Conifers, Angiosperm Trees, and Lianas: Growth, Whole-Plant Water and Nitrogen Use Efficiency, and Stable Isotope Composition (d13C and d18O) of Seedlings Grown in a Tropical Environment1[W][OA] Lucas A. Cernusak2*, Klaus Winter, Jorge Aranda, and Benjamin L. Turner Smithsonian Tropical Research Institute, Balboa, Ancon, Republic of Panama Seedlings of several species of gymnosperm trees, angiosperm trees, and angiosperm lianas were grown under tropical field conditions inthe Republic ofPanama; physiological processes controllingplant C and water fluxeswere assessed across this functionallydiverserangeofspecies.Relativegrowthrate,r,wasprimarilycontrolledbytheratioofleafareatoplantmass,of whichspecificleafareawasakeycomponent.Instantaneousphotosynthesis,whenexpressedonaleaf-massbasis,explained 69%ofvariationinr(P,0.0001,n594).Meanrofangiospermswassignificantlyhigherthanthatofthegymnosperms;within angiosperms,meanroflianaswashigherthanthatoftrees.Whole-plantnitrogenuseefficiencywasalsosignificantlyhigherin angiospermthaningymnospermspecies,andwasprimarilycontrolledbytherateofphotosynthesisforagivenamountofleaf nitrogen.Whole-plantwateruseefficiency,TE,variedsignificantlyamongspecies,andwasprimarilycontrolledbyc/c,the c i a ratio of intercellular to ambient CO partial pressures during photosynthesis. Instantaneous measurements of c/c explained 2 i a 51%ofvariationinTE (P,0.0001,n594).Whole-plant13Cdiscriminationalsovariedsignificantlyasafunctionofc/c (R25 c i a 0.57,P,0.0001,n594),andwas,accordingly,agoodpredictorofTE.The18Oenrichmentofstemdrymatterwasprimarily c controlledbythepredicted18Oenrichmentofevaporativesiteswithinleaves(R250.61,P,0.0001,n594),withsomeresidual variationexplainedbymeantranspirationrate.Measurementsofcarbonandoxygenstableisotoperatioscouldprovideauseful meansofparameterizingphysiologicalmodelsoftropicalforesttrees. Tropical forest ecosystems have been subject to ex- accumulation), nitrogen (N) use efficiency (NUE; the tensive perturbations associated with anthropogenic rate of C accumulation for a given N content), water activity in recent decades, and such perturbations will useefficiency(theratioofwhole-plantCgaintowater likely continue into the foreseeable future (Laurance loss),andstableisotopecomposition(d13Candd18O)in et al., 2004; Wright, 2005). Effective environmental seedlings of a diverse suite of species grown side- managementrequiresknowledgeofhowsuchpertur- by-sidein a tropical environment. bations impact upon cycling of carbon (C) and water Conifers dominated the world’s forests prior to the between forest trees and the atmosphere, and how Cretaceous radiation in angiosperm diversity. How- theseCandwaterfluxesrelatetoplantnutrientstatus. ever, conifers are largely absent from the lowland A sound, mechanistic understanding of the physio- tropical and subtropical forests of today. It has been logicalprocessesthatcontrolphotosynthesisandtran- suggested that one means by which angiosperm tree spiration in tropical trees is therefore essential for species are able to out-compete gymnosperm tree understandingandmanagingthehumanimpactupon species in tropical environments is through faster tropical forests. In this study, we analyzed the phys- seedling growth caused by improved hydraulic effi- iological controls over growth (the relative rate of C ciency(Bond,1989;Brodribbetal.,2005).Angiosperm xylem tissue contains vessels, specialized water- conducting cells that are generally larger in diameter, 1ThisworkwassupportedbytheSmithsonianTropicalResearch and therefore more conductive to water, than conifer Institute.L.A.C.wassupportedbyapostdoctoralfellowshipfrom tracheids (Sperry et al., 2006). Conifer tracheid diame- theSmithsonianInstitutionandaTupperResearchFellowshipfrom tersarebiomechanicallyconstrainedbecausethesecells theSmithsonianTropicalResearchInstitute. must perform the dual function of conducting water 2Present address: School of Environmental and Life Sciences, and providing structural support to woody tissues, CharlesDarwinUniversity,Darwin,NorthernTerritory0909,Australia. whereasvesselsneednotperformthelatterfunctionin *Correspondingauthor;[email protected]. angiosperm wood. Lianas are large woody vines that Theauthorresponsiblefordistributionofmaterialsintegraltothe occur predominantly in tropical forests; by attaching findingspresentedinthisarticleinaccordancewiththepolicydescribed themselvestoneighboringtrees, theyhaveevolvedan in the Instructions for Authors (www.plantphysiol.org) is: Lucas A. additional means of freeing their xylem tissues from Cernusak([email protected]). [W]TheonlineversionofthisarticlecontainsWeb-onlydata. structural constraints. Thus, angiosperm lianas may [OA]OpenAccessarticlecanbeviewedonlinewithoutasubscription. achieve further increases in hydraulic efficiency com- www.plantphysiol.org/cgi/doi/10.1104/pp.108.123521 paredtoangiospermtrees(Gartneretal.,1990). 642 Plant Physiology, September 2008, Vol. 148, pp. 642–659, www.plantphysiol.org (cid:2) 2008 American Society of Plant Biologists Growth, Water Use, and Stable Isotopes in Tropical Trees Inthisstudy,wegrewseedlingsofseveralspeciesof the diffusive fluxes of CO and water vapor into and 2 gymnospermtrees,angiospermtrees,andangiosperm out of the leaf, respectively (Farquhar and Richards, lianas in a tropical environment. We used this func- 1984): tionallydiverserangeofspeciestoquantifythephys- A c 2c iological controls over their C and water fluxes. We 5 a i ð3Þ also took advantage of the contrasting physiology of E 1:6n the study species to test the theoretical basis for var- where E is transpiration (mol H O m22 s21), c and c 2 a i iation in the C and oxygen (O) stable isotope compo- are CO partial pressures in ambient air and leaf 2 sition of plant dry matter. intercellular air spaces, respectively, v is the leaf-to- air vapor pressure difference, and 1.6 is the ratio of diffusivitiesofCO andH Oinair.Thevisdefinedas 2 2 THEORY e-e , where e and e are the intercellular and ambient i a i a vapor pressures, respectively. The ratio of C gain to Growth water loss can be scaled to the whole-plant level by FollowingMasleandFarquhar(1988),andbasedon taking into account respiratory C use and water loss earlier treatments (Blackman, 1919; Evans, 1972), we not associated with photosynthesis (Farquhar and writethefollowingexpressiontodescribefactorsthat Richards,1984; Hubickand Farquhar, 1989): influencetherelativerateofCaccumulationofaplant: (cid:2) (cid:3) c ð1 2f Þc 1 2 i r5 1 (cid:2)dmc 5 Alð1 2fcÞ ð1Þ TE 5 c a ca ð4Þ mc dt r c 1:6nð11fwÞ whererisrelativegrowthrate(molCmol21Cs21),mc whereTEcisthetranspirationefficiencyofCgain,and is plant C mass (mol C), t is time (s), A is leaf pho- f isunproductivewaterlossasaproportionofwater tosyntheticrate(molCm22s21),listhelightperiodas lowss associated with C uptake, the former mainly a fraction of 24 h, fc is the proportion of C gained in comprising water loss at night through partially photosynthesis that is subsequently used for respira- open stomata. Thus, f can be approximated as E / w n tionbyleavesatnightandbyroots andstemsduring E , where E is nighttime transpiration and E is d n d dayandnight,andristheratioofplantCmasstoleaf daytime transpiration. area (mol C m22). Equation 1 provides a useful tool We suggest that the leaf to air vapor pressure for examining sources of variation in r among plant difference, v, can be written as the product of the air speciesandindividualswithinaspecies.Itissimilarto vapor pressure deficit (D), and a second term, f , v the classical decomposition of r into net assimilation whichdescribesthemagnitudeofvrelativetoD,such rate(NAR;gm22s21)andleafarearatio(LAR;m2g21), that v 5Df . This allows Equation 3 tobe writtenas v but allows the assimilation term to be expressed as a (cid:2) (cid:3) net photosynthetic rate, such as would be measured c ð1 2f Þc 1 2 i using standard gas exchange techniques (Long et al., c a c 1996). TableIprovidesdefinitionsofallabbreviations D(cid:2)TEc5 1:6f ð11f Þa ð5Þ and symbols used in this article. v w Weighting TE by D facilitates comparison of the c transpiration efficiency of plants grown under differ- NUE ent environmental conditions by accounting for vari- Multiplying both sides of Equation 1 by the molar ationduetodifferencesinatmosphericvaporpressure ratioofplantCtoNyieldsanexpressionfortheNUE deficit(TannerandSinclair,1983;HubickandFarquhar, ofC accumulation: 1989). Thus, it accounts for variation in TE that is c purelyenvironmental.TheD(cid:2)TE hasunitsofPamolC 1 dm c NUE5 (cid:2) c 5A lð1 2f Þn ð2Þ mol21H O. mn dt n c l Photos2ynthetic discrimination against 13C (D13C) whereNUEiswhole-plantNUE(molCmol21Ns21), shares a common dependence with TEc on ci/ca. The mnisplantNmass(molN),AnisphotosyntheticNUE D13C in C3 plants relates to ci/ca according to the (molCmol21Ns21),andn istheproportionofplantN following equation (Farquhar et al., 1982; Farquhar l allocated to leaves. Equation 2 provides a basis for andRichards,1984; Hubick et al., 1986): ilninvkeisntgigAatni,oanst,rawitiothfteNnUqEu,anatnifiiendteignraetceodphmyesiaosluorgeicoafl D13C5a 2d1ðb 2aÞci ð6Þ c NUE atthe whole-plant level. a where a is the discrimination against 13C during dif- fusion through stomata (4.4&), b is discrimination Transpiration Efficiency and C Isotope Discrimination against 13C during carboxylation by Rubisco (29&), The ratio of C gain to water loss at the leaf level anddisacompositetermthatsummarizescollectively duringphotosynthesiscanbeexpressedastheratioof thediscriminationsassociatedwithdissolutionofCO , 2 Plant Physiol. Vol. 148, 2008 643 Cernusak et al. TableI. Abbreviationsandsymbolsusedinthetext A Area-basedphotosynthesisrate(mmolCO m22s21) 2 A Mass-basedphotosynthesisrate(nmolCO g21s21) m 2 A PhotosyntheticNUE(mmolCO mol21Ns21) n 2 a Discriminationagainst13Cduringdiffusionthroughstomata b Discriminationagainst13CduringcarboxylationbyRubisco C Molarconcentrationofwater(molm23) c PartialpressureofCO inambientair(Pa) a 2 c PartialpressureofCO inleafintercellularairspaces(Pa) i 2 D Vaporpressuredeficitofambientair(kPa) D Growth-weightedvaporpressuredeficitofambientair(kPa) g D Averagedaytimevaporpressuredeficitofambientairduringweeki(kPa) i D DiffusivityofH18Oinwater(m2s21) 18 2 d Discriminationagainst13CduringC photosynthesisnotassociatedwithaorb 3 E Transpirationrate(mmolm22s21) E Daytimetranspirationrate(mmolm22s21) d E Cumulativetranspirationoverthecourseoftheexperiment(mol) t E Nighttimetranspirationrate(mmolm22s21) n e Vaporpressureofambientair(kPa) a e Vaporpressureinleafintercellularairspaces(kPa) i g Stomatalconductance(molm22s21) s L Scaledeffectivepathlengthrelatingto18Oadvectionanddiffusion(m) LAR Leafarearatio(m2kg21) LA Leafareaattheinitiationoftheexperiment(m2) 1 LA Leafareaattheconclusionoftheexperiment(m2) 2 l Lightperiodasafractionof24h l MassofCinleaflitterabscisedduringtheexperiment(molC) c m PlantCmass(molC) c m PlantCmassattheinitiationoftheexperiment(molC) c1 m PlantCmassattheconclusionoftheexperiment(molC) c2 m PlantNmass(molN) n MTR Meantranspirationrateoverthecourseoftheexperiment(molm22d21) NAR Netassimilationrate(gdrymatterm22s21) NUE Whole-plantNuseefficiency(molCmol21Nd21) n LeafNasaproportionofwhole-plantN 1 P LeafPperunitarea(mmolm22) area p ProportionofOatomsexchangingwithmediumwaterduringcellulosesynthesis ex p Proportionofwaterindevelopingcellsnotsubjecttoevaporative18Oenrichment x R The18O/16Oratioofanywaterordrymattercomponentofinterest R The13C/12CratioofCO inambientair a 2 R The13C/12CratioofplantC p R The18O/16Oratioofirrigation(source)water s r Relativegrowthrate(molCmol21Cd21) SLA Specificleafarea(m2kg21) TE Whole-planttranspirationefficiencyofCgain(mmolCmol21HO) c 2 v Leaf-to-airvaporpressuredifference(kPa) v Growth-weightedleaf-to-airvaporpressuredifference(kPa) g v Averagedaytimeleaf-to-airvaporpressuredifferenceforweeki(kPa) i w Predicteddrymatterincrementforweeki(g) i D13C Discriminationagainst13C D13C Discriminationagainst13CindrymatterofthewholeplantrelativetoCO inair p 2 D18O The18Oenrichmentofevaporativesiteswithinleavescomparedtosourcewater e D18O Growth-weightedpredictionofD18O overthecourseoftheexperiment eg e D18O PredictedaveragedaytimeD18O forweeki ei e D18O The18Oenrichmentofaveragelaminaleafwatercomparedtosourcewater L D18O The18Oenrichmentofstemdrymattercomparedtosourcewater p D18O The18Oenrichmentofatmosphericwatervaporcomparedtosourcewater v d13C The13C/12CratiorelativetothePeeDeeBelmniteinternationalstandard d13C Thed13CofCO inambientair a 2 d13C Thed13CofplantC p d18O The18O/16OratiorelativetoViennaStandardMeanOceanWater d18O Thed18Oofstemdrymatter p d18O Thed18Oofirrigation(source)water s (Tablecontinuesonfollowingpage.) 644 Plant Physiol. Vol. 148, 2008 Growth, Water Use, and Stable Isotopes in Tropical Trees TableI. (Continuedfrompreviouspage.) e Thed18Odifferencebetweenplantdrymatterandcelluloseextractedfromit cp e KineticH18Ofractionationfordiffusionthroughstomataandleafboundarylayer k 2 e The18Oenrichmentofcellulosecomparedtothewaterinwhichitformed wc e1 EquilibriumH18Ofractionationduringthephasechangefromliquidtogas 2 f Proportionofnetphotosynthesissubsequentlyusedforrespiration c f ScalingfactortoconvertDtov(5v/D) v f Unproductivewaterlossasproportionofthatassociatedwithphotosynthesis w § Pe´cletnumber r PlantCmassperunitleafarea(molCm22) liquid phase diffusion, photorespiration, and dark can be calculated as a function of leaf temperature respiration (Farquhar et al., 1989a). The term d may (BottingaandCraig,1969),ande canbecalculatedby k be excluded from Equation 5, in which case the re- partitioning the resistance to water vapor diffusion duction in D13C caused by d is often accounted for by between stomata and boundary layer, with the two taking a lower value for b. The D13C is defined with weightedbyappropriatefractionationfactors(Farquhar respecttoCO inairasD13C5R /R 21,whereR is et al., 1989b; Cappa et al., 2003). The D18O can be 2 a p a v 13C/12C of CO in air and R is 13C/12C of plant calculated from measurements of the d18O of ambient 2 p C.CombiningEquations 5 and6 gives vaporandsourcewater.Ifsuchdataarenotavailable, a reasonable approximation is to estimate D18O as D(cid:2)TE 5 cað1 2fcÞðb 2d 2D13CÞ ð7Þ 2e1,whichmeansthatambientvaporisassumedtvobe c 1:6f ð11f Þðb 2aÞ in isotopic equilibrium with soil water. An up- v w to-date summary of equations necessary for parame- Equation 7 suggests a negative linear dependence of terizationofEquation8canbefoundinCernusaketal. TE (or D(cid:2)TE ) on D13C, although it can be seen that c c (2007b). therearemanyothertermsinEquation7thathavethe Averagelaminaleafwater18Oenrichment(D18O )is potentialtoinfluencetherelationshipbetweenthetwo. L generally less than that predicted for evaporative site water (Yakir et al., 1989; Flanagan, 1993; Farquhar O Isotope Enrichment et al., 2007), and carbohydrates exported from leaves havebeenobservedtocarrythesignalofD18O rather It has been suggested that measurements of the O thanD18O (Barbouretal.,2000b;CernusaketaLl.,2003, isotope enrichment of plant organic material (D18Op) 2005; Gesesler et al., 2007). The D18O has been sug- can provide complementary information to that in- gested to relate to D18O according tLo the following ferred from D13C in analyses of plant water-use effi- e relationship(FarquharandLloyd,1993;Farquharand ciency (Farquhar et al., 1989b, 1994; Sternberg et al., Gan, 2003): 1989;YakirandIsraeli,1995).Specifically,D18O could p provide information about the ratio of ambient to in- ð1 2e2§Þ tercellular vapor pressures, e /e, and thus about the D18O 5D18O ð9Þ a i L e § leaf-to-airvaporpressuredifference,e-e ,duringpho- i a tosynthesis. Note that ei-ea is equal to v in Equation 3. where § is a Pe´clet number, defined as EL/(CD ), In the steady state, water at the evaporative sites in where E is transpiration rate (mol m22 s21), L is18a leaves becomes enriched in 18O relative to water en- scaled effective path length (m), C is the molar con- teringtheplantfromthesoil,accordingtothefollow- centrationofwater(molm23),andD isthediffusiv- ing relationship (Craig and Gordon, 1965; Dongmann ityofH 18Oinwater(m2s21).TheCi1s8aconstant,and etal., 1974;Farquharand Lloyd, 1993): 2 D can be calculated from leaf temperature (Cuntz 18 e et al., 2007). The constancy of L, or otherwise, is D18O 5e11e 1ðD18O 2e Þ a ð8Þ e k v k e currently under investigation (Barbour and Farquhar, i 2004; Barbour, 2007; Kahmen et al., 2008; Ripullone where D18O is the 18O enrichment of evaporative site e etal.,2008). IfLisassumedrelativelyconstant, Equa- water relative to source water, e1 is the equilibrium tion 9 predicts that D18O will vary as a function of fractionationbetweenliquidwaterandvapor,ekisthe bothD18O andE.TotestfoLraninfluenceofEonD18O , kinetic fractionation that occurs during diffusion of it is necesesary to first account for variation in D18OL water vapor out of the leaf, and D18Ov is the discrim- causedbyD18O (Flanaganetal.,1994).Tothisend,theL inationofambientvaporwithrespecttosourcewater. relativedeviatioenofD18O fromD18O (12D18O /D18O) The D18O of any water or dry matter component is L e L e can be examined, in which case Equation 9 can be defined with respect to source water (water entering writtenas the roots from the soil) as D18O 5 R/R 2 1, where s D18Ois the18Oenrichment of thecomponent of inter- D18O ð1 2e2§Þ est and R and R arethe 18O/16O ratios of the compo- 1 2 L 51 2 ð10Þ nent of interest asnd sourcewater, respectively. The e1 D18Oe § Plant Physiol. Vol. 148, 2008 645 Cernusak et al. Equation 10 predicts that 1 2 D18O /D18O should trees had the lowest mean value of r, whereas angio- L e increase as E increases. spermlianashadthehighestmeanvalue;angiosperm The transfer of the leaf water 18O signal to plant treeshadameanvalueofrintermediatebetweenthat organic material can be described by the following ofgymnospermtreesandangiospermlianas(Fig.1A). equation (Barbourand Farquhar, 2000): Incontrast,therewaslessvariationamongspeciesand functional groups in instantaneous photosynthesis D18Op5D18OLð1 2pexpxÞ1ewc1ecp ð11Þ ratesexpressedonaleafareabasis(Fig.1B),although whereD18O is18Oenrichmentofplantdrymatter,p thespeciesPinuscaribaeaandStigmaphyllonhypargyreum p ex were notable for having relatively high values of A. is the proportion of O atoms that exchange with local Variation in r tended to be more closely associated waterduringsynthesisofcellulose,aprimaryconstit- with variation in 1/r than with variation in A. Gym- uent of plant dry matter, p is the proportion of x nosperm trees had the lowest mean value of 1/r, unenriched water at the site of tissue synthesis, e is wc whereas angiosperm trees had an intermediate mean the fractionation between organic oxygen and me- dium water, and e is the difference in D18O between value, and angiosperm lianas had the highest mean cp value (Fig. 1C). The liana species S. hypargyreum pos- tissuedrymatterandthecellulosecomponent.Fortree sessedswollen,tuberousroots,whichcausedittohave stems,p p hasbeenfoundtoberelativelyconstantat ex x aroot/shootratiomuchhigherthananyotherspecies about0.4(Rodenetal.,2000;Cernusaketal.,2005).The in the study (Table III), and to have a reduced 1/r e isrelativelyconstantatabout27&(Barbour,2007), wc relativeto the other two liana species (Fig. 1C). and for stem dry matter, e appears to be relatively cp Variation in instantaneous photosynthesis, when constant at about 25& (Borella et al., 1999; Barbour expressed on a leaf mass basis, was a good predictor etal.,2001;Cernusaketal.,2005).Iftheseassumptions are valid, variation in D18O should primarily reflect of variation in r (Fig. 2). The former was measured variationinD18O .Thus,comp biningEquations10and overseveralminutes,whereasthelatterwasmeasured L over several months. Mass-based photosynthesis, A , 11providesameansoftestingforaninfluenceofEon m D18O ,assumingthatD18O providesatime-integrated istheproductofA(molm22s21)andspecificleafarea recorpdof D18O (Barbourept al., 2004): (SLA;m2kg21).ThecorrelationbetweenAmandrwas L almostentirelydrivenbyvariationinSLA,becauseA on a leaf area basis was not significantly correlated (cid:2)D18Op 2ewc 2ecp(cid:3) with r(P 50.14,n 5 94). 1 2p p ð1 2e2§Þ The C and N concentrations of leaves, stems, roots, 1 2 ex x 51 2 ð12Þ D18O § and whole plants for each species are given in Sup- e plemental Table S1. For whole-plant C concentration, therewassignificantvariation,bothamongfunctional groups (P , 0.0001), and among species within func- RESULTS tional groups (P , 0.0001), as shown in Figure 3A. Gymnosperm trees had a mean C concentration of Growth,Photosynthesis,andElementalConcentrations 49.6%,significantlyhigherthanangiospermtreesand Daytime meteorological conditions over the course lianas.Angiospermtreesandlianasdidnotdifferfrom of the experiment are shown in Table II. Dates of each other in whole-plant C concentration, and had initiation of transpiration measurements and harvest mean values of 45.4% and 44.9%, respectively. For for each species are shown in Table III. Table III also wholeplantNconcentration,therewasalsosignificant shows the initial and final dry masses, in addition to variation among functional groups (P , 0.0001) and root/shoot ratios. Variation in relative growth rate, r, amongspecieswithinfunctionalgroups(P,0.0001),as amongspeciesisshowninFigure1A;variationinthe showninFigure3B.Angiospermlianashadthehighest components of r, A, and 1/r, is shown in Figure 1, B meanwhole-plantNconcentrationat1.22%,followed and C, respectively. The r varied significantly among by angiosperm trees at 1.01%, then by gymnosperm functional groups (P , 0.0001), and among species trees at 0.82%. Accordingly, there was significant within functional groups (P , 0.0001). Gymnosperm variation among functional groups (P , 0.0001) and TableII. Averagedaytimemeteorologicalconditionsatthestudysiteoverthecourseoftheexperiment Valuesaremonthlymeansofmeasurementstakenevery15minbetweenthehoursof7AMand5:30PMlocaltime.Wefocusedondaytimehoursto characterizeconditionsduringphotosyntheticgasexchange. 2005 2006 June July Aug Sept Oct Nov Dec Jan Feb March April May Airtemperature((cid:3)C) 28.0 26.7 28.6 29.6 30.0 27.7 30.7 28.1 27.9 29.1 28.7 28.2 Relativehumidity(%) 81.3 82.9 82.6 83.3 79.3 84.2 75.0 72.5 68.8 68.3 74.1 80.1 Vaporpressuredeficit(kPa) 0.73 0.62 0.71 0.70 0.89 0.60 1.12 1.06 1.18 1.31 1.03 0.78 Windspeed(ms21) 0.33 0.26 0.31 0.33 0.45 0.29 0.50 0.61 0.87 0.82 0.75 0.38 Photonfluxdensity(mmolm22s21) 685 655 649 628 750 602 743 808 859 925 854 685 646 Plant Physiol. Vol. 148, 2008 Growth, Water Use, and Stable Isotopes in Tropical Trees TableIII. Experimentaltimeperiod,initialandfinalplantdrymass,androottoshootratioforeachspeciesinthestudy Valuesforfinaldrymassandroottoshootratioaregivenasthemean,withtheSDinparentheses.AnSDisnotgivenforP.guatemalensisbecause onlyoneplantsurvivedforthisspecies.FullspeciesnamesaregiveninFigure2.NA,Notapplicable. No.of InitialDry FinalDry RoottoShoot Species Family StartDate EndDate Plants Mass Mass Ratio g g gg21 Gymnospermtreespecies C.lusitanica Cupressaceae May23,2005 Dec.13,2005 8 5.8 60.2(19.0) 0.36(0.06) P.caribaea Pinaceae May23,2005 Dec.13,2005 8 5.6 64.6(30.9) 0.27(0.09) P.guatemalensis Podocarpaceae May23,2005 May11,2006 1 0.8 107.3(NA) 0.16(NA) T.occidentalis Cupressaceae April26,2004 Dec.13,2005 8 4.7 41.4(16.5) 0.47(0.08) Angiospermtreespecies C.longifolium Clusiaceae July11,2005 March10,2006 6 3.1 62.6(27.0) 0.39(0.08) C.pratensis Clusiaceae Aug.22,2005 May11,2006 6 0.2 112.2(62.3) 0.76(0.26) H.alchorneoides Euphorbiaceae June20,2005 Dec.13,2005 6 0.2 36.6(30.2) 1.10(0.13) L.seemannii Tiliaceae Aug.29,2005 March10,2006 7 0.1 39.8(7.3) 0.61(0.10) P.pinnatum Fabaceae Nov.7,2005 March10,2006 2 0.6 53.9(16.6) 0.58(0.20) P.pinnatum Fabaceae Nov.7,2005 May11,2006 3 0.6 21.6(13.3) 0.36(0.14) S.macrophylla Meliaceae Nov.7,2005 May11,2006 7 0.8 24.3(11.8) 0.34(0.10) T.rosea Bignoniaceae June20,2005 Dec.13,2005 6 0.6 64.8(11.0) 0.92(0.12) T.grandis Verbenaceae May17,2004 May11,2006 7 0.1 45.5(11.3) 0.93(0.17) Angiospermlianaspecies G.lupuloides Rhamnaceae Aug.29,2005 March10,2006 6 0.01 31.5(15.6) 0.47(0.20) M.leiostachya Asteraceae Aug.29,2005 March10,2006 3 0.1 21.2(4.3) 0.37(0.27) M.leiostachya Asteraceae Nov.14,2005 May11,2006 3 0.3 23.2(4.8) 0.25(0.06) S.hypargyreum Malphigiaceae Aug.29,2005 March10,2006 7 0.1 50.3(20.3) 1.99(0.21) among species within functional groups (P , 0.0001) When expressed on a leaf area basis, the leaf P in whole-plant C/N mass ratio (Fig. 3C). Gymno- concentration was significantly correlated with mean spermtreeshadthehighestmeanwhole-plantC/Nat transpiration rate (MTR) across all individuals (R2 5 64.7gg21,followedbyangiospermtreesat49.6gg21, 0.24, P , 0.0001, n 5 94). The equation relating the then by angiosperm lianas at 39.0 g g21. two was P 5 0.096MTR 1 2.75, where P is in area area Concentrationsofphosphorus(P),calcium(Ca),and mmol m22, andMTR isin mol m22 d21. potassium(K),andtheN/Pmassratioinleafdrymat- ter for each species are shown in Table IV. There was significant variation among functional groups (P , NUE 0.0001) and among species within functional groups (P,0.0001)forallelementsandforN/P.Angiosperm Equation2presentsameansforanalyzingvariation lianastendedtohavehighermeanconcentrationsofP, among functional groups and species in whole-plant Ca,andKintheirleafdrymatterthanangiospermand NUE (mol C mol21 N s21). We calculated NUE as the gymnosperm trees.The mean leaf N/P was higher in productofrandm /m ,thewhole-plantCtoNmolar c n angiospermtreesthaningymnospermtreesorangio- ratio; thus, a relatively high C/N has the effect of spermlianas;meanvalueswere9.3,4.9,and4.7gg21, increasing NUE. Figure 4A shows variation among respectively. speciesinNUE.Therewassignificantvariationamong Figure1. AtoC,Variationamongspe- cies in mean relative growth rate (A), netphotosynthesis,expressedonaleaf area basis (B), and leaf area per unit plantCmass,1/r(C).Errorbarsrepre- sent1SE.Samplesizesforeachspecies aregiveninTableIII. Plant Physiol. Vol. 148, 2008 647 Cernusak et al. leaves in gymnosperm trees compensated to some extent for their much lower A . However, the A was n n stillthe dominant controlover NUE (Fig.5). Transpiration Efficiency Mean values for each species for TE , the whole- c plant transpiration efficiency of C gain, are shown in Table V. Also shown in Table V are the growth- weighted estimates of the daytime vapor pressure deficit,D ,byspecies.Therewassignificantvariation, g bothamongfunctionalgroups(P,0.0001),andamong specieswithinfunctionalgroups(P,0.0001),inboth TE andD .However,acrossthefulldataset,TE and c g c D werenotsignificantlycorrelated(P50.11,n594), g suggestingthatD wasnotaprimarycontroloverTE . g c Taking the product of D and TE allows analysis of g c variation in TE independently of variation in D , as c g articulated in Equation 5. The D (cid:2)TE also varied g c significantly among functional groups (P , 0.0001) and among species within functional groups (P , 0.0001). Angiosperm trees had the highest mean D (cid:2)TE at 1.58 Pa mol C mol21 H O, followed by an- g c 2 giosperm lianas at 1.30 Pa mol C mol21 H O, then by 2 gymnosperm trees at 1.11 Pa mol C mol21 H O. 2 Among all species, there was a 3.7-fold variation in D (cid:2)TE (i.e.thelargestspeciesmeanwas3.7timesthe g c smallest speciesmean). Table V summarizes for each species the compo- nentsofD (cid:2)TE thatwequantified:f ,theratioofleaf- g c v to-air vapor pressure difference to air vapor pressure deficit; f , the ratio of unproductive to productive w water loss; and c/c , the ratio of intercellular to i a ambientCO partialpressuresduringphotosynthesis. 2 Althoughtherewasa1.8-foldvariationamongspecies Figure 2. Mean relative growth rate plotted against instantaneous in f , this parameter did not appear to be a primary measurementsofphotosynthesisexpressedonaleafmassbasis.White v controloverD (cid:2)TE :theterm1/f explainedonly13% symbols with internal cross-hairs refer to gymnosperm tree species; g c v of variation in D (cid:2)TE (R2 5 0.13, P 5 0.0004, n 5 94); completely white symbols refer to angiosperm liana species; black g c symbols and black symbols with internal cross-hairs refer to angio- moreover,theslopeoftherelationshipbetweenDg(cid:2)TEc spermtreespecies. and 1/f was negative, opposite to that predicted by v Equation 5. The parameter f similarly did not ap- w pear to exert a strong control over D (cid:2)TE : although g c functional groups (P , 0.0001) and among species D (cid:2)TE waspositivelycorrelatedwith1/(11f )(R25 g c w within functional groups (P , 0.0001). However, un- 0.18, P , 0.0001, n 5 92), there was only a 1.1-fold like results for r, angiosperm trees and lianas did not variationinthistermamongspecies,suggestingthatit differfromeachotherwithrespecttoNUE(P50.84). could only account for a variation in D (cid:2)TE of ap- g c Incontrast,NUEofgymnospermtreeswaslowerthan proximately 10%. In contrast, thec/c appearedtobe i a thatofbothangiospermtrees(P,0.0001)andangio- theprimarycontroloverD (cid:2)TE .Amongspecies,there g c sperm lianas (P , 0.0001). Figure 4, B and C, shows was a 2.3-fold variation in instantaneous measure- variation among species in the NUE components, A mentsof12c/c ,andc/c explained46%ofvariation n i a i a and n. Variation in NUE among species tended to inD (cid:2)TE .RegressioncoefficientsaregiveninTableVI. l g c reflect variation in A , the photosynthetic NUE (Fig. Furthermore,instantaneousmeasurementsof12c/c n i a 4B). The A was also higher in angiosperm trees and explained 64% of variation in the composite term n lianas than in gymnosperm trees (Fig. 4B). This vari- v (cid:2)TE (11f )(R250.64,P,0.0001,n594).Taking g c w ationinA wasoffsettoalesserextentbyvariationin this product means that only the variables f , c , and n c a n, the proportion of plant N allocated to leaves (Fig. c/c remainon theright side of Equation 5. l i a 4C). Gymnosperm trees had highest mean n, at 0.69, Variationininstantaneousmeasurementsofc/c was l i a followed by angiosperm trees at 0.59, then by angio- largely driven by variation in stomatal conductance, spermlianas at 0.50.Thus, ahigher allocation of Nto g, rather than by variation in photosynthesis, A. If g s s 648 Plant Physiol. Vol. 148, 2008 Growth, Water Use, and Stable Isotopes in Tropical Trees Figure3. AtoC,Variationamongspe- ciesintheCconcentrationofdrymatter onawhole-plantbasis(A),theNcon- centration of dry matter on a whole- plantbasis(B),andtheC/Nmassratio ofdrymatteronawhole-plantbasis(C). Errorbars represent 1 SE.Samplesizes foreachspeciesaregiveninTableIII. controlsvariationinc/c ,thenc/c shoulddecreaseas favorablywiththeinstantaneousmeasurementsoff i a i a v 1/g increases. The 1/g is equivalent to the stomatal (R2 5 0.42, P , 0.0001, n 5 55), thus providing some s s resistance. In contrast, if A controls c/c , then c/c validation of theformer. i a i a should decrease as A increases. Figure 6A shows that instantaneous c/c decreased as a linear function of i a Stable Isotope Composition 1/g. In contrast, Figure 6B shows that there was a s weaktendencyforc/c toincreaseasAincreased,op- The C isotope composition of leaves, stems, roots, i a posite to the trend that would be expected if Awere and whole plants is shown for each species in Table controlling c/c . VII. Also shown is the difference in d13C between i a We used measurements of leaf temperature, taken leavesandthesumofstemsplusroots,theheterotro- with a hand-held infrared thermometer, to calculate phic component of the plant. Across all individuals, values of f for the species harvested on the second leaf d13C was more negative than stem d13C (P , v andthirdharvestdates(TableIII).Wethencompared 0.0001, n 5 94) and root d13C (P , 0.0001, n 5 94), these instantaneous measurements of f with our whereas stem d13C was more negative than root d13C, v time-integrated estimates for each plant based on but by a much smaller amount (P 5 0.0008, n 5 94); leaf energy balance predictions and meteorological mean values for leaf, stem, and root d13C were229.4, data. The time-integrated estimates of f compared 228.1,and 227.8&, respectively. v TableIV. LeafP,Ca,andKconcentrations,andN/Pratiosofexperimentalplants Values are given as the mean for each species, with the SD in parentheses. No SD is given for P.guatemalensisbecauseonlyoneindividualofthisspeciessurvived.Samplesizesfortheotherspecies rangedfromfivetoeightindividuals,asshowninTableIII.NA,Notapplicable. Species P Ca K N/P gkg21 gkg21 gkg21 gg21 Gymnospermtreespecies C.lusitanica 3.10(0.76) 8.3(0.9) 20.9(2.8) 3.6(0.8) P.caribaea 1.43(0.47) 3.6(1.4) 9.6(1.6) 7.6(1.4) P.guatemalensis 2.48(NA) 8.1(NA) 16.9(NA) 5.9(NA) T.occidentalis 4.24(0.86) 10.5(2.2) 16.8(1.5) 3.4(0.8) Angiospermtreespecies C.longifolium 0.94(0.23) 7.7(0.8) 8.4(1.4) 12.4(1.5) C.pratensis 1.38(0.47) 13.6(1.1) 14.4(5.2) 9.7(2.1) H.alchorneoides 2.22(0.59) 12.5(3.9) 24.0(4.3) 6.8(1.5) L.seemannii 3.59(0.81) 18.0(2.1) 14.2(1.6) 5.0(1.4) P.pinnatum 1.68(0.23) 10.2(2.1) 16.7(4.3) 18.2(5.7) S.macrophylla 1.33(0.35) 13.4(1.5) 20.3(3.0) 12.0(2.8) T.rosea 1.43(0.04) 14.4(4.3) 15.1(4.4) 11.1(2.2) T.grandis 5.96(0.42) 8.6(1.0) 13.3(2.1) 2.0(0.2) Angiospermlianaspecies G.lupuloides 5.33(1.15) 14.4(2.1) 24.5(2.4) 4.4(0.9) M.leiostachya 2.66(0.30) 13.2(1.4) 29.5(3.5) 6.6(1.2) S.hypargyreum 6.23(1.09) 23.4(3.3) 24.6(4.0) 3.5(0.5) Plant Physiol. Vol. 148, 2008 649 Cernusak et al. Figure4. AtoC,Variationamongspe- ciesinwhole-plantNUE(A),photosyn- thetic NUE (B), and n, the leaf N l contentasaproportionofwhole-plant Ncontent(C).Errorbarsrepresent1SE. Samplesizesforeachspeciesaregiven inTableIII. We converted plant d13C values to 13C discrimina- as a linear function of 1/g (Fig. 6C). In contrast, the s tionbyassumingd13CofatmosphericCO tobe28&. D13C showed a weak tendency to increase as a func- 2 p Whole-plantD13C,D13C ,coveredarangefrom18.8& tionofA(Fig.6D),oppositetothetrendthatwouldbe p to22.9&amongspecies,correspondingtod13C values expected if Acontrolled variation inc/c . p i a rangingfrom226.3&to230.2&(TableVII).Therewas Variation among species in the O isotope composi- significantvariationinD13C amongfunctionalgroups tion of stem dry matter is given in Table VII. We p (P 5 0.002) and among species within functional calculated the 18O enrichment above source water of groups(P,0.0001).TheD13C waslowerinangiosperm stemdrymatter,D18O ,fromthemeand18Oofirrigation p p treesthan ingymnosperm trees,whereasangiosperm waterof24.3&.TheobservedD18O wassignificantly p lianas did not differ significantly from angiosperm or correlatedwiththepredicted18Oenrichmentofevap- gymnosperm trees. Mean values were 21.0&, 21.2&, orative site water, D18O , weighted by predicted e and 21.5& for angiosperm trees, angiosperm lianas, weeklygrowthincrements(Fig.9).Wetestedwhether and gymnosperm trees, respectively. theresidualvariationinD18O ,afteraccountingforvar- p The D13C was significantly correlated with instan- iation in D18O , was related to transpiration rate by p e taneousmeasurementsofc/c (Fig.7),aspredictedby plotting 1 2 [(D18O 2 e 2 e )/(1 2 p p )]/D18O i a p wc cp ex x e Equation 6. Weestimated the term d of Equation 6by against the MTR. As shown in Equation 12, this least-squaresregressionbyassumingfixedvaluesfora term should increasewith anincreasing transpiration and b of 4.4& and 29&, respectively. This resulted in rate if there is a significant Pe´clet effect. Our analysis an estimate for d of 3.1&; the regression equation detected a significant relationship between the two explained57%ofvariationinD13C .Thus,thepredic- p tive power of this relationship was equivalent to that obtained with a standard linear regression, in which both the slope and intercept are free to vary (Fig. 7). Using the mean estimate of 3.1& for d, and values of 4.4&and29&foraandb,respectively,wecalculateda D13C -based estimate of c/c for each plant. Mean p i a valuesoftheseestimatesforeachspeciesaregiven in TableV.Therewasa 2.4-fold variation amongspecies in the D13C -based estimates of 1 2c/c . p i a Variation in D13C was significantly correlated with p variation in D (cid:2)TE (Fig. 8); the former explained 49% g c of variation in the latter. Regression coefficients and thecoefficientofdeterminationforleast-squareslinear regressions of TE , D (cid:2)TE , and v (cid:2)TE against D13C of c g c g c leaves, stems, roots, and whole plants are given in Table VI. In general, whole-plant D13C was a better predictorofvariationinTE thanD13Cofleaves,stems, c or roots individually. Additionally, weighting of TE c byD orv tendedtoresultinmodestincreasesinthe g g Figure 5. Whole-plant NUE plotted against photosynthetic NUE. proportion of variation explained by the regression Whole-plant NUE was calculated from mean relative growth rate, models(TableVI). measured over several months, whereas photosynthetic NUE was Correlations between D13C and 1/g and A further p s calculated from instantaneous photosynthesis measurements, taken supported the conclusion that variation in ci/ca was over several minutes. Different symbols refer to different species, as largelydriven by variationin g. The D13C decreased detailedinFigure2. s p 650 Plant Physiol. Vol. 148, 2008 Growth, Water Use, and Stable Isotopes in Tropical Trees TableV. Transpirationefficiencyandrelatedparametersforeachspecies Symboldefinitionsareasfollows:transpirationefficiencyofCuptake(TE);growth-weighteddaytimevaporpressuredeficit(D)andleaf-to-air c g vaporpressuredifference(v);theratioofnighttimetodaytimetranspiration(E/E);andtheratioofintercellulartoambientCO partialpressures g n d 2 (c/c).Thec/c isgivenasthevaluemeasuredwithaportablephotosynthesissystem(instantaneous),orasthevalueestimatedfromwhole-plant i a i a 13Cdiscrimination(D13Cp-based).Valuesaregivenasthemeanforeachspecies,withtheSDinparentheses.NoSDisgivenforP.guatemalensis becauseonlyoneindividualofthisspeciessurvived.Samplesizesfortheotherspeciesrangedfromfivetoeightindividuals,asshowninTableIII.NA, Notapplicable. Instantaneous D13C-Based Species TE D v f (5v/D) f (5E/E) p c g g v g g w n d c/c c/c i a i a mmolC kPa kPa mol21HO 2 Gymnospermtreespecies C.lusitanica 1.29(0.20) 0.76(0.01) 0.90(0.07) 1.17(0.08) 0.11(0.02) 0.78(0.04) 0.80(0.02) P.caribaea 1.20(0.19) 0.76(0.01) 0.72(0.05) 0.95(0.07) 0.04(0.02) 0.86(0.03) 0.88(0.02) P.guatemalensis 3.48(NA) 1.04(NA) 1.71(NA) 1.64(NA) 0.04(NA) 0.63(NA) 0.63(NA) T.occidentalis 1.85(0.19) 0.76(0.01) 0.92(0.03) 1.21(0.05) 0.03(0.03) 0.80(0.02) 0.78(0.01) Angiospermtreespecies C.longifolium 1.83(0.24) 1.05(0.02) 1.53(0.12) 1.46(0.10) 0.04(0.04) 0.76(0.04) 0.78(0.05) C.pratensis 2.16(0.22) 1.02(0.01) 1.42(0.06) 1.39(0.06) 0.01(0.01) 0.76(0.03) 0.73(0.01) H.alchorneoides 1.50(0.20) 0.83(0.01) 1.37(0.04) 1.65(0.05) 0.08(0.01) 0.80(0.02) 0.85(0.01) L.seemannii 0.98(0.15) 1.20(0.00) 1.43(0.20) 1.19(0.16) 0.05(0.01) 0.80(0.03) 0.81(0.02) P.pinnatum 2.76(0.62) 1.12(0.10) 1.53(0.28) 1.39(0.36) 0.02(0.01) 0.73(0.09) 0.71(0.06) S.macrophylla 1.07(0.22) 1.05(0.01) 1.21(0.03) 1.15(0.03) 0.07(0.01) 0.86(0.02) 0.88(0.02) T.rosea 1.92(0.24) 0.82(0.00) 1.37(0.18) 1.67(0.21) 0.11(0.02) 0.75(0.06) 0.79(0.03) T.grandis 0.84(0.07) 0.98(0.01) 1.20(0.08) 1.23(0.09) 0.08(0.01) 0.88(0.03) 0.81(0.03) Angiospermlianaspecies G.lupuloides 0.96(0.34) 1.25(0.01) 1.37(0.08) 1.10(0.06) 0.04(0.01) 0.84(0.03) 0.80(0.02) M.leiostachya 0.89(0.23) 1.11(0.10) 1.39(0.20) 1.25(0.12) 0.07(0.02) 0.79(0.05) 0.80(0.06) S.hypargyreum 1.35(0.14) 1.21(0.01) 1.18(0.06) 0.97(0.05) 0.03(0.01) 0.82(0.03) 0.82(0.02) parameters(R250.14,P50.0002,n594),supporting areafor a given plantbiomass (Fig. 1). Of the compo- thenotionofasignificantPe´cleteffect,althoughthere nents of1/r,theSLA playeda keyrole,suchthatthe wasconsiderablescatterintherelationship.Usingthe product of SLA and instantaneous photosynthesis, A, MTR and E /E , we calculated a daytime MTR, then was a strong predictor of variation in r (Fig. 2). The n d used the nonlinear regression routine in SYSTAT to whole-plant NUE was mainly controlled by A , the n solve for an average value of L, the scaled effective photosyntheticrate foragiven amountof leaf N(Fig. path length, across the full data set. This analysis 5). An increase in the proportional allocation of N to estimated a mean value of L for the full data set of leaves, n, in species with low A was observed; how- l n 53mm,withthe95%confidenceintervalrangingfrom ever,theincreasedn compensatedtoonlyarelatively l 43to 62 mm. modestextentforlowA (Fig.4).Theprimarycontrol n overthetranspirationefficiencyofCuptake,TE ,was c c/c , the ratio of intercellular to ambient CO partial i a 2 pressuresduringphotosyntheticgasexchange(Tables DISCUSSION VandVI).Thec/c wasalsotheprimarycontrolover i a Inthisarticle,wepresentacomprehensivecompar- whole-plant 13C discrimination, D13C (Fig. 7), such p isonofphysiologicalprocessesinseedlingsofconifers, that variation in D13C was closely correlated with p angiosperm trees, and angiosperm lianas under trop- variation in TE (Table VI; Fig. 8). The c/c , in turn, c i a ical field conditions. The comparison yielded novel was largely controlled by stomatal conductance, g s insights into physiological differences among these (Fig.6).The18Oenrichmentofstemdrymatter,D18O , p functional groups, when grown in a tropical environ- was primarily controlled by the predicted 18O enrich- ment.Forexample,weobservedthatlianaspecies,on ment of the evaporative sites within leaves, D18O , e average,hadhigher1/rthantreespecies,andthatthis during photosynthetic gas exchange (Fig. 9). Variation trait was associated with faster growth. We also ob- inleaftranspirationratefurtherexplainedsomeofthe servedthatgymnospermtreeshadsignificantlylower residual variation in D18O not accounted for by vari- p whole-plant NUE than both angiosperm trees and ationinD18O . e angiospermlianas.Inaddition,theresultsprovidedan integrated account of the physiological controls over Growth and NUE growth, whole-plant water and NUE, and stable iso- tope composition across the full range of species. We observed that the term 1/r was the primary Relative growth rate, r, was mainly controlled by controlovervariationinr,andthatAwasarelatively variation in 1/r, the amount of assimilative surface conservative parameter among species (Fig. 1). These Plant Physiol. Vol. 148, 2008 651