinated suite of life history traits including seed masstime to reproduction longevity and the number of seeds aplant can produce per year (Moles and Leishman 2008 andreferences therein) In particular a high reproductive heightenhances pollination andor efficiency of seed dispersal inherbaceous species (Waller 1988 Verbeek and Boasson

Light transmission (J cm-2)

2500TM

MMPHTM

TA a2000

ASDCAS BMCN CF

RPVP1500 GR

TC

VP

TC

1000PB

500IC

BPBE

DG

IC

0 10 20 30 40 500

Plant height (cm)

Soil transpirable water ( of maximum)

100GRRPTC

b

80

AS

BM DC

GR

MM PH

RPTC

TMTAVP

80PB

CFDCPH

60 CNDGIC

60

BPBE

40

BP

20

0 20 40 60 80 1000

Rooting depth (cm)0 20 40 60 80 100

Fig 6 a Light depletion as a function of plant height and b soiltranspirable water (an index of water content) as a function of rootingdepth measured in 18 monocultures grown in an experimental gardenSpecies are represented by their initials to illustrate the generality ofrelationship AR Arenaria serpyllifolia BP (photograph) Brachypo-dium phoenicoides BE Bromus erectus BM Bromus madritensis CNCalamentha nepeta CF Crepis foetida DG Dactylis glomerata DCDaucus carota GR Geranium rotundifolium IC (photograph) Inulaconyza MM (photograph) Medicago minima PH Picris hieracioidesPB Psoralea bituminosa RP Rubia peregrine TC Teucrium chamae-drys TM Tordylium maximum TA Trifolium angustifolium VPVeronica persica Coefficient of non-linear regression in a r2=093(Plt0001 n=18) coefficient of linear regression in b r2=076(Plt0001 n=18) Redrawn from Violle et al (2009)


1995 Lortie and Aarssen 1999 Soons et al 2004)explaining the large investment of many grassland plantspecies especially rosette plants in stalk length (Bazzaz etal 2000) In trees the height at which seeds are releasedhas a positive effect on the distance at which these seeds aredispersed and is one of the parameters taken into account inmost dispersal models (eg Greene and Johnson 1989Kuparinen 2006)

Beyond these positive correlations height carries severaltrade-offs which are not fully understood these include theupper limit on height the pace at which species growupwards the duration over which stems persist at theirupper height (King 1990 Westoby et al 2002)

33 The seed size dimension

Seed mass varies among species over a range of more thanten orders of magnitude (Harper et al 1970 Moles et al2005a) and appears to be one of the least plasticcomponents of plant structure (Harper et al 1970) Seedmass affects virtually all aspects of plant ecology includingdispersal seedling establishment and persistence (Harper etal 1970 Westoby 1998 Weiher et al 1999 Fenner andThompson 2005)

A well-documented trade-off pertaining to seed produc-tion concerns seed mass and number For a given carboninvestment to reproduction plants can either produce manysmall or few large seeds (Smith and Fretwell 1974 Shipleyand Dion 1992 Leishman et al 2000) (Fig 7) and seedmass appears as the best easy predictor of seed output persquare metre of canopy cover (Westoby 1998) Theseextremes therefore reflect two different strategies ofmaximising either output or competitive ability (Jakobssonand Eriksson 2000) The large output advantage of small-seeded species is actually counter-balanced later in the lifecycle mainly at the seedling stage large seeds tend toproduce large seedlings both within and among species(reviewed in Leishman et al 2000) which have a higherprobability of survival under different kinds of unfavoura-ble environmental conditions (Jakobsson and Eriksson2000 Leishman et al 2000 and references therein) Thisincludes situations where seedlings compete with estab-lished vegetation andor other seedlings germinate underdeep shade in the soil or under a deep litter layer (reviewedin Westoby et al 2002) Such advantages are the conse-quence of large seeds having larger amount of reservesallowing seedlings to sustain respiration for longer periodsunder carbon deficit (Westoby et al 2002)

The relationships between seed mass and speciesdispersal ability in space and time have also been thesubject of much research Dispersal in space has beenrelated to seed mass in that wind-dispersed seeds tend tobe light in mass and that small seeds are associated with

large seed production enhancing dispersal by itself(Weiher et al 1999) However dispersal distances whichalso depend on other attributes of the propagule anddispersal agents themselves have not been found to beeasily related to seed mass or any other plant attribute(Hughes et al 1994) Dispersal in time can be assessed asthe propagule longevity in the seed bank (Thompson et al1998) Thompson et al (1993) were the first to show that acombination of seed mass and shape could be used topredict seed persistence in the soil in the British florasmall and compact (round) seeds were found to persistlonger in the seed bank than large and elongated seedsThis has subsequently been found in other EuropeanSouth American and Iranian floras (references in Fennerand Thompson 2005) but not in the Australian flora(Leishman and Westoby 1998) Fenner and Thompson(2005) argued that this was due to the presence of manylarge hard-seeded species in the Australian data set whichwere likely to be persistent because of their physicaldormancy Whatever the case persistence in the seed bankmight well depend on seed attributes other than simplysize and shape

Finally seed mass is connected with other traits such asgrowth form and plant height On average seed mass ishigher in woody than in herbaceous species in trees than inshrubs and in herbaceous perennials than in annuals (Moleset al 2005a Tautenhahn et al 2008 and references therein)

Individual seed mass (g log scale)10-6 10-5 10-4 10-3 10-2 10-1 100

Number of seeds per ramet (log scale)

10-1

100

101

102

103

104

105

106

Fig 7 Relationship between number of seeds per ramet (egindividual vegetative unit) and the average mass of individual seedsof herbaceous angiosperms for three data sets Grey points data for285 ramets representing 57 species from wetlands roadsides oldfields ditches and woodlands (taken from Shipley and Dion 1992)black points data for 34 species growing in Mediterranean old fields(taken from Vile et al 2006b) white points data for 18 species fromMediterranean old fields grown in an experimental garden at twonitrogen levels (taken from Fortunel et al 2009b) Correlationcoefficient r=minus061 (Plt0001 n=347)


More generally large species tend to produce larger seedsthan small species (Rees 1997 Thompson et al 1998 Reesand Venable 2007) and there is currently intense debate asto what evolutionary models might account for this positiverelationship (Venable and Rees 2009 Westoby et al 2009)It is beyond the scope of this paper to review the differentarguments put forward and we will simply report heresome of the concluding sentences proposed by Venable andRees (2009) lsquoSo why do large plants have large seeds [hellip]The pattern is a weak one While very small plants neverproduce seeds as large as those of big seeded trees plantsof any size have seeds that vary approximately 400ndash650-fold between species In terms of the fitness consequencesof seed size this is a huge inter-specific range independentof adult sizersquo

34 Beyond a three-dimensional scheme

The restricted number of traits discussed above should beconsidered as capturing a shortlist of leading readily quanti-fiable dimensions of variation among species (Westoby et al2002) The list needs not stop here but if reasonably wideagreement can be achieved about a few traits worthmeasuring consistently then we may hope for considerablebenefits from using these traits as predictors of ecologicalbehaviour (Westoby 1999 Westoby et al 2002) These traitscould then be used to compare ecological communities oncommon grounds andor literature synthesis

It is clear however that the limited number ofdimensions captured by such a short list of traits cannotaccount for all aspects of plant and vegetation functioningFor example to understand vegetation response to distur-bance (McIntyre et al 1999 Weiher et al 1999) land usechange (Garnier et al 2007) or water availability (Ackerly2004) traits relevant to the particular environmentalvariable under investigation should be measured But thekey point is that traits describing the leading dimensionsidentified so far should be included as well as absolutevariables allowing one to compare species and experi-ments This was indeed the case in the four examplesmentioned above in which SLA plant height and seed masswere actually considered

Only above-ground traits which are relatively easilyaccessible have been discussed so far Recent studies haveendeavoured to provide the same kind of informationbelow-ground focusing on the functions of root traitstrade-offs among them and how they relate to above-ground traits that could then be used as proxies for below-ground function (Lavorel et al 2007 and referencestherein) Suites of correlated traits appear to also exist inroots and represent a trade-off between root longevity andgrowth rate (cf Roumet et al 2006 Craine 2009 Freschetet al 2010) However root traits are definitely not easy to

measure and a number of challenges still need to beaddressed (Lavorel et al 2007) (1) identify traits related tokey root functions such as nutrient acquisition anchoringrhizospheric activity decomposition rate (2) test relation-ships between leaf and root traits for later use of leaf traitsas potential proxies of root functions (3) standardise roottrait measurements for broad comparisons Further leadingdimensions pertaining to root functioning might well beidentified in the near future provided that root researchcatches up with that conducted on the other structures of theplant

4 A functional approach to plant community structure

41 Environmental gradients and response traits

Plant functional traits have been intensively used to identifygroups of species with similar response to a particularenvironmental factor (upper part of Fig 3) defined asfunctional response groups (Gitay and Noble 1997 Lavorelet al 1997) There is now a large body of data documentinghow plant traits vary in relation to major environmentalfactors Although most significant associations betweentraits and environmental factors have been obtained using acorrelative approach some of them appear robust enoughamong sites and vegetation types to form the basis offunctional classifications for global-scale modelling (Lavorelet al 2007 Harrison et al 2010)

Recent reviews document which traits vary consistentlyalong environmental gradients thereby identifying lsquore-sponse traitsrsquo (Lavorel and Garnier 2002) Leaf traitsrelated to resource economy including SLA vary withnutrient availability as do some root traits (Lavorel andGarnier 2002 Lavorel et al 2007 Ordontildeez et al 2009)Traits describing plant phenology leaf structure andor rootmorphology vary with water availability (Niinemets 2001Ackerly 2004 Mitchell et al 2008) Life cycle plantmorphology and re-growth ability are involved in plantresponse to grazing (Diaz et al 2007) andor to fire(Ackerly 2004 Pausas et al 2004 Keith et al 2007)Traitslinked to more complex environmental gradients have alsobeen identified For example traits responding to change intemperature or moisture availability such as life form leafsize plant height and root architecture can be used todescribe the response to climate (Thuiller et al 2004Fortunel et al 2009a Moles et al 2009) Similarly leaftraits related to resource economy plant phenology and lifeform have been found to vary significantly along gradientsof secondary succession (Garnier et al 2004 Navas et al2010) and land use (Garnier et al 2007 Queacutetier et al2007) The reader is referred to these various studies formore details


A key issue in this analysis of trait response toenvironmental factors is that the abundances of thosespecies which persist in a community are usually notidentical In general plant communities have a typicalstructure with a relatively small number of dominantspecies which account for a high proportion of the totalbiomass and a large number of minor species that accountfor a low proportion of the biomass (Whittaker 1965 Grime1998) It has been argued that this pattern of abundance wasreflecting differential adequacy between species and theirenvironment dominant species being better fit than lessabundant species (Vile et al 2006a Cingolani et al 2007)This generally results in a tighter relationship between traitsand environment when species abundance is taken intoaccount by calculating community-weighted means (CWM)of traits (see Section 45 and Fig 8a) (Garnier et al 2004Shipley et al 2006 Vile et al 2006a Cingolani et al 2007and see discussion in Pakeman et al 2008)

42 Assembly rules and environmental filters

The need to identify rules governing the assembly ofcommunities to predict which subset of the species poolfound in a particular geographical region will occur in aspecific habitat (first steps described in Fig 2) was firstrecognised by Diamond (1975) However rules originallysearched for by description of plant or animal associationswere highly contingent and failed to provide us withgeneral principles about actual multi-species communitiesMore recently assembly rules have been considered asresulting from the impact of filters imposed on a regionalspecies pool that filter out phenotypes with non-appropriatevalues of response traits whereas species with appropriatetrait values survive to reproduce (Keddy 1992a Weiher andKeddy 1999 McGill et al 2006)

It has been hypothesised that community organisationwas the result of three main filters operating at differentlevels (Belyea and Lancaster 1999 Lortie et al 2004) (1)a dispersal filter which determines the pool of potentialcolonists available at a particular time and place in relationto stochastic biogeographical events storage effects andlandscape structure (2) an abiotic filter which expressesthe impacts of environmental conditions (soil pH humid-ity temperature etc) resource availability and distur-bance (3) a biotic filter pertaining to positive andnegative interactions among organisms in the communityThe two latter filters restrict species establishment andmediate interactions among successful colonists individ-uals are filtered on the bases of their physiologicaltolerance to the environment and their ability to cope withorganisms they interact with Recently new experimentaland statistical procedures aiming at disentangling therelative effects of these different filters have demonstrated

a major impact of habitat filtering over stochastic eventsespecially due to dispersal (Gilbert and Lechowicz 2004Cornwell et al 2006) although the taxonomic identity of

CWM_SLA (m2 kg-1)

10 15 20 25

Species SLA (m2 kg-1)

0

10

20

30

(b)

Species trait value

CommunityweightedmeanvalueA B

Species1

Species2

Species3

αA1

αB1

β

(a)

Fig 8 Decomposition of the variation in CWM of traits into between-site (β) and within-site (α) variation a Trait values of threehypothetical species at two different sites are shown the size of thesymbol is proportional to the abundance of the species the blacksquares represent the CWM values in each community and the verticalarrows represent α variation for species 1 within the two sites withCWM values A and B (αA1 and αB1) while horizontal arrow showsthe variation of CWM values (β trait variation) between the two sites(derived from Ackerly and Cornwell 2007 Shipley 2010) b Actualdecomposition along a nutrient availability gradient in Mediterraneanrangelands where values on the X-axis show the variation ofcommunity weighted means of specific leaf area (CWM_SLA) on 6points along the gradient (β trait variation) The Y-axis shows the αtrait variation for a number of species (each symbol and line representa different species) present in the different situations along thegradient (the grass Bromus erectus is the only species present in the6 situations and is shown in red) Data taken from Fayolle (2008)


co-occurring species was found to depend on historicalcontingency species with similar trait values tended to co-occur more often than expected by chance

43 Assembly rules and trait distributionwithin communities

Identifying how community assembly processes shape thedistributions of trait values within and among communitiesis a major goal of community ecology (Weiher and Keddy1995 Weiher and Keddy 1999 McGill et al 2006) Traitdistribution at a site is said to be convergent (respectivelydivergent) when the range of trait values is narrower(respectively larger) than expected by chance Habitatfiltering generally leads to convergence of traits within acommunity (cf Fig 2) due to common adaptation to thephysical environment (Cornwell and Ackerly 2009 deBello et al 2009) However convergence is generally morestringent in vegetative traits responding to resource levelsthan in regenerative ones responding to disturbance (Grime2006) (cf the two hypotheses shown in Fig 2) By contrastniche differentiation that limits interactions among similarplantsmdashthe limiting similarity principlemdashleads to traitdivergence (Stubbs and Wilson 2004 Wilson 2007Cornwell and Ackerly 2009) Therefore the extent towhich trait values tend to converge or diverge withincommunities depends on the balance between processesrelated to habitat constraints and those leading to limitingsimilarity between individuals As a consequence attributeconvergence varies with spatial scale traits involved ininteractions with other plants such as those describingresource use are generally less convergent at the neighbour-hood scale than at the habitat scale (McKane et al 2002Stubbs and Wilson 2004 Cavender-Bares et al 2009)Convergence of traits may also vary with time divergenceis often promoted in the regenerative phase (Fukami et al2005 Kraft et al 2008) as a response to the unpredictabilityof disturbance events governing recruitment (Grime 2006)whereas traits used to characterise the established phase ofthe life cycle tend to converge as a result of the non-randommortality of juveniles

Changes in trait distribution among communities arenot only due to species replacement the intra-specificvariability of traits due to phenotypic plasticity orecotypic variation is far from being negligible (cfFig 8b) and may explain a substantial part of attributevariations along regional gradients (Cornwell and Ackerly2009) The relative importance of inter- vs intra-specificvariability of attributes is likely to depend on the situationunder study because of large differences in intra-specificvariability among traits along different types of gradients(Garnier et al 2001a Roche et al 2004) Characterisingthis relative importance on trait distribution is critical in

the context of global change organisms might firstrespond to shifts in environmental conditions throughplasticity of individual species while species replacementandor genetic shifts within the species might then follow(Ackerly 2003) with larger impact on functional charac-teristics of communities

Recognising the importance of intra-specific variationamong sites Ackerly and Cornwell (2007) proposed todecompose species mean trait value into two componentsbeta trait value refers to species mean position along anenvironmental gradient (ie a measure of niche position)and alpha trait value is the difference between a speciestrait value and the weighted mean of co-occurring taxawithin a community (Fig 8) This partitioning of traitvalues builds on alpha and beta niche of species (Pickettand Bazzaz 1978 Ackerly et al 2006) the alpha nicherefers to attributes that differentiate a species from co-occurring taxa and therefore may contribute to non-neutral maintenance of species diversity within a com-munity whereas the beta niche refers to species distri-bution across habitat or geographic gradients Beta traitvalues have generally lower divergence than expected atrandom (cf example of Mediterranean rangelands alonga nutrient gradient Fig 8b) an observation consistentwith the hypothesis of trait convergence due to habitatfiltering By contrast alpha trait values are generallyhighly variable which might be interpreted as a limitingsimilarity effect (Cornwell and Ackerly 2009 de Bello etal 2009)

44 Phylogeny traits and community assembly

A highly debated issue is to understand whether traitconvergence as discussed above is related to the position ofspecies in the phylogeny (Webb et al 2002) A frequentassumption is that closely related species are more similarin terms of environmental requirements than more distantlyrelated species (niche conservatism) According to thishypothesis a clustered phylogenetic structure of communi-ties is expected when traits converge because of habitatfiltering species should be more often phylogeneticallyrelated than expected at random However at the neigh-bourhood scale where trait divergence is due to processeslimiting similarity between plants the same hypothesisleads to the prediction of an over-dispersed phylogeneticstructure interacting species should be less often phyloge-netically related than expected at random By contrasthabitat filtering might lead to an over-dispersed phyloge-netic structure if traits important for habitat specialisationare labile and close relative species specialise for differentniches

Understanding the influence of phylogeny on the func-tional structure of communities has been the topic of


numerous experimental studies (for a recent review seeCavender-Bares et al 2009) A first conclusion is thatphylogenetic conservatism differs among traits for exampleseed mass is a highly conserved trait among AmericanQuercus species whereas plant height is not (Cavender-Bares et al 2006) A second conclusion is that thephylogenetic structure of communities is generally non-random at multiple spatial and taxonomic scales revealing amajor influence of habitat filtering and competitive inter-actions on species distribution Unravelling the impacts ofphylogenies on community assembly is still in its infancyand is a whole research field under development A centralissue will be to detect the importance of the phylogeneticsignals on traits (see example of the phylogeny of LMA inFig 3) which will be made possible by the increasingavailability of phylogenetic data computing power andinformatics tools (Cavender-Bares et al 2009) Possibleapplications include the prediction of ecosystem processesfrom the phylogenetic structure of communities providedthat effect traits as defined above are phylogeneticallyconserved (Cadotte et al 2009)

45 The functional structure of communities

The functional structure of a community has been originallydescribed by the diversity in functional types based on differ-ences in life form plant morphology or resource acquisitiontypes (eg N-fixing or non-fixing species C4 or C3 photosyn-thetic pathways etc) (Lavorel et al 1997) Because oflimitations of methods based on discrete and subjectiveclassifications the functional diversity (FD) of communitiesis currently defined as the value range and relative abundanceof traits found in a given community (Diacuteaz et al 2007a)

This functional structure can be described through twocomponents value and range of traits (cf Fig 3)corresponding to two types of quantitative variables Thefirst ones allow one to estimate the average value of traits inthe communities and can be written as

CFP frac14Xn

ifrac141

pi traiti eth1THORN

where CFP represents the functional parameter of thecommunity (Violle et al 2007b) pi and traiti arerespectively a weighting factor and the trait value forspecies i and n is the total number of species in thecommunity When only presenceabsence data for speciesare available all pi are fixed at the same value (1n) and thecalculated CFP are unweighted averages when speciesabundances are known a common procedure is to use therelative abundance of species as pi values therebycalculating CWM (Garnier et al 2004 for differentweighting procedures see also Hodgson et al 2005)CWM represents the most probable attribute that a species

taken at random in the community will display Thiscalculation is based on the massndashratio hypothesis (Grime1998 and detailed in Section 51) which proposes that thecontrols on function by species are proportional to theirabundance CWM is a robust index showing low suscep-tibility to experimental procedures used to evaluate speciesrelative abundance and trait values (Lavorel et al 2008)

The second type of variables aims at describing thefunctional dissimilarity among the species present in thecommunities Numerous indices have been proposed to doso (reviewed in Petchey and Gaston 2006) Some are basedon the sum of (functional attribute diversity Walker et al1999) or average (quadratic entropy Lepš et al 2006)functional distances between species pairs in a multivariatefunctional trait space distances between species along ahierarchical classification (FD Petchey et al 2004) or onthe distribution of abundances along functional trait axes(FDvar Mason et al 2003) More recently indices ofphylogenetic diversity that specifically take into accountspecies abundance in the community were proposed assurrogate of FD for traits supposed to be highly conservedthrough phylogeny (Cadotte et al 2010)

The design and use of these indices are currently highlydebated A critical point is that they must meet severalcriteria to be of general use they must be designed to dealwith several traits if necessary take into account speciesabundance and measure all facets of FD First the questionof how many and which traits to use is crucial to get ascomplete an identification of FD as possible traits must berelevant to account for mechanisms of species coexistenceas well as effects of species on ecosystem functioning (seeSection 51) Increasing the number of traits should thusmake the identification of FD clearer however this caninduce trivial relationships with species richness that shouldbe avoided These arguments led Lepš et al (2006) tosuggest that traits that represent different ecologicalstrategies or allow one to assess ad hoc response toenvironmental factors are the best candidates By contrastquantitative methods for selecting traits without a prioriarguments are currently being developed (Bernhardt-Romermann et al 2008 Sonnier et al submitted) Secondthe effect of each species in the calculation of indices has tobe weighted according to abundance in order to reflectspecies contribution to ecosystem functioning as stressedabove and in Section 51 This criterion whose importanceremains to be established in the case of communitydynamics is not fulfilled for most indices based onfunctional distances between species pairs Third FD is arather complex idea that can be partitioned into threecomponents functional richness defined as the volume ofthe functional space occupied by the species in thecommunity functional evenness defined as the regularityof the distribution of abundance in this volume functional


divergence (FDvg hereafter) represented by the divergencein the distribution of abundance in this volume (Fig 9)Initially calculated with single-trait indices (Mason et al2005) new procedures using several traits have beenrecently proposed (Villeger et al 2008) Among thesecomponents FDvg appears of particular interest to charac-terise the FD of communities by revealing the complemen-tary (high FDvg) or similarity (low FDvg) in species traitswithin a community corresponding to the divergence vsconvergence issue as discussed in Section 44

5 Traits and ecosystem properties

Chapin et al (2002) recognise that EP are controlled by twobroad categories of factors (1) independent controllingvariables (state factors) which are climate parent rockmaterial topography potential biota and time and (2)interactive controls which are factors that both control andare controlled by EP resources modulators (temperaturepH etc) disturbance biotic community and humanactivities The occurrence of different controls on EP atlarge scales is illustrated in Fig 10 showing the relation-ships between net primary productivity (NPP) of differentbiomes and either annual rainfall or mean height ofvegetation For a detailed account of these different controls

on various EP the reader is referred to eg Schulze andZwoumllfer (1987) Aber and Mellilo (2001) and Chapin et al(2002) Here we will focus on studies which haveexplicitly investigated the controls that plants as majorplayers in the biotic component might exert on EP alone orin combination with various other factors

Awareness that plant communities have substantialimpacts on EP has long been recognised (reviewed inVitousek and Hooper 1993 Grime 2001 Chapin et al2002 Wardle 2002 Eviner and Chapin 2003) The issue ofwhether and why more species andor predefined functionalgroups have significant effects on the functioning ofecosystems has been the subject of a lively debate overthe past two decades (for recent syntheses see eg Loreauet al 2001 Hooper et al 2005 Balvanera et al 2006Cardinale et al 2006 Cardinale et al 2007 Naeem et al2009) and there is now a growing consensus that EPdepends more on the functional characteristics of speciesthan on their number (Chapin 1993 Grime 1997 Chapin etal 2000 Diacuteaz and Cabido 2001 Lavorel and Garnier 2002Diacuteaz et al 2007a) As discussed in Section 45 thesefunctional characteristics translate into two components ofcommunity FD value and range of traits These compo-nents can be theoretically linked to two broad hypothesesconcerning the relation between diversity and EP theselection mechanism (ie the selection of the dominantspecies Grime 1998) and the niche complementarity effect(Petchey and Gaston 2006) respectively

Diacuteaz et al (2007b) proposed a framework to assess therelative importance of abiotic and biotic controls on EP andthe services they can deliver to humans which takes intoaccount these two components of FD In this frameworkthe effects of abiotic factors are tested first followed bythose of CWM and FD finally potential effects ofparticular species andor species groups not accounted byFD metrics are considered In the remaining part of thissection we will therefore focus on studies examining theimpact of FD components on EP thereby addressing thequestion of how organisms and communities affect ecosys-tem functioning

51 The biomass ratio hypothesis community-levelweighted means of traits

The selection effect forms the basis of the lsquobiomass ratiohypothesisrsquo (BRH Grime 1998) which states that partic-ular traits of locally abundant species will determine therate and magnitude of instantaneous ecosystem processesSuch traits have been called lsquoeffect traitsrsquo (Diacuteaz and Cabido2001 Lavorel and Garnier 2002 cf Fig 3) According tothe BRH EP therefore depend on CWM of effect traits(Garnier et al 2004 Diacuteaz et al 2007a Violle et al 2007b)An increasing number of studies show that this hypothesis

Relative abundanceof traits

High evennessLow divergence

Low evenness

Niche axis

High divergence

(trait value)

Mintrait value

Maxtrait value

Functional richness

Fig 9 Components of FD Functional richness corresponds to theportion of niche axis represented by the range of trait values found inthe community Functional evenness and divergence representrespectively the regularity and divergence of distribution of traitsacross the niche axis Here we represent two theoretical casesrepresenting contrasted patterns in functional evenness and divergencefor a same functional richness and average trait value


holds for some key biogeochemical processes such asprimary productivity components of the water balance leafdigestibility and litter decomposition

511 Components of primary productivity

Several traits related to plant growth andor to the LESdescribed above have been shown to affect ecosystem NPPor some of its components Early literature surveys whichdid not take species abundance into account have shownpositive associations between whole-plant RGR averagedacross species and ecosystem above-ground net primaryproductivity (ANPP) across biomes (Chapin 1993) Build-ing on Chapinrsquos work (1993) a formal relationship betweenspecies traits and NPP has been proposed by Lavorel andGarnier (2002) which formalises the BRH as applied toproductivity of ecosystems

NPP frac14Pn

ifrac141Ni Moi etheRGRi

ethtft0THORNi 1THORNΔT

eth2THORN

where n is the total number of species in the community Ni

is the number of individuals of species i per unit groundarea Moi is the initial average biomass of individuals ofspecies i RGRi and (tfminus t0)i are respectively the RGR andperiod of active growth of species i and ΔT is the periodover which NPP is assessed Due to the difficulty to accessthe underground compartment to date most tests of Eq 2have been conducted taking only above-ground plant partsinto account

Since realised RGR is particularly difficult to assess inthe field this relationship has been tested using either RGRdata measured in the lab (RGRmax) or independentmeasurements of traits involved in the LES andor assumedto be related to RGR (see Section 31) Leoni et al (2009)report that inter-specific differences in RGRmax measuredunder controlled conditions on dominant species of grazedand ungrazed plots of Uruguayan grasslands tend to scale

up with above-ground NPP (ANPP) measured in these plots(Altesor et al 2005) This suggests that species ranking forRGRmax and realised RGR remains comparable in spite ofthe complex differences in ontogeny plantndashenvironmentand plantndashplant interactions involved in the comparisonbetween laboratory and field-grown plants (see discussionin Vile et al 2006a) Different relationships were foundbetween the CWM of various leaf traits and ANPP ANPPwas found to be related to CWM_leaf angle in Australiangrasslands (Mokany et al 2008) to CWM_leaf area insubalpine grasslands in the French Alps (Gross et al 2008)

Fig 10 Examples of potential abiotic and biotic controls on ANPP ofterrestrial ecosystems Relationships between ANPP and a annualprecipitation (redrawn from Gower 2002) and b average height of thevegetation (data taken from Saugier et al 2001) for different biomesNote that classification of biomes and ANPP values for a given biomemight differ between the two studies in a BODBL boreal deciduousbroadleaved BODNL boreal deciduous needle-leaved BOENL borealevergreen needle-leaved DESRT deserts TEMGS temperate grass-lands TROGS tropical grasslands TEDBL temperate deciduousbroadleaved TEEBL temperate evergreen broadleaved TEENL tem-perate evergreen needle-leaved TRDBL tropical deciduous broad-leaved TREBL tropical evergreen broadleaved TUNDR tundraWDLND woodlands in b BORFT boreal forests CROPS cropsDESRT deserts MEDSH Mediterranean shrublands TEMFT temper-ate forests TEMGS temperate grasslands TROFT tropical forestsTROGS tropical savannas and grasslands TUNDR tundra Correlationcoefficient in a r=090 (Plt0001 n=13) and in b r=091 (Plt0001n=8)


and to a combination of CWM_LDMC and traits of rootsand rhizomes in mesocosms of mountain grasslands ofCentral France (Klumpp and Soussana 2009)

In studies conducted on herbaceous ecosystems wheremaximum above-ground standing biomass (AGBmass) wasused as an estimate of ANPP (for discussion see Scurlock etal 2002) CWM_leaf area was positively related to AGBmass

(Thompson et al 2005 Mokany et al 2008) while Garnieret al (2004) found weak negative relationships betweenAGBmass and CWM_SLA and CWM_LDMC and Diacuteaz et al(2007a) did not find any significant relationship betweenAGBmass and any CWM value

ANPP is a ground area-based variable (expressed in eggrammes per square metre per day) and in Eq 1 traits areweighted by the absolute biomass of species in thecommunity (NitimesMoi) This means that ANPP depends onthe initial standing biomass of the whole community (seealso Section 54) as much as on species traits To accountfor this mass effect and assess more directly the role oftraits as components of ecosystem productivity Reich et al(1992 1997) calculated the ANPPlive canopy dry massratio (lsquoecosystem production efficiencyrsquo) while Garnier etal (2004) built on Eq 2 to devise a lsquospecific net ecosystemproductivityrsquo as

SNPP frac14logeeth

Pn

ifrac141pi eRGRi

ethtft0THORNiTHORNΔT

eth3THORN

where pi is the initial proportion of species i in thecommunity instead of its biomass Ecosystem productionefficiency and SNPP are both in units of grammes pergramme of green biomass per unit time and can be viewedas the RGR of the whole community In the surveyconducted by Reich et al (1992) ecosystem productionefficiency (above-ground) was found to be positivelyrelated to SLA and negatively related to LLS in foreststands Testing Eq 3 in a successional sere of SouthernFrance Garnier et al (2004) found tight relationshipsbetween both CWM_ SLA and CWM_LNC and above-ground SNPP (see also Klumpp and Soussana 2009)Interestingly working in the same successional sere Vileet al (2006a) found a positive relationship between above-ground SNPP and CWM_RGRmax In these three studiestraits involved in the LES were thus shown to have asignificant impact on ecosystem productivity

Contrasting with these results in support of the BRH Diacuteazet al (2007a) found that ANPP and above-ground SNPPwere both unrelated to any CWM in French subalpinegrasslands but were dependent on an index of nitrogenlimitation Over a much broader spatial scale a study of 60plots spread across the Amazonian forest from the Pacific tothe Atlantic coasts and covering a 4000-km range fromnorth to south showed that coarse wood production was

unrelated to CWM of maximum height and wood densityand was primarily driven by environmental factors whichwere not explicitly identified (Baker et al 2009)

Clearly the impacts of abiotic factors vs those of thefunctional structure of the community on components ofecosystem NPP requires further testing on a wide range ofsituations

512 Decomposition and accumulation of litter

Given the close coupling between SNPP andor ecosystemproduction efficiency and litter decomposition across ecosys-tems (eg Cebriaacuten and Duarte 1995 Chapin et al 2002) theBRH is likely to apply for the process of decomposition aswell with traits involved in the LES There is still somedebate as to which characteristics of the litter have an effectof decomposition (see reviews in eg Aerts 1997 Aerts andChapin 2000 Chapin et al 2002 Wardle 2002) and whichtraits of living leaves best reflects the quality of the litter (seeSection 31) Although the BRH is therefore likely to applyfor decomposition as well an actual formalism comparableto that proposed above for SNPP is not straightforwardWhatever the form of the relationship empirical tests inseveral situations have shown that CWM of traits involvedin the LES tend to scale with community-level decomposi-tion rates (Garnier et al 2004 Cortez et al 2007 Quested etal 2007 Queacutetier et al 2007 Fortunel et al 2009a) Asfound at the species level CWM_LDMC correlates betterwith decomposition rates than either CWM_SLA (Garnier etal 2004 Quested et al 2007 Fortunel et al 2009a) orCWM_leaf nutrient concentrations (Garnier et al 2004Fortunel et al 2009a but see Diacuteaz et al 2007b) In a broad-scale study across 11 European sites CWM_LDMC was thecharacteristic of living leaves best related to the lignin tonitrogen ratio of the initial litter the latter explaining 44 ofthe variance in litter decomposition rate (Fortunel et al2009a see also Diacuteaz et al 2007b)

Since traits have a role both on the production anddecomposition rates we also expect effects on litteraccumulation CWM_LDMC was positively related toabove-ground total dead matter in Australian grasslands(Mokany et al 2008) and across 11 European sites (Garnieret al 2007) while Diacuteaz et al (2007a) found that a multipleregression involving CWM of vegetative plant height LNCand leaf tensile strength accounted for 75 of the variancein above-ground total dead matter

513 Soil water content

At any given time the soil water content of a particular soildepends on the balance between rainfall actual evapotrans-piration from the plant cover surface runoff and percolationthrough the soil (Larcher 2003) For the sake of simplicity


and because we have no information as to how speciesinfluence the two latter processes these will not beconsidered in the following Since actual evapotranspirationduring a time interval Δt is the sum of evaporation andwater uptake from the whole cover a simplified soil waterbalance model can be written as (cf eg Boulant et al2008)

Stthorn1 frac14 minfrac12St thorn RΔt EΔt UΔt FC eth4THORNwhere St+1 and St are soil water content at time t+1 and trespectively RΔt EΔt and UΔt are rainfall evaporation fromthe whole cover and water uptake by plants between t+1and t (Δt) respectively and FC is the soil field capacity (thesoil water content at saturation after the gravitational waterhas percolated) According to Eq 4 the decrease in soilwater content after rainfall events therefore depends onEΔt and UΔt (see also Eviner and Chapin 2003) EΔt

depends primarily on the availability of energy at theground surface except after rainfall when water interceptedby the canopy evaporates (cf Schulze et al 2005) It isinversely related to the leaf area index of the vegetation(Schulze et al 1995) and to litter accumulation (Eviner andChapin 2003) As traits related to plant growth and litteraccumulation have been discussed in Section 512 wehereafter focus on the modulation of water uptake (UΔt) byplant traits

Since root distribution and maximum rooting depth ofthe different species as well as soil properties vary withsoil depth (Schwinning and Ehleringer 2001 and refer-ences therein) water uptake over the entire soil profile canbe conveniently subdivided into uptake from differentdiscrete layers Water uptake from soil layer i by species jduring a short time interval dt (Udt(ij)) over which allvariables may be considered constant follows Darcyrsquos lawand can be written (cf Schwinning and Ehleringer 2001Larcher 2003)

UdtethijTHORN frac14 MrootethijTHORN SRAij kij ethΨ soili Ψ jTHORN eth5THORN

where Mroot(ij) SRAij and kij are respectively thebiomass of roots the specific root area (SRA the ratioof root area to root dry mass) and the root area-basedhydraulic conductance to water transport of species j inlayer i while Ψsoili is the bulk soil water potential in layeri and Ψj is the water potential of species j If we assumeadditivity of the process among species the decrease insoil water content over the entire soil profile can be writtenas

Udt frac14Xm

ifrac141

Xn

jfrac141

frac12MrootethijTHORN SRAij kij ethΨ soili Ψ jTHORN

eth6THORN

where n is the number of species in the community and mis the number of soil layers considered Assuming thatroots have a cylindrical shape SRA can be written as

SRA frac14 RAroot

Mrootfrac14 Lroot p droot

Mrooteth7THORN

where RAroot Lroot and droot are the root area the rootlength and the average root diameter respectively Com-bining Eqs 5 and 6 leads to

Udt frac14Xm

ifrac141

Xn

jfrac141

frac12LrootethijTHORN p drootethijTHORN kij ethΨ soili Ψ jTHORN

eth8THORN

Equations 6 and 8 show that rooting depth and rootdistribution of the different species in the different soil layerswill have impacts on the water content of these layers It alsoshows that soilndashplant hydraulic properties and plant waterpotential might affect soil water content beyond the effectsof biomass andor length (Eviner and Chapin 2003)

To our knowledge Eqs 6 and 8 have not been testeddirectly in complex communities but some studies haveassessed the effects of individual traits appearing in eitherof these on soil water content Working in alpine grass-lands Gross et al (2008) have found a significant effect ofCWM_Lroot on the water content of the upper 15 cm soillayer while in Australian grasslands Mokany et al (2008)found a strong negative effect of CWM_SRA on the watercontent of the entire soil profile averaged over the growingseason

Recognising that whole-plant water uptake and transpira-tion are tightly coupled (cf Schwinning and Ehleringer 2001Larcher 2003) we also expect traits controlling water fluxesfrom leaves to have an impact on soil water content This isthe case not only for water potential (Eqs 6 and 8) but alsofor stomatal conductance whose sensitivity to air water vapourdeficit is known to vary among species (Oren et al 1999)

Finally the time dependence of water uptake should alsobe considered which corresponds to the integration ofEqs 6 and 8 over time In particular changes in rootbiomass andor length in relation to the phenology of thedifferent species which compose the communities will haveimpacts on soil water content (Eviner and Chapin 2003)For example a species which completes its life cycleduring spring will have little effect on soil water contentduring summer unless no rainfall event occurs (Schwinningand Ehleringer 2001 Gross et al 2008)

514 Other components of biogeochemical cycles

Numerous other components of biogeochemical cycles arelikely to be affected by plant traits (reviewed in Wardle


2002 Eviner and Chapin 2003) Among those soilcarbon sequestration has recently received increasedattention (de Deyn et al 2008 Klumpp and Soussana2009) By contrast with the properties discussed so far anumber of intermediate complex steps have to be takeninto account to understand the links between plant traitsand soil C Carbon sequestration in soils represents thebalance between carbon inputs in and outputs from thesoil and many traits potentially affect these differentfluxes as well as potential species interactions (Fig 11)Traits pertaining to plant growth rate litter production andquality as discussed above play a substantial role in thisrespect

Beyond the qualitative identification of traits affectingsoil C sequestration (Eviner and Chapin 2003 de Deyn etal 2008) Klumpp and Soussana (2009) assessed quantita-tively the relative impacts of various plant traits on below-ground net carbon storage (BNCS) and top soil organiccarbon content (SOC) of semi-natural grasslands subjectedto different grazing regimes They found that both BNCSand SOC were negatively related to CWM_specific rootlength (SRL the ratio of root length to root mass) whileBNCS was positively related to CWM_root diameterGenerally low SRL is associated with thick dense rootswith low nitrogen and high lignin concentrations (ref inLavorel et al 2007) which are likely to produce slowlydecomposing litter (Hobbie et al 2010) ultimately leadingto SOC stabilisation (Fig 11) Similarly total soil carboncontent was found to be negatively related to CWM_SLAand CWM_LNC and positively to CWM_LDMC in aMediterranean old field succession (Garnier et al 2004) Asdiscussed above these traits lead to slow decompositionrate of leaves (Cortez et al 2007 Quested et al 2007) and

accumulation of litter (Garnier et al 2007) two variablesaffecting the rate of carbon input in soils

52 Complementarity among species incorporatingfunctional divergence

So far only few studies have examined the impacts ofFDvg on EP either independently or in combination withCWM and abiotic effects These were all conducted inperennial herbaceous communities either from Europe(Thompson et al 2005 Diacuteaz et al 2007b Klumpp andSoussana 2009 Schumacher and Roscher 2009) or Australia(Mokany et al 2008) A synthesis is currently difficult toconduct for at least two reasons First the restricted numberof studies available for a particular EP limits our potential togeneralise safely Second as discussed in Schumacher andRoscher (2009) the outcomes of the analyses stronglydepend on the way the different variables are considered inthe stepwise procedures In particular conclusions appearhighly sensitive to the fact that variables are treatedindependently rather than in combination with the othersand whether the selection procedure is based on individualcorrelations as proposed by Diacuteaz et al (2007b) or allvariables are treated in combination in a global analysis assuggested by Schumacher and Roscher (2009) Anotherissue pertains to whether traits are given equal weight in theanalyses (Schumacher and Roscher 2009)

A striking example is given by the study conducted bySchumacher and Roscher (2009) in semi-natural grasslandsof Germany This study showed that when a stepwiseprocedure that did not place any constraint on the selectionof variables introduced in the multiple regression was usedAGBmass was best explained by a combination of total soil

Fig 11 Potential plant traitcomposition effects on soilcarbon sequestration throughinfluencing the ratio betweencarbon gains (C-in) and losses(C-out) Taken from De Deyn etal (2008) with the agreement ofWiley


nitrogen CWM of life cycle vertical distribution of canopystructure SLA and rooting depth and FDvg of life cyclearea of leaves and vegetative height (all positive) Howeverusing the selection procedure proposed by Diacuteaz et al(2007b) they found negative effects of FDvg for manytraits This points to the strong dependency of theconclusions on the statistical procedure used

In most studies assessing the effects of FDvg on EPthese statistical effects were not clearly considered Thismight explain why in a number of studies the effects ofFDvg of different traits or combinations thereof were foundto be either negative (Thompson et al 2005 Mokany et al2008) or non-significant (Diacuteaz et al 2007b) for severalproperties Only in the study by Klumpp and Soussana(2009) investigating the impact of abiotic (disturbance) andbiotic factors in a temperate grassland on six EP wereeffects of FDvg for several traits found to be all positive

A clear assessment of the impact of FDvg on EP willrequire that statistical procedures be validated and stand-ardised Only when this is done will we be able to detectthe importance of complementarity among species for thefunctioning of ecosystems in natural and semi-naturalenvironments

53 The role of plant traits on ecosystem propertiesin controlled experiments

The framework proposed by Diacuteaz et al (2007b) aims atunderstanding the relative roles of abiotic and bioticcomponents on EP in semi-natural and natural environ-ments as discussed in the previous section Although thisapproach has the advantage of being relevant for realecosystems it also carries an inherent inability to identifythe direction of causation between the variables measured(but see the structural equation approach as applied byGrace et al 2007) Manipulative experiments in whichspecies with contrasting attributes are grown in experimen-tal gardens under identical common conditions have thusbeen designed to overcome this problem

In some of these experiments simple plant traits werefound to correlate with several EP In the study by Pontes etal (2007) on 13 perennial C3 pasture species grown over2 years at two levels of nitrogen supply under two cuttingregimes ANPP was found to be negatively related to LNCexpressed on a fresh mass basis while digestibility wasfound to increase (respectively decrease) with SLA (re-spectively LDMC) In monocultures of species fromMediterranean old fields grown at two levels of nitrogensupply a multiple regression combining SLA and LNCexplained 59 of the variance in ANPP in the first year ofgrowth (Fig 12) while in the low nitrogen treatment of thesame experiment the soil water content assessed over theentire growing season (cf Violle et al 2007a) was found to

be negatively related to species rooting depth (Violle et al2009 Fig 6b)

The large-scale lsquoJena experimentrsquo where 61 speciescommon to Central European grasslands were grown inmonocultures under similar environmental conditionsyielded more complex results after 4 months of growth(Heisse et al 2007) Once the effects of seedling densitywere accounted for differences in above-ground biomassproduction were based on variable combinations of traitsincluding time to seedling emergence the vertical distri-bution of biomass along the stem the size of individualshoots the allocation of biomass between roots and shootsand some of the traits involved in the LES No singlecombination of traits led to a high biomass productionFinally Craine et al (2002) grew 33 grasslands species for5 years under low nitrogen availability and found thatamong non-legume species which produced and main-tained large amounts of biomass were those with lowernitrogen concentrations in tissues greater longevity oforgans and higher root to shoot ratios The two formercorrespond to traits identified in Berendse and Aertsrsquos(1987) model as those allowing species to dominate

Fig 12 Relationship between observed and predicted ANPP in acommon garden experiment where 12 Mediterranean species fromthree successional stages were grown at high (black circles) and low(white circles) levels of nitrogen supply (Garnier et al unpublishedresults) Predicted values are calculated from SLA and LNC of thedifferent species as ANPP=minus586+022timesSLA+020timesLNC (correla-tion coefficient of regression between predicted and observed valuesr2=059 Plt0001 n=24) Details of the experiment can be found inKazakou et al (2007)


nutrient-poor communities Under these conditions of lownitrogen availability Craine et al (2002) also found thatnitrogen fixation allowed legumes to produce greaterbiomass above-ground than other species with high stembiomass and low fine root productions

These few experiments conducted under comparableconditions for a given set of species show that thecombinations of traits influencing EP are likely to varyaccording to the environmental conditions chosen and thetime scale of the experiment In particular results obtainedover the short term should not be extrapolated on a longerterm especially in perennial species It is likely thatanswers to this complex issue will be better tackled bycombining such controlled experiments with observationalmanipulative experiments in the field and modelling

54 A different perspective ecosystem allometry

Ecosystem allometry is a recent perspective to the relation-ship between organisms and ecosystem functioning whichfocuses on size distribution within communities (Enquist etal 2003 Kerkhoff and Enquist 2006) It is clearly a trait-based approach as size is one of the performance traitsdefined in Section 21 but here the analysis is carried outwithout distinguishing the different plant species thatcompose the communities The rationale as applied to thestudy of primary productivity of plant communities(Kerkhoff and Enquist 2006) is briefly presented below

By definition NPP is the sum of the growth of allindividuals present in the community The allometricapproach proposes that whole-plant individual growth rate(GR) can be written as

GR frac14 bg Mplant3=4 eth9THORN

where βg is the allometric coefficient relating growth rate tototal plant mass (Mplant) The 34 scaling exponent reflectsconstraints on resource supply to individual cells (West etal 1997) Empirical tests suggest that βg is constant acrossa wide range of photosynthetic organisms (Niklas andEnquist 2001) whatever the species considered Thisimplies that when the mass range is large the size effectoverrides any inter-specific differences in plant functioningKerkhoff and Enquist (2006) then suggested to groupindividuals of plant communities into K size classes ofaverage mass Mplant(k) with similar growth rate EcosystemNPP can thus be written as

NPP frac14XK

kfrac140

nk bg MplantethkTHORN3=4 eth10THORN

where nk is the number of individuals in size class k persquare metre Further development of this equation shows

that NPP should be proportional to the 34 power of totalcommunity phytomass (Kerkhoff and Enquist 2006) whichis the sum of Mplant(k) for all size classes An empirical testacross plant communities ranging from arctic tundra totropical forests yielded as expected a log linear relation-ship between NPP and total community phytomass but thescaling coefficient was substantially lower than predictedfrom theory (046 vs 075 Kerkhoff and Enquist 2006)This was interpreted as a failure of the model to account forprocesses occurring within plant communities such asasymmetry in light capture resulting from size hierarchiesIn the same study Kerkhoff and Enquist (2006) developedpredictions for the scaling between community phytomassand nutrient mass (nitrogen and phosphorus) which were incloser agreement with empirical data

Further testing of this approach is warranted especiallyin situations where the biomass range of the communitiescompared is narrower from arctic tundra to tropical forestscommunity biomass varies approximately 600-fold whilein many cases a much narrower range is of interest Withingrasslands for example differences of annual rainfall from250 to 1400 mmyear across the central grassland region ofthe USA induced a sevenfold range in biomass produced(Sala et al 1988) and the continuous fertilisation of plots inthe Park Grass Experiment (UK) during 150 years induceda threefold difference between the most and least produc-tive treatments (Crawley et al 2005) Over such reducedranges intrinsic differences in plant physiological proper-ties among species might have impacts on ecosystemfunctioning beyond those of size

6 Plant traits in agricultural and forestry contexts

In this section we give examples of the use of traits inagricultural and forestry contexts In doing so we willidentify the environmental gradients that are generated byhuman practices and examine (1) the response traits tothese gradients be they generic or specific to the particularenvironmental conditions found in agricultural contexts (2)the rules of community assembly and whether they differfrom the general case (3) how traits relate to ecosystemprocesses of agricultural importance

61 Grasslands

Most experimental studies documenting the response ofthe functional structure of communities to gradients andreferred to so far in this paper have been performed ongrasslands In this section we selected only studies withan explicit agricultural perspective In the following itshould be kept in mind that all types of grasslands havenot been equally studied in the literature there is


currently an over-representation of grazed situations overmown fields and very few studies have been conductedin grasslands characterised by an intensive use offertilisers and sown species (Plantureux et al 2009)

In grasslands the defoliation regime (either grazing ormowing for hay making) and the supply of fertilisers appearas the main management practices affecting vegetationstructure and functioning The quantitative assessment ofthe impacts of defoliation is relatively difficult to character-ise (cf Garnier et al 2007) it varies with the date ofgrazing or cutting the time interval between two successivecutting or grazing periods especially during favourableseasons for growth and the amount of harvested orconsumed biomass (Duru et al 1998) Assessment of thisimpact requires regular measurement of standing biomassto take into account the seasonal variation of growth theimpact of dates of defoliation and compensatory re-growthafter defoliation (McNaughton et al 1996) These experi-mental limits explain why data from different studies arehardly comparable except when treatments contrast inten-sive grazing to very low or absence of grazing (Louault etal 2005 Evju et al 2009)

Although the assessment of nutrient limitation couldtheoretically be done through soil measurements there iscurrently no simple method to assess the amount ofnutrients a given soil can release and make available toplants This has led to the development of nutrient indicesfor nitrogen and phosphorus (Duru and Ducrocq 1997Lemaire and Gastal 1997) which are the main limitingnutrients for grassland production These indices whichreflect the growth limitation by nutrient availability allowone to quantify changes in nutrient level due to fertilisationor grazing management

The traits of species that respond to managementpractices in grasslands are the same as those respondingto the same selective pressures in non-managed areas Forexample leaf traits such as SLA or LDMC are stronglymodulated by nutrient availability (Duru et al 1998McIntyre 2008) Grazing affects a large series of traits ahigh grazing pressure generally favours annual overperennial plants short plants over tall plants prostrate overerect plants and species with stoloniferous and rosettearchitecture over those with a tussock architecture thetiming of defoliation also impacts on phenology includingleaf senescence (Duru et al 1998 Louault et al 2005 Diazet al 2007 Kahmen and Poschlod 2008 Romermann et al2008)

A recent meta-analysis examined whether the well-documented negative effect of land use intensification onspecies richness results in a similar decline in the FD ofplant communities (Flynn et al 2009) These authorsassumed that there were three possibilities a similardecrease in species richness and FD with intensification a

faster decrease in FD than in species richness revealing thatfunctionally unique species are lost first a lower decreasein FD than in species richness showing that functionallyredundant species are lost first The outcome of the analysisrevealed that most grasslands showed a faster decline in FDthan in species richness However this relationship seemsto depend on environment as illustrated by a comparativestudy performed in Mongolia (Sasaki et al 2009a b) Thisstudy suggests that faster loss in FD than in species richnessoccurs with increasing grazing intensity in harsh environ-ments whereas functionally redundant species are lost firstin benign conditions

A trait-based approach to grassland management alsoappears to be promising In particular biomass productionthe date of peak production and the digestibility of herbagecan be predicted from leaf traits andor plant height (Al HajKhaled et al 2006 Pontes et al 2007 Ansquer et al 2009b)Furthermore the FDvg of LDMC correlates closely with thevariation in biomass production around the peak of herbagemass a variable of importance for managers which givesinformation on the temporal flexibility of herbage use(Ansquer et al 2009b) On the basis of these results a setof easily recognised species with similar functional traits(LDMC and phenology) can be used to infer ecosystemprocesses of agricultural importance (Ansquer et al 2009a)

Other functions of grasslands can also be deduced fromtrait values For example in Australia the role of grass-lands to limit soil erosion has been negatively related totraits determining high production achieved there byreduced plant cover with low persistence (McIntyre 2008)

A major question for research is to define whichpractices may favour the constancy of ecosystem processesof grasslands under global changes This challenge is ofimportance knowing that a significant number of traditionallsquostress-tolerantrsquo grassland species have been replaced overthe last century by more ruderal species with faster growthand shorter life span (Van Calster et al 2008 Walker et al2009) with probable major impact on future primaryproduction of grasslands

62 Crop weed communities

Over the last decades the weed flora had dramaticallychanged in response to shift in management practices Onthe one hand new rules for integrated weed managementaiming at reducing the use of herbicides and preserving thebiodiversity of agroecosystems is currently leading to theemergence of more diverse communities On the otherhand there are still highly specialised crops in which aspecialised weed flora occurs in response to specifictechniques and pesticide use Since a decrease in the useof herbicides is likely to occur in such systems for bothtechnical and social reasons (Aubertot et al 2007 Wezel et


al 2009) the switch to practices other than pesticideapplications may well result in an increasing diversity ofweed species The use of traits may be an opportunity todetect groups of weedy species which similarly respond toa set of practices or affect crop yield

Major variations in weed composition between fieldsare associated with human management factors currentcrop type preceding crop type soil pH and texture landdrainage (Fried et al 2008 Butler et al 2009 Hawes etal 2009) Crop type includes the effects of sowing seasonherbicide families and fertilisation regimes which inducespecific light conditions affecting growth phenology(Lososova et al 2008) It is noticeable that differences inweed diversity between genetically modified herbicide-tolerant crop varieties and conventional varieties are muchlower than between different crop species (Freckleton etal 2003) Within individual crop types the landscapeorganisation (occurrence of hedges and meadows) andortillage depth are also of importance (Smith 2006 Butler etal 2009) Across all phases of the rotation however weedcommunity structure is affected more by the currentidentity of the crop than by the number of crops in therotation (Smith and Gross 2007)

New flora responding to current change in managementcan be characterised using the same set of traits as thosemeasured on spontaneous vegetation (Booth and Swanton2002) Weed species selected against over the past 60 yearsin cereal and soybean crops in UK have a short staturelarge seeds and flower late (Storkey et al 2010) As aconsequence exclusion from intensive systems over time isexplained by low competitive ability in high fertilitytreatment in relation to low stature and low resilience toherbicide influence due to low reproductive output Bycontrast the current weed flora occurring in intensivesystems is highly specialised in terms of resource usephenology and tolerance to herbicides For example weedsincreasing in abundance in sunflower crops since the 1970sbelong to a lsquosunflower-mimickingrsquo functional group theyare more nitrophilous and heliophilous less sensitive tosunflower herbicides and share a rapid summer life cycleindependently of phylogeny (Fried et al 2009) Character-isation of these responses by traits is not always straight-forward Herbicide tolerance is partly explained by leafanatomy and architecture cuticle thickness and phenologybut also depends on physiological characteristics that are noteasily measurable Plant response to sowing time or tillageincludes germination syndrome and life cycle both of whichinfluence how species respond to changes in soil resourcelevels and light availability driven by seasonal disturbanceregime (Smith 2006) On the other hand the weed flora inextensively managed systems where disturbances are irreg-ular and of varying spatial extent is composed of less-specialised species sharing characteristics with colonisers or

ruderal species (Lososova et al 2006) they resemble thespecies with short stature large seeds and late flowering thatoccurred in cultivated areas years ago and have beenexcluded by intensification (Storkey et al 2010)

The different flora found under contrasted weed man-agement also differ in FD the reduction of FD in intensivecropping systems may occur through increased specialisa-tion rather than by further enhanced spread of the mostgeneralist species (Fried et al 2009) whereas FD inextensive places increases in relation to the occurrence ofless-specialised species (Lososova et al 2006)

Combining weed and crop characteristics is a furtherstep to assess the impact of the FD of the weedndashcropcommunity on EP Weed size relative to that of the cropis a major parameter impacting final crop yield in cropndashweed models (eg Kropff et al 1992 Booth and Swanton2002) In addition recent research suggests that the abilityof weeds to draw on resource pools other than that ofcrops which can be assessed by differences in root depthand architecture (Deen et al 2003 Smith et al 2009) maybe related to lower weedndashcrop competition (Smith et al2009) This means that the FDvg of the cropndashweedcommunity defined with traits related to resource usecould be of major importance to predict crop yield a highFDvg-inducing complementary in resource use by weedsand crop across time or space should result in a reducedimpact of weeds on crops

Weeds can also have an impact on the local biodiversityIn extensively managed areas the diversity of the localweed flora has been shown to correlate negatively withherbivore load and outbreaks of pests and positively withnumber of non-pest invertebrate species (Hawes et al2005) A recent study showed that functional types ofweeds defined by categorical traits can be used to assessthat of invertebrates (Hawes et al 2009) However asimilar study performed in UK showed that a number ofweeds that have a high importance for invertebrates andbirds are classified as noxious considering their largenegative impact on crops (Storkey 2006) which shouldrefrain farmers to conserve them in non-extensive areas

63 Forests

The utility of traits for forest management in temperateareas has not been clearly established yet as most forestsare composed of one or a few species taken among a set ofapproximately ten well-known species The recent empha-sis on other services than biomass production (regulation ofclimate and flood support for biogeochemical cycles andbiodiversity supplying of recreational values etc) whichcannot be achieved through traditionally managed forestsinitiated the proposal of more lsquonature-orientatedrsquo manage-ment This on-going change is characterised by the


increasing importance of natural regeneration that can leadto the establishment of mixed forests As a consequencefuture management of forests will require knowledge onspecies different from the original set in order to predictsuccessful establishment and biomass production of mixedstands

The identification of traits linked to the successfulestablishment of mixed forests from natural regenerationis an on-going effort and shows that shade tolerancedetected by wood density and SLA are probably of majorimportance (Janse-ten Klooster et al 2007) In trees mostcomparative experiments have been performed in tropicalspecies (Poorter and Rose 2005 Muller-Landau et al2008) This implies that relationships between seedlingestablishment and adult traits are not clearly established formost species found in temperate environments (but seeKaelke et al 2001 Stancioiu and OrsquoHara 2006)

The recent damages caused to forests by extremeclimatic events such as storms and severe droughtsdemonstrate the urgent need for large comparisons ofspecies according to functions which have not beencharacterised so far The identification of relevant andeasily measured traits related to drought wind toleranceand root anchorage that could be used to monitor routineselection of species appears as an important research needin this respect (Cochard et al 2008 Hallik et al 2009Stokes et al 2009)

The likely increasing importance of mixed forests andthe search for higher resistance to disturbances (Vila et al2007) is an opportunity to test the usefulness of traits forforest management This could be done by building onprevious work performed in tropical areas where the highspecies diversity of forests was a strong incentive toidentify functional groups for management of biomassproduction and maintenance of species diversity (Vanclayet al 1997 Gourlet-Fleury et al 2005 Kooyman andRossetto 2008)

7 Trait databases and ecoinformatics

71 Availability of trait data

The methodology of screening for traits which takes itsroots in the seminal work by Grime and Hunt (1975) forRGR has progressively built up over the last three decadesThis has led to the constitution of large data sets of planttraits allowing us to address general questions pertaining toplant functioning and evolution This includes the detectionof (1) broad relationships among traits such as allometricscaling (West et al 1999 Enquist and Niklas 2002 Reichet al 2006) and trade-offs (Wright et al 2004 Bolmgrenand Cowan 2008) at various levels of organisation (2) the

phylogenetic signal on traits (Moles et al 2005a Watanabeet al 2007 Flores et al submitted manuscript) (3) therelationships between plant traits and environmental factorsat continental or global scales (Reich and Oleksyn 2004Moles et al 2005a Pakeman et al 2008 Moles et al 2009Ordontildeez et al 2009)

Although these data sets constitute an invaluable wealthfor the scientific community as a whole their compatibilityaccessibility and further dissemination remain problematicSome data sets are available as appendices to originalpublications (eg the Glopnet data set Wright et al 2004the wood data set Chave et al 2009) or as ecologicalarchives (eg fire-related traits of species from theMediterranean basin Paula et al 2009) and severalattempts have been made to develop publicly accessibleplant trait databases (CLO-PLA2 Klimes and Klimesova1999 BiolFlor Kuumlhn et al 2004 LEDA Kleyer et al2008) However this concerns only a small proportion ofthe data potentially available and a comprehensive ap-proach to data storing management and retrieval for plantbased research is currently lacking This involves devotingspecific efforts to data integration whose aim is to organiseand structure the data currently stored in distributedrepositories and in heterogeneous formats and make themavailable in easily and widely usable format Beyondintegration such efforts would provide substantial addedvalue to the data which could then also be used in contextsdifferent from those in which these were initially collectedfor or for other more general questions These issues haveled to the recent development of ecoinformatics and therealisation of the importance of metadata and ontologies asapplied to the field of ecology

72 Towards a trait-based ecoinformatics

This issue as applied to traits falls under the broaderperspective of what is now called lsquoecoinformaticsrsquo definedas lsquoa field of research and development focused on theinterface between ecology computer science and informa-tion technologyrsquo and considered as the new bioinformatics(Jones et al 2006) The recent development of ecoinfor-matics results from the acknowledgement that data sharingand integration should become leading priorities in ecologyas this has been recognised three decades ago in molecularbiology (Stein 2008)

Madin et al (2008) have proposed an architecture forecological and environmental data management involvingthree levels The first level is composed of data storedwithin distributed data repositories while the third higherlevel is composed of software tools used by scientists anddata managers to work on the data Madin et al (2008)stress that an intermediate level is required between thesetwo in order to develop tools that are more effective for


querying integrating analysing and publishing data fromthe raw material available in data repositories Metadatastandards and ontologies are two key elements of this level

Metadata may be defined as lsquoinformation about datarsquo iethe information required to understand data including data setcontents context structure and accessibility (Michener et al1997) Examples of metadata include the identification ofwho conducted an experiment (name of observer institutionstorage location and form etc) where (latitude longitudealtitude etc) and how it was conducted (measuring deviceprecision etc) and so on A number of benefits are attachedto the production of metadata in addition to data (Scurlock etal 2002 Michener 2006) These include increased datalongevity facilitation of data reuse by those who haveproduced the data data sharing with others data mining forpurposes and questions other than those for which datacollection was initially designed for This is all the moreimportant that databases may outlive the original investigatorand that ecological research requires integration and synthe-sis from a wide range of disciplines (Jones et al 2006Michener 2006) Several metadata content specificationscould be used to document ecological data (Michener et al1997) but at this time there is no generally accepted formatfor metadata standards in ecology (Scurlock et al 2002) andtheir use is not widespread This certainly applies to the fieldof trait-based research which needs specific efforts toimprove its practices in this respect

A second aspect concerns the development of core andspecialised ontologies a key step in the formalisation ofconcepts in a particular field of research In the informationsciences ontology is a formal model of knowledgecovering a specified discipline in which (1) each element(eg field name or column in a database) is preciselydefined and (2) each possible relationship between dataelements is parameterised or constrained (Schuurman andLeszczynski 2008) Owing to the formal logical structure ofontologies computer-based reasoning systems can usethem to draw inferences or conclusions enabling ontologiesto help identify important aspects of a data set that werehidden or implicit (Jones et al 2006) For example if one islooking for data on SLA it might be stored in distributeddatabases under a variable named lsquoSLArsquo or its inverselsquoLMArsquo Both variables will first have to be preciselydefined in a controlled vocabulary and an ontology wouldcapture the idea that SLA=1LMA Any query addressed tothe distributed databases through the ontology filter willthen extract automatically data stored under the differentvariable headings Some general ontologies have beendeveloped for ecology such as eg OBOE (ExtensibleObservation Ontology Madin et al 2007) which providesecologists with a formal framework for ecological obser-vations and associated measures and protocols A primaryobjective is thus to develop a domain ontology relevant for

a trait-based approach to ecology compatible with high-level ontologies such as OBOE This can be done usingrelated specialised ontologies such as plant ontology(Jaiswal et al 2002) or plant growth stage ontology (Pujaret al 2006) as starting points

This whole approach to data management constitutes afirst step towards the long-term objective of ecological dataintegration and sharing through the development ofinformation systems comparable to those currently usedfor gene sequences and protein data (GenBank EMBL) ortaxonomic diversity (GBIF Guralnick et al 2007)

8 Conclusions and perspectives

In this review we have highlighted a number of strengthsof the trait-based approach to address questions pertainingto plant and ecosystem functioning community dynamicsand species distribution in ecological contexts and someselected agronomical situations Although the focus was onplants there is now increasing recognition that traits mightbe useful tools to address questions relative to otherorganisms as well be they invertebrate or vertebrateanimals (cf de Bello et al 2010) or microorganisms (Greenet al 2008) This functional characterisation of biodiversitythat traits make possible enables us to bring insights whichgo well beyond what is possible when species are merelydescribed by Latin binomials

Several types of factors were decisive to the majorbreakthroughs in plant trait research that have taken placeover the past 15 years First the imperious need to predictchanges in vegetation in response to global change drivershas triggered the search for simplification of floristiccomplexity second the identification of a small set ofeasily measured traits relevant to plant functioning anddistribution that could be measured on a wide array ofspecies and situations worldwide was decisive third a setof consistent concepts and hypotheses allowing us to scalefrom the individual to higher levels of organisation provedvery powerful fourth the design of protocols to measuretraits in a standardised way and the recognition that sharingdata and information among scientists working in differentsystems were essential to progress

Challenges ahead to trait-based research pertain to anumber of aspects In terms of plant functioning whethereg further axes of variation can be detected in plantsincluding some involving root dimensions and whether anyeasily measured traits can be related to the various aspectsof the regeneration niche (sensu Grubb 1977) and todemographic parameters such as population birth and deathrate appear as central issues At the community levelunderstanding the mechanisms through which species traitsas driven by environmental factors and biotic interactions


determine its functional structure will require theoreticalexperimental and modelling approaches Whether phylog-eny has a prominent role in these mechanisms is currently ahighly debated topic (eg Cavender-Bares et al 2009) Thiswill also require a search for a better identification andquantification of the environmental gradients driving traitresponses There are some initiatives in this direction (egKnevel et al 2005 Garnier et al 2007 Bartholomeus et al2008 Ordontildeez et al 2009) but a general agreement on aminimum set of required environmental variables andstandardised protocols to measure them is still lacking Atthe ecosystem level the understanding of how and whichtraits (in terms of values and ranges) affect processesproperties and services remains very preliminary It is alsoessential to recognise that simultaneous effects on multiplelinked ecosystem processes and services are involved(reviewed in de Bello et al 2010) and that scaling fromtraits to services requires innovating approaches allowingus to bridge the gap between biological and social sciences(cf de Chazal et al 2008) Predictions at all scales will alsorequire an increased coupling to modelling approachesfrom ecophysiologicalecosystem levels (eg Martineau andSaugier 2007 van Wijk 2007) to that of the wholebiosphere (eg Prentice et al 2007) Last but not leastfurther progress will depend on the continued developmentof trait databases and information systems allowingscientists from different disciplines (taxonomy ecologyenvironmental sciences etc) to share and combine theirdata to tackle broad-scale and complex questions inecology

The prospects to apply a comparative approach based ontraits in the field of agronomy are wide Intensiveagricultural systems where crop diversity is reduced toone or very few species that are generally geneticallyhomogeneous and where external inputs are often suppliedin large quantities are increasingly criticised for theirnegative environmental impacts (Giller et al 1997 Auber-tot et al 2007 Le Roux et al 2009 Wezel et al 2009Griffon 2010) Reducing these impacts will involvereducing eg the use of fertilisers andor pesticides andrely more on regulations involving a wider range of diverseorganisms (eg soil organisms for the control of fertilityand non-pest insectsmicroorganisms for the biologicalcontrol of pests) As shown in the case of grasslands(Section 61) trait-based tools might be used for themanagement of these more complex systems where adetailed mechanistic and modelling approach based on in-depth knowledge of all organisms involved will probablynot be tractable At the other end of the spectrumMaleacutezieux et al (2009) advocated the use of functionaltraits (or groups) to characterise and model complex multi-species cropping systems such as agroforestry systems onwhich a majority of the worldrsquos farmer depend particularly

in tropical regions (Swift and Anderson 1993 Maleacutezieux etal 2009) For this transfer from ecological to agronomicalscience to succeed the specificities of these variousagricultural systems such as the identity and strength ofthe selective pressures acting in these environments (egtypes and return times of specific disturbances such astillage andor selective pesticide applications) will have tobe identified at the same time as the responses of theorganisms themselves which may not have been describedas yet in less artificial situations Le Roux et al (2009chapter 1) have discussed these issues in the special case ofthe impacts of agricultural practices on biodiversity

As nicely put by Westoby et al (2002) lsquoThere is muchto be done There is also a real hope that we may be gettingsomewherersquo And getting somewhere will involve contin-ued and enhanced collaboration across disciplines rangingfrom ecophysiology to macroecology crossing experimen-tal and theoretical approaches as applied to natural andman-modified systems

Acknowledgments We thank Guillaume Fried Hendrik Davy andJean-Claude Geacutegout for providing us with references illustrating theuse of traits for weed and forest management Isabelle Mougenot andCyrille Violle critically reviewed several parts of the manuscript Thispublication has benefited from the continuous exchanges of peopleand ideas in the framework of GDR 2574 lsquoTRAITSrsquo (CNRS France)

References

Aber JD Mellilo JM (2001) Terrestrial ecosystems BrooksColePacific Grove

Ackerly DD (2003) Community assembly niche conservatism andadaptive evolution in changing environments Int J Plant Sci 164(3)S165ndashS184 doi101086368401

Ackerly D (2004) Functional strategies of chaparral shrubs in relationto seasonal water deficit and disturbance Ecol Monog 7425ndash44doi10189003-4022

Ackerly DD Cornwell WK (2007) A trait-based approach tocommunity assembly partitioning of species trait values intowithin- and among-community components Ecol Lett 10135ndash145 doi101111j1461-0248200601006x

Ackerly DD Monson RK (2003) Waking the sleeping giant theevolutionary foundations of plant functions Int J Plant Sci 164(3)S1ndashS6 doi101086374729

Ackerly DD Schwilk DW Webb CO (2006) Niche evolution andadaptive radiation testing the order of trait divergence Ecology87(7)S50ndashS61 doi1018900012-9658(2006)87[50NEAART]20CO2

Aerts R (1997) Climate leaf litter chemistry and leaf litterdecomposition in terrestrial ecosystems a triangular relationshipOikos 79439ndash449

Aerts R Chapin FS III (2000) The mineral nutrition of wild plantsrevisited a re-evaluation of processes and patterns Adv Ecol Res301ndash67 doi101016S0065-2504(08)60016-1

Al Haj Khaled R Duru M Decruyenaere V Jouany C Cruz P (2006)Using leaf traits to rank native grasses according to their nutritivevalue Rangeland Ecol Manage 59648ndash654 doi10211105-031R21


Albert CH Thuiller W Yoccoz NG Soudant A Boucher F SacconeP Lavorel S (2010) Intraspecific functional variability extentstructure and sources of variation J Ecol 98604ndash613doi101111j1365-2745201001651x

Altesor A Oesterheld M Leoni E Lezama F Rodrigraveguez C (2005)Effect of grazing exclosure on community structure andproductivity of a Uruguayan grassland Plant Ecol 17983ndash91doi101007s11258-004-5800-5

Ansquer P Duru M Theau JP Cruz P (2009a) Convergence in planttraits between species within grassland communities simplifiestheir monitoring Ecol Indic 9(5)1020ndash1029 doi101016jecolind200812002

Ansquer P Duru M Theau JP Cruz P (2009b) Functional traits asindicators of fodder provision over a short time scale in species-rich grasslands Ann Bot 103(1)117ndash126 doi101093aobmcn215

Arnold SJ (1983) Morphology performance and fitness Am Zool23347ndash361 doi101093icb232347

Aubertot J-N Barbier J-M Carpentier A Gril J-J Guichard L LucasP Savary S Voltz M (eds) (2007) Pesticides agriculture etenvironnement Reacuteduire lrsquoutilisation des pesticides et en limiterles impacts environnementaux Quae Versailles

Baker TR Phillips OL Laurance WF Pitman NCA Almeida S ArroyoL DiFiore A Erwin T Higuchi N Killeen TJ Laurance SGNascimento H Monteagudo A Neill DA Silva JNM Malhi YLogravepez GG Peacock J Quesada CA Lewis SL Lloyd J (2009) Dospecies traits determine patterns of wood production in Amazonianforests Biogeosci 6297ndash307 doi105194bg-6-297-2009

Balvanera P Pfisterer AB Buchmann N He J-S Nakashizuka TRaffaelli D Schmid B (2006) Quantifying the evidence forbiodiversity effects on ecosystem functioning and services EcolLett 91146ndash1156 doi101111j1461-0248200600963x

Bartholomeus RP Witte J-PM van Bodegom PM Aerts R (2008) Theneed of data harmonization to derive robust empirical relation-ships between soil conditions and vegetation J Veg Sci 19799ndash808 doi1031702008-8-18450

Bazzaz FA Ackerly DD Reekie EG (2000) Reproductive allocationin plants In Fenner M (ed) Seeds The ecology of regenerationin plant communities CAB International Wallingford pp 1ndash29

Belyea LR Lancaster J (1999) Assembly rules within a contingentecology Oikos 86402ndash416

Berendse F Aerts R (1987) Nitrogen-use-efficiency a biologicallymeaningful definition Funct Ecol 1293ndash296

Bernhardt-Romermann M Romermann C Nuske R Parth A Klotz SSchmidt W Stadler J (2008) On the identification of the mostsuitable traits for plant functional trait analyses Oikos 117(10)1533ndash1541 doi101111j0030-1299200816776x

Bolmgren K Cowan PD (2008) Time-size tradeoffs a phylogeneticcomparative study of flowering time plant height ans seed massin a north-temperate flora Oikos 117424ndash429 doi101111j20070030-129916142x

Booth BD Swanton CJ (2002) Assembly theory applied to weedcommunities Weed Sci 50(1)2ndash13 doi1016140043-1745(2002)050[0002AIATAT]20CO2

Boulant N Kunstler G Rambal S Lepart J (2008) Seed supplydrought and grazing determine spatio-temporal patterns ofrecruitment for native and introduced invasive pines in grasslandsDivers Distrib 14862ndash874 doi101111j1472-4642200800494x

Bradshaw AD (1987) Functional ecology comparative ecologyFunct Ecol 171

Butler SJ Brooks D Feber RE Storkey J Vickery JA Norris K(2009) A cross-taxonomic index for quantifying the health offarmland biodiversity J Appl Ecol 46(6)1154ndash1162doi101111j1365-2664200901709x

Cadotte MW Cavender-Bares J Tilman D Oakley TH (2009) Usingphylogenetic functional and trait diversity to understand patterns

of plant community productivity PLoS ONE 4(5)e5695doi101371journalpone0005695

Cadotte MW Davies TJ Regetz J Kembel SW Cleland E OakleyTH (2010) Phylogenetic diversity metrics for ecological commu-nities integrating species richness abundance and evolutionaryhistory Ecol Lett 13(1)96ndash105 doi101111j1461-0248200901405x

Calow P (1987) Towards a definition of functional ecology FunctEcol 157ndash61

Cardinale BJ Srivastava DS Duffy JE Wright JP Downing ALSankaran M Jouseau C (2006) Effects of biodiversity on thefunctioning of trophic groups and ecosystems Nature 443989ndash992 doi101038nature05202

Cardinale BJ Wright JP Cadotte MW Carroll IT Hector ASrivastava DS Loreau M Weis JJ (2007) Impacts of plantdiversity on biomass production increase through time because ofspecies complementarity P Natl Acad Sci USA 10418123ndash18128 doi101073pnas0709069104

Cavender-Bares J Keen A Miles B (2006) Phylogenetic structure ofFloridian plant communities depends on taxonomic and spatialscale Ecology 87(7)S109ndashS122 doi1018900012-9658(2006)87[109PSOFPC]20CO2]

Cavender-Bares J Kozak KH Fine PVA Kembel SW (2009) Themerging of community ecology and phylogenetic biology EcolLett 12693ndash715 doi101111j1461-0248200901314x

Cebriaacuten J Duarte CM (1995) Plant growth-rate dependence of detritalcarbon storage in ecosystems Science 2681606ndash1608doi101126science26852171606

Chapin FS III (1980) The mineral nutrition of wild plants Annu RevEcol Syst 11233ndash260

Chapin FS III (1993) Functional role of growth forms in ecosystemand global processes In Ehleringer JR Field CB (eds) Scalingphysiological processes Leaf to globe Academic San Diego pp287ndash312

Chapin FS III Autumn K Pugnaire F (1993) Evolution of suites oftraits in response to environmental stress Am Nat 142S78ndashS92doi101086285524

Chapin FS III Zavaleta ES Eviner VT Naylor RL Vitousek PMReynolds HL Hooper DU Lavorel S Sala OE Hobbie SEMack MC Diacuteaz S (2000) Consequences of changing biodiver-sity Nature 405234ndash242 doi10103835012241

Chapin FS III Matson PA Mooney HA (2002) Principles of terrestrialecosystem ecology Springer New York

Chave J Andalo C Brown S Cairns MA Chambers JQ Eamus DFoumllster H Fromard F Higuchi N Kira T Lescure J-P NelsonBW Ogawa H Puig H Rieacutera B Yamakura T (2005) Treeallometry and improved estimation of carbon stocks and balancein tropical forests Oecologia 14587ndash99 doi101007s00442-005-0100-x

Chave J Coomes D Jansen S Lewis SL Swenson NG Zanne AE(2009) Towards a worldwide wood economics spectrum EcolLett 12351ndash366 doi101111j1461-0248200901285x

Cingolani AM Cabido M Gurvich DE Renison D Diacuteaz S (2007)Filtering processes in the assembly of plant communities arespecies presence and adundance driven by the same traits J VegSci 18911ndash920 doi01111j1654-11032007tb02607x

Cochard H Barigah ST Kleinhentz M Eshel A (2008) Is xylemcavitation resistance a relevant criterion for screening droughtresistance among Prunus species J Plant Physiol 165(9)976ndash982 doi101016jjplph200707020

Cornelissen JHC Thompson K (1997) Functional leaf attributespredict litter decomposition rate in herbaceous plants NewPhytol 135109ndash114 doi101046j1469-8137199700628x

Cornelissen JHC Peacuterez-Harguindeguy N Diacuteaz S Grime JP MarzanoB Cabido M Vendramini F Cerabolini B (1999) Leaf structureand defence control litter decomposition rate across species and


life forms in regional flora on two continents New Phytol143191ndash200 doi101046j1469-8137199900430x

Cornelissen JHC Cerabolini B Castro-Diacuteez P Villar-Salvador PMontserrat-Martiacute G Puyravaud JP Maestro M Werger MJAAerts R (2003a) Functional traits of woody plants correspon-dence of species rankings between field adults and laboratory-grown seedlings J Veg Sci 14311ndash322 doi101111j1654-11032003tb02157x

Cornelissen JHC Lavorel S Garnier E Diacuteaz S Buchmann NGurvich DE Reich PB ter Steege H Morgan HD van derHeijden MGA Pausas JG Poorter H (2003b) A handbook ofprotocols for standardised and easy measurement of plantfunctional traits worldwide Aust J Bot 51335ndash380doi101071BT02124

Cornwell WK Ackerly DD (2009) Community assembly and shifts inplant trait distributions across an environmental gradient incoastal California Ecol Monog 79109ndash126 doi10189007-11341

Cornwell WK Schwilk DW Ackerly DD (2006) A trait-based test forhabitat filtering convex hull volume Ecology 87(6)1465ndash1471doi1018900012-9658(2006)87[1465ATTFHF]20CO2

Cornwell WK Cornelissen JHC Amatangelo K Dorrepaal E EvinerVT Godoy O Hobbie SE Hoorens B Kurokawa H Peacuterez-Harguindeguy N Quested HM Santiago LS Wardle DA WrightIJ Aerts R Allison SD van Bodegom P Brovkin V Chatain ACallaghan TV Diacuteaz S Garnier E Gurvich DE Kazakou E KleinJA Read J Reich PB Soudzilovskaia NA Vaieretti MVWestoby M (2008) Plant traits are the dominant control on litterdecomposition rates within biomes worldwide Ecol Lett111065ndash1071 doi101111j1461-0248200801219x

Cortez J Garnier E Peacuterez-Harguindeguy N Debussche M Gillon D(2007) Plant traits litter quality and decomposition in aMediterranean old-field succession Plant Soil 29619ndash34doi101007s11104-007-9285-6

Craine JM (2009) Resource strategies of wild plants PrincetonUniversity Press Princeton

Craine JM Tilman D Wedin D Reich P Tjoelker M Knops J (2002)Functional traits productivity and effects on nitrogen cycling of33 grassland species Funct Ecol 16563ndash574 doi101046j1365-2435200200660x

Crawley MJ Johnston AE Silvertown J Dodd M de Mazencourt CHeard MS Henman DF Edwards GR (2005) Determinants ofspecies richness in the Park Grass Experiment Am Nat 165179ndash192 doi101086427270

de Bello F Thuiller W Leps J Choler P Clement JC Macek PSebastia MT Lavorel S (2009) Partitioning of functionaldiversity reveals the scale and extent of trait convergence anddivergence J Veg Sci 20(3)475ndash486 doi101111j1654-1103200901042x

de Bello F Lavorel S Diacuteaz S Harrington R Cornelissen JHCBardgett RD Berg MP Cipriotti P Feld CK Hering DMartins da Silva P Potts SG Sandin L Sousa JP Storkey JWardle DA Harrison PA (2010) Towards an assessment ofmultiple ecosystem processes and services via functional traitsBiodivers Conserv 192873ndash2893 doi101007s10531-010-9850-9

de Chazal J Queacutetier F Lavorel S Van Doorn A (2008) Includingmultiple differing stakeholder values into vulnerability assess-ments of socio-ecological systems Glob Environ Chang 18508ndash520 doi101016jgloenvcha200804005

de Deyn GB Cornelissen JHC Bardgett RD (2008) Plant functionaltraits and soil carbon sequestration in contrasting biomes EcolLett 11516ndash531 doi101111j1461-0248200801164x

Deen W Cousens R Warringa J Bastiaans L Carberry P Rebel KRiha S Murphy C Benjamin LR Cloughley C Cussans JForcella F Hunt T Jamieson P Lindquist J Wang E (2003) An

evaluation of four crop weed competition models using acommon data set Weed Res 43(2)116ndash129 doi101046j1365-3180200300323x

Diamond JM (1975) Assembly of species communities In Cody MLDiamond JM (eds) Ecology and evolution of communitiesBelknap Cambridge pp 342ndash444

Diacuteaz S Cabido M (2001) Vive la diffeacuterence plant functional diversitymatters to ecosystem processes Trends Ecol Evol 16646ndash655doi10123412345678

Diaz S Lavorel S McIntyre S Falczuk V Casanoves F MilchunasDG Skarpe C Rusch G Sternberg M Noy-Meir I Landsberg JZhang W Clark H Campbell BD (2007) Plant trait responses tograzingmdasha global synthesis Glob Change Biol 13(2)313ndash341doi101111j1365-2486200601288

Diacuteaz S Fargione J Chapin FS III Tilman D (2006) Biodiversity lossthreatens human well-being PLoS Biol 4(8)e277 doi101371journalpbio0040277

Diacuteaz S Lavorel S Chapin FS III Tecco PA Gurvich DE Grigulis K(2007a) Functional diversitymdashat the crossroads between ecosys-tem functioning and environmental filters In Canadell JGPataki DE Pitelka LF (eds) Terrestrial ecosystems in a changingworld Springer Berlin pp 81ndash91

Diacuteaz S Lavorel S de Bello F Queacutetier F Grigulis K Robson M(2007b) Incorporating plant functional diversity effects inecosystem service assessments P Natl Acad Sci USA10420684ndash20689 doi101073pnas0704716104

Duarte CM Sand-Jensen K Nielsen SL Enriacutequez S Agustiacute S(1995) Comparative functional ecology rationale and poten-tials Trends Ecol Evol 10418ndash421 doi101016S0169-5347(00)89163-6

Duru M Ducrocq H (1997) A nitrogen and phosphorus herbagenutrient index as a tool for assessing the effect of N and P supplyon the dry matter yield of permanent pastures Nutr CyclAgroecosyst 4759ndash69

Duru M Balent G Gibon A Magda D Theau JP Cruz P Jouany C(1998) Fonctionnement et dynamique des prairies permanentesExemple des pyrenees centrales Fourrages 15397ndash113

Eckstein RL Karlsson PS Weih M (1999) Leaf life span and nutrientresorption as determinants of plant nutrient conservation intemperatendasharctic regions New Phytol 143177ndash189doi101046j1469-8137199900429x

Enquist BJ Niklas KJ (2002) Global allocation rules for patterns ofbiomass partitioning in seed plants Science 2951517ndash1520doi101126science1066360

Enquist BJ Economo EP Huxman TE Allen AP Ignace DD GillolyJF (2003) Scaling metabolism from organisms to ecosystemsNature 423639ndash642 doi101038nature01671

Eviner VT Chapin FS III (2003) Functional matrix a conceptualframework for predicting multiple plant effects on ecosystemsAnnu Rev Ecol Syst 34455ndash485 doi101146annurevecolsys34011802132342

Evju M Austrheim G Halvorsen R Mysterud A (2009) Grazingresponses in herbs in relation to herbivore selectivity and planttraits in an alpine ecosystem Oecologia 16177ndash85 doi101007s00442-009-1358-1

Fayolle A (2008) Structure des communauteacutes de plantes herbaceacutees surles Grands Causses strateacutegies fonctionnelles des espegraveces etinteractions speacutecifiques PhD thesis SupAgro Montpellier (EDSIBAGHE) Montpellier p 225

Fenner M Thompson K (2005) The ecology of seeds CambridgeUniversity Press Cambridge

Flynn DFB Gogol-Prokurat M Nogeire T Molinari N RichersBT Lin BB Simpson N Mayfield MM DeClerck F (2009)Loss of functional diversity under land use intensificationacross multiple taxa Ecol Lett 12(1)22ndash33 doi101111j1461-0248200801255x


Fortunel C Garnier E Joffre R Kazakou E Quested H Grigulis KLavorel S Ansquer P Castro H Cruz P Dolezal J Eriksson OFreitas H Golodets C Jouany C Kigel J Kleyer M Lehsten VLepš J Meier T Pakeman R Papadimitriou M Papanastasis VPQueacutetier F Robson M Sternberg M Theau J-P Theacutebault AZarovali M (2009a) Leaf traits capture the effects of land usechanges and climate on litter decomposability of grasslandsacross Europe Ecology 90598ndash611 doi10189008-04181

Fortunel C Violle C Roumet C Buatois B Navas M-L Garnier E(2009b) Allocation strategies and seed traits are hardly affectedby nitrogen supply in 18 species differing in successional statusPerspect Plant Ecol Evol Syst 11267ndash283 doi101016jppees200904003

Freckleton RP Sutherland WJ Watkinson AR (2003) Deciding thefuture of GM crops in Europe Science 302(5647)994ndash996doi101126science1092671

Freschet GT Cornelissen JHC van Logtestijn RSP Aerts R (2010)Evidence of the lsquoplant economics spectrumrsquo in a subarctic flora JEcol 98362ndash373 doi101111j1365-2745200901615x

Fried G Norton LR Reboud X (2008) Environmental and manage-ment factors determining weed species composition and diversityin France Agr Ecosyst Environ 12868ndash76 doi101016jagee200805003

Fried G Chauvel B Reboud X (2009) A functional analysis of large-scale temporal shifts from 1970 to 2000 in weed assemblages ofsunflower crops in France J Veg Sci 20(1)49ndash58 doi101111j1654-1103200905284x

Fukami T Bezemer TM Mortimer SR van der Putten WH (2005)Species divergence and trait convergence in experimental plantcommunity assembly Ecol Lett 8(12)1283ndash1290 doi101111j1461-0248200500829x

Garnier E Aronson J (1998) Nitrogen use efficiency from leaf to standlevel clarifying the concept In Lambers H Poorter H vanVuuren MMI (eds) Inherent variation in plant growth Physio-logical mechanisms and ecological consequences BackhuysLeiden pp 515ndash538

Garnier E Laurent G (1994) Leaf anatomy specific mass and watercontent in congeneric annual and perennial grass species NewPhytol 128725ndash736 doi101111j1469-81371994tb04036x

Garnier E Laurent G Bellmann A Debain S Berthelier P Ducout BRoumet C Navas M-L (2001a) Consistency of species rankingbased on functional leaf traits New Phytol 15269ndash83doi101046j0028-646x200100239x

Garnier E Shipley B Roumet C Laurent G (2001b) A standardizedprotocol for the determination of specific leaf area and leaf drymatter content Funct Ecol 15688ndash695 doi101046j0269-8463200100563x

Garnier E Cortez J Billegraves G Navas M-L Roumet C Debussche MLaurent G Blanchard A Aubry D Bellmann A Neill CToussaint J-P (2004) Plant functional markers capture ecosystemproperties during secondary succession Ecology 85(9)2630ndash2637 doi10189003-0799

Garnier E Lavorel S Ansquer P Castro H Cruz P Dolezal JEriksson O Fortunel C Freitas H Golodets C Grigulis KJouany C Kazakou E Kigel J Kleyer M Lehsten V Lepš JMeier T Pakeman R Papadimitriou M Papanastasis VPQuested H Queacutetier F Robson M Roumet C Rusch G SkarpeM Sternberg M Theau J-P Theacutebault A Vile D Zarovali M(2007) Assessing the effects of land use change on plant traitscommunities and ecosystem functioning in grasslands astandardized methodology and lessons from an application to11 European sites Ann Bot 99967ndash985 doi101093aobmcl215

Gaudet CL Keddy PA (1988) A comparative approach to predictingcompetitive ability from plant traits Nature 334242ndash243doi101038334242a0

Gilbert B Lechowicz MJ (2004) Neutrality niches and dispersal in atemperate forest understory P Natl Acad Sci USA 101(20)7651ndash7656 doi101073pnas0400814101

Giller KE Beare MH Lavelle P Izac MN Swift MJ (1997) Agriculturalintensification soil biodiversity and agroecosystem function ApplSoil Ecol 63ndash16 doi101016S0929-1393(96)00149-7

Gitay H Noble IR (1997) What are funtional types and how shouldwe seek them In Smith TM Shugart HH Woodward FI (eds)Plant funtional types Their relevance to ecosystem propertiesand global change Cambridge University Press Cambridge pp3-19

Gourlet-Fleury S Blanc L Picard N Sist P Dick J Nasi R SwaineMD Forni E (2005) Grouping species for predicting mixedtropical forest dynamics looking for a strategy Ann For Sci62785ndash796 doi101051forest2005084

Gower ST (2002) Productivity of terrestrial ecosystems In MooneyHA Canadell JG (eds) The Earth system biological andecological dimensions of global environmental change WileyNew York pp 516ndash521

Grace JB Anderson TM Smith MD Seabloom E Andelman SJMeche G Weiher E Allain LK Jutila H Sankaran M Knops JRitchie M Willig MR (2007) Does species diversity limitproductivity in natural grassland communities Ecol Lett10680ndash689 doi101111j1461-0248200701058x

Green JL Bohannan BJM Whitaker RJ (2008) Microbial biogeogra-phy from taxonomy to traits Science 3201039ndash1043doi101126science1153475

Greene DF Johnson EA (1989) A model of wind dispersal of wingedor plumed seeds Ecology 70339ndash347 doi1023071937538

Griffon M (2010) Pour des agricultures eacutecologiquement intensivesEditions de lrsquoAube La Tour drsquoAigues

Grime JP Hunt R (1975) Relative growth-rate its range and adaptivesignificance in a local flora J Ecol 63393-422 httpwwwjstororgstable2258728

Grime JP (1977) Evidence for the existence of three primary strategiesin plants and its relevance to ecological and evolutionary theoryAm Nat 1111169ndash1194 doi101086283244

Grime JP (1979) Plant strategies and vegetation processes WileyChichester

Grime JP (1997) Biodiversity and ecosystem function the debate deepensScience 2771260ndash1261 doi101126science27753301260

Grime JP (1998) Benefits of plant diversity to ecosystems immediatefilter and founder effects J Ecol 86902ndash910 doi101046j1365-2745199800306x

Grime JP (2001) Plant strategies vegetation processes and ecosystemproperties Wiley Chichester

Grime JP (2006) Trait convergence and trait divergence in herbaceousplant communities mechanisms and consequences J Veg Sci 17(2)255ndash260 doi101111j1654-11032006tb02444x

Grime JP Anderson JM (1986) Environmental controls over organismactivity In Van Cleve K Chapin FSI Flanagan PW Viereck LADyrness CT (eds) Forest ecosystems in Alaskan taiumlga a synthesisof structure and function Springer Berlin pp 89ndash95

Grime JP Thompson K Hunt R Hodgson JG Cornelissen JHCRorison IH Hendry GAF Ashenden TW Askew AP Band SRBooth RE Bossard CC Campbell BD Cooper JEL DavisonAW Gupta PL Hall W Hand DW Hannah MA Hillier SHHodkinson DJ Jalili A Liu Z Mackey JML Matthews NMowforth MA Neal AM Reader RJ Reiling K Ross-Fraser WSpencer RE Sutton F Tasker DE Thorpe PC Whitehouse J(1997) Integrated screening validates primary axes of specialisa-tion in plants Oikos 79259ndash281

Gross N Robson TM Lavorel S Albert C Le Bagousse-Pinguet YGuillemin R (2008) Plant response traits mediate the effects ofsubalpine grasslands on soil moisture New Phytol 180652ndash662doi101111j1469-8137200802577x


Grubb PJ (1977) The maintenance of species-richness in plantcommunities the importance of the regeneration niche BiolRev 52107ndash145 doi101111j1469-185X1977tb01347x

Guralnick RP Hill AP Lane M (2007) Towards a collaborative globalinfrastructure for biodiversity assessment Ecol Lett 10663ndash672doi101111j1461-0248200701063x

Hallik L Niinemets U Wright IJ (2009) Are species shade anddrought tolerance reflected in leaf-level structural and func-tional differentiation in Northern Hemisphere temperate woodyflora New Phytol 184(1)257ndash274 doi101111j1469-8137200902918x

Harper JL Lovell PH Moore KG (1970) The shapes and sizes ofseeds Annu Rev Ecol Syst 1327ndash356

Harrison SP Prentice IC Barboni D Kohfeld KE Ni J Sutra JP(2010) Ecophysiological and bioclimatic foundations for a globalplant functional classification J Veg Sci 21(2)300ndash317doi101111j1654-1103200901144x

Hawes C Begg GS Squire GR Iannetta PPM (2005) Individuals asthe basic accounting unit in studies of ecosystem functionfunctional diversity in shepherdrsquos purse Capsella Oikos 109(3)521ndash534 doi101111j0030-1299200513853x

Hawes C Haughton AJ Bohan DA Squire GR (2009) Functionalapproaches for assessing plant and invertebrate abundancepatterns in arable systems Basic Appl Ecol 10(1)34ndash42doi101016jbaae200711007

Heisse K Roscher C Schuhmacher J Schulze E-D (2007) Establish-ment of grassland species in monocultures different strategieslead to success Oecologia 152435ndash447 doi101007s00442-007-0666-6

Hendry GAF Grime JP (eds) (1993) Methods in comparative plantecology Chapman amp Hall London

Hobbie SE Oleksyn J Eissenstat DM Reich PB (2010) Fine rootdecomposition rates do not mirror those of leaf litter amongtemperate tree species Oecologia 162505ndash513 doi101007s00442-009-1479-6

Hodgson JG Montserrat-Martiacute G Cerabolini B Ceriani RMMaestro-Martiacute M Peco B Wilson PJ Thompson K Grime JPBand SR Bogard A Castro-Diacuteez P Charles M Jones G Peacuterez-Rontomeacute MC Caccianiga M Alard D Bakker JP CornelissenJHC Dutoit T Grootjans AP Guerrero-Campo J Gupta PLHynde A Kahmen S Poschlod P Romo-Diacuteez A Rorison IHRoseacuten E Schreiber K-F Tallowin J de Torres EL Villar-Salvador P (2005) A functional method for classifying Europeangrasslands for use in joint ecological and economic studies BasicAppl Ecol 6119ndash131 doi101016jbaae200501006

Hooper DU Chapin FS III Ewel JJ Hector A Inchausti P Lavorel SLawton JH Lodge DM Loreau M Naeem S Schmid B SetaumllaumlH Symstad AJ Vandermeer J Wardle DA (2005) Effects ofbiodiversity on ecosystem functioning a consensus of currentknowledge Ecol Monog 753ndash35 doi10189004-0922

Hughes L Dunlop M French K Leishman MR Rice B Rodgerson LWestoby M (1994) Predicting dispersal spectra a minimal set ofhypotheses based on plant attributes J Ecol 82933ndash950

Hummel I Vile D Violle C Devaux J Ricci B Blanchard A GarnierE Roumet C (2007) Relating root structure and anatomy towhole plant functioning the case of fourteen herbaceousMediterranean species New Phytol 173313ndash321 doi101111j1469-8137200601912x

Jaiswal P Ware D Ni J Chang K Zhao W Schmidt S Pan X ClarkK Teytelman L Cartinhour S Stein L McCouch SR (2002)Gramene development and integration of trait and geneontologies for rice Comp Funct Genomics 3132ndash136doi101002cfg156

Jakobsson A Eriksson O (2000) A comparative study of seed numberseed size seedling size and recruitment in grassland plants Oikos88494ndash502 doi101034j1600-07062000880304x

Janse-ten Klooster SH Thomas EJP Sterck FJ (2007) Explaininginterspecific differences in sapling growth and shade tolerance intemperate forests J Ecol 951250ndash1260 doi101111j1365-2745200701299x

Jones MB Schildhauer MP Reichman OJ Bowers S (2006) The newbioinformatics integrating ecological data from the gene to thebiosphere Annu Rev Ecol Syst 37519ndash544 doi101146annurevecolsys37091305110031

Kaelke CM Kruger EL Reich PB (2001) Trade-offs in seedlingsurvival growth and physiology among hardwood species ofcontrasting successional status along a light-availability gradientCan J Bot 31(9)1602ndash1616 doi101139cjfr-31-9-1602

Kahmen S Poschlod P (2008) Effects of grassland management onplant functional trait composition Agr Ecosyst Environ 128(3)137ndash145 doi101016jagee200805016

Kazakou E Garnier E Navas M-L Roumet C Collin C Laurent G(2007) Components of nutrient residence time and the leafeconomics spectrum in species from Mediterranean old-fieldsdiffering in successional status Funct Ecol 21235ndash245doi101111j1365-2435200601242x

Kazakou E Violle C Roumet C Pintor C Gimenez O Garnier E (2009)Litter quality and decomposability of species from a Mediterraneansuccession depend on leaf traits but not on nitrogen supply Ann Bot1041151ndash1161 doi101093aobmcp202

Keddy P (1992a) Assembly and response rules two goals forpredictive community ecology J Veg Sci 3157ndash164doi1023073235676

Keddy PA (1992b) A pragmatic approach to functional ecology FunctEcol 6621ndash626

Keith DA Holman L Rodoreda S Lemmon J Bedward M (2007)Plant functional types can predict decade-scale changes in fire-prone vegetation J Ecol 95(6)1324ndash1337 doi101111j1365-2745200701302x

Kerkhoff AJ Enquist BJ (2006) Ecosystem allometry the scalingof nutrient stocks and primary productivity across plantcommunities Ecol Lett 9419ndash427 doi101111j1461-0248200600888x

King DA (1990) The adaptive significance of tree height Am Nat135809ndash828 doi101086285075

Kleyer M Bekker RM Knevel IC Bakker JP Thompson KSonnenschein M Poschlod P van Groenendael JM Klimes LKlimesova J Klotz S Rusch GM Hermy M Adriaens DBoedeltje G Bossuyt B Dannemann A Endels P GotzenbergerL Hodgson JG Jackel AK Kuhn I Kunzmann D Ozinga WARomermann C Stadler M Schlegelmilch J Steendam HJTackenberg O Wilmann B Cornelissen JHC Eriksson OGarnier E Peco B (2008) The LEDA Traitbase a database oflife-history traits of the Northwest European flora J Ecol 96(6)1266ndash1274 doi101111j1365-2745200801430x

Klimes L Klimesova J (1999) CLO-PLA2mdasha database of clonalplants in Central Europe Plant Ecol 1419ndash19

Klumpp K Soussana J-F (2009) Using functional traits to predictgrassland ecosystem change a mathematical test of the response-and-effect trait approach Glob Change Biol 152921ndash2934doi101111j1365-2486200901905x

Knevel IC Bekker RM Kunzmann D Stadler M Thompson K (eds)(2005) The LEDA traitbase Collecting and measuring standardsof life-history traits of the Northern European flora University ofGroningen Groningen

Kooyman R Rossetto M (2008) Definition of plant functional groupsfor informing implementation scenarios in resource-limitedmulti-species recovery planning Biodivers Conserv 17(12)2917ndash2937 doi101007s10531-008-9405-5

Kraft NJB Valencia R Ackerly D (2008) Functional traits and niche-based tree community assembly in an Amazonian forest Science322580ndash582 doi101126science1160662


Kropff MJ Weaver SE Smits MA (1992) Use of ecophysiologicalmodels for cropndashweed interference relations amongst weeddensity relative time of weed emergence relative leaf area andyield loss Weed Sci 40296ndash301

Kuumlhn I Durka W Klotz S (2004) BiolFlormdasha new plant-trait databaseas a tool for plant invasion ecology Divers Distrib 10363ndash365doi1031702007-8-18349

Kuparinen A (2006) Mechanistic models for wind dispersal TrendsPlant Sci 11296ndash301 doi101016jtplants200604006

Larcher W (2003) Physiological plant ecology ecophysiology andstress physiology of functional groups Springer Berlin

Lavorel S Garnier E (2002) Predicting changes in communitycomposition and ecosystem functioning from plant traits revisit-ing the Holy Grail Funct Ecol 16545ndash556 doi101046j1365-2435200200664x

Lavorel S Mcintyre S Landsberg J Forbes TDA (1997) Plantfunctional classifications from general groups to specific groupsbased on response to disturbance Trends Ecol Evol 12(12)474ndash478 doi101016S0169-5347(97)01219-6

Lavorel S Diacuteaz S Cornelissen JHC Garnier E Harrison SP McIntyreS Pausas J Peacuterez-Harguindeguy N Roumet C Urcelay C (2007)Plant functional types are we getting any closer to the HolyGrail In Canadell J Pataki D Pitelka L (eds) Terrestrialecosystems in a changing world Springer Berlin pp 149ndash164

Lavorel S Grigulis K McIntyre S Williams NSG Garden D DorroughJ Berman S Quetier F Thebault A Bonis A (2008) Assessingfunctional diversity in the fieldmdashmethodology matters Funct Ecol22(1)134ndash147 doi101111j1365-2435200701339x

Le Roux X Barbault R Baudry J Burel F Doussan I Garnier EHerzog F Lavorel S Lifran R Roger-Estrade J Sarthou J-PTrommetter M (eds) (2009) Agriculture et biodiversiteacute Valoriserles synergies Quae Versailles

Leishman MR Westoby M (1998) Seed size and shape are not relatedto persistence in soil in Australia in the same way as in BritainFunct Ecol 12480ndash485 doi101046j1365-2435199800215x

Leishman MR Wright IJ Moles AT Westoby M (2000) Theevolutionary ecology of seed size In Fenner M (ed) Seeds theecology of regeneration in plant communities CAB Internation-al Wallingford pp 31ndash57

Lemaire G Gastal F (1997) N uptake and distribution in plantcanopies In Lemaire G (ed) Diagnosis of the nitrogen status inthe crops Springer Berlin pp 3ndash44

Leoni E Altesor A Paruelo JM (2009) Explaining patterns of primaryproduction from individual level traits J Veg Sci 20612ndash619doi101111j1654-1103200901080x

Lepš J de Bello F Lavorel S Berman S (2006) Quantifying andinterpreting functional diversity of natural communities practicalconsiderations matter Preslia 78481ndash501

Loreau M Naeem S Inchausti P Bengtsson J Grime JP Hector AHooper DU Huston MA Raffaelli D Schmid B Tilman DWardle DA (2001) Biodiversity and ecosystem functioningcurrent knowledge and future challenges Science 294804ndash808doi101126science1064088

Lortie CJ Aarssen LW (1999) The advantage of being tall higherflowers receive more pollen in Verbascum thapsus L (Scrophu-lariaceae) EcoScience 668ndash71

Lortie CJ Brooker RW Choler P Kikvidze Z Michalet RPugnaire FI Callaway RM (2004) Rethinking plant commu-nity theory Oikos 107(2)433ndash438 doi101111j0030-1299200413250x

Lososova Z Chytry M Kuhn I Hajek O Horakova V Pysek P TichyL (2006) Patterns of plant traits in annual vegetation of man-made habitats in central Europe Perspect Plant Ecol Evol 8(2)69ndash81 doi101016jppees200607001

Lososova Z Chytry M Kuehn I (2008) Plant attributes determin-ing the regional abundance of weeds on central European

arable land J Biogeogr 35(1)177ndash187 doi101111j1365-2699200701778x

Louault F Pillar VD Aufregravere J Garnier E Soussana J-F (2005) Planttraits and functional types in response to reduced disturbance in asemi-natural grassland J Veg Sci 16151ndash160 doi1016581100-9233(2005)016[0151PTAFTI]20CO2

Madin J Bowers S Schildhauer M Krivov S Pennington D Villa F(2007) An ontology for describing and synthesizing ecologicalobservation data Ecol Inf 2279ndash296 doi101016jecoinf200705004

Madin JS Bowers S Schildhauer MP Jones MB (2008) Advancingecological research with ontologies Trends Ecol Evol 23159ndash168 doi101016jtree200711007

Maleacutezieux E Crozat Y Dupraz C Laurans M Makowski D Ozier-Lafontaine H Rapidel B de Tourdonnet S Valantin-Morison M(2009) Mixing plant species in cropping systems concepts toolsand models A review Agron Sustain Dev 2943ndash62 doi101051agro2007057

Martineau Y Saugier B (2007) A process-based model of old fieldsuccession linking ecosystem and community ecology EcolModel 204399ndash419 doi101016jecolmodel200701023

Mason NWH MacGillivray K Steel JB Wilson JB (2003) An indexof functional diversity J Veg Sci 14(4)571ndash578 doi101111j1654-11032003tb02184x

Mason NWH Mouillot D Lee WG Wilson JB (2005) Functionalrichness functional evenness and functional divergence theprimary components of functional diversity Oikos 111112ndash118doi101111j0030-1299200513886x

McGill BJ Enquist BJ Weiher E Westoby M (2006) Rebuildingcommunity ecology from functional traits Trends Ecol Evol21178ndash185 doi101016jtree200602002

McIntyre S (2008) The role of plant leaf attributes in linking land use toecosystem function in temperate grassy vegetation Agr EcosystEnviron 128(4)251ndash258 doi101016jagee200806015

McIntyre S Lavorel S Landsberg J Forbes TDA (1999)Disturbance response in vegetationmdashtowards a global per-spective on functional traits J Veg Sci 10621ndash630doi1023073237077

McKane RB Johnson LC Shaver GR Nadelhoffer KJ Rastetter EBFry B Giblin AE Kielland K Kwiatkowski BL Laundre JAMurray G (2002) Resource-based niches provide a basis for plantspecies diversity and dominance in arctic tundra Nature 415(6867)68ndash71 doi101038415068a

McNaughton SJ Milchunas DG Franck DA (1996) How can netprimary productivity be measured in grazing ecosystemsEcology 77(3)974ndash977 doi1023072265518

Meinzer FC Andrade JL Goldstein G Holbrook NM Cavelier JWright SJ (1999) Partitioning of soil water among canopy treesin a seasonally dry tropical forest Oecologia 121293ndash301

Michener WK (2006) Meta-information concepts for ecological datamanagement Ecol Inf 13ndash7 doi101016jecoinf200508004

Michener WK Brunt JW Helly JJ Kirchner TB Stafford SG (1997)Nongeospatial metadata for the ecological sciences Ecol Appl7330ndash342 doi1018901051-0761(1997)007[0330NMFTES]20CO2]

Mitchell P Veneklaas E Lambers H Burgess S (2008) Using multipletrait associations to define hydraulic functional types in plantcommunities of south-western Australia Oecologia 158(3)385ndash397 doi101007s00442-008-1152-5

Mokany K Ash J (2008) Are traits measured on pot grown plantsrepresentative of those in natural communities J Veg Sci19119ndash126 doi1031702007-8-18340

Mokany K Ash J Roxburgh S (2008) Functional identity is moreimportant than diversity in influencing ecosystem processes in atemperate native grassland J Ecol 96884ndash893 doi101111j1365-2745200801395x


Moles AT Leishman MR (2008) The seedling as part of a plantrsquos lifehistory strategy In Leck MA Parker VT Simpson RL (eds)Seedling ecology and evolution Cambridge University PressCambridge pp 217ndash238

Moles AT Ackerly DD Webb CO Tweddle JC Dickie JB PitmanAJ Westoby M (2005a) Factors that shape seed mass evolutionP Natl Acad Sci USA 10210540ndash10544 doi101073pnas0501473102

Moles AT Ackerly DD Webb CO Tweddle JC Dickie JB WestobyM (2005b) A brief history of seed size Science 307576ndash580doi101126science1104863

Moles AT Warton DI Warman L Swenson NG Laffan SW ZanneAE Pitman A Hemmings FA Leishman MR (2009) Globalpatterns in plant height J Ecol 97(5)923ndash932 doi101111j1365-2745200901526x

Muller-Landau HC Wright SJ Calderon O Condit R Hubbell SP(2008) Interspecific variation in primary seed dispersal in atropical forest J Ecol 96(4)653ndash667 doi101111j1365-2745200801399x

Naeem S Bunker DA Hector A Loreau M Perrings C (eds) (2009)Biodiversity ecosystem functioning and human wellbeingmdashanecological and economic perspective Oxford University PressNew York City

Navas ML Violle C (2009) Plant traits related to competition how dothey shape the functional diversity of communities CommunityEcol 10(1)131ndash137 doi101556ComEc102009115

Navas ML Roumet C Bellmann A Laurent G Garnier E (2010)Suites of plant traits in species from different stages of aMediterranean secondary succession Plant Biol Stuttg 12(1)183ndash196 doi101111j1438-8677200900208x

Niinemets Uuml (2001) Global-scale climatic controls of leaf dry mass perarea density and thickness in trees and shrubs Ecology 82(2)453ndash469 doi1018900012-9658(2001)082[0453GSCCOL]20CO2]

Niklas KJ (1995a) Plant height and the properties of some herbaceousstems Ann Bot 75133ndash142 doi101006anbo19951004

Niklas KJ (1995b) Size-dependent allometry of tree height diameterand trunk-taper Ann Bot 75217ndash227 doi101006anbo19951015

Niklas KJ Enquist BJ (2001) Invariant scaling relationships forinterspecific plant biomass production rates and body size P NatlAcad Sci USA 982922ndash2997 doi101073pnas041590298

Ordontildeez JC van Bodegom PM Witte J-PM Wright IJ Reich PBAerts R (2009) A global study of relationships between leaftraits climate and soil measures of nutrient fertility Glob EcolBiogeogr 18137ndash149 doi101111j1466-8238200800441x

Oren O Sperry JS Katul GG Pataki DE Ewers BE Phillips N SchaumlferKVR (1999) Survey and synthesis of intra- and interspecific variationin stomatal sensitivity to vapour pressure deficit Plant Cell Environ221515ndash1526 doi101046j1365-3040199900513x

Pakeman RJ Garnier E Lavorel S Ansquer P Castro H Cruz PDoležal J Eriksson O Freitas H Golodets C Kigel J Kleyer MLepš J Meier T Papadimitriou M Papanastasis VP Quested HQueacutetier F Rusch G Sternberg M Theau J-P Theacutebault A Vile D(2008) Impact of abundance weighing on the response of seedtraits to climate and land use change J Ecol 96355ndash366doi101111j1365-2745200701336x

Paula S Arianoutsou M Kazanis D Tavsanoglu C Lloret F Buhk COjeda F Luna B Moreno JM Rodrigo A Espelta JM Palacio SFernaacutendez-Santos B Fernandes PM Pausas JG (2009) Fire-related traits for plant species of the Mediterranean BasinEcology 901420 doi10189008-13091

Pausas JG Bradstock RA Keith DA Keeley JE Fire Network GCTE(2004) Plant functional traits in relation to fire in crown-fireecosystems Ecology 85(4)1085ndash1100 doi10189002-4094

Peacuterez-Harguindeguy N Diacuteaz S Cornelissen JHC Vendramini FCabido M Castellanos A (2000) Chemistry and toughness

predict leaf litter decomposition rates over a wide spectrum offunctional types and taxa in central Argentina Plant Soil 21821ndash30 doi101023A1014981715532

Petchey OL Gaston KJ (2006) Functional diversity back to basicsand looking forward Ecol Lett 9741ndash758 doi101111j1461-0248200600924x

Petchey OL Hector A Gaston KJ (2004) How do different measuresof functional diversity perform Ecology 85(3)847ndash857doi10189003-0226

Pickett STA Bazzaz FA (1978) Organization of an assemblage ofearly successional species on a soil moisture gradient Ecology591248ndash1255 doi1023071938238

Plantureux S Bellon S Burel F Chauvel B Dajoz I Guy P LelievreV Ranjard L Roger-Estrade J Sarthou J-P Viaux P (2009)Prospective Agriculture Biodiversiteacute INRA Deacutepartement Envi-ronnement Agronomie Montpellier p 310

Pontes LDS Soussana J-F Louault F Andueza D Carregravere P (2007)Leaf traits affect the above-ground productivity and quality ofpasture grasses Funct Ecol 21844ndash853 doi101111j1365-2435200701316x

Poorter H Garnier E (2007) Ecological significance of inherentvariation in relative growth rate and its components In PugnaireFI Valladares F (eds) Functional plant ecology 2nd edn CRCBoca Raton pp 67ndash100

Poorter L Rose S (2005) Light-dependent changes in the relationshipbetween seed mass and seedling traits a meta-analysis for rainforest tree species Oecologia 142(3)378ndash387 doi101007s00442-004-1732-y

Prentice IC Bondeau A Cramer W Harrison SP Hickler T Lucht WSitch S Smith B Sykes MT (2007) Dynamic global vegetationmodeling quantifying terrestrial ecosystem responses to large-scale environmental change In Canadell J Pitelka LF Pataki D(eds) Terrestrial ecosystems in a changing world SpringerBerlin pp 175ndash192

Prior LD Eamus D Bowman DMJS (2003) Leaf attributes in theseasonally dry tropics a comparison of four habitats innorthern Australia Funct Ecol 17504ndash515 doi101046j1365-2435200300761x

Pugnaire F Valladares F (eds) (2007) Functional plant ecology CRCBoca Raton

Pujar A Jaiswal P Kellogg EA Ilic K Vincent L Avraham SStevens P Zapata F Reiser L Rhee SY Sachs MM Schaeffer MStein L Ware D McCouch SR (2006) Whole plant growth stageontology for angiosperms and its application in plant biologyPlant Physiol 142414ndash428 doi101104pp106085720

Quested H Eriksson O Fortunel C Garnier E (2007) Plant traits relateto whole-community litter quality and decomposition followingland use change Funct Ecol 211016ndash1026 doi101111j1365-2435200701324x

Queacutetier F Theacutebault A Lavorel S (2007) Plant traits in a state andtransition framework as markers of ecosystem response to land-use change Ecol Monog 7733ndash52 doi10189006-0054

Raunkiaer C (1934) The life forms of plants and statistical plantgeography Oxford University Press Oxford

ReesM (1997) Evolutionary ecology of seed dormancy and seed size InSilvertown J Franco M Harper JL (eds) Plant life historiesEcology phylogeny and evolution The Royal Society Cambridgepp 121ndash142

Rees M Venable DL (2007) Why do big plants make big seeds JEcol 95926ndash936 doi101111j1365-2745200701277x

Reich PB Oleksyn J (2004) Global patterns of plant leaf N and P inrelation to temperature and latitude P Natl Acad Sci USA10111001ndash11006 doi101073pnas0403588101

Reich PB Walters MB Ellsworth DS (1992) Leaf life-span in relationto leaf plant and stand characteristics among diverse ecosys-tems Ecol Monog 62365ndash392 doi1023072937116


Reich PB Walters MB Ellsworth DS (1997) From tropics to tundraglobal convergence in plant functioning Proc Nat Acad Sci USA9413730ndash13734

Reich PB Tjoelker MG Machado J-L Oleksyn J (2006) Universalscaling of respiratory metabolism size and nitrogen in plantsNature 439457ndash461 doi101038nature04282

Roche P Diacuteaz BN Gachet S (2004) Congruency analysis of speciesranking based on leaf traits which traits are the more reliablePlant Ecol 17437ndash48

Roderick ML Berry SL Noble IR Farquhar GD (1999) A theoreticalapproach to linking the composition and morphology with thefunction of leaves Funct Ecol 13683ndash695 doi101046j1365-2435199900368x

Romermann C Tackenberg O Jackel AK Poschlod P (2008)Eutrophication and fragmentation are related to speciesrsquo rate ofdecline but not to species rarity results from a functionalapproach Biodivers Conserv 17(3)591ndash604 doi101007s10531-007-9283-21

Roumet C Urcelay C Diacuteaz S (2006) Suites of root traits differbetween annual and perennial species growing in the field NewPhytol 170357ndash358 doi101111j1469-8137200601667x

Ryser P Urbas P (2000) Ecological significance of leaf life spanamong Central European grass species Oikos 9141ndash50doi101034j1600-07062000910104x

Sala OE Parton WJ Joyce LA Lauenroth WK (1988) Primaryproduction of the central grassland region of the United StatesEcology 6940ndash45 doi1023071943158

Sasaki T Okubo S Okayasu T Jamsran U Ohkuro T Takeuchi K(2009a) Management applicability of the intermediate distur-bance hypothesis across Mongolian rangeland ecosystems EcolAppl 19(2)423ndash432 doi10189008-18501

Sasaki T Okubo S Okayasu T Jamsran U Ohkuro T Takeuchi K(2009b) Two-phase functional redundancy in plant communitiesalong a grazing gradient in Mongolian rangelands Ecology 90(9)2598ndash2608 doi10189008-01441

Saugier B Roy J Mooney HA (2001) Estimations of global terrestrialproductivity converging towards a single number In Roy JSaugier B Mooney HA (eds) Terrestrial global productivityAcademic San Diego pp 543ndash557

Schenk HJ Jackson RB (2002) Rooting depths lateral root spreadsand below-groundabove-ground allometries of plants in water-limited ecosystems J Ecol 90480ndash494 doi101046j1365-2745200200682x

Schulze E-D Zwoumllfer H (eds) (1987) Potential and limitations inecosystem analysis Springer Berlin

Schulze E-D Leuning R Kelliher FM (1995) Environmentalregulation of surface conductance for evaporation from vegeta-tion Vegetatio 12179ndash87

Schulze E-D Beck E Muumlller-Hohenstein K (2005) Plant ecologySpringer Berlin

Schumacher J Roscher C (2009) Differential effects of functionaltraits on aboveground biomass in semi-natural grasslands Oikos118(11)1659ndash1668 doi101111j1600-0706200917711x

Schuurman N Leszczynski A (2008) Ontologies for bioinformaticsBioinf Biol 2187ndash200

Schwinning S Ehleringer JR (2001) Water use trade-offs and optimaladaptations to pulse-driven arid ecosystems J Ecol 89464ndash480doi101046j1365-2745200100576x

Scurlock JMO Johnson K Olson RJ (2002) Estimating net primaryproductivity from grassland biomass dynamics measurements GlobChange Biol 8736ndash753 doi101046j1365-2486200200512x

Shipley B (2010) From plant traits to vegetation structure Chance andselection in the assembly of ecological communities CambridgeUniversity Press Cambridge

Shipley B Dion J (1992) The allometry of seed production in herbaceousangiosperms Am Nat 139467ndash483 doi101086285339

Shipley B Vile D Garnier E (2006) From plant traits to plantcommunities a statistical mechanistic approach to biodiversityScience 314812ndash814 doi101126science1131344

Silvertown J Franco M Harper JL (eds) (1997) Plant life historiesEcology phylogeny and evolution Cambridge University PressCambridge

Small E (1972) Photosynthetic rates in relation to nitrogen recyclingas an adaptation to nutrient deficiency in peat bog plants Can JBot 502227ndash2233

Smith RG (2006) Timing of tillage is an important filter on theassembly of weed communities Weed Sci 54(4)705ndash712doi101614WS-05-177R11

Smith CC Fretwell SD (1974) The optimal balance between size andnumber of offspring Am Nat 108499ndash506 doi101086282929

Smith RG Gross KL (2007) Assembly of weed communities along acrop diversity gradient J Appl Ecol 44(5)1046ndash1056doi101111j1365-2664200701335x

Smith RG Mortensen DA Ryan MR (2009) A new hypothesis for thefunctional role of diversity in mediating resource pools andweed-crop competition in agroecosystems Weed Res 5037ndash48doi101111j1365-3180200900745x

Soons MB Heil GW Nathan R Katul GG (2004) Determinants oflong-distance seed dispersal by wind in grasslands Ecology853056ndash3068 doi10189003-0522

Stancioiu PT OrsquoHara KL (2006) Leaf area and growth efficiency ofregeneration in mixed species multiaged forests of the RomanianCarpathians For Ecol Manag 222(1ndash3)55ndash66 doi101016jforeco200510018

Stein LD (2008) Towards a cyberinfrastructure for the biologicalsciences progress visions and challenges Nat Rev Genet 9678ndash688 doi101038nrg2414

Stokes A Atger C Bengough AG Fourcaud T Sidle RC (2009)Desirable plant root traits for protecting natural and engineeredslopes against landslides Plant Soil 324(1ndash2)1ndash30 doi101007s11104-009-0159-y

Storkey J (2006) A functional group approach to the management ofUK arable weeds to support biological diversity Weed Res 46(6)513ndash522 doi101111j1365-3180200600528x

Storkey J Moss SR Cussans JW (2010) Using assembly theory toexplain changes in a weed flora in response to agriculturalintensification Weed Sci 58(1)39ndash46 doi101614ws-09-0961

Stubbs WJ Wilson JB (2004) Evidence for limiting similarity in asand dune community J Ecol 92(4)557ndash567 doi101111j0022-0477200400898x

Suding KN Lavorel S Chapin FS III Cornelissen JHC Diacuteaz SGarnier E Goldberg DE Hooper DU Jackson ST Navas M-L(2008) Scaling environmental change through the community-level a trait-based response-and-effect framework for plantsGlob Change Biol 141125ndash1140 doi101111j1365-2486200801557x

Swift MJ Anderson JM (1993) Biodiversity and ecosystem functionin agricultural systems In Schulze E-D Mooney HA (eds)Biodiversity and ecosystem function Springer Berlin pp 15ndash41

Tautenhahn S Heilmeier H Goumltzenberger L Klotz S Wirth C Kuumlhn I(2008) On the biogeography of seed mass in Germanymdashdistribution patterns and environmental correlates Ecography31457ndash468 doi101111j20080906-759005439x

Thompson K Band SR Hodgson JG (1993) Seed size and shapepredict persistence in soil Funct Ecol 7236ndash241

Thompson K Bakker JP Bekker RM Hodgson JG (1998) Ecologicalcorrelates of seed persistence in soil in the NW European flora JEcol 86163ndash169 doi101046j1365-2745199800240x

Thompson K Askew AP Grime JP Dunnett NP Willis AJ (2005)Biodiverstiy ecosystem function and plant traits in mature andimmature plant communities Funct Ecol 19355ndash358doi101111j1365-2435200500936x


Thuiller W Lavorel S Midgley G Lavergne S Rebelo T (2004)Relating plant traits and species distributions along bioclimaticgradients for 88 Leucadendron taxa Ecology 85(6)1688ndash1699doi10189003-0148

Van Calster H Vandenberghe R Ruysen M Verheyen K Hermy MDecocq G (2008) Unexpectedly high 20th century floristic lossesin a rural landscape in northern France J Ecol 96(5)927ndash936doi101111j1365-2745200801412x

van Wijk MT (2007) Predicting ecosystem functioning from plant traitsresults from a multi-scale ecophysiological modeling approachEcol Model 203453ndash463 doi101016jecolmodel200612007

Vanclay JK Gillison AN Keenan RJ (1997) Using plant functionalattributes to quantify site productivity and growth patterns inmixed forests For Ecol Manag 94(1ndash3)149ndash163 doi101016S0378-1127(96)03972-2

Venable DL Rees M (2009) The scaling of seed size J Ecol 9727ndash31 doi101111j1365-2745200801461x

Verbeek NAM Boasson R (1995) Flowering height and postfloralelongation of flower stalks in 13 species of angiosperms Can JBot 73723ndash727

Vila M Vayreda J Comas L Ibanez JJ Mata T Obon B (2007)Species richness and wood production a positive association inMediterranean forests Ecol Lett 10(3)241ndash250 doi101111j1461-0248200701016x

Vile D Garnier E Shipley B Laurent G Navas M-L Roumet CLavorel S Diacuteaz S Hodgson JG Lloret F Midgley GF Poorter HRutherford MC Wilson PJ Wright IJ (2005) Specific leaf areaand dry matter content estimate thickness in laminar leaves AnnBot 961129ndash1136 doi101093aobmci264

Vile D Shipley B Garnier E (2006a) Ecosystem productivity can bepredicted from potential relative growth rate and species abundanceEcol Lett 91061ndash1067 doi101111j1461-0248200600958x

Vile D Shipley B Garnier E (2006b) A structural equation model tointegrate changes in functional strategies during old-fieldsuccession Ecology 87504ndash517 doi10189005-0822

Villeger S Mason NWH Mouillot D (2008) New multidimensionalfunctional diversity indices for amultifaceted framework in functionalecology Ecology 89(8)2290ndash2301 doi10189007-12061

Violle C Lecoeur J Navas M-L (2007a) How relevant areinstantaneous measurements for assessing resource depletionunder plant cover A test on light and soil water availability in18 herbaceous communities Funct Ecol 21185ndash190doi101111j1365-2435200601241x

Violle C Navas M-L Vile D Kazakou E Fortunel C Hummel IGarnier E (2007b) Let the concept of trait be functional Oikos116882ndash892 doi101111j20070030-129915559x

Violle C Garnier E Lecœur J Roumet C Podeur C Blanchard ANavas M-L (2009) Competition traits and resource depletion inplant communities Oecologia 160747ndash755 doi101007s00442-009-1333-x

Vitousek PM Hooper DU (1993) Biological diversity and terrestrialecosystem biogeochemistry In Schulze E-D Mooney HA (eds)Biodiversity and ecosystem function Springer Berlin pp 3ndash14

Wahl S Ryser P (2000) Root tissue structure is linked to ecologicalstrategies of grasses New Phytol 148459ndash471 doi101046j1469-8137200000775x

Walker B Kinzig A Langridge J (1999) Plant attribute diversityresilience and ecosystem function the nature and significance ofdominant and minor species Ecosystems 295ndash113

Walker KJ Preston CD Boon CR (2009) Fifty years of change inan area of intensive agriculture plant trait responses tohabitat modification and conservation Bedfordshire EnglBiodivers Conserv 18(13)3597ndash3613 doi101007s10531-009-9662-y

Waller DM (1988) Plant morphology and reproduction In LovettDoust J Lovett Doust L (eds) Plant reproductive ecology

patterns and strategies Oxford University Press New York pp203ndash227

Wardle DA (2002) Communities and ecosystems Linking theaboveground and belowground components Princeton Univer-sity Press Princeton

Wardle DA Barker GM Bonner KI Nicholson KS (1998) Cancomparative approaches based on plant ecophysiological traitspredict the nature of biotic interactions and individual plantspecies effects in ecosystems J Ecol 86405ndash420 doi101046j1365-2745199800268x

Watanabe T Broadley MR Jansen S White PJ Takada J Satake KTakamatsu T Tuah SJ Osaki M (2007) Evolutionary control ofleaf element composition in plants New Phytol 174516ndash523doi101111j1469-8137200702078x

Webb CO Ackerly DD McPeek MA Donoghue MJ (2002)Phylogenies and community ecology Annu Rev Ecol Syst33475ndash505 doi101146annurevecolsys33010802150448

Weedon JT Cornwell WK Cornelissen JHC Zanne AE Wirth CCoomes DA (2009) Global meta-analysis of wood decompositionrates a role for trait variation among tree species Ecol Lett1245ndash56 doi101111j1461-0248200801259x

Weiher E Keddy PA (1995) Assembly rules null models and traitdispersion new questions front old patterns Oikos 74(1)159ndash164

Weiher E Keddy P (1999) Ecological assembly rules Perspectivesadvances retreats Cambridge University Press Cambridge

Weiher E van der Werf A Thompson K Roderick M Garnier EEriksson O (1999) Challenging Theophrastus a common corelist of plant traits for functional ecology J Veg Sci 10609ndash620doi1023073237076

West GB Brown JH Enquist BJ (1997) A general model for theorigin of allometric scaling laws in biology Science 276122ndash126 doi101126science2765309122

West GB Brown JH Enquist BJ (1999) A general model for thestructure and allometry of plant vascular systems Nature400664ndash667 doi10103823251

Westoby M (1998) A leaf-height-seed (LHS) plant ecology strategys ch eme P l a n t So i l 199 213ndash227 do i 10 1023 A1004327224729

Westoby M (1999) Generalization in functional plant ecology thespecies sampling problem plant ecology strategy schemes andphylogeny In Pugnaire FI Valladares F (eds) Handbook offunctional plant ecology Marcel Dekker Inc New York pp847ndash872

Westoby M Falster DS Moles AT Vesk PA Wright IJ (2002) Plantecological strategies some leading dimensions of variationbetween species Annu Rev Ecol Syst 33125ndash159doi101146annurevecolsys33010802150452

Westoby M Moles AT Falster DS (2009) Evolutionary coordinationbetween offspring size at independence and adult size J Ecol 97(1)23ndash26 doi101111j1365-2745200801396x

Wezel A Bellon S Doreacute T Francis C Vallod D David C (2009)Agroecology as a science a movement and a practice A reviewAgron Sustain Devt 29503ndash515 doi101051agro2009004

Whittaker RH (1965) Dominance and diversity in land plant communi-ties Science 147250ndash260 doi101126science1473655

Wilson JB (2007) Trait-divergence assembly rules have beendemonstrated limiting similarity lives A reply to Grime J VegSci 18(3)451ndash452 doi1016581100-9233(2007)18[451TARHBD]20CO2

Wilson PJ Thompson K Hodgson JG (1999) Specific leaf area andleaf dry matter content as alternative predictors of plantstrategies New Phytol 143155ndash162 doi101046j1469-8137199900427x

Wirth C Schumacher J Schulze E-D (2004) Generic biomassfunctions for Norway spruce in Central Europemdasha meta-


analysis approach toward prediction and uncertainty estimationTree Physiol 24121ndash139 doi101093treephys242121

Woodward FI Diament AD (1991) Functional approaches to predictingthe ecological effects of global change Funct Ecol 5202ndash212

Wright IJ (2001) Leaf economics of perennial species from sitescontrasted on rainfall and soil nutrients Division of Environmentand Life Sciences Macquarie University Sydney p 217

Wright IJ Reich PB Westoby M Ackerly DD Baruch Z Bongers FCavender-Bares J Chapin FS III Cornelissen JHC Diemer MFlexas J Garnier E Groom PK Gulias J Hikosaka K Lamont

BB Lee T Lee W Lusk C Midgley JJ Navas M-L NiinemetsUuml Oleksyn J Osada N Poorter H Poot P Prior L Pyankov VIRoumet C Thomas SC Tjoelker MG Veneklaas E Villar R(2004) The worldwide leaf economics spectrum Nature428821ndash827 doi101038nature02403

Wright IJ Reich PB Cornelissen JHC Falster DS Garnier EHikosaka K Lamont BB Lee W Oleksyn J Osada N PoorterH Villar R Warton DI Westoby M (2005) Assessing thegenerality of global leaf trait relationships New Phytol 166485ndash496 doi101111j1469-8137200501349x


A trait-based approach to comparative functional plant ecology concepts methods and applications for agroecology A review

Abstract

Introduction a trait-based approach to comparative functional ecology

Definitions and a framework

What is a trait

A response-and-effect framework

A functional characterisation of plants

The resource use dimension

The height dimension

The seed size dimension

Beyond a three-dimensional scheme

A functional approach to plant community structure

Environmental gradients and response traits

Assembly rules and environmental filters

Assembly rules and trait distribution within communities

Phylogeny traits and community assembly

The functional structure of communities

Traits and ecosystem properties

The biomass ratio hypothesis community-level weighted means of traits

Components of primary productivity

Decomposition and accumulation of litter

Soil water content

Other components of biogeochemical cycles

Complementarity among species incorporating functional divergence

The role of plant traits on ecosystem properties in controlled experiments

A different perspective ecosystem allometry

Plant traits in agricultural and forestry contexts

Grasslands

Crop weed communities

Forests

Trait databases and ecoinformatics

Availability of trait data

Towards a trait-based ecoinformatics

Conclusions and perspectives

References

unanswered questions and directions for research usingthe functional approach to biodiversity made possible bythe use of traits will be discussed in the contexts ofecological and agronomical systems The latter indeedcover a wide range of environmental conditions andbiological diversity and the prospect for reducingenvironmental impacts in highly productive low-diversity systems will certainly imply improving ourskills for the management of more diverse systems proneto a trait-based approach as reviewed here

Keywords Agroecology Biodiversity Communitystructure Comparative ecology Ecoinformatics

Ecological strategy Ecosystem properties Environmentalconditions Functional diversity Plant functional trait

Contents

1 Introduction a trait-based approach to comparativefunctional ecology 2

2 Definitions and a framework 321 What is a trait 322 A response-and-effect framework 4

3 A functional characterization of plants 531 The resource use dimension 632 The height dimension 833 The seed size dimension 1034 Beyond a three dimensional scheme 11

4 A functional approach to plant communitystructure 11

41 Environmental gradients and response traits 1142 Assembly rules and environmental filters 1243 Assembly rules and trait distribution within

communities 1344 Phylogeny traits and community assembly 13

5 Traits and ecosystem properties 1551 The biomass ratio hypothesis community-level

weighed means of traits 15511 Components of primary productivity 16512 Decomposition and accumulation of

litter 17513 Soil water content 17514 Other components of biogeochemical

cycles 1852 Complementarity among species incorporating

functional divergence 1953 The role of plant traits on ecosystem properties in

controlled experiments 2054 A different perspective ecosystem allometry 21

6 Plant traits in agricultural and forestry contexts 2161 Grasslands 2162 Crop weed communities 22

63 Forests 237 Trait databases and ecoinformatics 24

71 Availability of trait data 2472 Towards a trait-based ecoinformatics 24

8 Conclusions and perspectives 25

1 Introduction a trait-based approach to comparativefunctional ecology

A prominent field of research in functional ecology seeks tounderstand why and how ecological systems and theircomponents operate differently across environments (egCalow 1987 Duarte et al 1995 Westoby 1999 Pugnaireand Valladares 2007) The comparative approach is thusconsubstantial to this field of ecology (eg Bradshaw 1987Keddy 1992b Westoby 1999) If we are to understandquestions such as why and how are functions coordinatedwithin organisms why and how do different species performdifferently along ecological gradients why and how dospecies interact within a community and why and how dospecies affect the functioning of ecosystems comparingspecies is an absolute necessity Keddy (1992b) and Westoby(1999) also argue that for at least two reasons suchcomparisons should be conducted on a large number ofspecies First in the struggle to find broad generalisations wehave to make sure that the patterns observed are not simplepeculiarities of the species examined Second the tests ofrelationships between species and environment andor eco-system functioning need to be statistically powerful androbust As stressed by Keddy (1992b) this approach isliterally orthogonal to that followed in population ecologywhere many aspects pertaining to population dynamics areassessed in relatively few species There is also no reasonwhy the comparative approach could not apply to the largevariety of agricultural systems encountered across the world(eg Swift and Anderson 1993 Maleacutezieux et al 2009) Thesesystems indeed cover a wide range of environmentalconditions and biological diversity and the prospect forreducing environmental impacts in highly productive low-diversity systems (Giller et al 1997 Wezel et al 2009Griffon 2010) will certainly imply improving our skills forthe management of more diverse systems

There is a growing consensus that a trait-based approachhas a strong potential to address several of the issuesintroduced above A non-exhaustive list includes (1) thefunctioning of organisms and how it relates to the environment(eg Grime 1979 Chapin et al 1993 Grime et al 1997Westoby et al 2002) (2) the understanding of unsolvedquestions in community ecology such as the identification ofrules governing the assembly of communities (McGill et al2006 Suding et al 2008 Shipley 2010) and (3) the






21 What is a trait

















Trait values

Species pool

Habitat filter


similarity















LLS Leaf life span



Lroot Root length















































2500TM

MMPHTM

TA a2000

ASDCAS BMCN CF

RPVP1500 GR

TC

VP

TC

1000PB

500IC

BPBE

DG

IC

0 10 20 30 40 500

Plant height (cm)


100GRRPTC

b

80

AS

BM DC

GR

MM PH

RPTC

TMTAVP

80PB

CFDCPH

60 CNDGIC

60

BPBE

40

BP

20

0 20 40 60 80 1000














10-1

100

101

102

103

104

105

106



















CWM_SLA (m2 kg-1)

10 15 20 25


0

10

20

30

(b)

Species trait value


Species1

Species2

Species3

αA1

αB1

β

(a)

















CFP frac14Xn

ifrac141

pi traiti eth1THORN
















Low evenness

Niche axis

High divergence

(trait value)

Mintrait value

Maxtrait value

Functional richness






NPP frac14Pn



eth2THORN











SNPP frac14logeeth

Pn

ifrac141pi eRGRi


eth3THORN

















Udt frac14Xm

ifrac141

Xn

jfrac141


eth6THORN


SRA frac14 RAroot


Mrooteth7THORN


Udt frac14Xm

ifrac141

Xn

jfrac141


eth8THORN

































NPP frac14XK

kfrac140







61 Grasslands






















63 Forests































References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References





21 What is a trait

















Trait values

Species pool

Habitat filter


similarity















LLS Leaf life span



Lroot Root length















































2500TM

MMPHTM

TA a2000

ASDCAS BMCN CF

RPVP1500 GR

TC

VP

TC

1000PB

500IC

BPBE

DG

IC

0 10 20 30 40 500

Plant height (cm)


100GRRPTC

b

80

AS

BM DC

GR

MM PH

RPTC

TMTAVP

80PB

CFDCPH

60 CNDGIC

60

BPBE

40

BP

20

0 20 40 60 80 1000














10-1

100

101

102

103

104

105

106



















CWM_SLA (m2 kg-1)

10 15 20 25


0

10

20

30

(b)

Species trait value


Species1

Species2

Species3

αA1

αB1

β

(a)

















CFP frac14Xn

ifrac141

pi traiti eth1THORN
















Low evenness

Niche axis

High divergence

(trait value)

Mintrait value

Maxtrait value

Functional richness






NPP frac14Pn



eth2THORN











SNPP frac14logeeth

Pn

ifrac141pi eRGRi


eth3THORN

















Udt frac14Xm

ifrac141

Xn

jfrac141


eth6THORN


SRA frac14 RAroot


Mrooteth7THORN


Udt frac14Xm

ifrac141

Xn

jfrac141


eth8THORN

































NPP frac14XK

kfrac140







61 Grasslands






















63 Forests































References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References









Trait values

Species pool

Habitat filter


similarity















LLS Leaf life span



Lroot Root length















































2500TM

MMPHTM

TA a2000

ASDCAS BMCN CF

RPVP1500 GR

TC

VP

TC

1000PB

500IC

BPBE

DG

IC

0 10 20 30 40 500

Plant height (cm)


100GRRPTC

b

80

AS

BM DC

GR

MM PH

RPTC

TMTAVP

80PB

CFDCPH

60 CNDGIC

60

BPBE

40

BP

20

0 20 40 60 80 1000














10-1

100

101

102

103

104

105

106



















CWM_SLA (m2 kg-1)

10 15 20 25


0

10

20

30

(b)

Species trait value


Species1

Species2

Species3

αA1

αB1

β

(a)

















CFP frac14Xn

ifrac141

pi traiti eth1THORN
















Low evenness

Niche axis

High divergence

(trait value)

Mintrait value

Maxtrait value

Functional richness






NPP frac14Pn



eth2THORN











SNPP frac14logeeth

Pn

ifrac141pi eRGRi


eth3THORN

















Udt frac14Xm

ifrac141

Xn

jfrac141


eth6THORN


SRA frac14 RAroot


Mrooteth7THORN


Udt frac14Xm

ifrac141

Xn

jfrac141


eth8THORN

































NPP frac14XK

kfrac140







61 Grasslands






















63 Forests































References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References









LLS Leaf life span



Lroot Root length















































2500TM

MMPHTM

TA a2000

ASDCAS BMCN CF

RPVP1500 GR

TC

VP

TC

1000PB

500IC

BPBE

DG

IC

0 10 20 30 40 500

Plant height (cm)


100GRRPTC

b

80

AS

BM DC

GR

MM PH

RPTC

TMTAVP

80PB

CFDCPH

60 CNDGIC

60

BPBE

40

BP

20

0 20 40 60 80 1000














10-1

100

101

102

103

104

105

106



















CWM_SLA (m2 kg-1)

10 15 20 25


0

10

20

30

(b)

Species trait value


Species1

Species2

Species3

αA1

αB1

β

(a)

















CFP frac14Xn

ifrac141

pi traiti eth1THORN
















Low evenness

Niche axis

High divergence

(trait value)

Mintrait value

Maxtrait value

Functional richness






NPP frac14Pn



eth2THORN











SNPP frac14logeeth

Pn

ifrac141pi eRGRi


eth3THORN

















Udt frac14Xm

ifrac141

Xn

jfrac141


eth6THORN


SRA frac14 RAroot


Mrooteth7THORN


Udt frac14Xm

ifrac141

Xn

jfrac141


eth8THORN

































NPP frac14XK

kfrac140







61 Grasslands






















63 Forests































References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References
























2500TM

MMPHTM

TA a2000

ASDCAS BMCN CF

RPVP1500 GR

TC

VP

TC

1000PB

500IC

BPBE

DG

IC

0 10 20 30 40 500

Plant height (cm)


100GRRPTC

b

80

AS

BM DC

GR

MM PH

RPTC

TMTAVP

80PB

CFDCPH

60 CNDGIC

60

BPBE

40

BP

20

0 20 40 60 80 1000














10-1

100

101

102

103

104

105

106



















CWM_SLA (m2 kg-1)

10 15 20 25


0

10

20

30

(b)

Species trait value


Species1

Species2

Species3

αA1

αB1

β

(a)

















CFP frac14Xn

ifrac141

pi traiti eth1THORN
















Low evenness

Niche axis

High divergence

(trait value)

Mintrait value

Maxtrait value

Functional richness






NPP frac14Pn



eth2THORN











SNPP frac14logeeth

Pn

ifrac141pi eRGRi


eth3THORN

















Udt frac14Xm

ifrac141

Xn

jfrac141


eth6THORN


SRA frac14 RAroot


Mrooteth7THORN


Udt frac14Xm

ifrac141

Xn

jfrac141


eth8THORN

































NPP frac14XK

kfrac140







61 Grasslands






















63 Forests































References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References











10-1

100

101

102

103

104

105

106



















CWM_SLA (m2 kg-1)

10 15 20 25


0

10

20

30

(b)

Species trait value


Species1

Species2

Species3

αA1

αB1

β

(a)

















CFP frac14Xn

ifrac141

pi traiti eth1THORN
















Low evenness

Niche axis

High divergence

(trait value)

Mintrait value

Maxtrait value

Functional richness






NPP frac14Pn



eth2THORN











SNPP frac14logeeth

Pn

ifrac141pi eRGRi


eth3THORN

















Udt frac14Xm

ifrac141

Xn

jfrac141


eth6THORN


SRA frac14 RAroot


Mrooteth7THORN


Udt frac14Xm

ifrac141

Xn

jfrac141


eth8THORN

































NPP frac14XK

kfrac140







61 Grasslands






















63 Forests































References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References

















CWM_SLA (m2 kg-1)

10 15 20 25


0

10

20

30

(b)

Species trait value


Species1

Species2

Species3

αA1

αB1

β

(a)

















CFP frac14Xn

ifrac141

pi traiti eth1THORN
















Low evenness

Niche axis

High divergence

(trait value)

Mintrait value

Maxtrait value

Functional richness






NPP frac14Pn



eth2THORN











SNPP frac14logeeth

Pn

ifrac141pi eRGRi


eth3THORN

















Udt frac14Xm

ifrac141

Xn

jfrac141


eth6THORN


SRA frac14 RAroot


Mrooteth7THORN


Udt frac14Xm

ifrac141

Xn

jfrac141


eth8THORN

































NPP frac14XK

kfrac140







61 Grasslands






















63 Forests































References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References















CFP frac14Xn

ifrac141

pi traiti eth1THORN
















Low evenness

Niche axis

High divergence

(trait value)

Mintrait value

Maxtrait value

Functional richness






NPP frac14Pn



eth2THORN











SNPP frac14logeeth

Pn

ifrac141pi eRGRi


eth3THORN

















Udt frac14Xm

ifrac141

Xn

jfrac141


eth6THORN


SRA frac14 RAroot


Mrooteth7THORN


Udt frac14Xm

ifrac141

Xn

jfrac141


eth8THORN

































NPP frac14XK

kfrac140







61 Grasslands






















63 Forests































References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References











Low evenness

Niche axis

High divergence

(trait value)

Mintrait value

Maxtrait value

Functional richness






NPP frac14Pn



eth2THORN











SNPP frac14logeeth

Pn

ifrac141pi eRGRi


eth3THORN

















Udt frac14Xm

ifrac141

Xn

jfrac141


eth6THORN


SRA frac14 RAroot


Mrooteth7THORN


Udt frac14Xm

ifrac141

Xn

jfrac141


eth8THORN

































NPP frac14XK

kfrac140







61 Grasslands






















63 Forests































References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References




NPP frac14Pn



eth2THORN











SNPP frac14logeeth

Pn

ifrac141pi eRGRi


eth3THORN

















Udt frac14Xm

ifrac141

Xn

jfrac141


eth6THORN


SRA frac14 RAroot


Mrooteth7THORN


Udt frac14Xm

ifrac141

Xn

jfrac141


eth8THORN

































NPP frac14XK

kfrac140







61 Grasslands






















63 Forests































References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References





SNPP frac14logeeth

Pn

ifrac141pi eRGRi


eth3THORN

















Udt frac14Xm

ifrac141

Xn

jfrac141


eth6THORN


SRA frac14 RAroot


Mrooteth7THORN


Udt frac14Xm

ifrac141

Xn

jfrac141


eth8THORN

































NPP frac14XK

kfrac140







61 Grasslands






















63 Forests































References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References







Udt frac14Xm

ifrac141

Xn

jfrac141


eth6THORN


SRA frac14 RAroot


Mrooteth7THORN


Udt frac14Xm

ifrac141

Xn

jfrac141


eth8THORN

































NPP frac14XK

kfrac140







61 Grasslands






















63 Forests































References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References


























NPP frac14XK

kfrac140







61 Grasslands






















63 Forests































References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References




















63 Forests































References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References





























References































































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References





















































































































































































































































































































Abstract



What is a trait

















Soil water content






Grasslands


Forests





References

a trait-based approach to comparative functional plant ecology ... › uploads › membres ›...

Documents