Ecotypic responses of switchgrass to altered precipitation

11
Ecotypic responses of switchgrass to altered precipitation Jeffrey C. Hartman A,C , Jesse B. Nippert A and Clint J. Springer B A Division of Biology, Kansas State University, 116 Ackert Hall, Manhattan, KS 66506, USA. B Department of Biology, St. Josephs University, 5600 City Avenue, Philadelphia, PA 19131, USA. C Corresponding author. Email: [email protected] Abstract. Anthropogenic climate change is projected to alter precipitation patterns, resulting in novel environments for plants. The responses of dominant plant species (e.g. Panicum virgatum L. (switchgrass)) to climate changes can drive broader ecosystem processes such as primary productivity. Using a rainfall mesocosm facility, three ecotypes of P. virgatum (collected from Kansas, Oklahoma and Texas, USA) were subjected to three precipitation regimes (average, –25%, +25%) to determine the physiological and growth responses to altered precipitation in a common garden setting. Results showed mean maximum photosynthetic rates, stomatal conductance, transpiration, midday water potential and dark-adapted chlorophyll fluorescence were lowest in the Kansas ecotypes. Increased precipitation treatments raised the mean midday water potentials and lowered water-use efficiency. Aboveground biomass responded positively to changes in precipitation, but flowering initiation was later and rates were lower for Texas ecotypes. In general, ecotype origin was a better predictor of differences in physiological responses and flowering, whereas the precipitation treatments had greater control over biomass production. Depending on the growth variable measured, these results show responses for P. virgatum are under varying ecotypic or environmental control with few interactions, suggesting that future predictions to climate change need not inherently consider localised adaptations in this economically important and widely distributed species. Additional keywords: aboveground biomass, chlorophyll fluorescence, climate change, ecotype, gas exchange, Panicum virgatum. Received 12 October 2011, accepted 30 December 2011, published online 9 February 2012 Introduction Anthropogenic climate change is creating novel environments for plants in terrestrial ecosystems. Mean annual air temperatures for the central North American Great Plains have been projected to increase by 4 C by 2100 (Christensen et al. 2007), along with greater interannual and intra-annual variability in rainfall. Climate models for the Great Plains region predict shifts in precipitation from summer to winter months, resulting in an increased likelihood of growing season drought (Easterling et al. 2000; Meehl et al. 2005; Christensen et al. 2007). Longer drought periods are likely to result in highly variable soil moisture, especially in the upper soil layers (Knapp et al. 2002). Summer precipitation and soil water availability are key predictors of grassland annual net primary productivity (ANPP; Nippert et al. 2006); therefore, to understand the impacts of these forecasted changes on ecosystem processes, potential plant responses to experimental manipulations of climate conditions must be assessed. Within ecosystems, dominant species often influence community structure, dynamics, invasibility and ecosystem function (Grime 1998; Smith and Knapp 2003; Emery and Gross 2007). Dominant species generally have larger population sizes and higher intraspecific genetic variation than species with smaller populations in the same communities (McNaughton and Wolf 1970). Because of this, they will contribute to and affect ecosystems more than rarer species (Hillebrand et al. 2008; Hughes et al. 2008). Populations of dominant species with significant intraspecific genetic variation enable a broad range of physiological and growth responses to environmental change, making these species more capable of responding to alterations in environmental conditions (Norberg et al. 2001; Jump and Peñuelas 2005). Thus, in order to accurately predict the responses of ecosystems to future climate change, multiple ecotypes of a dominant species within a resource gradient should be examined (Callaway et al. 2003). Responses to short-term disturbances are largely governed by the plasticity of the species to environmental conditions, but if the variation in these plastic responses provides any selective advantage, then the plasticity has evolutionary consequences. Examining ecotypic variation in response to environmental manipulation in a dominant species, such as Panicum virgatum L. (switchgrass), provides insight into the traits most likely to respond to natural selection as the climate changes, as well as the impact that these changes may have on ecosystem functioning (Ward and Kelly 2004). P. virgatum is a common perennial C 4 grass of the tallgrass prairie ecosystems of the Central Great Plains, with a broad range of adaptation to growing conditions across North America (Parrish and Fike 2005; Hartman et al. 2011). Across North America, P. virgatum possesses large genotypic and phenotypic CSIRO PUBLISHING Functional Plant Biology, 2012, 39, 126–136 http://dx.doi.org/10.1071/FP11229 Journal compilation Ó CSIRO 2012 www.publish.csiro.au/journals/fpb

Transcript of Ecotypic responses of switchgrass to altered precipitation

Page 1: Ecotypic responses of switchgrass to altered precipitation

Ecotypic responses of switchgrass to altered precipitation

Jeffrey C HartmanAC Jesse B NippertA and Clint J SpringerB

ADivision of Biology Kansas State University 116 Ackert Hall Manhattan KS 66506 USABDepartment of Biology St Josephs University 5600 City Avenue Philadelphia PA 19131 USACCorresponding author Email jhartmanhuskersunledu

Abstract Anthropogenic climate change is projected to alter precipitation patterns resulting in novel environments forplants The responses of dominant plant species (eg Panicum virgatum L (switchgrass)) to climate changes can drivebroader ecosystemprocesses such as primary productivity Using a rainfallmesocosm facility three ecotypes ofP virgatum(collected fromKansasOklahomaandTexasUSA)were subjected to three precipitation regimes (averagendash25+25) todetermine the physiological and growth responses to altered precipitation in a commongarden setting Results showedmeanmaximum photosynthetic rates stomatal conductance transpiration midday water potential and dark-adapted chlorophyllfluorescencewere lowest in theKansas ecotypes Increased precipitation treatments raised themeanmiddaywater potentialsand lowered water-use efficiency Aboveground biomass responded positively to changes in precipitation but floweringinitiationwas later and rateswere lower for Texas ecotypes In general ecotype originwas a better predictor of differences inphysiological responses and flowering whereas the precipitation treatments had greater control over biomass productionDepending on the growth variable measured these results show responses for P virgatum are under varying ecotypic orenvironmental control with few interactions suggesting that future predictions to climate change need not inherentlyconsider localised adaptations in this economically important and widely distributed species

Additional keywords aboveground biomass chlorophyll fluorescence climate change ecotype gas exchangePanicum virgatum

Received 12 October 2011 accepted 30 December 2011 published online 9 February 2012

Introduction

Anthropogenic climate change is creating novel environments forplants in terrestrial ecosystemsMean annual air temperatures forthe central North American Great Plains have been projected toincrease by 4C by 2100 (Christensen et al 2007) along withgreater interannual and intra-annual variability in rainfallClimate models for the Great Plains region predict shifts inprecipitation from summer to winter months resulting in anincreased likelihood of growing season drought (Easterlinget al 2000 Meehl et al 2005 Christensen et al 2007)Longer drought periods are likely to result in highly variablesoil moisture especially in the upper soil layers (Knapp et al2002) Summer precipitation and soil water availability are keypredictors of grassland annual net primary productivity (ANPPNippert et al 2006) therefore to understand the impacts of theseforecasted changes on ecosystem processes potential plantresponses to experimental manipulations of climate conditionsmust be assessed

Within ecosystems dominant species often influencecommunity structure dynamics invasibility and ecosystemfunction (Grime 1998 Smith and Knapp 2003 Emery andGross 2007) Dominant species generally have largerpopulation sizes and higher intraspecific genetic variation thanspecies with smaller populations in the same communities(McNaughton and Wolf 1970) Because of this they will

contribute to and affect ecosystems more than rarer species(Hillebrand et al 2008 Hughes et al 2008) Populations ofdominant species with significant intraspecific genetic variationenable a broad range of physiological and growth responses toenvironmental change making these species more capable ofresponding to alterations in environmental conditions (Norberget al 2001 Jump andPentildeuelas 2005) Thus in order to accuratelypredict the responses of ecosystems to future climate changemultiple ecotypes of a dominant species within a resourcegradient should be examined (Callaway et al 2003)Responses to short-term disturbances are largely governed bythe plasticity of the species to environmental conditions but ifthe variation in these plastic responses provides any selectiveadvantage then the plasticity has evolutionary consequencesExamining ecotypic variation in response to environmentalmanipulation in a dominant species such as Panicumvirgatum L (switchgrass) provides insight into the traits mostlikely to respond to natural selection as the climate changes aswell as the impact that these changes may have on ecosystemfunctioning (Ward and Kelly 2004) P virgatum is a commonperennial C4 grass of the tallgrass prairie ecosystems of theCentral Great Plains with a broad range of adaptation togrowing conditions across North America (Parrish and Fike2005 Hartman et al 2011) Across North AmericaP virgatum possesses large genotypic and phenotypic

CSIRO PUBLISHING

Functional Plant Biology 2012 39 126ndash136httpdxdoiorg101071FP11229

Journal compilation CSIRO 2012 wwwpublishcsiroaujournalsfpb

variability (Casler et al 2004 Das et al 2004) that allows it to bebroadly adapted to a wide range of environmental conditions(Parrish and Fike 2005) In the past P virgatum has been used asforage and in the 1970s agronomic work focussed on increasingits forage value and yield (Berg 1971) In the last 15 years it hasbeen identified and studied extensively as a biofuel cropspecies (McLaughlin and Kszos 2005)Variable precipitationsignificantly affects physiological processes in individualplants with physiological responses that translate to ecosystemprocesses (Silletti and Knapp 2001 Knapp et al 2002 Fayet al 2008 Nippert et al 2009 Fay et al 2011) P virgatumproductivity is often co-limited by nitrogen andwater availability(Heaton et al 2004) Response to water availability variesacross ecotypes but in general P virgatum is less droughttolerant compared with other co-occurring C4 grass speciessuch as Andropogon gerardii Vitman Axonopus scoparius(Flugge) Kuhlm Dactylis glomerata (L) Sorghastrum nutans(L) Nash and Schizachyrium scoparium (Michx) Nash (Knapp1984Knapp1985Stout 1992Stout et al 1988Muir et al 2001Tucker et al 2011) Water availability is the dominantenvironmental control over individual plant tiller growthwhere individuals of P virgatum under water stress have beenshown to delay inflorescence development and decreasedreproductive output (Sanderson and Reed 2000) Water stresshas also been shown to reduce the biomass production ofP virgatum by up to 80 (Barney et al 2009) Because soilwater availability affects the rate of development of P virgatum(Sanderson 1992) this species is a likely candidate to exhibitaltered growth responses to a variable and probably drier futureclimate ForPvirgatum previous studies have focussedon singlegenotypes or agronomic cultivars (Evers and Parsons 2003Berdahl et al 2005 Barney et al 2009 Wang et al 2010)with little understanding of the differences among naturalpopulations to gradients in precipitation conditions

Therefore we conducted an experiment to characterise thephysiological responses and growth of three naturally occurringecotypes of P virgatum to changes in precipitation amount Byusing different ecotypes growing in a gradient of soil moistureconditions our objective was to assess the physiologicalresponses and differences in aboveground biomass amongecotypes across a precipitation gradient and determine theplasticity of responses These objectives will address whetherphysiological and growth responses vary according to ecotypeprecipitation treatment or ecotype precipitation interactions

Materials and methodsStudy site and mesocosm facilityThis research was conducted in the Rainfall Mesocosm Facilityat the Konza Prairie Biological Station (KPBS) KPBS is a3487 ha Long Term Ecological Research (LTER) site locatedin north-eastern Kansas USA (391N 969W) characterisedby a mid-continental climate with cool dry winters and warmwet summers Long-term annual precipitation is 860mm(1891ndash2006) with 75 of the rainfall occurring during thegrowing season (AprilndashSeptember) The rainfall mesocosmfacility contains 64 isolated 26m3 mesocosms Themesocosms were constructed in 2003 using plastic-linedplywood sheets and arranged in a pair of 2 16m arrays

underneath an 11 25m rainout shelter (Rainbow PlusStuppy Greenhouse Manufacturing North Kansas City MOUSA) The shelter has open walls and ends 24-m high eavesthat maximise air movement and heat dissipation and a roof ofclear corrugated polycarbonate (DynaGlas Plus SPSInternational San Jose CA USA) that allows gt90 lighttransmission Each mesocosm lsquocellrsquo (144m2 18m deep)contains a reconstructed soil profile from soil collected atKPBS Previously this facility was used for a precipitationvariability experiment with native tallgrass prairie plantcommunities (Fay et al 2008) During the summer (June toAugust) of 2008 all relic above- and belowground plantbiomass was removed by hand and the top 30 cm of soil washomogenised Subsequent volunteer plants that germinated wereweeded by hand without the application of herbicide

Panicum virgatum L rhizomes were randomly collected inthe early summer of 2008 from three geographically distinctnatural populations (hereafter ecotypes) in native tallgrassprairie These locations span a latitudinal gradient and includethe Konza Prairie in north-east Kansas the Tallgrass PrairiePreserve in north-east Oklahoma and native tallgrass prairie ineast central Texas (United States Department of Agriculture ndash

Agriculture Research Service landholdings near Temple TX)These locations have similar mean annual precipitationamounts and similar environmental histories (Table 1) Ateach site ~50 rhizomes of P virgatum were collected from 10distinct locations encompassing a range of site conditions(elevation aspect soil type and burning and grazing historieswhen applicable) This method of collection provided thegreatest amount of representative genetic variation withinthe range of site conditions across each of the three locationsThe three ecotypes were randomly assigned to the mesocosmfacility with each mesocosm cell containing rhizomescollected from the 10 sample locations of a single geographicecotype Rhizomes were planted with 40 cm spacing a distancethat facilitates a high number of tillers per plant but stillallows competition with neighbouring individuals (Sandersonand Reed 2000) During 2008 all mesocosms were wateredevery 3ndash5 days to promote establishment and minimise waterstress During May and June of 2009 additional individualsof P virgatum from Kansas and Oklahoma were planted toaugment the number of cells with switchgrass ecotypes Themesocosms were frequently weeded throughout the growingseason to maintain P virgatum species in each cell andwere watered as needed Of the 64 cells 21 containedindividuals from Konza Prairie (Kansas) 16 from

Table 1 General information for the sites from which each ecotypewas sampled

Konza Prairie precipitation and temperature averaged from 1891 to 2006Temple from 1914 to 2010 and Tallgrass Prairie Preserve from 1895 to 2010

Site MAP(mm)

Meantemperaturerange (C)

Latitude andlongitude

Area(ha)

Konza Prairie Kansas 860 ndash18ndash265 391N 969W 3487Tallgrass PrairiePreserve Oklahoma

877 12ndash278 3650N 9625W 15410

Temple Texas 878 92ndash289 3105N 9734W 178

Responses of switchgrass to altered precipitation Functional Plant Biology 127

Tallgrass Prairie Preserve (Oklahoma) and 16 from Temple(Texas)

Three precipitation treatments were implemented reflectingclimate change predictions for this region (Alley et al 2007)contemporary average for the growing season (62625mm) a25 increase and a 25 decrease Hereafter these treatments arereferred to as average increased and decreased For Kansasecotypes the three treatments had an equal number of cellseach (7) For the Oklahoma and Texas ecotypes there werefive cells each allocated to the increased and averagetreatments and six to the decreased treatment The unbalancedtreatment design resulted from the removal of a previouslyplanted (and unsuccessfully grown) northern ecotype Thetiming of rain events was not altered and precipitation wasapplied every 6 days which is the long-term average ambientrainfall interval Rainfall was applied through a metered handsprayer (model TM075 Great Plains Industries Wichita KSUSA) using water from an onsite well This water was applied atthe soil surface to minimise losses to canopy interception orrunoff Precipitation treatments were applied from late April toearly October 2010

Sampling procedure and variables measured

Sampling of the physiological variables was conducted on 10dates over the course of the growing season and classified intothree seasonal periods early season mid-season and late seasonData were collected on 28 May 31 May and 9 June for the earlyseason sampling period (Period 1) 30 June 22 July 23 July27 July and 30 July for the mid-season sampling period(Period 2) and 11 August and 26 August for the late seasonsampling period (Period 3) Sampling periods were chosenrather than monthly responses because previous work atKonza has shown the greatest increase in biomass inMayndashearly June (Period 1) allocation to reproductive growthduring late JunendashJuly (Period 2) and the lowest rates of growth inAugustndashSeptember (Period 3) (Knapp et al1998)

For each sampling date one individual was randomly selectedfrom each mesocosm cell and gas exchange dark-adaptedchlorophyll fluorescence (FvFm) and midday water potential(Ymid) were measured These physiological measurementswere conducted on the newest fully expanded mature leaf onthe widest portion of the grass blade For each sampling periodindividuals were tagged to ensure that all physiologicalmeasurements were conducted on the same individual leaf tominimise within-plant variability Because physiologicalmeasurements vary based on the time of day measurementorder for each population was randomised for each samplingperiod Gas exchange measurements were conducted between0900 hours and 1600 hours Central Standard Time when solarradiation was typically above 70 of full sun levels Soilmoisture (0ndash10 cm) data was collected concurrently with gasexchange and water potential measurements using a HydraProbe II Soil Sensor (Stevens Water Monitoring SystemsPortland OR USA) in units of water fraction by volume (wfv)

Gas exchange measurements were conducted using a LI-COR 6400 IRGA with an artificial redndashblue light-emittingdiode light source (6400ndash02B LI-COR Lincoln NE USA)Plants were placed inside the leaf chamber and allowed to reachsteady-state photosynthesis at mean ambient carbon (Ca)

(400mmolmol1) and at a saturating light intensity(2000mmolm2 s1) Leaf temperature was allowed to varywith ambient air temperature Relative humidity in the cuvettewas adjusted to reflect ambient environmental conditions(generally ranging from 30 to 50) The variables measuredincluded CO2 assimilation at ambient Ca (Amax) stomatalconductance to water vapour (gs) leaf-level transpiration (E)and instantaneouswater use efficiency (WUEAmaxE

ndash1)Middaywater potential (Ymid) was measured concurrently with gasexchange using a Scholander-type pressure bomb (PMSInstruments Albany OR USA) Ymid measurements wereconducted on the same selected individual used for gasexchange measurements however a different leaf was usedDark-adapted maximum photochemical efficiency (FvFm) wasassessed using a MINI-PAM photosynthesis yield analyser(Heinz Walz GmbH Effeltrich Germany) Measurementswere conducted on the same individuals within the samplingdate using the same leaf that was used for gas exchangemeasurements FvFm measurements were recorded during thenight Individuals were allowed to adapt for a minimum of 1 hafter complete darkness before measurement

Aboveground biomass was harvested dried at 65C for atleast 48 h and weighed for each individual at the conclusion ofthe growing season (SeptemberndashOctober) Individuals weremonitored for flowering tillers starting in mid-July and thetotal number of flowering tillers was counted on seven datesAs measures of fitness the reproductive biomass and tillernumbers were measured The flowering tillers from each plantwere separated and weighed The percent biomass allocated toreproduction was determined and used as a measure of fitnessOther characteristics such as flowering and non-flowering tillerheight and number of leaves per tiller were also measured

Statistical analyses

Changes in P virgatum physiological responses betweenecotypes and precipitation treatments were analysed using amixed effects model ANOVA (Proc Mixed SAS ver 92SAS Institute Cary NC USA) with the precipitationtreatments ecotypes and sampling period as fixed effects andthe specific mesocosm cell as a random effect Multiplecomparison tests between ecotypes precipitation treatmentsand sampling period were performed using Tukeyrsquos HonestlySignificant Difference test

Non-linear regression analysis was conducted to assess therelationship of gs toFvFm among ecotypes The relationship wasfitted using a two parameter logarithmic equation (y= y0 + a ln(abs(x)) Sigmaplot ver 110 Systat Software San Jose CAUSA) Residual analysis was performed to determine if themodelrsquos fit produced statistically significant regression modelsFitted models were then tested with ANOVA to determine ifsignificant differences existed among the model fits acrossecotypes (Sokal and Rohlf 1995)

Results

The precipitation treatments successfully altered the soilmoisture in this experiment Mean soil moisture wassignificantly higher in the increased precipitation treatment(024 0006wfv) compared with the average (P = 00012

128 Functional Plant Biology J C Hartman et al

021 0006wfv) and decreased treatments (Plt 00001018 0006wfv) and the average treatment was significantlygreater than the decreased treatment (P= 00358) Mean soilmoisture for all treatments at the beginning of the seasonranged from 030 to 040wfv at the end of the season therange was 010ndash015wfv The soil moisture for all treatmentsdecreased over the course of the season as ambient temperatureincreased

Physiology

Mean physiological responses varied according to ecotype(Table 2) Kansas ecotypes displayed significantly lower Amaxgs E Ymid and FvFm (Fig 1) compared with the Texas andOklahoma ecotypes To examine the potential for correlationbetween leaf-level gas exchange and leaf photochemistry asenvironmental conditions changed over the summer FvFmndashgsresponse curves were compared by ecotype (Fig 2) Fittedmodels were significantly different from zero for Kansas(P lt 00001) Oklahoma (P = 00011) and Texas (P lt 00001)FvFmndashgs response curves Kansas ecotypes exhibitedsignificantly lower (P = 00402) FvFm values as gs decreasedcompared with the Texas ecotype Fitted models did notsignificantly differ between Kansas and Oklahoma ecotypes(P = 01337) or between Texas and Oklahoma ecotypes(P = 05545) At an FvFm below 077 Texas and Oklahomaecotypes had similar rates of gs (near 0molmndash2 sndash1) but as gsincreased Texas ecotypes responded with higher FvFm

Among the physiological parameters only WUE and Ymid

varied due to precipitation treatments (Fig 1) MeanWUE in theincreased treatment was only 80 of the WUE in the decreasedtreatment and 72 of the WUE in the average precipitationtreatmentYmid was closely associated with soil moisture wherethe Ymid levels of individuals in the decreased treatment were15 lower than in the increased treatment and 5 lower thanthe average treatment Mean Ymid responses for the averageprecipitation treatment was 10 lower compared with theincreased treatment

All of the physiological parameters displayed significantdecreases over the course of the growing season (Table 2)Photosynthetic rates declined by 86 gs by 87 and Ymid by82 over the course of the growing season The effect ofprecipitation on Amax gs and E did depend on the samplingperiod (Table 2) Increased precipitation only increased Amaxgs andE in the early season period (Fig 3) In the mid-season andlate season sampling periods increased precipitation did notexhibit the same effect on gas exchange responses as it did inthe early season andonlydecreasedWUEduring themid-seasonLater in the season high temperatures negated the potentialpositive physiological responses of increased waterEcotype precipitation interactions were not observed in anyof the physiological responses (Table 2)

Aboveground biomass

Aboveground biomass responses varied by both ecotype andprecipitation individually with few interactions (Table 3) Totalbiomass total tillers flowering tiller biomass flowering tillerheight leaves per flowering tiller and the percent biomassallocated to reproductive tillers all showed a significant

Tab

le2

Mixed-effectsmod

elANOVAof

precipitationtreatm

entsan

decotyp

e(fixed

precipitation

ecotyp

eperiod

ran

domcell)on

physiologicalv

ariables

F-and

P-valuesareprovided

with

significantrespon

sesPlt005Plt010A

maxC

O2assimilatio

natam

bientC

O2g

sstom

atalconductancetowatervapourW

UEw

ater-useefficiencyF

vF

mdark-adapted

chloroph

yllfl

uorescenceY

midm

iddaywater

potential

Response

Precipitatio

nEcotype

Period

Precipitatio

n

Precipitatio

n

Ecotype

Precipitatio

necotyp

e

FP

FP

FP

ecotyp

eperiod

period

period

FP

FP

FP

FP

Amax

12

030

1065

00002

73

71

lt00001

15

019

285

00001

08

068

09

052

g s04

063

859

00007

63

64

lt00001

08

048

243

00013

05

090

10

045

WUE(A

maxEndash1)

441

00175

18

016

975

lt00001

16

016

156

00704

13

018

07

082

E13

026

783

00012

45

85

lt00001

11

035

201

00099

13

016

09

052

FvF

m13

027

791

00012

30

22

lt00001

01

095

08

055

14

016

04

098

Ymid

101

00003

495

00116

23

444

lt0

0001

18

014

10

042

156

007

88

09

052

Responses of switchgrass to altered precipitation Functional Plant Biology 129

response to the precipitation treatments (Table 4) Individuals inthe increased treatment had on average 15 more tillers anddouble the total aboveground and flowering tiller biomassThose individuals in the increased treatment also displayed a22 and 28 increase in mean tiller height compared with theaverage and decreased treatments respectively The increasedtreatment contained individuals with a greater number of leavesper tiller and that allocated over 94 of their biomass toreproductive tillers (Table 4)

Ecotype was a significant predictor of some abovegroundbiomass responses (Table 3) with differences in the total number

of tillers biomass per non-flowering tiller non-flowering tillerbiomass leaves per non-flowering tiller and percent biomassallocated to reproduction (Table 4) Texas ecotypes had amean ofless than 50 tillers per individual and over double the non-flowering tiller biomass and biomass per non-flowering tillercompared with Kansas and Oklahoma ecotypes Texas ecotypesalso displayed significantly less biomass allocation toreproductive tillers (75) compared with Kansas (87) andOklahoma ecotypes (94)

There were few ecotype precipitation interactions(Table 3) but the total biomass per tiller biomass per

EcotypeKansas Oklahoma Texas

Kansas Oklahoma Texas

Am

ax (

μ mol

mndash2

sndash1

)

0

5

10

15

20

25

(a)

Ecotype

g s (

mol

H2O

mndash2

sndash1

)E

(m

mol

mndash2

sndash1

)

WU

E (

μmol

mm

olndash1

)

000

002

004

006

008

010

012

014

Ecotype

0

1

2

3

4

5

6

Treatment

Decreased Average Increased0

1

2

3

4

5

Ecotype

FvF

m

073

074

075

076

077

078

079

080

081

Decre

ased

Avera

ge

Incr

ease

d

Kansa

s

Oklaho

ma

Texas

ψm

id (

MP

a)

ndash25

ndash20

ndash15

ndash10

ndash05

00

a

bb

a

b

b

a

bb

a

b b

b b

a

ab

c

ab b

Kansas Oklahoma Texas

Kansas Oklahoma Texas

(b)

(c) (d )

(e) (f )

Fig 1 Physiological responses for Panicum virgatum ecotypes across precipitation treatments Each point is the overall meanresponse (1 se) of the ecotype within the precipitation treatment (a) carbon assimilation at ambient CO2 (b) dark-adaptedchlorophyllfluorescence (c) stomatal conductance (d) water use efficiency (e) transpiration ( f )midday leaf-levelwater potentialIn (a) (b) (c) and (e) statistically significant (Plt 005) ecotype effects were present (varying letters among ecotypes denotesPlt 005) Significant precipitation treatment effects are present in (d) In ( f ) statistically significant ecotype and treatment effectswere present and significant differences are denoted by varying letters within each section of the panel

130 Functional Plant Biology J C Hartman et al

flowering tiller and the number of non-flowering tillers showedinteractions among ecotypes and precipitation treatments(Table 5) Texas ecotypes had significantly greater biomassper tiller in the increased treatments with the same results forbiomass per flowering tiller Kansas and Texas ecotypesdecreased the amount of non-flowering tillers as the soilmoisture increased Oklahoma ecotypes displayed the oppositetrend Oklahoma ecotypes had significantly fewer non-flowering

tillers in the decreased treatment compared with the other twoecotypes (Table 5)

Flowering

Flowering phenology was affected by both ecotype andprecipitation individually (Table 3) Texas ecotypes had halfthe number of flowering tillers at the end of season comparedwith Kansas and Oklahoma ecotypes (Table 4 Fig 4) Texasecotypes also initiated flowering a month later than the otherecotypes and flowered at a slower rate (Fig 4) Floweringphenology varied significantly with precipitation treatmentsIndividuals in the increased treatment had 40 and 38 moreflowering tillers at the end of the season compared with theaverage and decreased treatments respectively (Table 4Fig 4) Precipitation treatments affected flowering rates withthe increased treatment flowering at a greater rate compared withthe average and decreased treatment (Fig 4)

Discussion

The primary objective of this study was to characterise theecotype-specific physiological responses and growth ofP virgatum to changes in forecasted precipitation amount forthe Central Plains region of the United States by the year 2100Our results show that the precipitation treatments significantlyaffected soil moisture and the corresponding responses ofplant growth and physiology Significant effects of ecotype

gs (mol H2O mndash2 sndash1)

00 01 02 03 04 05

Fv

Fm

072

074

076

078

080

082

KansasOklahoma Texas

y = y0 + aln(abs(x))R-adj = 02442

Fig 2 Representative FvFmndashgs curves for all ecotypes Curves were fitfor all data points for each ecotype The R-adjusted value is for the overall fitof all ecotype data points

Early Mid Late Early Mid Late

Am

ax (

μmol

mndash2

sndash1

)E

(m

mol

mndash2

sndash1

)

0

10

20

30

40DecreasedAverageIncreased

000

005

010

015

020

025

Summer sampling period

0

2

4

6

8

10

Summer sampling period

WU

E (

μmol

mm

olndash1

)

0

1

2

3

4

5

6

a ab

a a

b

a a a

a a

b

a abb

a aa

a abb

aab

b

aab b

a a aa a

ba

aa

(a)

Early Mid LateEarly Mid Late

g s (

mol

H2O

mndash2

sndash1

)

(b)

(c) (d )

Fig 3 Gas exchange responses by treatment during the earlymid-season and late periods of the growing season (a)Carbon assimilationat ambient CO2 (b) stomatal conductance (c) transpiration (d) water use efficiency (Amax E

ndash1) Bars are means (1 se) and significantdifferences are indicated by letters within each sampling period

Responses of switchgrass to altered precipitation Functional Plant Biology 131

and precipitation were present but there were fewecotype precipitation interactions In general physiologicaldifferences reflected ecotype origin whereas differences inaboveground biomass largely reflected differences in soilwater availability from the precipitation treatments

Ecotype effects

Differences among means in the physiological variablesmeasured largely reflected differences by ecotype andcorresponding adaptation to the local environment of originWithin each ecotype similar changes in Amax gs and E reflectthe coupled gas exchange relationships documented previouslyfor other dominant prairie grasses (Polley et al 1992) It hasalso been shown that carbon assimilation in switchgrass iscontrolled mostly by population responses to localenvironmental conditions (Wullschleger et al 1996) In this

study the Kansas ecotypes had the lowest carbon assimilationrates providing support for the local environmental conditionresponse (Wullschleger et al 1996)

The FvFm ndash gs relationship developed (Fig 2) shows astatistical relationship between responses of leafphotochemistry to changes in stomatal conductance that variesby ecotypeReducedFvFmvalues indicate a decreased efficiencyof non-photochemical quenching and are reflective of thefunctioning of the light reactions of photosynthesis (Maxwelland Johnson 2000) Changes in FvFm can be especially useful instudies where physiological performance is strongly decreasedfrom longperiods of drought (Resco et al 2008) These data showthat nearly one-quarter of the variation themaximumefficiencyofPSII can be statistically attributed to changes in stomatalconductance (Fig 2) From first principles one would assumethat a reduction in stomatal water loss would result in increased

Table 3 Mixed-effects model ANOVA of precipitation treatments and ecotype (fixed precipitation ecotype random cell) effects on abovegroundbiomass variables

F- and P-values are provided with significant responses P lt 005 Plt 010

Response Precipitation Ecotype Ecotype precipitationF P F P F P

Tillers (per individual) 371 00326 338 00429 11 035Biomass (g per individual) 948 00004 21 012 15 021Biomass per tiller (g per individual) 449 00168 1109 00001 431 00050 allocated to reproduction (per individual) 483 00127 599 00050 191 012Flowering tillers (per individual) 754 00015 848 00008 05 068Flowering tiller biomass (g per individual) 1146 lt00001 08 044 15 019Biomass per flowering tiller (g per individual) 286 00681 1333 lt00001 351 00142Flowering tiller height (cm per individual) 512 00100 18 016 16 017Leaves per flowering tiller (per individual) 547 00075 22 011 11 035Non-flowering tillers (per individual) 325 00481 291 00651 287 00337Non-flowering tiller biomass (g per individual) 11 032 484 00126 05 067Biomass per non-flowering tiller (g per individual) 007 093 1585 lt00001 17 015Non-flowering tiller height (cm per individual) 09 038 305 00576 11 036Leaves per non-flowering tiller (per individual) 02 080 670 00029 18 013

Table 4 Mean responses (1 se) of various aboveground biomass measurements for ecotype and precipitation treatment effectsWithin the ecotype or treatment category bolded numbers show significance at the P 005 level and italicised numbers show marginal significance at theP 010 level Superscripts display significant differences between precipitation treatments or ecotypes within a category NS signifies no significance between

the ecotypes or the precipitation treatments

Variables Precipitation EcotypeDecreased Average Increased Kansas Oklahoma Texas

Tillers (per individual) 56 plusmn 5ab 47 plusmn 6b 71 plusmn 6a 69 plusmn 5a 58 plusmn 6ab 48 plusmn 6b

Biomass (g per individual) 202plusmn 28b 181 plusmn 30b 352 plusmn 30a NS NS NSBiomass per tiller (g per individual) 36 plusmn 03b 40 plusmn 03ab 50 plusmn 03a 30 plusmn 03b 44 plusmn 03a 52 plusmn 03a

Flowering tillers (per individual) 35 plusmn 4b 36 plusmn 4b 58 plusmn 4a 53 plusmn 4a 48 plusmn 5a 27 plusmn 5b

Flowering tiller biomass (g per individual) 157plusmn 27b 157 plusmn 28b 325 plusmn 28a NS NS NSBiomass per flowering tiller (g per individual) 41plusmn 04b 49plusmn 07ab 56plusmn 04a 32 plusmn 04b 50 plusmn 04a 64 plusmn 04a

Flowering tiller height (cm per individual) 94 plusmn 7b 101 plusmn 8b 129 plusmn 8a NS NS NSLeaves per flowering tiller (per individual) 47 plusmn 01b 51 plusmn 02ab 56 plusmn 02a NS NS NSNon-flowering tillers (per individual) 21 plusmn 2a 11 plusmn 3b 12 plusmn 3ab 15plusmn 2ab 9plusmn 3b 20plusmn 3a

Non-flowering tiller biomass (g per individual) NS NS NS 20 plusmn 9b 14 plusmn 11b 59 plusmn 11a

Biomass per non-flowering tiller (g per individual) NS NS NS 07 plusmn 01b 08 plusmn 01b 20 plusmn 01a

Non-flowering tiller height (cm per individual) NS NS NS 27plusmn 3ab 24plusmn 4b 38plusmn 4a

Leaves per non-flowering tiller (per individual) NS NS NS 32 plusmn 02b 30 plusmn 03b 44 plusmn 03a

allocated to reproduction (per individual) 77 plusmn 3b 85 plusmn 3ab 94 plusmn 3a 87 plusmn 3a 94 plusmn 4a 75 plusmn 4b

132 Functional Plant Biology J C Hartman et al

leaf temperature with the potential for corresponding decreasedphotochemistry during hot or dry summer conditions For allecotypesFvFm decreased nonlinearly with stomatal closure buttherewere clear differences in this relationship between ecotypes

with a general trend of increasing FvFm values for a given rate ofstomatal conductance from Kansas to Oklahoma to Texasecotypes Texas ecotypes had higher maximum FvFm valuescompared with Kansas ecotypes with values near the acceptedmaximal value of 083 (Krause and Weis 1991) As gs valuesdecreased Texas and Oklahoma ecotypes were able to maintainhigher FvFm values compared with the Kansas ecotypesindicating a greater decrease in the photochemical efficiencyof PSII in Kansas ecotypes Interpretation of the relationshipbetween FvFm and gs requires further investigation but thesedata show promise for linking changes in stomatal conductancewith themaintenance of leaf photochemistry as the ability to coolthe leaf through latent heat exchange becomes very limited withlow gs (Nippert et al 2009) and photochemistry is subsequentlyreducedby increased leaf temperatures (Albert et al 2011)Basedon theseFvFmndashgs response curves (Fig 2) Oklahoma and Texasecotypes are potentially more tolerant of higher leaf temperaturewith higher FvFm at low rates of gas exchange compared withKansas ecotypes

P virgatum is a highly photoperiod sensitive species(Benedict 1940) and differences in flowering dates have beendocumented for various ecotype and cultivars (McMillan 1965VanEsbroeck et al 2003Casler et al 2004) Sanderson andWolf(1995) recorded similar flowering dates for Alamo and Cave-in-Rock cultivars compared with the ecotypes in this study Alamocultivars flowered around the end of September and Cave-in-Rock cultivars flowered around the first week of July Theflowering time and rates seen in this study correspond withknown ecotype responses to changes in latitude

Many of the aboveground biomass responses with significantecotypic differences were a result of the Texas ecotypesproducing fewer tillers but with an increased biomass pertiller and number of leaves per tiller while the Oklahoma andKansas ecotypes had higher tiller density (Table 4) Typically thebest estimates to predict changes in biomass yield for this speciesare increased biomass per tiller and number of leaves per tiller(seen in Texas ecotypes) or higher tiller density (noted in the

Table 5 Mean responses (1 se) of various aboveground biomass measurements for interactions between ecotypes and precipitation treatmentsBolded numbers show significance at theP 005 level Superscripts display significant differences between ecotype precipitation combinations NS signifies

no significant difference between the means

Variables Ecotype precipitationDecreased Average Increased

Kansas Oklahoma Texas Kansas Oklahoma Texas Kansas Oklahoma Texas

Tillers (per individual) NS NS NS NS NS NS NS NS NSBiomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per tiller (g per individual) 28 plusmn 05b 42 plusmn 05b 38 plusmn 05b 29 plusmn 05b 50 plusmn 06ab 42 plusmn 06b 34 plusmn 05b 39 plusmn 06b 76 plusmn 06a

Flowering tillers (per individual) NS NS NS NS NS NS NS NS NSFlowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per flowering tiller (g per individual) 28 plusmn 07bc 44 plusmn 07bc 51 plusmn 07bc 31 plusmn 07bc 65 plusmn 08ab 52 plusmn 08abc 36 plusmn 07bc 42 plusmn 08bc 90 plusmn 08a

Flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per flowering tiller (per individual) NS NS NS NS NS NS NS NS NSNon-flowering tillers (per individual) 31 plusmn 4a 5 plusmn 5b 28 plusmn 5ab 8 plusmn 4b 7 plusmn 5ab 18 plusmn 5ab 7 plusmn 4b 15 plusmn 5ab 15 plusmn 5ab

Non-flowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per non-flowering tiller (g per individual) NS NS NS NS NS NS NS NS NSNon-flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per non-flowering tiller (per individual) NS NS NS NS NS NS NS NS NS allocated to reproduction (per individual) NS NS NS NS NS NS NS NS NS

Nov

Till

ers

flow

ered

(in

divndash1

)T

iller

s flo

wer

ed (

indi

vndash1)

0

10

20

30

40

50

60IncreasedAverageDecreased

0

10

20

30

40

50

60 Kansas Oklahoma Texas

(a)

(b)

Jul Aug Sep Oct

Fig 4 Total tillers flowered per individual over the course of the growingseason (a) Flowering tillers by ecotype (b) flowering tillers by precipitationtreatment Data points are means 1se Asterisks indicate statisticalsignificance (Plt 005) between groups within each sampling date

Responses of switchgrass to altered precipitation Functional Plant Biology 133

Oklahoma and Kansas ecotypes Boe and Beck 2008) Althoughtotal biomass was not significantly different among ecotypes theTexas ecotypes contain components (higher biomass and leavesper tiller) that if heritablemay lead to greater biomass inmultiplegeneration studies

Precipitation effects

The altered environmental conditions resulting from theprecipitation treatments significantly altered responses inaboveground biomass measurements with statisticallysignificant responses in the physiological variables largelyreflecting changes across the sampling periods (Fig 3)Flowering tiller heights were greatest in the increasedprecipitation treatments reaching mean heights of 129mThese heights are reflective of those observed in prairieremnant populations although they are lower than those foundin P virgatum cultivars (Das et al 2004 Casler 2005Alexopoulou et al 2008) but it should be noted that plantheights are variable from year to year (Alexopoulou et al2008) The average number of leaves per flowering tiller waslower than those found in agronomic P virgatum cultivars (VanEsbroeck et al 1997) however these cultivars have been bred tooptimise yield These data illustrate individual biomass responsesto precipitation of a dominant species in the tallgrass prairieecosystem further supporting the conclusions of others thatfuture changes in precipitation are likely to significantly alterecosystem productivity (Knapp 1984 Fay et al 2003 Fay et al2008 Zhou et al 2009)

Previous attempts to link leaf-level to whole-plant responsesfor grassland species have shown positive relationships betweenphotosynthesis and plant success (McAllister et al 1998) butleaf-level to whole-plant relationships in grasslands arecommonly weak (Nippert et al 2007) In this study no clearrelationships between leaf-level photosynthetic responses andtotal aboveground biomass were present for these ecotypesAlthough the increases in leaf-level photosynthesis did notcorrespond to total biomass increases other measurements ofbiomass increased from the increased precipitation partiallysupporting the positive relationship between higher rates ofgas exchange and plant success previously noted byMcAllister et al (1998)

The only physiological responses affected by precipitationtreatments were Ymid and WUE with WUE being lowest inthe increased treatments andYmid highest in increased treatmentsAlthough ANOVA results showed significant precipitationsample period interactions (Table 2) many of the effects wereonly seen in the early season period (Fig 3) Later in the seasonincreasing summer temperatures negated the positive impacts ofthe added water Decreased WUE across the growing seasoncorresponding to higher air temperatures has been previouslynoted for C4 species where WUE can be reduced by 40 indrought conditions (Taylor et al 2011) As water availabilitydecreasedYmid decreased to amean of ndash22MPa The decreasedprecipitation treatment did not lower soil water availability tothe point where the critical water potential (leaf pressurepotential when stomatal closure is less than 5 of themaximum) was reached for switchgrass (Ycrit = ndash32MPaTucker et al 2011) Indeed previous studies have shown that

switchgrass can maintain growth and functioning at lower waterpotentials than those recorded here (Knapp 1984 Stroup et al2003) With more severe decreases in water availability YmidAmax gs E and WUE should continue to decline with thepotential for increased cuticular conductance when drought-stressed (Manzoni et al 2011)

Ecotype precipitation interactions

Surprisingly few ecotype precipitation interactions werepresent for the variablesmeasured in this study (Tables 2 3) withthe primary exception being greater biomass per tiller in theincreased treatment for Texas ecotypes These differences largelyreflect the southern Texas ecotype being adapted to a longergrowing season (Newell 1968) Growth of the Texas ecotype in alocation with a shorter growing season delayed reproductivematurity and although the overall biomass yield was notsignificantly greater for Texas ecotypes compared withOklahoma and Kansas under greater precipitation differencesthe trends observed in the current study may become statisticallysignificant Because Texas ecotypes had a longer period in whichthey grew vegetatively theywere better able to take advantage ofthe increased precipitation treatment to produce higherbiomass per tiller greater biomass per flowering tiller and highnon-flowering tiller counts in all precipitation treatments Theseresults from a southern Texas ecotype growing in amore northernlocation are similar to other studies that show increases inbiomass longer photosynthetic activity and delayed floweringfor southern ecotypes grown in more northern latitudes (Quinn1969 Lemus et al 2002 Casler et al 2004 Berdahl et al 2005Casler et al 2007)

Conclusion

The physiological responses of Amax gs E FvFm and Ymid inP virgatum were explained by ecotype differences howeverrobust responses to altered precipitationwere seen inWUEYmidflowering times and many aboveground biomass variablesEcotype differences were also seen in several abovegroundbiomass variables and most strikingly in flowering initiationand rates In this study there were few interactions betweenecotype and precipitation amount suggesting that althoughprecipitation is a strong driver of biomass production localadaptations in the ecotypes used in the current study have lessimpact on these responses than initially expected It was expectedthat the environmental differences among the origins of theecotypes mainly the effects of higher temperature coupledwith similar mean annual precipitation (860ndash878mm) wouldlead to significant ecotype precipitation interactions in theparameters measured with the Texas ecotypes being the mosttolerant of the changes The lack of interactions between geneticbackground and precipitation also has broader evolutionaryimplications gene environment interactions are a primaryindicator of the potential for natural selection in a given trait(Schlichting 1989) The broader implications of the fewecotype precipitation interactions for this widely-distributedspecies suggest that it is not likely toundergo selection in responseto changes in precipitation As many have correctly suggested itis necessary to examine the role of genetic variation in theresponse of plant species to future changes in climate

134 Functional Plant Biology J C Hartman et al

(Callaway et al 2003 Ward and Kelly 2004) however asindicated by this study it may not always be necessary toconsider localised adaptation when working to project theimpacts of climate change on the performance of a species likeswitchgrass across its geographical distribution

Acknowledgements

We thankDr JohnBlair andDr TedMorgan for their advice and comments onthemanuscriptWealso thank theNatureConservancy for allowingcollectionof switchgrass rhizomes from the Tallgrass Prairie Preserve inOklahoma andlandholdingsnear Temple TXGracieOrozcoRebeccaOrozco JacobCarterand Teall Culbertson provided assistance with fieldwork and the Division ofBiology at Kansas State University Konza Prairie Biological Station KonzaPrairie LTER (DEB-0823341) and the US Department of Energy(EEE000314) provided funding and access to facilities

References

Albert KR Mikkelsen TN Michelsen A Ro-Poulsen H van der Linden L(2011) Interactive effects of drought elevated CO2 and warming onphotosynthetic capacity and photosystem performance in temperateheath plants Journal of Plant Physiology 168 1550ndash1561 doi101016jjplph201102011

Alexopoulou E Sharma N Papatheohari Y Christou M Piscioneri IPanoutsou D Pignatelli V (2008) Biomass yields for upland andlowland switchgrass varieties grown in the Mediterranean regionBiomass and Bioenergy 32 926ndash933 doi101016jbiombioe200801015

AlleyRBBerntsen T Bindoff NL ChenZ ChidthaisongA et al (2007) Hephysical science basis summary for policy makersrsquo (IPCC SecretariatGeneva)

Barney JN Mann JJ Kyser GB Blumwald B Deynze AV DiTomaso JM(2009)Toleranceof switchgrass to extremesoilmoisture stress ecologicalimplications Plant Science 177 724ndash732 doi101016jplantsci200909003

Benedict HM (1940) Effect of day length and temperature on the floweringand growth of four species of grasses Journal of Agricultural Research61 661ndash671

Berdahl JD Frank AB Krupinsky JM Carr PM Hanson JD Johnson HA(2005) Biomass yield phenology and survival of diverse switchgrasscultivars and experimental strains in western North Dakota AgronomyJournal 97 549ndash555 doi102134agronj20050549

Berg CC (1971) Forage yield of switchgrass (Panicum virgatum) inPennsylvania Agronomy Journal 63 785ndash786 doi102134agronj197100021962006300050038x

Boe A Beck DL (2008) Yield components of biomass in switchgrass CropScience 48 1306ndash1311 doi102135cropsci2007080482

Callaway RM Pennings SC Richards CL (2003) Phenotypic plasticity andinteractions among plants Ecology 84 1115ndash1128 doi1018900012-9658(2003)084[1115PPAIAP]20CO2

Casler MD (2005) Ecotypic variation among switchgrass populations fromthe northern USA Crop Science 45 388ndash398 doi102135cropsci20050388

Casler MD Vogel KP Taliaferro CM Wynia RL (2004) Latitudinaladaptation of switchgrass populations Crop Science 44 293ndash303

Casler MD Vogel KP Taliaferro CM Ehlke NJ Berdahl JD Brummer ECKallenbach RL West CP Mitchell RB (2007) Latitudinal andlongitudinal adaptation of switchgrass populations Crop Science 472249ndash2260 doi102135cropsci2006120780

Christensen JH Hewitson B Busuioc A Chen A Gao X Held I Jones RKolli RK KwonWT Laprise R Magantildea Rueda VMearns L MeneacutendezCG Raumlisaumlnen J Rinke A Sarr A (2007) Regional climate projectionsIn lsquoClimate change 2007 the physical science basis Contributionof Working Group I to the Fourth Assessment Report of the

Intergovernmental Panel on Climate Change (eds S Solomon D QinM Manning Z Chen M Marquis KB Averyt M Tignor HL Miller)pp 887ndash892 (Cambridge University Press Cambridge UK)

Das MK Fuentes RG Taliaferro CM (2004) Genetic variability and traitrelationships in switchgrass Crop Science 44 443ndash448

Easterling DR Meehl GA Parmesan C Changnon SA Karl TR Mearns LO(2000) Climate extremes observations modeling and impacts Science289 2068ndash2074 doi101126science28954872068

Emery SM Gross KL (2007) Dominant species identity not communityevenness regulates invasion in experimental grassland communitiesEcology 88 954ndash964 doi10189006-0568

Evers GW Parsons MJ (2003) Soil type and moisture level influence onAlamo switchgrass emergence and seedling growth Crop Science 43288ndash294 doi102135cropsci20030288

Fay PA Carlisle JD Knapp AK Blair JM Collins SL (2003) Productivityresponses to altered rainfall patterns in a C4-dominated grasslandOecologia 137 245ndash251 doi101007s00442-003-1331-3

FayPAKaufmanDMNippert JBCarlisle JDHarperCW(2008)Changes ingrassland ecosystem function due to extreme rainfall events implicationsfor responses to climate changeGlobal Change Biology 14 1600ndash1608doi101111j1365-2486200801605x

Fay PA Blair JM Smith MD Nippert JB Carlisle JD Knapp AK (2011)Relative effects of precipitation variability and warming on tallgrassprairie ecosystem function Biogeosciences 8 3053ndash3068 doi105194bg-8-3053-2011

Grime JP (1998) Benefits of plant diversity to ecosystems immediatefilter and founder effects Journal of Ecology 86 902ndash910doi101046j1365-2745199800306x

Hartman JC Nippert JB OrozcoRA Springer CJ (2011) Potential ecologicalimpacts of switchgrass (Panicum virgatum L) biofuel cultivation in theCentral Great Plains USA Biomass and Bioenergy 35 3415ndash3421doi101016jbiombioe201104055

Heaton E Voigt T Long SP (2004) A quantitative review comparing theyields of two candidate C4 perennial biomass crops in relation to nitrogentemperature and water Biomass and Bioenergy 27 21ndash30 doi101016jbiombioe200310005

HillebrandHBennettDMCadotteMW(2008)Consequences of dominancea review of evenness effects on local and regional ecosystem processesEcology 89 1510ndash1520 doi10189007-10531

Hughes AR Inouye BD Johnson MTJ Underwood N Vellend M (2008)Ecological consequences of genetic diversity Ecology Letters 11609ndash623 doi101111j1461-0248200801179x

Jump AS Pentildeuelas J (2005) Running to stand still adaptation and responseof plants to rapid climate change Ecology Letters 8 1010ndash1020doi101111j1461-0248200500796x

KnappAK (1984)Water relations and growth of three grasses duringwet anddrought years in a tallgrass prairie Oecologia 65 35ndash43 doi101007BF00384460

Knapp AK (1985) Effect of fire and drought on the ecophysiology ofAndropogon gerardii and Panicum virgatum in a tallgrass prairieEcology 66 1309ndash1320 doi1023071939184

KnappAKBriggs JMHartnett DCCollins SC (1998) lsquoGrassland dynamicslong-term ecological research in tallgrass prairiersquo (Oxford UniversityPress New York)

KnappAKFayPABlair JMCollinsSLSmithMDCarlisle JDHarperCWDanner BT Lett MS McCarron JK (2002) Rainfall variability carboncycling and plant species diversity in mesic grassland Science 2982202ndash2205 doi101126science1076347

Krause GH Weis E (1991) Chlorophyll fluorescence and photosynthesisthe basics Annual Review of Plant Physiology and Plant MolecularBiology 42 313ndash349 doi101146annurevpp42060191001525

Lemus R Brummer EC Moore KJ Molstad ME Burras CE Barker MF(2002) Biomass yield and quality of 20 switchgrass populations insouthern Iowa USA Biomass and Bioenergy 23 433ndash442doi101016S0961-9534(02)00073-9

Responses of switchgrass to altered precipitation Functional Plant Biology 135

Manzoni S Vico G Katul G Fay PA Polley W Palmroth S Porporato A(2011) Optimizing stomatal conductance for maximum carbon gainunder water stress a meta-analysis across plant functional types andclimates Functional Ecology 25 456ndash467 doi101111j1365-2435201001822x

Maxwell K JohnsonGN (2000) Chlorophyll fluorescence ndash a practical guideJournal of Experimental Botany 51 659ndash668 doi101093jexbot51345659

McAllister CA Knapp AK Maragni LA (1998) Is leaf-level photosynthesisrelated to plant success in a highly productive grasslandOecologia 11740ndash46 doi101007s004420050629

McLaughlin SB Kszos LN (2005) Development of switchgrass (Panicumvirgatum) as a bioenergy feedstock in the United States Biomass andBioenergy 28 515ndash535 doi101016jbiombioe200405006

McMillan C (1965) Ecotypic differentiation within four NorthAmerican prairie grasses II Behavioral variation within transplantedcommunity fractions American Journal of Botany 52 55ndash65doi1023072439975

McNaughton SJ Wolf LL (1970) Dominance and the niche in ecologicalsystems Science 167 131ndash139 doi101126science1673915131

Meehl GA Arblaster JM Tebaldi C (2005) Understanding future patterns ofincreased precipitation intensity in climate model simulationsGeophysical Research Letters 32 L18719 doi1010292005GL023680

Muir JP Sanderson MA Ocumpaugh WR Jones RM Reed RL (2001)Biomass production of lsquoAlamorsquo switchgrass in response to nitrogenphosphorus and row spacing Agronomy Journal 93 896ndash901doi102134agronj2001934896x

Newell LC (1968) Effects of strain source and management practice onforage yields of two warm-season prairie grasses Crop Science 8205ndash210 doi102135cropsci19680011183X000800020022x

Nippert JBKnappAKBriggs JM (2006) Intra-annual rainfall variability andgrassland productivity can the past predict the futurePlant Ecology 18465ndash74 doi101007s11258-005-9052-9

Nippert JB Fay PA Knapp AK (2007) Photosynthetic traits in C3 and C4

grassland species in mesocosm and field environments Environmentaland Experimental Botany 60 412ndash420 doi101016jenvexpbot200612012

Nippert JB Fay PA Carlisle JD Knapp AK Smith MD (2009)Ecophysiological responses of two dominant grasses to alteredtemperature and precipitation regimes Acta Oecologica 35 400ndash408doi101016jactao200901010

Norberg J Swaney DP Dushoff J Lin J Casagrandi R Levin SA (2001)Phenotypic diversity and ecosystem functioning in changingenvironments a theoretical framework Proceedings of the NationalAcademy of Sciences of the United States of America 9811376ndash11381 doi101073pnas171315998

Parrish DJ Fike JH (2005) The biology and agronomy of switchgrass forbiofuels Critical Reviews in Plant Sciences 24 423ndash459 doi10108007352680500316433

Polley HW Norman JM Arkebauer TJ Walter-Shea EA Greegor DH JrBramer B (1992) Leaf gas exchange of Andropogon gerardii VitmanPanicum virgatum L and Sorghastrum nutans (L) Nash in a tallgrassprairie Journal of Geophysical Research 97 18 837ndash18 844

Quinn JA (1969) Variability among High Plains populations of Panicumvirgatum Bulletin of the Torrey Botanical Club 96 20ndash41 doi1023072484005

Resco V IgnanceDD SunWHuxman TEWeltzin JFWilliams DG (2008)Chlorophyll fluorescence predawn water potential and photosynthesisin pulse-drive ecosystems ndash implications for ecological studiesFunctional Ecology 22 479ndash483 doi101111j1365-2435200801396x

Sanderson MA (1992) Morphological development of switchgrass andkleingrass Agronomy Journal 84 415ndash419 doi102134agronj199200021962008400030013x

Sanderson MA Reed RL (2000) Switchgrass growth and developmentwater nitrogen and plant density effects Journal of RangeManagement 53 221ndash227 doi1023074003287

SandersonMAWolf DD (1995)Morphological development of switchgrassin diverse environments Agronomy Journal 87 908ndash915 doi102134agronj199500021962008700050022x

Schlichting CD (1989) Phenotypic integration and environmental changeBioscience 39 460ndash464 doi1023071311138

SillettiAMKnappAK(2001)Responsesof the codominantgrassland speciesAndropogon gerardii and Sorghastrum nutans to long-termmanipulations of nitrogen and water American Midland Naturalist145 159ndash167 doi1016740003-0031(2001)145[0159ROTCGS]20CO2

SmithMDKnappAK (2003)Dominant speciesmaintain ecosystem functionwith non-random species loss Ecology Letters 6 509ndash517 doi101046j1461-0248200300454x

Sokal RRRohlf FJ (1995) lsquoBiometry the principles and practices of statisticsin biological research 3rd ednrsquo (WH Freeman New York)

StoutWL (1992)Water-use efficiency of grasses as affected by soil nitrogenand temperature Soil Science Society of America Journal 56 897ndash902doi102136sssaj199203615995005600030036x

Stout WL Jung GA Shaffer JA (1988) Effects of soil and nitrogen on wateruse efficiency of tall fescue and switchgrass under humid conditions SoilScience Society of America Journal 52 429ndash434 doi102136sssaj198803615995005200020023x

Stroup JA Sanderson MA Muir JP McFarland MJ Reed RL (2003)Comparison of growth and performance in upland and lowlandswitchgrass types to water and nitrogen stress Bioresource Technology86 65ndash72 doi101016S0960-8524(02)00102-5

Taylor SH Ripley BSWoodward FI Osborne CP (2011) Drought limitationofphotosynthesis differs betweenC3andC4grass species in a comparativeexperiment Plant Cell amp Environment 34 65ndash75 doi101111j1365-3040201002226x

Tucker SS Craine JM Nippert JB (2011) Physiological drought toleranceand the structuring of tallgrass prairie assemblages Ecosphere 2 1ndash19doi101890ES11-000231

Van Esbroeck GA Hussey MA Sanderson MA (1997) Leaf appearancerate and final leaf number of switchgrass cultivars Crop Science 37864ndash870 doi102135cropsci19970011183X003700030028x

Van Esbroeck GA Hussey MA Sanderson MA (2003) Variation betweenAlamo and Cave-in- Rock switchgrass in response to photoperiodextension Crop Science 43 639ndash643

WangDLeBauerDSDietzeMC(2010)Aquantitative reviewcomparing theyield of switchgrass in monocultures and mixtures in relation to climateand management factors GCB Bioenergy 2 16ndash25 doi101111j1757-1707201001035x

Ward JK Kelly JK (2004) Scaling up evolutionary responses to CO2 lessonsfrom Arabidopsis Ecology Letters 7 427ndash440 doi101111j1461-0248200400589x

Wullschleger SD SandersonMAMcLaughlin SB Biradar DP Rayburn AL(1996) Photosynthetic rates and ploidy levels among populations ofswitchgrassCrop Science 36 306ndash312 doi102135cropsci19960011183X003600020016x

Zhou X Talley M Luo Y (2009) Biomass litter and soil respiration along aprecipitation gradient in southern Great Plains USA Ecosystems(New York NY) 12 1369ndash1380 doi101007s10021-009-9296-7

136 Functional Plant Biology J C Hartman et al

wwwpublishcsiroaujournalsfpb

Page 2: Ecotypic responses of switchgrass to altered precipitation

variability (Casler et al 2004 Das et al 2004) that allows it to bebroadly adapted to a wide range of environmental conditions(Parrish and Fike 2005) In the past P virgatum has been used asforage and in the 1970s agronomic work focussed on increasingits forage value and yield (Berg 1971) In the last 15 years it hasbeen identified and studied extensively as a biofuel cropspecies (McLaughlin and Kszos 2005)Variable precipitationsignificantly affects physiological processes in individualplants with physiological responses that translate to ecosystemprocesses (Silletti and Knapp 2001 Knapp et al 2002 Fayet al 2008 Nippert et al 2009 Fay et al 2011) P virgatumproductivity is often co-limited by nitrogen andwater availability(Heaton et al 2004) Response to water availability variesacross ecotypes but in general P virgatum is less droughttolerant compared with other co-occurring C4 grass speciessuch as Andropogon gerardii Vitman Axonopus scoparius(Flugge) Kuhlm Dactylis glomerata (L) Sorghastrum nutans(L) Nash and Schizachyrium scoparium (Michx) Nash (Knapp1984Knapp1985Stout 1992Stout et al 1988Muir et al 2001Tucker et al 2011) Water availability is the dominantenvironmental control over individual plant tiller growthwhere individuals of P virgatum under water stress have beenshown to delay inflorescence development and decreasedreproductive output (Sanderson and Reed 2000) Water stresshas also been shown to reduce the biomass production ofP virgatum by up to 80 (Barney et al 2009) Because soilwater availability affects the rate of development of P virgatum(Sanderson 1992) this species is a likely candidate to exhibitaltered growth responses to a variable and probably drier futureclimate ForPvirgatum previous studies have focussedon singlegenotypes or agronomic cultivars (Evers and Parsons 2003Berdahl et al 2005 Barney et al 2009 Wang et al 2010)with little understanding of the differences among naturalpopulations to gradients in precipitation conditions

Therefore we conducted an experiment to characterise thephysiological responses and growth of three naturally occurringecotypes of P virgatum to changes in precipitation amount Byusing different ecotypes growing in a gradient of soil moistureconditions our objective was to assess the physiologicalresponses and differences in aboveground biomass amongecotypes across a precipitation gradient and determine theplasticity of responses These objectives will address whetherphysiological and growth responses vary according to ecotypeprecipitation treatment or ecotype precipitation interactions

Materials and methodsStudy site and mesocosm facilityThis research was conducted in the Rainfall Mesocosm Facilityat the Konza Prairie Biological Station (KPBS) KPBS is a3487 ha Long Term Ecological Research (LTER) site locatedin north-eastern Kansas USA (391N 969W) characterisedby a mid-continental climate with cool dry winters and warmwet summers Long-term annual precipitation is 860mm(1891ndash2006) with 75 of the rainfall occurring during thegrowing season (AprilndashSeptember) The rainfall mesocosmfacility contains 64 isolated 26m3 mesocosms Themesocosms were constructed in 2003 using plastic-linedplywood sheets and arranged in a pair of 2 16m arrays

underneath an 11 25m rainout shelter (Rainbow PlusStuppy Greenhouse Manufacturing North Kansas City MOUSA) The shelter has open walls and ends 24-m high eavesthat maximise air movement and heat dissipation and a roof ofclear corrugated polycarbonate (DynaGlas Plus SPSInternational San Jose CA USA) that allows gt90 lighttransmission Each mesocosm lsquocellrsquo (144m2 18m deep)contains a reconstructed soil profile from soil collected atKPBS Previously this facility was used for a precipitationvariability experiment with native tallgrass prairie plantcommunities (Fay et al 2008) During the summer (June toAugust) of 2008 all relic above- and belowground plantbiomass was removed by hand and the top 30 cm of soil washomogenised Subsequent volunteer plants that germinated wereweeded by hand without the application of herbicide

Panicum virgatum L rhizomes were randomly collected inthe early summer of 2008 from three geographically distinctnatural populations (hereafter ecotypes) in native tallgrassprairie These locations span a latitudinal gradient and includethe Konza Prairie in north-east Kansas the Tallgrass PrairiePreserve in north-east Oklahoma and native tallgrass prairie ineast central Texas (United States Department of Agriculture ndash

Agriculture Research Service landholdings near Temple TX)These locations have similar mean annual precipitationamounts and similar environmental histories (Table 1) Ateach site ~50 rhizomes of P virgatum were collected from 10distinct locations encompassing a range of site conditions(elevation aspect soil type and burning and grazing historieswhen applicable) This method of collection provided thegreatest amount of representative genetic variation withinthe range of site conditions across each of the three locationsThe three ecotypes were randomly assigned to the mesocosmfacility with each mesocosm cell containing rhizomescollected from the 10 sample locations of a single geographicecotype Rhizomes were planted with 40 cm spacing a distancethat facilitates a high number of tillers per plant but stillallows competition with neighbouring individuals (Sandersonand Reed 2000) During 2008 all mesocosms were wateredevery 3ndash5 days to promote establishment and minimise waterstress During May and June of 2009 additional individualsof P virgatum from Kansas and Oklahoma were planted toaugment the number of cells with switchgrass ecotypes Themesocosms were frequently weeded throughout the growingseason to maintain P virgatum species in each cell andwere watered as needed Of the 64 cells 21 containedindividuals from Konza Prairie (Kansas) 16 from

Table 1 General information for the sites from which each ecotypewas sampled

Konza Prairie precipitation and temperature averaged from 1891 to 2006Temple from 1914 to 2010 and Tallgrass Prairie Preserve from 1895 to 2010

Site MAP(mm)

Meantemperaturerange (C)

Latitude andlongitude

Area(ha)

Konza Prairie Kansas 860 ndash18ndash265 391N 969W 3487Tallgrass PrairiePreserve Oklahoma

877 12ndash278 3650N 9625W 15410

Temple Texas 878 92ndash289 3105N 9734W 178

Responses of switchgrass to altered precipitation Functional Plant Biology 127

Tallgrass Prairie Preserve (Oklahoma) and 16 from Temple(Texas)

Three precipitation treatments were implemented reflectingclimate change predictions for this region (Alley et al 2007)contemporary average for the growing season (62625mm) a25 increase and a 25 decrease Hereafter these treatments arereferred to as average increased and decreased For Kansasecotypes the three treatments had an equal number of cellseach (7) For the Oklahoma and Texas ecotypes there werefive cells each allocated to the increased and averagetreatments and six to the decreased treatment The unbalancedtreatment design resulted from the removal of a previouslyplanted (and unsuccessfully grown) northern ecotype Thetiming of rain events was not altered and precipitation wasapplied every 6 days which is the long-term average ambientrainfall interval Rainfall was applied through a metered handsprayer (model TM075 Great Plains Industries Wichita KSUSA) using water from an onsite well This water was applied atthe soil surface to minimise losses to canopy interception orrunoff Precipitation treatments were applied from late April toearly October 2010

Sampling procedure and variables measured

Sampling of the physiological variables was conducted on 10dates over the course of the growing season and classified intothree seasonal periods early season mid-season and late seasonData were collected on 28 May 31 May and 9 June for the earlyseason sampling period (Period 1) 30 June 22 July 23 July27 July and 30 July for the mid-season sampling period(Period 2) and 11 August and 26 August for the late seasonsampling period (Period 3) Sampling periods were chosenrather than monthly responses because previous work atKonza has shown the greatest increase in biomass inMayndashearly June (Period 1) allocation to reproductive growthduring late JunendashJuly (Period 2) and the lowest rates of growth inAugustndashSeptember (Period 3) (Knapp et al1998)

For each sampling date one individual was randomly selectedfrom each mesocosm cell and gas exchange dark-adaptedchlorophyll fluorescence (FvFm) and midday water potential(Ymid) were measured These physiological measurementswere conducted on the newest fully expanded mature leaf onthe widest portion of the grass blade For each sampling periodindividuals were tagged to ensure that all physiologicalmeasurements were conducted on the same individual leaf tominimise within-plant variability Because physiologicalmeasurements vary based on the time of day measurementorder for each population was randomised for each samplingperiod Gas exchange measurements were conducted between0900 hours and 1600 hours Central Standard Time when solarradiation was typically above 70 of full sun levels Soilmoisture (0ndash10 cm) data was collected concurrently with gasexchange and water potential measurements using a HydraProbe II Soil Sensor (Stevens Water Monitoring SystemsPortland OR USA) in units of water fraction by volume (wfv)

Gas exchange measurements were conducted using a LI-COR 6400 IRGA with an artificial redndashblue light-emittingdiode light source (6400ndash02B LI-COR Lincoln NE USA)Plants were placed inside the leaf chamber and allowed to reachsteady-state photosynthesis at mean ambient carbon (Ca)

(400mmolmol1) and at a saturating light intensity(2000mmolm2 s1) Leaf temperature was allowed to varywith ambient air temperature Relative humidity in the cuvettewas adjusted to reflect ambient environmental conditions(generally ranging from 30 to 50) The variables measuredincluded CO2 assimilation at ambient Ca (Amax) stomatalconductance to water vapour (gs) leaf-level transpiration (E)and instantaneouswater use efficiency (WUEAmaxE

ndash1)Middaywater potential (Ymid) was measured concurrently with gasexchange using a Scholander-type pressure bomb (PMSInstruments Albany OR USA) Ymid measurements wereconducted on the same selected individual used for gasexchange measurements however a different leaf was usedDark-adapted maximum photochemical efficiency (FvFm) wasassessed using a MINI-PAM photosynthesis yield analyser(Heinz Walz GmbH Effeltrich Germany) Measurementswere conducted on the same individuals within the samplingdate using the same leaf that was used for gas exchangemeasurements FvFm measurements were recorded during thenight Individuals were allowed to adapt for a minimum of 1 hafter complete darkness before measurement

Aboveground biomass was harvested dried at 65C for atleast 48 h and weighed for each individual at the conclusion ofthe growing season (SeptemberndashOctober) Individuals weremonitored for flowering tillers starting in mid-July and thetotal number of flowering tillers was counted on seven datesAs measures of fitness the reproductive biomass and tillernumbers were measured The flowering tillers from each plantwere separated and weighed The percent biomass allocated toreproduction was determined and used as a measure of fitnessOther characteristics such as flowering and non-flowering tillerheight and number of leaves per tiller were also measured

Statistical analyses

Changes in P virgatum physiological responses betweenecotypes and precipitation treatments were analysed using amixed effects model ANOVA (Proc Mixed SAS ver 92SAS Institute Cary NC USA) with the precipitationtreatments ecotypes and sampling period as fixed effects andthe specific mesocosm cell as a random effect Multiplecomparison tests between ecotypes precipitation treatmentsand sampling period were performed using Tukeyrsquos HonestlySignificant Difference test

Non-linear regression analysis was conducted to assess therelationship of gs toFvFm among ecotypes The relationship wasfitted using a two parameter logarithmic equation (y= y0 + a ln(abs(x)) Sigmaplot ver 110 Systat Software San Jose CAUSA) Residual analysis was performed to determine if themodelrsquos fit produced statistically significant regression modelsFitted models were then tested with ANOVA to determine ifsignificant differences existed among the model fits acrossecotypes (Sokal and Rohlf 1995)

Results

The precipitation treatments successfully altered the soilmoisture in this experiment Mean soil moisture wassignificantly higher in the increased precipitation treatment(024 0006wfv) compared with the average (P = 00012

128 Functional Plant Biology J C Hartman et al

021 0006wfv) and decreased treatments (Plt 00001018 0006wfv) and the average treatment was significantlygreater than the decreased treatment (P= 00358) Mean soilmoisture for all treatments at the beginning of the seasonranged from 030 to 040wfv at the end of the season therange was 010ndash015wfv The soil moisture for all treatmentsdecreased over the course of the season as ambient temperatureincreased

Physiology

Mean physiological responses varied according to ecotype(Table 2) Kansas ecotypes displayed significantly lower Amaxgs E Ymid and FvFm (Fig 1) compared with the Texas andOklahoma ecotypes To examine the potential for correlationbetween leaf-level gas exchange and leaf photochemistry asenvironmental conditions changed over the summer FvFmndashgsresponse curves were compared by ecotype (Fig 2) Fittedmodels were significantly different from zero for Kansas(P lt 00001) Oklahoma (P = 00011) and Texas (P lt 00001)FvFmndashgs response curves Kansas ecotypes exhibitedsignificantly lower (P = 00402) FvFm values as gs decreasedcompared with the Texas ecotype Fitted models did notsignificantly differ between Kansas and Oklahoma ecotypes(P = 01337) or between Texas and Oklahoma ecotypes(P = 05545) At an FvFm below 077 Texas and Oklahomaecotypes had similar rates of gs (near 0molmndash2 sndash1) but as gsincreased Texas ecotypes responded with higher FvFm

Among the physiological parameters only WUE and Ymid

varied due to precipitation treatments (Fig 1) MeanWUE in theincreased treatment was only 80 of the WUE in the decreasedtreatment and 72 of the WUE in the average precipitationtreatmentYmid was closely associated with soil moisture wherethe Ymid levels of individuals in the decreased treatment were15 lower than in the increased treatment and 5 lower thanthe average treatment Mean Ymid responses for the averageprecipitation treatment was 10 lower compared with theincreased treatment

All of the physiological parameters displayed significantdecreases over the course of the growing season (Table 2)Photosynthetic rates declined by 86 gs by 87 and Ymid by82 over the course of the growing season The effect ofprecipitation on Amax gs and E did depend on the samplingperiod (Table 2) Increased precipitation only increased Amaxgs andE in the early season period (Fig 3) In the mid-season andlate season sampling periods increased precipitation did notexhibit the same effect on gas exchange responses as it did inthe early season andonlydecreasedWUEduring themid-seasonLater in the season high temperatures negated the potentialpositive physiological responses of increased waterEcotype precipitation interactions were not observed in anyof the physiological responses (Table 2)

Aboveground biomass

Aboveground biomass responses varied by both ecotype andprecipitation individually with few interactions (Table 3) Totalbiomass total tillers flowering tiller biomass flowering tillerheight leaves per flowering tiller and the percent biomassallocated to reproductive tillers all showed a significant

Tab

le2

Mixed-effectsmod

elANOVAof

precipitationtreatm

entsan

decotyp

e(fixed

precipitation

ecotyp

eperiod

ran

domcell)on

physiologicalv

ariables

F-and

P-valuesareprovided

with

significantrespon

sesPlt005Plt010A

maxC

O2assimilatio

natam

bientC

O2g

sstom

atalconductancetowatervapourW

UEw

ater-useefficiencyF

vF

mdark-adapted

chloroph

yllfl

uorescenceY

midm

iddaywater

potential

Response

Precipitatio

nEcotype

Period

Precipitatio

n

Precipitatio

n

Ecotype

Precipitatio

necotyp

e

FP

FP

FP

ecotyp

eperiod

period

period

FP

FP

FP

FP

Amax

12

030

1065

00002

73

71

lt00001

15

019

285

00001

08

068

09

052

g s04

063

859

00007

63

64

lt00001

08

048

243

00013

05

090

10

045

WUE(A

maxEndash1)

441

00175

18

016

975

lt00001

16

016

156

00704

13

018

07

082

E13

026

783

00012

45

85

lt00001

11

035

201

00099

13

016

09

052

FvF

m13

027

791

00012

30

22

lt00001

01

095

08

055

14

016

04

098

Ymid

101

00003

495

00116

23

444

lt0

0001

18

014

10

042

156

007

88

09

052

Responses of switchgrass to altered precipitation Functional Plant Biology 129

response to the precipitation treatments (Table 4) Individuals inthe increased treatment had on average 15 more tillers anddouble the total aboveground and flowering tiller biomassThose individuals in the increased treatment also displayed a22 and 28 increase in mean tiller height compared with theaverage and decreased treatments respectively The increasedtreatment contained individuals with a greater number of leavesper tiller and that allocated over 94 of their biomass toreproductive tillers (Table 4)

Ecotype was a significant predictor of some abovegroundbiomass responses (Table 3) with differences in the total number

of tillers biomass per non-flowering tiller non-flowering tillerbiomass leaves per non-flowering tiller and percent biomassallocated to reproduction (Table 4) Texas ecotypes had amean ofless than 50 tillers per individual and over double the non-flowering tiller biomass and biomass per non-flowering tillercompared with Kansas and Oklahoma ecotypes Texas ecotypesalso displayed significantly less biomass allocation toreproductive tillers (75) compared with Kansas (87) andOklahoma ecotypes (94)

There were few ecotype precipitation interactions(Table 3) but the total biomass per tiller biomass per

EcotypeKansas Oklahoma Texas

Kansas Oklahoma Texas

Am

ax (

μ mol

mndash2

sndash1

)

0

5

10

15

20

25

(a)

Ecotype

g s (

mol

H2O

mndash2

sndash1

)E

(m

mol

mndash2

sndash1

)

WU

E (

μmol

mm

olndash1

)

000

002

004

006

008

010

012

014

Ecotype

0

1

2

3

4

5

6

Treatment

Decreased Average Increased0

1

2

3

4

5

Ecotype

FvF

m

073

074

075

076

077

078

079

080

081

Decre

ased

Avera

ge

Incr

ease

d

Kansa

s

Oklaho

ma

Texas

ψm

id (

MP

a)

ndash25

ndash20

ndash15

ndash10

ndash05

00

a

bb

a

b

b

a

bb

a

b b

b b

a

ab

c

ab b

Kansas Oklahoma Texas

Kansas Oklahoma Texas

(b)

(c) (d )

(e) (f )

Fig 1 Physiological responses for Panicum virgatum ecotypes across precipitation treatments Each point is the overall meanresponse (1 se) of the ecotype within the precipitation treatment (a) carbon assimilation at ambient CO2 (b) dark-adaptedchlorophyllfluorescence (c) stomatal conductance (d) water use efficiency (e) transpiration ( f )midday leaf-levelwater potentialIn (a) (b) (c) and (e) statistically significant (Plt 005) ecotype effects were present (varying letters among ecotypes denotesPlt 005) Significant precipitation treatment effects are present in (d) In ( f ) statistically significant ecotype and treatment effectswere present and significant differences are denoted by varying letters within each section of the panel

130 Functional Plant Biology J C Hartman et al

flowering tiller and the number of non-flowering tillers showedinteractions among ecotypes and precipitation treatments(Table 5) Texas ecotypes had significantly greater biomassper tiller in the increased treatments with the same results forbiomass per flowering tiller Kansas and Texas ecotypesdecreased the amount of non-flowering tillers as the soilmoisture increased Oklahoma ecotypes displayed the oppositetrend Oklahoma ecotypes had significantly fewer non-flowering

tillers in the decreased treatment compared with the other twoecotypes (Table 5)

Flowering

Flowering phenology was affected by both ecotype andprecipitation individually (Table 3) Texas ecotypes had halfthe number of flowering tillers at the end of season comparedwith Kansas and Oklahoma ecotypes (Table 4 Fig 4) Texasecotypes also initiated flowering a month later than the otherecotypes and flowered at a slower rate (Fig 4) Floweringphenology varied significantly with precipitation treatmentsIndividuals in the increased treatment had 40 and 38 moreflowering tillers at the end of the season compared with theaverage and decreased treatments respectively (Table 4Fig 4) Precipitation treatments affected flowering rates withthe increased treatment flowering at a greater rate compared withthe average and decreased treatment (Fig 4)

Discussion

The primary objective of this study was to characterise theecotype-specific physiological responses and growth ofP virgatum to changes in forecasted precipitation amount forthe Central Plains region of the United States by the year 2100Our results show that the precipitation treatments significantlyaffected soil moisture and the corresponding responses ofplant growth and physiology Significant effects of ecotype

gs (mol H2O mndash2 sndash1)

00 01 02 03 04 05

Fv

Fm

072

074

076

078

080

082

KansasOklahoma Texas

y = y0 + aln(abs(x))R-adj = 02442

Fig 2 Representative FvFmndashgs curves for all ecotypes Curves were fitfor all data points for each ecotype The R-adjusted value is for the overall fitof all ecotype data points

Early Mid Late Early Mid Late

Am

ax (

μmol

mndash2

sndash1

)E

(m

mol

mndash2

sndash1

)

0

10

20

30

40DecreasedAverageIncreased

000

005

010

015

020

025

Summer sampling period

0

2

4

6

8

10

Summer sampling period

WU

E (

μmol

mm

olndash1

)

0

1

2

3

4

5

6

a ab

a a

b

a a a

a a

b

a abb

a aa

a abb

aab

b

aab b

a a aa a

ba

aa

(a)

Early Mid LateEarly Mid Late

g s (

mol

H2O

mndash2

sndash1

)

(b)

(c) (d )

Fig 3 Gas exchange responses by treatment during the earlymid-season and late periods of the growing season (a)Carbon assimilationat ambient CO2 (b) stomatal conductance (c) transpiration (d) water use efficiency (Amax E

ndash1) Bars are means (1 se) and significantdifferences are indicated by letters within each sampling period

Responses of switchgrass to altered precipitation Functional Plant Biology 131

and precipitation were present but there were fewecotype precipitation interactions In general physiologicaldifferences reflected ecotype origin whereas differences inaboveground biomass largely reflected differences in soilwater availability from the precipitation treatments

Ecotype effects

Differences among means in the physiological variablesmeasured largely reflected differences by ecotype andcorresponding adaptation to the local environment of originWithin each ecotype similar changes in Amax gs and E reflectthe coupled gas exchange relationships documented previouslyfor other dominant prairie grasses (Polley et al 1992) It hasalso been shown that carbon assimilation in switchgrass iscontrolled mostly by population responses to localenvironmental conditions (Wullschleger et al 1996) In this

study the Kansas ecotypes had the lowest carbon assimilationrates providing support for the local environmental conditionresponse (Wullschleger et al 1996)

The FvFm ndash gs relationship developed (Fig 2) shows astatistical relationship between responses of leafphotochemistry to changes in stomatal conductance that variesby ecotypeReducedFvFmvalues indicate a decreased efficiencyof non-photochemical quenching and are reflective of thefunctioning of the light reactions of photosynthesis (Maxwelland Johnson 2000) Changes in FvFm can be especially useful instudies where physiological performance is strongly decreasedfrom longperiods of drought (Resco et al 2008) These data showthat nearly one-quarter of the variation themaximumefficiencyofPSII can be statistically attributed to changes in stomatalconductance (Fig 2) From first principles one would assumethat a reduction in stomatal water loss would result in increased

Table 3 Mixed-effects model ANOVA of precipitation treatments and ecotype (fixed precipitation ecotype random cell) effects on abovegroundbiomass variables

F- and P-values are provided with significant responses P lt 005 Plt 010

Response Precipitation Ecotype Ecotype precipitationF P F P F P

Tillers (per individual) 371 00326 338 00429 11 035Biomass (g per individual) 948 00004 21 012 15 021Biomass per tiller (g per individual) 449 00168 1109 00001 431 00050 allocated to reproduction (per individual) 483 00127 599 00050 191 012Flowering tillers (per individual) 754 00015 848 00008 05 068Flowering tiller biomass (g per individual) 1146 lt00001 08 044 15 019Biomass per flowering tiller (g per individual) 286 00681 1333 lt00001 351 00142Flowering tiller height (cm per individual) 512 00100 18 016 16 017Leaves per flowering tiller (per individual) 547 00075 22 011 11 035Non-flowering tillers (per individual) 325 00481 291 00651 287 00337Non-flowering tiller biomass (g per individual) 11 032 484 00126 05 067Biomass per non-flowering tiller (g per individual) 007 093 1585 lt00001 17 015Non-flowering tiller height (cm per individual) 09 038 305 00576 11 036Leaves per non-flowering tiller (per individual) 02 080 670 00029 18 013

Table 4 Mean responses (1 se) of various aboveground biomass measurements for ecotype and precipitation treatment effectsWithin the ecotype or treatment category bolded numbers show significance at the P 005 level and italicised numbers show marginal significance at theP 010 level Superscripts display significant differences between precipitation treatments or ecotypes within a category NS signifies no significance between

the ecotypes or the precipitation treatments

Variables Precipitation EcotypeDecreased Average Increased Kansas Oklahoma Texas

Tillers (per individual) 56 plusmn 5ab 47 plusmn 6b 71 plusmn 6a 69 plusmn 5a 58 plusmn 6ab 48 plusmn 6b

Biomass (g per individual) 202plusmn 28b 181 plusmn 30b 352 plusmn 30a NS NS NSBiomass per tiller (g per individual) 36 plusmn 03b 40 plusmn 03ab 50 plusmn 03a 30 plusmn 03b 44 plusmn 03a 52 plusmn 03a

Flowering tillers (per individual) 35 plusmn 4b 36 plusmn 4b 58 plusmn 4a 53 plusmn 4a 48 plusmn 5a 27 plusmn 5b

Flowering tiller biomass (g per individual) 157plusmn 27b 157 plusmn 28b 325 plusmn 28a NS NS NSBiomass per flowering tiller (g per individual) 41plusmn 04b 49plusmn 07ab 56plusmn 04a 32 plusmn 04b 50 plusmn 04a 64 plusmn 04a

Flowering tiller height (cm per individual) 94 plusmn 7b 101 plusmn 8b 129 plusmn 8a NS NS NSLeaves per flowering tiller (per individual) 47 plusmn 01b 51 plusmn 02ab 56 plusmn 02a NS NS NSNon-flowering tillers (per individual) 21 plusmn 2a 11 plusmn 3b 12 plusmn 3ab 15plusmn 2ab 9plusmn 3b 20plusmn 3a

Non-flowering tiller biomass (g per individual) NS NS NS 20 plusmn 9b 14 plusmn 11b 59 plusmn 11a

Biomass per non-flowering tiller (g per individual) NS NS NS 07 plusmn 01b 08 plusmn 01b 20 plusmn 01a

Non-flowering tiller height (cm per individual) NS NS NS 27plusmn 3ab 24plusmn 4b 38plusmn 4a

Leaves per non-flowering tiller (per individual) NS NS NS 32 plusmn 02b 30 plusmn 03b 44 plusmn 03a

allocated to reproduction (per individual) 77 plusmn 3b 85 plusmn 3ab 94 plusmn 3a 87 plusmn 3a 94 plusmn 4a 75 plusmn 4b

132 Functional Plant Biology J C Hartman et al

leaf temperature with the potential for corresponding decreasedphotochemistry during hot or dry summer conditions For allecotypesFvFm decreased nonlinearly with stomatal closure buttherewere clear differences in this relationship between ecotypes

with a general trend of increasing FvFm values for a given rate ofstomatal conductance from Kansas to Oklahoma to Texasecotypes Texas ecotypes had higher maximum FvFm valuescompared with Kansas ecotypes with values near the acceptedmaximal value of 083 (Krause and Weis 1991) As gs valuesdecreased Texas and Oklahoma ecotypes were able to maintainhigher FvFm values compared with the Kansas ecotypesindicating a greater decrease in the photochemical efficiencyof PSII in Kansas ecotypes Interpretation of the relationshipbetween FvFm and gs requires further investigation but thesedata show promise for linking changes in stomatal conductancewith themaintenance of leaf photochemistry as the ability to coolthe leaf through latent heat exchange becomes very limited withlow gs (Nippert et al 2009) and photochemistry is subsequentlyreducedby increased leaf temperatures (Albert et al 2011)Basedon theseFvFmndashgs response curves (Fig 2) Oklahoma and Texasecotypes are potentially more tolerant of higher leaf temperaturewith higher FvFm at low rates of gas exchange compared withKansas ecotypes

P virgatum is a highly photoperiod sensitive species(Benedict 1940) and differences in flowering dates have beendocumented for various ecotype and cultivars (McMillan 1965VanEsbroeck et al 2003Casler et al 2004) Sanderson andWolf(1995) recorded similar flowering dates for Alamo and Cave-in-Rock cultivars compared with the ecotypes in this study Alamocultivars flowered around the end of September and Cave-in-Rock cultivars flowered around the first week of July Theflowering time and rates seen in this study correspond withknown ecotype responses to changes in latitude

Many of the aboveground biomass responses with significantecotypic differences were a result of the Texas ecotypesproducing fewer tillers but with an increased biomass pertiller and number of leaves per tiller while the Oklahoma andKansas ecotypes had higher tiller density (Table 4) Typically thebest estimates to predict changes in biomass yield for this speciesare increased biomass per tiller and number of leaves per tiller(seen in Texas ecotypes) or higher tiller density (noted in the

Table 5 Mean responses (1 se) of various aboveground biomass measurements for interactions between ecotypes and precipitation treatmentsBolded numbers show significance at theP 005 level Superscripts display significant differences between ecotype precipitation combinations NS signifies

no significant difference between the means

Variables Ecotype precipitationDecreased Average Increased

Kansas Oklahoma Texas Kansas Oklahoma Texas Kansas Oklahoma Texas

Tillers (per individual) NS NS NS NS NS NS NS NS NSBiomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per tiller (g per individual) 28 plusmn 05b 42 plusmn 05b 38 plusmn 05b 29 plusmn 05b 50 plusmn 06ab 42 plusmn 06b 34 plusmn 05b 39 plusmn 06b 76 plusmn 06a

Flowering tillers (per individual) NS NS NS NS NS NS NS NS NSFlowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per flowering tiller (g per individual) 28 plusmn 07bc 44 plusmn 07bc 51 plusmn 07bc 31 plusmn 07bc 65 plusmn 08ab 52 plusmn 08abc 36 plusmn 07bc 42 plusmn 08bc 90 plusmn 08a

Flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per flowering tiller (per individual) NS NS NS NS NS NS NS NS NSNon-flowering tillers (per individual) 31 plusmn 4a 5 plusmn 5b 28 plusmn 5ab 8 plusmn 4b 7 plusmn 5ab 18 plusmn 5ab 7 plusmn 4b 15 plusmn 5ab 15 plusmn 5ab

Non-flowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per non-flowering tiller (g per individual) NS NS NS NS NS NS NS NS NSNon-flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per non-flowering tiller (per individual) NS NS NS NS NS NS NS NS NS allocated to reproduction (per individual) NS NS NS NS NS NS NS NS NS

Nov

Till

ers

flow

ered

(in

divndash1

)T

iller

s flo

wer

ed (

indi

vndash1)

0

10

20

30

40

50

60IncreasedAverageDecreased

0

10

20

30

40

50

60 Kansas Oklahoma Texas

(a)

(b)

Jul Aug Sep Oct

Fig 4 Total tillers flowered per individual over the course of the growingseason (a) Flowering tillers by ecotype (b) flowering tillers by precipitationtreatment Data points are means 1se Asterisks indicate statisticalsignificance (Plt 005) between groups within each sampling date

Responses of switchgrass to altered precipitation Functional Plant Biology 133

Oklahoma and Kansas ecotypes Boe and Beck 2008) Althoughtotal biomass was not significantly different among ecotypes theTexas ecotypes contain components (higher biomass and leavesper tiller) that if heritablemay lead to greater biomass inmultiplegeneration studies

Precipitation effects

The altered environmental conditions resulting from theprecipitation treatments significantly altered responses inaboveground biomass measurements with statisticallysignificant responses in the physiological variables largelyreflecting changes across the sampling periods (Fig 3)Flowering tiller heights were greatest in the increasedprecipitation treatments reaching mean heights of 129mThese heights are reflective of those observed in prairieremnant populations although they are lower than those foundin P virgatum cultivars (Das et al 2004 Casler 2005Alexopoulou et al 2008) but it should be noted that plantheights are variable from year to year (Alexopoulou et al2008) The average number of leaves per flowering tiller waslower than those found in agronomic P virgatum cultivars (VanEsbroeck et al 1997) however these cultivars have been bred tooptimise yield These data illustrate individual biomass responsesto precipitation of a dominant species in the tallgrass prairieecosystem further supporting the conclusions of others thatfuture changes in precipitation are likely to significantly alterecosystem productivity (Knapp 1984 Fay et al 2003 Fay et al2008 Zhou et al 2009)

Previous attempts to link leaf-level to whole-plant responsesfor grassland species have shown positive relationships betweenphotosynthesis and plant success (McAllister et al 1998) butleaf-level to whole-plant relationships in grasslands arecommonly weak (Nippert et al 2007) In this study no clearrelationships between leaf-level photosynthetic responses andtotal aboveground biomass were present for these ecotypesAlthough the increases in leaf-level photosynthesis did notcorrespond to total biomass increases other measurements ofbiomass increased from the increased precipitation partiallysupporting the positive relationship between higher rates ofgas exchange and plant success previously noted byMcAllister et al (1998)

The only physiological responses affected by precipitationtreatments were Ymid and WUE with WUE being lowest inthe increased treatments andYmid highest in increased treatmentsAlthough ANOVA results showed significant precipitationsample period interactions (Table 2) many of the effects wereonly seen in the early season period (Fig 3) Later in the seasonincreasing summer temperatures negated the positive impacts ofthe added water Decreased WUE across the growing seasoncorresponding to higher air temperatures has been previouslynoted for C4 species where WUE can be reduced by 40 indrought conditions (Taylor et al 2011) As water availabilitydecreasedYmid decreased to amean of ndash22MPa The decreasedprecipitation treatment did not lower soil water availability tothe point where the critical water potential (leaf pressurepotential when stomatal closure is less than 5 of themaximum) was reached for switchgrass (Ycrit = ndash32MPaTucker et al 2011) Indeed previous studies have shown that

switchgrass can maintain growth and functioning at lower waterpotentials than those recorded here (Knapp 1984 Stroup et al2003) With more severe decreases in water availability YmidAmax gs E and WUE should continue to decline with thepotential for increased cuticular conductance when drought-stressed (Manzoni et al 2011)

Ecotype precipitation interactions

Surprisingly few ecotype precipitation interactions werepresent for the variablesmeasured in this study (Tables 2 3) withthe primary exception being greater biomass per tiller in theincreased treatment for Texas ecotypes These differences largelyreflect the southern Texas ecotype being adapted to a longergrowing season (Newell 1968) Growth of the Texas ecotype in alocation with a shorter growing season delayed reproductivematurity and although the overall biomass yield was notsignificantly greater for Texas ecotypes compared withOklahoma and Kansas under greater precipitation differencesthe trends observed in the current study may become statisticallysignificant Because Texas ecotypes had a longer period in whichthey grew vegetatively theywere better able to take advantage ofthe increased precipitation treatment to produce higherbiomass per tiller greater biomass per flowering tiller and highnon-flowering tiller counts in all precipitation treatments Theseresults from a southern Texas ecotype growing in amore northernlocation are similar to other studies that show increases inbiomass longer photosynthetic activity and delayed floweringfor southern ecotypes grown in more northern latitudes (Quinn1969 Lemus et al 2002 Casler et al 2004 Berdahl et al 2005Casler et al 2007)

Conclusion

The physiological responses of Amax gs E FvFm and Ymid inP virgatum were explained by ecotype differences howeverrobust responses to altered precipitationwere seen inWUEYmidflowering times and many aboveground biomass variablesEcotype differences were also seen in several abovegroundbiomass variables and most strikingly in flowering initiationand rates In this study there were few interactions betweenecotype and precipitation amount suggesting that althoughprecipitation is a strong driver of biomass production localadaptations in the ecotypes used in the current study have lessimpact on these responses than initially expected It was expectedthat the environmental differences among the origins of theecotypes mainly the effects of higher temperature coupledwith similar mean annual precipitation (860ndash878mm) wouldlead to significant ecotype precipitation interactions in theparameters measured with the Texas ecotypes being the mosttolerant of the changes The lack of interactions between geneticbackground and precipitation also has broader evolutionaryimplications gene environment interactions are a primaryindicator of the potential for natural selection in a given trait(Schlichting 1989) The broader implications of the fewecotype precipitation interactions for this widely-distributedspecies suggest that it is not likely toundergo selection in responseto changes in precipitation As many have correctly suggested itis necessary to examine the role of genetic variation in theresponse of plant species to future changes in climate

134 Functional Plant Biology J C Hartman et al

(Callaway et al 2003 Ward and Kelly 2004) however asindicated by this study it may not always be necessary toconsider localised adaptation when working to project theimpacts of climate change on the performance of a species likeswitchgrass across its geographical distribution

Acknowledgements

We thankDr JohnBlair andDr TedMorgan for their advice and comments onthemanuscriptWealso thank theNatureConservancy for allowingcollectionof switchgrass rhizomes from the Tallgrass Prairie Preserve inOklahoma andlandholdingsnear Temple TXGracieOrozcoRebeccaOrozco JacobCarterand Teall Culbertson provided assistance with fieldwork and the Division ofBiology at Kansas State University Konza Prairie Biological Station KonzaPrairie LTER (DEB-0823341) and the US Department of Energy(EEE000314) provided funding and access to facilities

References

Albert KR Mikkelsen TN Michelsen A Ro-Poulsen H van der Linden L(2011) Interactive effects of drought elevated CO2 and warming onphotosynthetic capacity and photosystem performance in temperateheath plants Journal of Plant Physiology 168 1550ndash1561 doi101016jjplph201102011

Alexopoulou E Sharma N Papatheohari Y Christou M Piscioneri IPanoutsou D Pignatelli V (2008) Biomass yields for upland andlowland switchgrass varieties grown in the Mediterranean regionBiomass and Bioenergy 32 926ndash933 doi101016jbiombioe200801015

AlleyRBBerntsen T Bindoff NL ChenZ ChidthaisongA et al (2007) Hephysical science basis summary for policy makersrsquo (IPCC SecretariatGeneva)

Barney JN Mann JJ Kyser GB Blumwald B Deynze AV DiTomaso JM(2009)Toleranceof switchgrass to extremesoilmoisture stress ecologicalimplications Plant Science 177 724ndash732 doi101016jplantsci200909003

Benedict HM (1940) Effect of day length and temperature on the floweringand growth of four species of grasses Journal of Agricultural Research61 661ndash671

Berdahl JD Frank AB Krupinsky JM Carr PM Hanson JD Johnson HA(2005) Biomass yield phenology and survival of diverse switchgrasscultivars and experimental strains in western North Dakota AgronomyJournal 97 549ndash555 doi102134agronj20050549

Berg CC (1971) Forage yield of switchgrass (Panicum virgatum) inPennsylvania Agronomy Journal 63 785ndash786 doi102134agronj197100021962006300050038x

Boe A Beck DL (2008) Yield components of biomass in switchgrass CropScience 48 1306ndash1311 doi102135cropsci2007080482

Callaway RM Pennings SC Richards CL (2003) Phenotypic plasticity andinteractions among plants Ecology 84 1115ndash1128 doi1018900012-9658(2003)084[1115PPAIAP]20CO2

Casler MD (2005) Ecotypic variation among switchgrass populations fromthe northern USA Crop Science 45 388ndash398 doi102135cropsci20050388

Casler MD Vogel KP Taliaferro CM Wynia RL (2004) Latitudinaladaptation of switchgrass populations Crop Science 44 293ndash303

Casler MD Vogel KP Taliaferro CM Ehlke NJ Berdahl JD Brummer ECKallenbach RL West CP Mitchell RB (2007) Latitudinal andlongitudinal adaptation of switchgrass populations Crop Science 472249ndash2260 doi102135cropsci2006120780

Christensen JH Hewitson B Busuioc A Chen A Gao X Held I Jones RKolli RK KwonWT Laprise R Magantildea Rueda VMearns L MeneacutendezCG Raumlisaumlnen J Rinke A Sarr A (2007) Regional climate projectionsIn lsquoClimate change 2007 the physical science basis Contributionof Working Group I to the Fourth Assessment Report of the

Intergovernmental Panel on Climate Change (eds S Solomon D QinM Manning Z Chen M Marquis KB Averyt M Tignor HL Miller)pp 887ndash892 (Cambridge University Press Cambridge UK)

Das MK Fuentes RG Taliaferro CM (2004) Genetic variability and traitrelationships in switchgrass Crop Science 44 443ndash448

Easterling DR Meehl GA Parmesan C Changnon SA Karl TR Mearns LO(2000) Climate extremes observations modeling and impacts Science289 2068ndash2074 doi101126science28954872068

Emery SM Gross KL (2007) Dominant species identity not communityevenness regulates invasion in experimental grassland communitiesEcology 88 954ndash964 doi10189006-0568

Evers GW Parsons MJ (2003) Soil type and moisture level influence onAlamo switchgrass emergence and seedling growth Crop Science 43288ndash294 doi102135cropsci20030288

Fay PA Carlisle JD Knapp AK Blair JM Collins SL (2003) Productivityresponses to altered rainfall patterns in a C4-dominated grasslandOecologia 137 245ndash251 doi101007s00442-003-1331-3

FayPAKaufmanDMNippert JBCarlisle JDHarperCW(2008)Changes ingrassland ecosystem function due to extreme rainfall events implicationsfor responses to climate changeGlobal Change Biology 14 1600ndash1608doi101111j1365-2486200801605x

Fay PA Blair JM Smith MD Nippert JB Carlisle JD Knapp AK (2011)Relative effects of precipitation variability and warming on tallgrassprairie ecosystem function Biogeosciences 8 3053ndash3068 doi105194bg-8-3053-2011

Grime JP (1998) Benefits of plant diversity to ecosystems immediatefilter and founder effects Journal of Ecology 86 902ndash910doi101046j1365-2745199800306x

Hartman JC Nippert JB OrozcoRA Springer CJ (2011) Potential ecologicalimpacts of switchgrass (Panicum virgatum L) biofuel cultivation in theCentral Great Plains USA Biomass and Bioenergy 35 3415ndash3421doi101016jbiombioe201104055

Heaton E Voigt T Long SP (2004) A quantitative review comparing theyields of two candidate C4 perennial biomass crops in relation to nitrogentemperature and water Biomass and Bioenergy 27 21ndash30 doi101016jbiombioe200310005

HillebrandHBennettDMCadotteMW(2008)Consequences of dominancea review of evenness effects on local and regional ecosystem processesEcology 89 1510ndash1520 doi10189007-10531

Hughes AR Inouye BD Johnson MTJ Underwood N Vellend M (2008)Ecological consequences of genetic diversity Ecology Letters 11609ndash623 doi101111j1461-0248200801179x

Jump AS Pentildeuelas J (2005) Running to stand still adaptation and responseof plants to rapid climate change Ecology Letters 8 1010ndash1020doi101111j1461-0248200500796x

KnappAK (1984)Water relations and growth of three grasses duringwet anddrought years in a tallgrass prairie Oecologia 65 35ndash43 doi101007BF00384460

Knapp AK (1985) Effect of fire and drought on the ecophysiology ofAndropogon gerardii and Panicum virgatum in a tallgrass prairieEcology 66 1309ndash1320 doi1023071939184

KnappAKBriggs JMHartnett DCCollins SC (1998) lsquoGrassland dynamicslong-term ecological research in tallgrass prairiersquo (Oxford UniversityPress New York)

KnappAKFayPABlair JMCollinsSLSmithMDCarlisle JDHarperCWDanner BT Lett MS McCarron JK (2002) Rainfall variability carboncycling and plant species diversity in mesic grassland Science 2982202ndash2205 doi101126science1076347

Krause GH Weis E (1991) Chlorophyll fluorescence and photosynthesisthe basics Annual Review of Plant Physiology and Plant MolecularBiology 42 313ndash349 doi101146annurevpp42060191001525

Lemus R Brummer EC Moore KJ Molstad ME Burras CE Barker MF(2002) Biomass yield and quality of 20 switchgrass populations insouthern Iowa USA Biomass and Bioenergy 23 433ndash442doi101016S0961-9534(02)00073-9

Responses of switchgrass to altered precipitation Functional Plant Biology 135

Manzoni S Vico G Katul G Fay PA Polley W Palmroth S Porporato A(2011) Optimizing stomatal conductance for maximum carbon gainunder water stress a meta-analysis across plant functional types andclimates Functional Ecology 25 456ndash467 doi101111j1365-2435201001822x

Maxwell K JohnsonGN (2000) Chlorophyll fluorescence ndash a practical guideJournal of Experimental Botany 51 659ndash668 doi101093jexbot51345659

McAllister CA Knapp AK Maragni LA (1998) Is leaf-level photosynthesisrelated to plant success in a highly productive grasslandOecologia 11740ndash46 doi101007s004420050629

McLaughlin SB Kszos LN (2005) Development of switchgrass (Panicumvirgatum) as a bioenergy feedstock in the United States Biomass andBioenergy 28 515ndash535 doi101016jbiombioe200405006

McMillan C (1965) Ecotypic differentiation within four NorthAmerican prairie grasses II Behavioral variation within transplantedcommunity fractions American Journal of Botany 52 55ndash65doi1023072439975

McNaughton SJ Wolf LL (1970) Dominance and the niche in ecologicalsystems Science 167 131ndash139 doi101126science1673915131

Meehl GA Arblaster JM Tebaldi C (2005) Understanding future patterns ofincreased precipitation intensity in climate model simulationsGeophysical Research Letters 32 L18719 doi1010292005GL023680

Muir JP Sanderson MA Ocumpaugh WR Jones RM Reed RL (2001)Biomass production of lsquoAlamorsquo switchgrass in response to nitrogenphosphorus and row spacing Agronomy Journal 93 896ndash901doi102134agronj2001934896x

Newell LC (1968) Effects of strain source and management practice onforage yields of two warm-season prairie grasses Crop Science 8205ndash210 doi102135cropsci19680011183X000800020022x

Nippert JBKnappAKBriggs JM (2006) Intra-annual rainfall variability andgrassland productivity can the past predict the futurePlant Ecology 18465ndash74 doi101007s11258-005-9052-9

Nippert JB Fay PA Knapp AK (2007) Photosynthetic traits in C3 and C4

grassland species in mesocosm and field environments Environmentaland Experimental Botany 60 412ndash420 doi101016jenvexpbot200612012

Nippert JB Fay PA Carlisle JD Knapp AK Smith MD (2009)Ecophysiological responses of two dominant grasses to alteredtemperature and precipitation regimes Acta Oecologica 35 400ndash408doi101016jactao200901010

Norberg J Swaney DP Dushoff J Lin J Casagrandi R Levin SA (2001)Phenotypic diversity and ecosystem functioning in changingenvironments a theoretical framework Proceedings of the NationalAcademy of Sciences of the United States of America 9811376ndash11381 doi101073pnas171315998

Parrish DJ Fike JH (2005) The biology and agronomy of switchgrass forbiofuels Critical Reviews in Plant Sciences 24 423ndash459 doi10108007352680500316433

Polley HW Norman JM Arkebauer TJ Walter-Shea EA Greegor DH JrBramer B (1992) Leaf gas exchange of Andropogon gerardii VitmanPanicum virgatum L and Sorghastrum nutans (L) Nash in a tallgrassprairie Journal of Geophysical Research 97 18 837ndash18 844

Quinn JA (1969) Variability among High Plains populations of Panicumvirgatum Bulletin of the Torrey Botanical Club 96 20ndash41 doi1023072484005

Resco V IgnanceDD SunWHuxman TEWeltzin JFWilliams DG (2008)Chlorophyll fluorescence predawn water potential and photosynthesisin pulse-drive ecosystems ndash implications for ecological studiesFunctional Ecology 22 479ndash483 doi101111j1365-2435200801396x

Sanderson MA (1992) Morphological development of switchgrass andkleingrass Agronomy Journal 84 415ndash419 doi102134agronj199200021962008400030013x

Sanderson MA Reed RL (2000) Switchgrass growth and developmentwater nitrogen and plant density effects Journal of RangeManagement 53 221ndash227 doi1023074003287

SandersonMAWolf DD (1995)Morphological development of switchgrassin diverse environments Agronomy Journal 87 908ndash915 doi102134agronj199500021962008700050022x

Schlichting CD (1989) Phenotypic integration and environmental changeBioscience 39 460ndash464 doi1023071311138

SillettiAMKnappAK(2001)Responsesof the codominantgrassland speciesAndropogon gerardii and Sorghastrum nutans to long-termmanipulations of nitrogen and water American Midland Naturalist145 159ndash167 doi1016740003-0031(2001)145[0159ROTCGS]20CO2

SmithMDKnappAK (2003)Dominant speciesmaintain ecosystem functionwith non-random species loss Ecology Letters 6 509ndash517 doi101046j1461-0248200300454x

Sokal RRRohlf FJ (1995) lsquoBiometry the principles and practices of statisticsin biological research 3rd ednrsquo (WH Freeman New York)

StoutWL (1992)Water-use efficiency of grasses as affected by soil nitrogenand temperature Soil Science Society of America Journal 56 897ndash902doi102136sssaj199203615995005600030036x

Stout WL Jung GA Shaffer JA (1988) Effects of soil and nitrogen on wateruse efficiency of tall fescue and switchgrass under humid conditions SoilScience Society of America Journal 52 429ndash434 doi102136sssaj198803615995005200020023x

Stroup JA Sanderson MA Muir JP McFarland MJ Reed RL (2003)Comparison of growth and performance in upland and lowlandswitchgrass types to water and nitrogen stress Bioresource Technology86 65ndash72 doi101016S0960-8524(02)00102-5

Taylor SH Ripley BSWoodward FI Osborne CP (2011) Drought limitationofphotosynthesis differs betweenC3andC4grass species in a comparativeexperiment Plant Cell amp Environment 34 65ndash75 doi101111j1365-3040201002226x

Tucker SS Craine JM Nippert JB (2011) Physiological drought toleranceand the structuring of tallgrass prairie assemblages Ecosphere 2 1ndash19doi101890ES11-000231

Van Esbroeck GA Hussey MA Sanderson MA (1997) Leaf appearancerate and final leaf number of switchgrass cultivars Crop Science 37864ndash870 doi102135cropsci19970011183X003700030028x

Van Esbroeck GA Hussey MA Sanderson MA (2003) Variation betweenAlamo and Cave-in- Rock switchgrass in response to photoperiodextension Crop Science 43 639ndash643

WangDLeBauerDSDietzeMC(2010)Aquantitative reviewcomparing theyield of switchgrass in monocultures and mixtures in relation to climateand management factors GCB Bioenergy 2 16ndash25 doi101111j1757-1707201001035x

Ward JK Kelly JK (2004) Scaling up evolutionary responses to CO2 lessonsfrom Arabidopsis Ecology Letters 7 427ndash440 doi101111j1461-0248200400589x

Wullschleger SD SandersonMAMcLaughlin SB Biradar DP Rayburn AL(1996) Photosynthetic rates and ploidy levels among populations ofswitchgrassCrop Science 36 306ndash312 doi102135cropsci19960011183X003600020016x

Zhou X Talley M Luo Y (2009) Biomass litter and soil respiration along aprecipitation gradient in southern Great Plains USA Ecosystems(New York NY) 12 1369ndash1380 doi101007s10021-009-9296-7

136 Functional Plant Biology J C Hartman et al

wwwpublishcsiroaujournalsfpb

Page 3: Ecotypic responses of switchgrass to altered precipitation

Tallgrass Prairie Preserve (Oklahoma) and 16 from Temple(Texas)

Three precipitation treatments were implemented reflectingclimate change predictions for this region (Alley et al 2007)contemporary average for the growing season (62625mm) a25 increase and a 25 decrease Hereafter these treatments arereferred to as average increased and decreased For Kansasecotypes the three treatments had an equal number of cellseach (7) For the Oklahoma and Texas ecotypes there werefive cells each allocated to the increased and averagetreatments and six to the decreased treatment The unbalancedtreatment design resulted from the removal of a previouslyplanted (and unsuccessfully grown) northern ecotype Thetiming of rain events was not altered and precipitation wasapplied every 6 days which is the long-term average ambientrainfall interval Rainfall was applied through a metered handsprayer (model TM075 Great Plains Industries Wichita KSUSA) using water from an onsite well This water was applied atthe soil surface to minimise losses to canopy interception orrunoff Precipitation treatments were applied from late April toearly October 2010

Sampling procedure and variables measured

Sampling of the physiological variables was conducted on 10dates over the course of the growing season and classified intothree seasonal periods early season mid-season and late seasonData were collected on 28 May 31 May and 9 June for the earlyseason sampling period (Period 1) 30 June 22 July 23 July27 July and 30 July for the mid-season sampling period(Period 2) and 11 August and 26 August for the late seasonsampling period (Period 3) Sampling periods were chosenrather than monthly responses because previous work atKonza has shown the greatest increase in biomass inMayndashearly June (Period 1) allocation to reproductive growthduring late JunendashJuly (Period 2) and the lowest rates of growth inAugustndashSeptember (Period 3) (Knapp et al1998)

For each sampling date one individual was randomly selectedfrom each mesocosm cell and gas exchange dark-adaptedchlorophyll fluorescence (FvFm) and midday water potential(Ymid) were measured These physiological measurementswere conducted on the newest fully expanded mature leaf onthe widest portion of the grass blade For each sampling periodindividuals were tagged to ensure that all physiologicalmeasurements were conducted on the same individual leaf tominimise within-plant variability Because physiologicalmeasurements vary based on the time of day measurementorder for each population was randomised for each samplingperiod Gas exchange measurements were conducted between0900 hours and 1600 hours Central Standard Time when solarradiation was typically above 70 of full sun levels Soilmoisture (0ndash10 cm) data was collected concurrently with gasexchange and water potential measurements using a HydraProbe II Soil Sensor (Stevens Water Monitoring SystemsPortland OR USA) in units of water fraction by volume (wfv)

Gas exchange measurements were conducted using a LI-COR 6400 IRGA with an artificial redndashblue light-emittingdiode light source (6400ndash02B LI-COR Lincoln NE USA)Plants were placed inside the leaf chamber and allowed to reachsteady-state photosynthesis at mean ambient carbon (Ca)

(400mmolmol1) and at a saturating light intensity(2000mmolm2 s1) Leaf temperature was allowed to varywith ambient air temperature Relative humidity in the cuvettewas adjusted to reflect ambient environmental conditions(generally ranging from 30 to 50) The variables measuredincluded CO2 assimilation at ambient Ca (Amax) stomatalconductance to water vapour (gs) leaf-level transpiration (E)and instantaneouswater use efficiency (WUEAmaxE

ndash1)Middaywater potential (Ymid) was measured concurrently with gasexchange using a Scholander-type pressure bomb (PMSInstruments Albany OR USA) Ymid measurements wereconducted on the same selected individual used for gasexchange measurements however a different leaf was usedDark-adapted maximum photochemical efficiency (FvFm) wasassessed using a MINI-PAM photosynthesis yield analyser(Heinz Walz GmbH Effeltrich Germany) Measurementswere conducted on the same individuals within the samplingdate using the same leaf that was used for gas exchangemeasurements FvFm measurements were recorded during thenight Individuals were allowed to adapt for a minimum of 1 hafter complete darkness before measurement

Aboveground biomass was harvested dried at 65C for atleast 48 h and weighed for each individual at the conclusion ofthe growing season (SeptemberndashOctober) Individuals weremonitored for flowering tillers starting in mid-July and thetotal number of flowering tillers was counted on seven datesAs measures of fitness the reproductive biomass and tillernumbers were measured The flowering tillers from each plantwere separated and weighed The percent biomass allocated toreproduction was determined and used as a measure of fitnessOther characteristics such as flowering and non-flowering tillerheight and number of leaves per tiller were also measured

Statistical analyses

Changes in P virgatum physiological responses betweenecotypes and precipitation treatments were analysed using amixed effects model ANOVA (Proc Mixed SAS ver 92SAS Institute Cary NC USA) with the precipitationtreatments ecotypes and sampling period as fixed effects andthe specific mesocosm cell as a random effect Multiplecomparison tests between ecotypes precipitation treatmentsand sampling period were performed using Tukeyrsquos HonestlySignificant Difference test

Non-linear regression analysis was conducted to assess therelationship of gs toFvFm among ecotypes The relationship wasfitted using a two parameter logarithmic equation (y= y0 + a ln(abs(x)) Sigmaplot ver 110 Systat Software San Jose CAUSA) Residual analysis was performed to determine if themodelrsquos fit produced statistically significant regression modelsFitted models were then tested with ANOVA to determine ifsignificant differences existed among the model fits acrossecotypes (Sokal and Rohlf 1995)

Results

The precipitation treatments successfully altered the soilmoisture in this experiment Mean soil moisture wassignificantly higher in the increased precipitation treatment(024 0006wfv) compared with the average (P = 00012

128 Functional Plant Biology J C Hartman et al

021 0006wfv) and decreased treatments (Plt 00001018 0006wfv) and the average treatment was significantlygreater than the decreased treatment (P= 00358) Mean soilmoisture for all treatments at the beginning of the seasonranged from 030 to 040wfv at the end of the season therange was 010ndash015wfv The soil moisture for all treatmentsdecreased over the course of the season as ambient temperatureincreased

Physiology

Mean physiological responses varied according to ecotype(Table 2) Kansas ecotypes displayed significantly lower Amaxgs E Ymid and FvFm (Fig 1) compared with the Texas andOklahoma ecotypes To examine the potential for correlationbetween leaf-level gas exchange and leaf photochemistry asenvironmental conditions changed over the summer FvFmndashgsresponse curves were compared by ecotype (Fig 2) Fittedmodels were significantly different from zero for Kansas(P lt 00001) Oklahoma (P = 00011) and Texas (P lt 00001)FvFmndashgs response curves Kansas ecotypes exhibitedsignificantly lower (P = 00402) FvFm values as gs decreasedcompared with the Texas ecotype Fitted models did notsignificantly differ between Kansas and Oklahoma ecotypes(P = 01337) or between Texas and Oklahoma ecotypes(P = 05545) At an FvFm below 077 Texas and Oklahomaecotypes had similar rates of gs (near 0molmndash2 sndash1) but as gsincreased Texas ecotypes responded with higher FvFm

Among the physiological parameters only WUE and Ymid

varied due to precipitation treatments (Fig 1) MeanWUE in theincreased treatment was only 80 of the WUE in the decreasedtreatment and 72 of the WUE in the average precipitationtreatmentYmid was closely associated with soil moisture wherethe Ymid levels of individuals in the decreased treatment were15 lower than in the increased treatment and 5 lower thanthe average treatment Mean Ymid responses for the averageprecipitation treatment was 10 lower compared with theincreased treatment

All of the physiological parameters displayed significantdecreases over the course of the growing season (Table 2)Photosynthetic rates declined by 86 gs by 87 and Ymid by82 over the course of the growing season The effect ofprecipitation on Amax gs and E did depend on the samplingperiod (Table 2) Increased precipitation only increased Amaxgs andE in the early season period (Fig 3) In the mid-season andlate season sampling periods increased precipitation did notexhibit the same effect on gas exchange responses as it did inthe early season andonlydecreasedWUEduring themid-seasonLater in the season high temperatures negated the potentialpositive physiological responses of increased waterEcotype precipitation interactions were not observed in anyof the physiological responses (Table 2)

Aboveground biomass

Aboveground biomass responses varied by both ecotype andprecipitation individually with few interactions (Table 3) Totalbiomass total tillers flowering tiller biomass flowering tillerheight leaves per flowering tiller and the percent biomassallocated to reproductive tillers all showed a significant

Tab

le2

Mixed-effectsmod

elANOVAof

precipitationtreatm

entsan

decotyp

e(fixed

precipitation

ecotyp

eperiod

ran

domcell)on

physiologicalv

ariables

F-and

P-valuesareprovided

with

significantrespon

sesPlt005Plt010A

maxC

O2assimilatio

natam

bientC

O2g

sstom

atalconductancetowatervapourW

UEw

ater-useefficiencyF

vF

mdark-adapted

chloroph

yllfl

uorescenceY

midm

iddaywater

potential

Response

Precipitatio

nEcotype

Period

Precipitatio

n

Precipitatio

n

Ecotype

Precipitatio

necotyp

e

FP

FP

FP

ecotyp

eperiod

period

period

FP

FP

FP

FP

Amax

12

030

1065

00002

73

71

lt00001

15

019

285

00001

08

068

09

052

g s04

063

859

00007

63

64

lt00001

08

048

243

00013

05

090

10

045

WUE(A

maxEndash1)

441

00175

18

016

975

lt00001

16

016

156

00704

13

018

07

082

E13

026

783

00012

45

85

lt00001

11

035

201

00099

13

016

09

052

FvF

m13

027

791

00012

30

22

lt00001

01

095

08

055

14

016

04

098

Ymid

101

00003

495

00116

23

444

lt0

0001

18

014

10

042

156

007

88

09

052

Responses of switchgrass to altered precipitation Functional Plant Biology 129

response to the precipitation treatments (Table 4) Individuals inthe increased treatment had on average 15 more tillers anddouble the total aboveground and flowering tiller biomassThose individuals in the increased treatment also displayed a22 and 28 increase in mean tiller height compared with theaverage and decreased treatments respectively The increasedtreatment contained individuals with a greater number of leavesper tiller and that allocated over 94 of their biomass toreproductive tillers (Table 4)

Ecotype was a significant predictor of some abovegroundbiomass responses (Table 3) with differences in the total number

of tillers biomass per non-flowering tiller non-flowering tillerbiomass leaves per non-flowering tiller and percent biomassallocated to reproduction (Table 4) Texas ecotypes had amean ofless than 50 tillers per individual and over double the non-flowering tiller biomass and biomass per non-flowering tillercompared with Kansas and Oklahoma ecotypes Texas ecotypesalso displayed significantly less biomass allocation toreproductive tillers (75) compared with Kansas (87) andOklahoma ecotypes (94)

There were few ecotype precipitation interactions(Table 3) but the total biomass per tiller biomass per

EcotypeKansas Oklahoma Texas

Kansas Oklahoma Texas

Am

ax (

μ mol

mndash2

sndash1

)

0

5

10

15

20

25

(a)

Ecotype

g s (

mol

H2O

mndash2

sndash1

)E

(m

mol

mndash2

sndash1

)

WU

E (

μmol

mm

olndash1

)

000

002

004

006

008

010

012

014

Ecotype

0

1

2

3

4

5

6

Treatment

Decreased Average Increased0

1

2

3

4

5

Ecotype

FvF

m

073

074

075

076

077

078

079

080

081

Decre

ased

Avera

ge

Incr

ease

d

Kansa

s

Oklaho

ma

Texas

ψm

id (

MP

a)

ndash25

ndash20

ndash15

ndash10

ndash05

00

a

bb

a

b

b

a

bb

a

b b

b b

a

ab

c

ab b

Kansas Oklahoma Texas

Kansas Oklahoma Texas

(b)

(c) (d )

(e) (f )

Fig 1 Physiological responses for Panicum virgatum ecotypes across precipitation treatments Each point is the overall meanresponse (1 se) of the ecotype within the precipitation treatment (a) carbon assimilation at ambient CO2 (b) dark-adaptedchlorophyllfluorescence (c) stomatal conductance (d) water use efficiency (e) transpiration ( f )midday leaf-levelwater potentialIn (a) (b) (c) and (e) statistically significant (Plt 005) ecotype effects were present (varying letters among ecotypes denotesPlt 005) Significant precipitation treatment effects are present in (d) In ( f ) statistically significant ecotype and treatment effectswere present and significant differences are denoted by varying letters within each section of the panel

130 Functional Plant Biology J C Hartman et al

flowering tiller and the number of non-flowering tillers showedinteractions among ecotypes and precipitation treatments(Table 5) Texas ecotypes had significantly greater biomassper tiller in the increased treatments with the same results forbiomass per flowering tiller Kansas and Texas ecotypesdecreased the amount of non-flowering tillers as the soilmoisture increased Oklahoma ecotypes displayed the oppositetrend Oklahoma ecotypes had significantly fewer non-flowering

tillers in the decreased treatment compared with the other twoecotypes (Table 5)

Flowering

Flowering phenology was affected by both ecotype andprecipitation individually (Table 3) Texas ecotypes had halfthe number of flowering tillers at the end of season comparedwith Kansas and Oklahoma ecotypes (Table 4 Fig 4) Texasecotypes also initiated flowering a month later than the otherecotypes and flowered at a slower rate (Fig 4) Floweringphenology varied significantly with precipitation treatmentsIndividuals in the increased treatment had 40 and 38 moreflowering tillers at the end of the season compared with theaverage and decreased treatments respectively (Table 4Fig 4) Precipitation treatments affected flowering rates withthe increased treatment flowering at a greater rate compared withthe average and decreased treatment (Fig 4)

Discussion

The primary objective of this study was to characterise theecotype-specific physiological responses and growth ofP virgatum to changes in forecasted precipitation amount forthe Central Plains region of the United States by the year 2100Our results show that the precipitation treatments significantlyaffected soil moisture and the corresponding responses ofplant growth and physiology Significant effects of ecotype

gs (mol H2O mndash2 sndash1)

00 01 02 03 04 05

Fv

Fm

072

074

076

078

080

082

KansasOklahoma Texas

y = y0 + aln(abs(x))R-adj = 02442

Fig 2 Representative FvFmndashgs curves for all ecotypes Curves were fitfor all data points for each ecotype The R-adjusted value is for the overall fitof all ecotype data points

Early Mid Late Early Mid Late

Am

ax (

μmol

mndash2

sndash1

)E

(m

mol

mndash2

sndash1

)

0

10

20

30

40DecreasedAverageIncreased

000

005

010

015

020

025

Summer sampling period

0

2

4

6

8

10

Summer sampling period

WU

E (

μmol

mm

olndash1

)

0

1

2

3

4

5

6

a ab

a a

b

a a a

a a

b

a abb

a aa

a abb

aab

b

aab b

a a aa a

ba

aa

(a)

Early Mid LateEarly Mid Late

g s (

mol

H2O

mndash2

sndash1

)

(b)

(c) (d )

Fig 3 Gas exchange responses by treatment during the earlymid-season and late periods of the growing season (a)Carbon assimilationat ambient CO2 (b) stomatal conductance (c) transpiration (d) water use efficiency (Amax E

ndash1) Bars are means (1 se) and significantdifferences are indicated by letters within each sampling period

Responses of switchgrass to altered precipitation Functional Plant Biology 131

and precipitation were present but there were fewecotype precipitation interactions In general physiologicaldifferences reflected ecotype origin whereas differences inaboveground biomass largely reflected differences in soilwater availability from the precipitation treatments

Ecotype effects

Differences among means in the physiological variablesmeasured largely reflected differences by ecotype andcorresponding adaptation to the local environment of originWithin each ecotype similar changes in Amax gs and E reflectthe coupled gas exchange relationships documented previouslyfor other dominant prairie grasses (Polley et al 1992) It hasalso been shown that carbon assimilation in switchgrass iscontrolled mostly by population responses to localenvironmental conditions (Wullschleger et al 1996) In this

study the Kansas ecotypes had the lowest carbon assimilationrates providing support for the local environmental conditionresponse (Wullschleger et al 1996)

The FvFm ndash gs relationship developed (Fig 2) shows astatistical relationship between responses of leafphotochemistry to changes in stomatal conductance that variesby ecotypeReducedFvFmvalues indicate a decreased efficiencyof non-photochemical quenching and are reflective of thefunctioning of the light reactions of photosynthesis (Maxwelland Johnson 2000) Changes in FvFm can be especially useful instudies where physiological performance is strongly decreasedfrom longperiods of drought (Resco et al 2008) These data showthat nearly one-quarter of the variation themaximumefficiencyofPSII can be statistically attributed to changes in stomatalconductance (Fig 2) From first principles one would assumethat a reduction in stomatal water loss would result in increased

Table 3 Mixed-effects model ANOVA of precipitation treatments and ecotype (fixed precipitation ecotype random cell) effects on abovegroundbiomass variables

F- and P-values are provided with significant responses P lt 005 Plt 010

Response Precipitation Ecotype Ecotype precipitationF P F P F P

Tillers (per individual) 371 00326 338 00429 11 035Biomass (g per individual) 948 00004 21 012 15 021Biomass per tiller (g per individual) 449 00168 1109 00001 431 00050 allocated to reproduction (per individual) 483 00127 599 00050 191 012Flowering tillers (per individual) 754 00015 848 00008 05 068Flowering tiller biomass (g per individual) 1146 lt00001 08 044 15 019Biomass per flowering tiller (g per individual) 286 00681 1333 lt00001 351 00142Flowering tiller height (cm per individual) 512 00100 18 016 16 017Leaves per flowering tiller (per individual) 547 00075 22 011 11 035Non-flowering tillers (per individual) 325 00481 291 00651 287 00337Non-flowering tiller biomass (g per individual) 11 032 484 00126 05 067Biomass per non-flowering tiller (g per individual) 007 093 1585 lt00001 17 015Non-flowering tiller height (cm per individual) 09 038 305 00576 11 036Leaves per non-flowering tiller (per individual) 02 080 670 00029 18 013

Table 4 Mean responses (1 se) of various aboveground biomass measurements for ecotype and precipitation treatment effectsWithin the ecotype or treatment category bolded numbers show significance at the P 005 level and italicised numbers show marginal significance at theP 010 level Superscripts display significant differences between precipitation treatments or ecotypes within a category NS signifies no significance between

the ecotypes or the precipitation treatments

Variables Precipitation EcotypeDecreased Average Increased Kansas Oklahoma Texas

Tillers (per individual) 56 plusmn 5ab 47 plusmn 6b 71 plusmn 6a 69 plusmn 5a 58 plusmn 6ab 48 plusmn 6b

Biomass (g per individual) 202plusmn 28b 181 plusmn 30b 352 plusmn 30a NS NS NSBiomass per tiller (g per individual) 36 plusmn 03b 40 plusmn 03ab 50 plusmn 03a 30 plusmn 03b 44 plusmn 03a 52 plusmn 03a

Flowering tillers (per individual) 35 plusmn 4b 36 plusmn 4b 58 plusmn 4a 53 plusmn 4a 48 plusmn 5a 27 plusmn 5b

Flowering tiller biomass (g per individual) 157plusmn 27b 157 plusmn 28b 325 plusmn 28a NS NS NSBiomass per flowering tiller (g per individual) 41plusmn 04b 49plusmn 07ab 56plusmn 04a 32 plusmn 04b 50 plusmn 04a 64 plusmn 04a

Flowering tiller height (cm per individual) 94 plusmn 7b 101 plusmn 8b 129 plusmn 8a NS NS NSLeaves per flowering tiller (per individual) 47 plusmn 01b 51 plusmn 02ab 56 plusmn 02a NS NS NSNon-flowering tillers (per individual) 21 plusmn 2a 11 plusmn 3b 12 plusmn 3ab 15plusmn 2ab 9plusmn 3b 20plusmn 3a

Non-flowering tiller biomass (g per individual) NS NS NS 20 plusmn 9b 14 plusmn 11b 59 plusmn 11a

Biomass per non-flowering tiller (g per individual) NS NS NS 07 plusmn 01b 08 plusmn 01b 20 plusmn 01a

Non-flowering tiller height (cm per individual) NS NS NS 27plusmn 3ab 24plusmn 4b 38plusmn 4a

Leaves per non-flowering tiller (per individual) NS NS NS 32 plusmn 02b 30 plusmn 03b 44 plusmn 03a

allocated to reproduction (per individual) 77 plusmn 3b 85 plusmn 3ab 94 plusmn 3a 87 plusmn 3a 94 plusmn 4a 75 plusmn 4b

132 Functional Plant Biology J C Hartman et al

leaf temperature with the potential for corresponding decreasedphotochemistry during hot or dry summer conditions For allecotypesFvFm decreased nonlinearly with stomatal closure buttherewere clear differences in this relationship between ecotypes

with a general trend of increasing FvFm values for a given rate ofstomatal conductance from Kansas to Oklahoma to Texasecotypes Texas ecotypes had higher maximum FvFm valuescompared with Kansas ecotypes with values near the acceptedmaximal value of 083 (Krause and Weis 1991) As gs valuesdecreased Texas and Oklahoma ecotypes were able to maintainhigher FvFm values compared with the Kansas ecotypesindicating a greater decrease in the photochemical efficiencyof PSII in Kansas ecotypes Interpretation of the relationshipbetween FvFm and gs requires further investigation but thesedata show promise for linking changes in stomatal conductancewith themaintenance of leaf photochemistry as the ability to coolthe leaf through latent heat exchange becomes very limited withlow gs (Nippert et al 2009) and photochemistry is subsequentlyreducedby increased leaf temperatures (Albert et al 2011)Basedon theseFvFmndashgs response curves (Fig 2) Oklahoma and Texasecotypes are potentially more tolerant of higher leaf temperaturewith higher FvFm at low rates of gas exchange compared withKansas ecotypes

P virgatum is a highly photoperiod sensitive species(Benedict 1940) and differences in flowering dates have beendocumented for various ecotype and cultivars (McMillan 1965VanEsbroeck et al 2003Casler et al 2004) Sanderson andWolf(1995) recorded similar flowering dates for Alamo and Cave-in-Rock cultivars compared with the ecotypes in this study Alamocultivars flowered around the end of September and Cave-in-Rock cultivars flowered around the first week of July Theflowering time and rates seen in this study correspond withknown ecotype responses to changes in latitude

Many of the aboveground biomass responses with significantecotypic differences were a result of the Texas ecotypesproducing fewer tillers but with an increased biomass pertiller and number of leaves per tiller while the Oklahoma andKansas ecotypes had higher tiller density (Table 4) Typically thebest estimates to predict changes in biomass yield for this speciesare increased biomass per tiller and number of leaves per tiller(seen in Texas ecotypes) or higher tiller density (noted in the

Table 5 Mean responses (1 se) of various aboveground biomass measurements for interactions between ecotypes and precipitation treatmentsBolded numbers show significance at theP 005 level Superscripts display significant differences between ecotype precipitation combinations NS signifies

no significant difference between the means

Variables Ecotype precipitationDecreased Average Increased

Kansas Oklahoma Texas Kansas Oklahoma Texas Kansas Oklahoma Texas

Tillers (per individual) NS NS NS NS NS NS NS NS NSBiomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per tiller (g per individual) 28 plusmn 05b 42 plusmn 05b 38 plusmn 05b 29 plusmn 05b 50 plusmn 06ab 42 plusmn 06b 34 plusmn 05b 39 plusmn 06b 76 plusmn 06a

Flowering tillers (per individual) NS NS NS NS NS NS NS NS NSFlowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per flowering tiller (g per individual) 28 plusmn 07bc 44 plusmn 07bc 51 plusmn 07bc 31 plusmn 07bc 65 plusmn 08ab 52 plusmn 08abc 36 plusmn 07bc 42 plusmn 08bc 90 plusmn 08a

Flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per flowering tiller (per individual) NS NS NS NS NS NS NS NS NSNon-flowering tillers (per individual) 31 plusmn 4a 5 plusmn 5b 28 plusmn 5ab 8 plusmn 4b 7 plusmn 5ab 18 plusmn 5ab 7 plusmn 4b 15 plusmn 5ab 15 plusmn 5ab

Non-flowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per non-flowering tiller (g per individual) NS NS NS NS NS NS NS NS NSNon-flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per non-flowering tiller (per individual) NS NS NS NS NS NS NS NS NS allocated to reproduction (per individual) NS NS NS NS NS NS NS NS NS

Nov

Till

ers

flow

ered

(in

divndash1

)T

iller

s flo

wer

ed (

indi

vndash1)

0

10

20

30

40

50

60IncreasedAverageDecreased

0

10

20

30

40

50

60 Kansas Oklahoma Texas

(a)

(b)

Jul Aug Sep Oct

Fig 4 Total tillers flowered per individual over the course of the growingseason (a) Flowering tillers by ecotype (b) flowering tillers by precipitationtreatment Data points are means 1se Asterisks indicate statisticalsignificance (Plt 005) between groups within each sampling date

Responses of switchgrass to altered precipitation Functional Plant Biology 133

Oklahoma and Kansas ecotypes Boe and Beck 2008) Althoughtotal biomass was not significantly different among ecotypes theTexas ecotypes contain components (higher biomass and leavesper tiller) that if heritablemay lead to greater biomass inmultiplegeneration studies

Precipitation effects

The altered environmental conditions resulting from theprecipitation treatments significantly altered responses inaboveground biomass measurements with statisticallysignificant responses in the physiological variables largelyreflecting changes across the sampling periods (Fig 3)Flowering tiller heights were greatest in the increasedprecipitation treatments reaching mean heights of 129mThese heights are reflective of those observed in prairieremnant populations although they are lower than those foundin P virgatum cultivars (Das et al 2004 Casler 2005Alexopoulou et al 2008) but it should be noted that plantheights are variable from year to year (Alexopoulou et al2008) The average number of leaves per flowering tiller waslower than those found in agronomic P virgatum cultivars (VanEsbroeck et al 1997) however these cultivars have been bred tooptimise yield These data illustrate individual biomass responsesto precipitation of a dominant species in the tallgrass prairieecosystem further supporting the conclusions of others thatfuture changes in precipitation are likely to significantly alterecosystem productivity (Knapp 1984 Fay et al 2003 Fay et al2008 Zhou et al 2009)

Previous attempts to link leaf-level to whole-plant responsesfor grassland species have shown positive relationships betweenphotosynthesis and plant success (McAllister et al 1998) butleaf-level to whole-plant relationships in grasslands arecommonly weak (Nippert et al 2007) In this study no clearrelationships between leaf-level photosynthetic responses andtotal aboveground biomass were present for these ecotypesAlthough the increases in leaf-level photosynthesis did notcorrespond to total biomass increases other measurements ofbiomass increased from the increased precipitation partiallysupporting the positive relationship between higher rates ofgas exchange and plant success previously noted byMcAllister et al (1998)

The only physiological responses affected by precipitationtreatments were Ymid and WUE with WUE being lowest inthe increased treatments andYmid highest in increased treatmentsAlthough ANOVA results showed significant precipitationsample period interactions (Table 2) many of the effects wereonly seen in the early season period (Fig 3) Later in the seasonincreasing summer temperatures negated the positive impacts ofthe added water Decreased WUE across the growing seasoncorresponding to higher air temperatures has been previouslynoted for C4 species where WUE can be reduced by 40 indrought conditions (Taylor et al 2011) As water availabilitydecreasedYmid decreased to amean of ndash22MPa The decreasedprecipitation treatment did not lower soil water availability tothe point where the critical water potential (leaf pressurepotential when stomatal closure is less than 5 of themaximum) was reached for switchgrass (Ycrit = ndash32MPaTucker et al 2011) Indeed previous studies have shown that

switchgrass can maintain growth and functioning at lower waterpotentials than those recorded here (Knapp 1984 Stroup et al2003) With more severe decreases in water availability YmidAmax gs E and WUE should continue to decline with thepotential for increased cuticular conductance when drought-stressed (Manzoni et al 2011)

Ecotype precipitation interactions

Surprisingly few ecotype precipitation interactions werepresent for the variablesmeasured in this study (Tables 2 3) withthe primary exception being greater biomass per tiller in theincreased treatment for Texas ecotypes These differences largelyreflect the southern Texas ecotype being adapted to a longergrowing season (Newell 1968) Growth of the Texas ecotype in alocation with a shorter growing season delayed reproductivematurity and although the overall biomass yield was notsignificantly greater for Texas ecotypes compared withOklahoma and Kansas under greater precipitation differencesthe trends observed in the current study may become statisticallysignificant Because Texas ecotypes had a longer period in whichthey grew vegetatively theywere better able to take advantage ofthe increased precipitation treatment to produce higherbiomass per tiller greater biomass per flowering tiller and highnon-flowering tiller counts in all precipitation treatments Theseresults from a southern Texas ecotype growing in amore northernlocation are similar to other studies that show increases inbiomass longer photosynthetic activity and delayed floweringfor southern ecotypes grown in more northern latitudes (Quinn1969 Lemus et al 2002 Casler et al 2004 Berdahl et al 2005Casler et al 2007)

Conclusion

The physiological responses of Amax gs E FvFm and Ymid inP virgatum were explained by ecotype differences howeverrobust responses to altered precipitationwere seen inWUEYmidflowering times and many aboveground biomass variablesEcotype differences were also seen in several abovegroundbiomass variables and most strikingly in flowering initiationand rates In this study there were few interactions betweenecotype and precipitation amount suggesting that althoughprecipitation is a strong driver of biomass production localadaptations in the ecotypes used in the current study have lessimpact on these responses than initially expected It was expectedthat the environmental differences among the origins of theecotypes mainly the effects of higher temperature coupledwith similar mean annual precipitation (860ndash878mm) wouldlead to significant ecotype precipitation interactions in theparameters measured with the Texas ecotypes being the mosttolerant of the changes The lack of interactions between geneticbackground and precipitation also has broader evolutionaryimplications gene environment interactions are a primaryindicator of the potential for natural selection in a given trait(Schlichting 1989) The broader implications of the fewecotype precipitation interactions for this widely-distributedspecies suggest that it is not likely toundergo selection in responseto changes in precipitation As many have correctly suggested itis necessary to examine the role of genetic variation in theresponse of plant species to future changes in climate

134 Functional Plant Biology J C Hartman et al

(Callaway et al 2003 Ward and Kelly 2004) however asindicated by this study it may not always be necessary toconsider localised adaptation when working to project theimpacts of climate change on the performance of a species likeswitchgrass across its geographical distribution

Acknowledgements

We thankDr JohnBlair andDr TedMorgan for their advice and comments onthemanuscriptWealso thank theNatureConservancy for allowingcollectionof switchgrass rhizomes from the Tallgrass Prairie Preserve inOklahoma andlandholdingsnear Temple TXGracieOrozcoRebeccaOrozco JacobCarterand Teall Culbertson provided assistance with fieldwork and the Division ofBiology at Kansas State University Konza Prairie Biological Station KonzaPrairie LTER (DEB-0823341) and the US Department of Energy(EEE000314) provided funding and access to facilities

References

Albert KR Mikkelsen TN Michelsen A Ro-Poulsen H van der Linden L(2011) Interactive effects of drought elevated CO2 and warming onphotosynthetic capacity and photosystem performance in temperateheath plants Journal of Plant Physiology 168 1550ndash1561 doi101016jjplph201102011

Alexopoulou E Sharma N Papatheohari Y Christou M Piscioneri IPanoutsou D Pignatelli V (2008) Biomass yields for upland andlowland switchgrass varieties grown in the Mediterranean regionBiomass and Bioenergy 32 926ndash933 doi101016jbiombioe200801015

AlleyRBBerntsen T Bindoff NL ChenZ ChidthaisongA et al (2007) Hephysical science basis summary for policy makersrsquo (IPCC SecretariatGeneva)

Barney JN Mann JJ Kyser GB Blumwald B Deynze AV DiTomaso JM(2009)Toleranceof switchgrass to extremesoilmoisture stress ecologicalimplications Plant Science 177 724ndash732 doi101016jplantsci200909003

Benedict HM (1940) Effect of day length and temperature on the floweringand growth of four species of grasses Journal of Agricultural Research61 661ndash671

Berdahl JD Frank AB Krupinsky JM Carr PM Hanson JD Johnson HA(2005) Biomass yield phenology and survival of diverse switchgrasscultivars and experimental strains in western North Dakota AgronomyJournal 97 549ndash555 doi102134agronj20050549

Berg CC (1971) Forage yield of switchgrass (Panicum virgatum) inPennsylvania Agronomy Journal 63 785ndash786 doi102134agronj197100021962006300050038x

Boe A Beck DL (2008) Yield components of biomass in switchgrass CropScience 48 1306ndash1311 doi102135cropsci2007080482

Callaway RM Pennings SC Richards CL (2003) Phenotypic plasticity andinteractions among plants Ecology 84 1115ndash1128 doi1018900012-9658(2003)084[1115PPAIAP]20CO2

Casler MD (2005) Ecotypic variation among switchgrass populations fromthe northern USA Crop Science 45 388ndash398 doi102135cropsci20050388

Casler MD Vogel KP Taliaferro CM Wynia RL (2004) Latitudinaladaptation of switchgrass populations Crop Science 44 293ndash303

Casler MD Vogel KP Taliaferro CM Ehlke NJ Berdahl JD Brummer ECKallenbach RL West CP Mitchell RB (2007) Latitudinal andlongitudinal adaptation of switchgrass populations Crop Science 472249ndash2260 doi102135cropsci2006120780

Christensen JH Hewitson B Busuioc A Chen A Gao X Held I Jones RKolli RK KwonWT Laprise R Magantildea Rueda VMearns L MeneacutendezCG Raumlisaumlnen J Rinke A Sarr A (2007) Regional climate projectionsIn lsquoClimate change 2007 the physical science basis Contributionof Working Group I to the Fourth Assessment Report of the

Intergovernmental Panel on Climate Change (eds S Solomon D QinM Manning Z Chen M Marquis KB Averyt M Tignor HL Miller)pp 887ndash892 (Cambridge University Press Cambridge UK)

Das MK Fuentes RG Taliaferro CM (2004) Genetic variability and traitrelationships in switchgrass Crop Science 44 443ndash448

Easterling DR Meehl GA Parmesan C Changnon SA Karl TR Mearns LO(2000) Climate extremes observations modeling and impacts Science289 2068ndash2074 doi101126science28954872068

Emery SM Gross KL (2007) Dominant species identity not communityevenness regulates invasion in experimental grassland communitiesEcology 88 954ndash964 doi10189006-0568

Evers GW Parsons MJ (2003) Soil type and moisture level influence onAlamo switchgrass emergence and seedling growth Crop Science 43288ndash294 doi102135cropsci20030288

Fay PA Carlisle JD Knapp AK Blair JM Collins SL (2003) Productivityresponses to altered rainfall patterns in a C4-dominated grasslandOecologia 137 245ndash251 doi101007s00442-003-1331-3

FayPAKaufmanDMNippert JBCarlisle JDHarperCW(2008)Changes ingrassland ecosystem function due to extreme rainfall events implicationsfor responses to climate changeGlobal Change Biology 14 1600ndash1608doi101111j1365-2486200801605x

Fay PA Blair JM Smith MD Nippert JB Carlisle JD Knapp AK (2011)Relative effects of precipitation variability and warming on tallgrassprairie ecosystem function Biogeosciences 8 3053ndash3068 doi105194bg-8-3053-2011

Grime JP (1998) Benefits of plant diversity to ecosystems immediatefilter and founder effects Journal of Ecology 86 902ndash910doi101046j1365-2745199800306x

Hartman JC Nippert JB OrozcoRA Springer CJ (2011) Potential ecologicalimpacts of switchgrass (Panicum virgatum L) biofuel cultivation in theCentral Great Plains USA Biomass and Bioenergy 35 3415ndash3421doi101016jbiombioe201104055

Heaton E Voigt T Long SP (2004) A quantitative review comparing theyields of two candidate C4 perennial biomass crops in relation to nitrogentemperature and water Biomass and Bioenergy 27 21ndash30 doi101016jbiombioe200310005

HillebrandHBennettDMCadotteMW(2008)Consequences of dominancea review of evenness effects on local and regional ecosystem processesEcology 89 1510ndash1520 doi10189007-10531

Hughes AR Inouye BD Johnson MTJ Underwood N Vellend M (2008)Ecological consequences of genetic diversity Ecology Letters 11609ndash623 doi101111j1461-0248200801179x

Jump AS Pentildeuelas J (2005) Running to stand still adaptation and responseof plants to rapid climate change Ecology Letters 8 1010ndash1020doi101111j1461-0248200500796x

KnappAK (1984)Water relations and growth of three grasses duringwet anddrought years in a tallgrass prairie Oecologia 65 35ndash43 doi101007BF00384460

Knapp AK (1985) Effect of fire and drought on the ecophysiology ofAndropogon gerardii and Panicum virgatum in a tallgrass prairieEcology 66 1309ndash1320 doi1023071939184

KnappAKBriggs JMHartnett DCCollins SC (1998) lsquoGrassland dynamicslong-term ecological research in tallgrass prairiersquo (Oxford UniversityPress New York)

KnappAKFayPABlair JMCollinsSLSmithMDCarlisle JDHarperCWDanner BT Lett MS McCarron JK (2002) Rainfall variability carboncycling and plant species diversity in mesic grassland Science 2982202ndash2205 doi101126science1076347

Krause GH Weis E (1991) Chlorophyll fluorescence and photosynthesisthe basics Annual Review of Plant Physiology and Plant MolecularBiology 42 313ndash349 doi101146annurevpp42060191001525

Lemus R Brummer EC Moore KJ Molstad ME Burras CE Barker MF(2002) Biomass yield and quality of 20 switchgrass populations insouthern Iowa USA Biomass and Bioenergy 23 433ndash442doi101016S0961-9534(02)00073-9

Responses of switchgrass to altered precipitation Functional Plant Biology 135

Manzoni S Vico G Katul G Fay PA Polley W Palmroth S Porporato A(2011) Optimizing stomatal conductance for maximum carbon gainunder water stress a meta-analysis across plant functional types andclimates Functional Ecology 25 456ndash467 doi101111j1365-2435201001822x

Maxwell K JohnsonGN (2000) Chlorophyll fluorescence ndash a practical guideJournal of Experimental Botany 51 659ndash668 doi101093jexbot51345659

McAllister CA Knapp AK Maragni LA (1998) Is leaf-level photosynthesisrelated to plant success in a highly productive grasslandOecologia 11740ndash46 doi101007s004420050629

McLaughlin SB Kszos LN (2005) Development of switchgrass (Panicumvirgatum) as a bioenergy feedstock in the United States Biomass andBioenergy 28 515ndash535 doi101016jbiombioe200405006

McMillan C (1965) Ecotypic differentiation within four NorthAmerican prairie grasses II Behavioral variation within transplantedcommunity fractions American Journal of Botany 52 55ndash65doi1023072439975

McNaughton SJ Wolf LL (1970) Dominance and the niche in ecologicalsystems Science 167 131ndash139 doi101126science1673915131

Meehl GA Arblaster JM Tebaldi C (2005) Understanding future patterns ofincreased precipitation intensity in climate model simulationsGeophysical Research Letters 32 L18719 doi1010292005GL023680

Muir JP Sanderson MA Ocumpaugh WR Jones RM Reed RL (2001)Biomass production of lsquoAlamorsquo switchgrass in response to nitrogenphosphorus and row spacing Agronomy Journal 93 896ndash901doi102134agronj2001934896x

Newell LC (1968) Effects of strain source and management practice onforage yields of two warm-season prairie grasses Crop Science 8205ndash210 doi102135cropsci19680011183X000800020022x

Nippert JBKnappAKBriggs JM (2006) Intra-annual rainfall variability andgrassland productivity can the past predict the futurePlant Ecology 18465ndash74 doi101007s11258-005-9052-9

Nippert JB Fay PA Knapp AK (2007) Photosynthetic traits in C3 and C4

grassland species in mesocosm and field environments Environmentaland Experimental Botany 60 412ndash420 doi101016jenvexpbot200612012

Nippert JB Fay PA Carlisle JD Knapp AK Smith MD (2009)Ecophysiological responses of two dominant grasses to alteredtemperature and precipitation regimes Acta Oecologica 35 400ndash408doi101016jactao200901010

Norberg J Swaney DP Dushoff J Lin J Casagrandi R Levin SA (2001)Phenotypic diversity and ecosystem functioning in changingenvironments a theoretical framework Proceedings of the NationalAcademy of Sciences of the United States of America 9811376ndash11381 doi101073pnas171315998

Parrish DJ Fike JH (2005) The biology and agronomy of switchgrass forbiofuels Critical Reviews in Plant Sciences 24 423ndash459 doi10108007352680500316433

Polley HW Norman JM Arkebauer TJ Walter-Shea EA Greegor DH JrBramer B (1992) Leaf gas exchange of Andropogon gerardii VitmanPanicum virgatum L and Sorghastrum nutans (L) Nash in a tallgrassprairie Journal of Geophysical Research 97 18 837ndash18 844

Quinn JA (1969) Variability among High Plains populations of Panicumvirgatum Bulletin of the Torrey Botanical Club 96 20ndash41 doi1023072484005

Resco V IgnanceDD SunWHuxman TEWeltzin JFWilliams DG (2008)Chlorophyll fluorescence predawn water potential and photosynthesisin pulse-drive ecosystems ndash implications for ecological studiesFunctional Ecology 22 479ndash483 doi101111j1365-2435200801396x

Sanderson MA (1992) Morphological development of switchgrass andkleingrass Agronomy Journal 84 415ndash419 doi102134agronj199200021962008400030013x

Sanderson MA Reed RL (2000) Switchgrass growth and developmentwater nitrogen and plant density effects Journal of RangeManagement 53 221ndash227 doi1023074003287

SandersonMAWolf DD (1995)Morphological development of switchgrassin diverse environments Agronomy Journal 87 908ndash915 doi102134agronj199500021962008700050022x

Schlichting CD (1989) Phenotypic integration and environmental changeBioscience 39 460ndash464 doi1023071311138

SillettiAMKnappAK(2001)Responsesof the codominantgrassland speciesAndropogon gerardii and Sorghastrum nutans to long-termmanipulations of nitrogen and water American Midland Naturalist145 159ndash167 doi1016740003-0031(2001)145[0159ROTCGS]20CO2

SmithMDKnappAK (2003)Dominant speciesmaintain ecosystem functionwith non-random species loss Ecology Letters 6 509ndash517 doi101046j1461-0248200300454x

Sokal RRRohlf FJ (1995) lsquoBiometry the principles and practices of statisticsin biological research 3rd ednrsquo (WH Freeman New York)

StoutWL (1992)Water-use efficiency of grasses as affected by soil nitrogenand temperature Soil Science Society of America Journal 56 897ndash902doi102136sssaj199203615995005600030036x

Stout WL Jung GA Shaffer JA (1988) Effects of soil and nitrogen on wateruse efficiency of tall fescue and switchgrass under humid conditions SoilScience Society of America Journal 52 429ndash434 doi102136sssaj198803615995005200020023x

Stroup JA Sanderson MA Muir JP McFarland MJ Reed RL (2003)Comparison of growth and performance in upland and lowlandswitchgrass types to water and nitrogen stress Bioresource Technology86 65ndash72 doi101016S0960-8524(02)00102-5

Taylor SH Ripley BSWoodward FI Osborne CP (2011) Drought limitationofphotosynthesis differs betweenC3andC4grass species in a comparativeexperiment Plant Cell amp Environment 34 65ndash75 doi101111j1365-3040201002226x

Tucker SS Craine JM Nippert JB (2011) Physiological drought toleranceand the structuring of tallgrass prairie assemblages Ecosphere 2 1ndash19doi101890ES11-000231

Van Esbroeck GA Hussey MA Sanderson MA (1997) Leaf appearancerate and final leaf number of switchgrass cultivars Crop Science 37864ndash870 doi102135cropsci19970011183X003700030028x

Van Esbroeck GA Hussey MA Sanderson MA (2003) Variation betweenAlamo and Cave-in- Rock switchgrass in response to photoperiodextension Crop Science 43 639ndash643

WangDLeBauerDSDietzeMC(2010)Aquantitative reviewcomparing theyield of switchgrass in monocultures and mixtures in relation to climateand management factors GCB Bioenergy 2 16ndash25 doi101111j1757-1707201001035x

Ward JK Kelly JK (2004) Scaling up evolutionary responses to CO2 lessonsfrom Arabidopsis Ecology Letters 7 427ndash440 doi101111j1461-0248200400589x

Wullschleger SD SandersonMAMcLaughlin SB Biradar DP Rayburn AL(1996) Photosynthetic rates and ploidy levels among populations ofswitchgrassCrop Science 36 306ndash312 doi102135cropsci19960011183X003600020016x

Zhou X Talley M Luo Y (2009) Biomass litter and soil respiration along aprecipitation gradient in southern Great Plains USA Ecosystems(New York NY) 12 1369ndash1380 doi101007s10021-009-9296-7

136 Functional Plant Biology J C Hartman et al

wwwpublishcsiroaujournalsfpb

Page 4: Ecotypic responses of switchgrass to altered precipitation

021 0006wfv) and decreased treatments (Plt 00001018 0006wfv) and the average treatment was significantlygreater than the decreased treatment (P= 00358) Mean soilmoisture for all treatments at the beginning of the seasonranged from 030 to 040wfv at the end of the season therange was 010ndash015wfv The soil moisture for all treatmentsdecreased over the course of the season as ambient temperatureincreased

Physiology

Mean physiological responses varied according to ecotype(Table 2) Kansas ecotypes displayed significantly lower Amaxgs E Ymid and FvFm (Fig 1) compared with the Texas andOklahoma ecotypes To examine the potential for correlationbetween leaf-level gas exchange and leaf photochemistry asenvironmental conditions changed over the summer FvFmndashgsresponse curves were compared by ecotype (Fig 2) Fittedmodels were significantly different from zero for Kansas(P lt 00001) Oklahoma (P = 00011) and Texas (P lt 00001)FvFmndashgs response curves Kansas ecotypes exhibitedsignificantly lower (P = 00402) FvFm values as gs decreasedcompared with the Texas ecotype Fitted models did notsignificantly differ between Kansas and Oklahoma ecotypes(P = 01337) or between Texas and Oklahoma ecotypes(P = 05545) At an FvFm below 077 Texas and Oklahomaecotypes had similar rates of gs (near 0molmndash2 sndash1) but as gsincreased Texas ecotypes responded with higher FvFm

Among the physiological parameters only WUE and Ymid

varied due to precipitation treatments (Fig 1) MeanWUE in theincreased treatment was only 80 of the WUE in the decreasedtreatment and 72 of the WUE in the average precipitationtreatmentYmid was closely associated with soil moisture wherethe Ymid levels of individuals in the decreased treatment were15 lower than in the increased treatment and 5 lower thanthe average treatment Mean Ymid responses for the averageprecipitation treatment was 10 lower compared with theincreased treatment

All of the physiological parameters displayed significantdecreases over the course of the growing season (Table 2)Photosynthetic rates declined by 86 gs by 87 and Ymid by82 over the course of the growing season The effect ofprecipitation on Amax gs and E did depend on the samplingperiod (Table 2) Increased precipitation only increased Amaxgs andE in the early season period (Fig 3) In the mid-season andlate season sampling periods increased precipitation did notexhibit the same effect on gas exchange responses as it did inthe early season andonlydecreasedWUEduring themid-seasonLater in the season high temperatures negated the potentialpositive physiological responses of increased waterEcotype precipitation interactions were not observed in anyof the physiological responses (Table 2)

Aboveground biomass

Aboveground biomass responses varied by both ecotype andprecipitation individually with few interactions (Table 3) Totalbiomass total tillers flowering tiller biomass flowering tillerheight leaves per flowering tiller and the percent biomassallocated to reproductive tillers all showed a significant

Tab

le2

Mixed-effectsmod

elANOVAof

precipitationtreatm

entsan

decotyp

e(fixed

precipitation

ecotyp

eperiod

ran

domcell)on

physiologicalv

ariables

F-and

P-valuesareprovided

with

significantrespon

sesPlt005Plt010A

maxC

O2assimilatio

natam

bientC

O2g

sstom

atalconductancetowatervapourW

UEw

ater-useefficiencyF

vF

mdark-adapted

chloroph

yllfl

uorescenceY

midm

iddaywater

potential

Response

Precipitatio

nEcotype

Period

Precipitatio

n

Precipitatio

n

Ecotype

Precipitatio

necotyp

e

FP

FP

FP

ecotyp

eperiod

period

period

FP

FP

FP

FP

Amax

12

030

1065

00002

73

71

lt00001

15

019

285

00001

08

068

09

052

g s04

063

859

00007

63

64

lt00001

08

048

243

00013

05

090

10

045

WUE(A

maxEndash1)

441

00175

18

016

975

lt00001

16

016

156

00704

13

018

07

082

E13

026

783

00012

45

85

lt00001

11

035

201

00099

13

016

09

052

FvF

m13

027

791

00012

30

22

lt00001

01

095

08

055

14

016

04

098

Ymid

101

00003

495

00116

23

444

lt0

0001

18

014

10

042

156

007

88

09

052

Responses of switchgrass to altered precipitation Functional Plant Biology 129

response to the precipitation treatments (Table 4) Individuals inthe increased treatment had on average 15 more tillers anddouble the total aboveground and flowering tiller biomassThose individuals in the increased treatment also displayed a22 and 28 increase in mean tiller height compared with theaverage and decreased treatments respectively The increasedtreatment contained individuals with a greater number of leavesper tiller and that allocated over 94 of their biomass toreproductive tillers (Table 4)

Ecotype was a significant predictor of some abovegroundbiomass responses (Table 3) with differences in the total number

of tillers biomass per non-flowering tiller non-flowering tillerbiomass leaves per non-flowering tiller and percent biomassallocated to reproduction (Table 4) Texas ecotypes had amean ofless than 50 tillers per individual and over double the non-flowering tiller biomass and biomass per non-flowering tillercompared with Kansas and Oklahoma ecotypes Texas ecotypesalso displayed significantly less biomass allocation toreproductive tillers (75) compared with Kansas (87) andOklahoma ecotypes (94)

There were few ecotype precipitation interactions(Table 3) but the total biomass per tiller biomass per

EcotypeKansas Oklahoma Texas

Kansas Oklahoma Texas

Am

ax (

μ mol

mndash2

sndash1

)

0

5

10

15

20

25

(a)

Ecotype

g s (

mol

H2O

mndash2

sndash1

)E

(m

mol

mndash2

sndash1

)

WU

E (

μmol

mm

olndash1

)

000

002

004

006

008

010

012

014

Ecotype

0

1

2

3

4

5

6

Treatment

Decreased Average Increased0

1

2

3

4

5

Ecotype

FvF

m

073

074

075

076

077

078

079

080

081

Decre

ased

Avera

ge

Incr

ease

d

Kansa

s

Oklaho

ma

Texas

ψm

id (

MP

a)

ndash25

ndash20

ndash15

ndash10

ndash05

00

a

bb

a

b

b

a

bb

a

b b

b b

a

ab

c

ab b

Kansas Oklahoma Texas

Kansas Oklahoma Texas

(b)

(c) (d )

(e) (f )

Fig 1 Physiological responses for Panicum virgatum ecotypes across precipitation treatments Each point is the overall meanresponse (1 se) of the ecotype within the precipitation treatment (a) carbon assimilation at ambient CO2 (b) dark-adaptedchlorophyllfluorescence (c) stomatal conductance (d) water use efficiency (e) transpiration ( f )midday leaf-levelwater potentialIn (a) (b) (c) and (e) statistically significant (Plt 005) ecotype effects were present (varying letters among ecotypes denotesPlt 005) Significant precipitation treatment effects are present in (d) In ( f ) statistically significant ecotype and treatment effectswere present and significant differences are denoted by varying letters within each section of the panel

130 Functional Plant Biology J C Hartman et al

flowering tiller and the number of non-flowering tillers showedinteractions among ecotypes and precipitation treatments(Table 5) Texas ecotypes had significantly greater biomassper tiller in the increased treatments with the same results forbiomass per flowering tiller Kansas and Texas ecotypesdecreased the amount of non-flowering tillers as the soilmoisture increased Oklahoma ecotypes displayed the oppositetrend Oklahoma ecotypes had significantly fewer non-flowering

tillers in the decreased treatment compared with the other twoecotypes (Table 5)

Flowering

Flowering phenology was affected by both ecotype andprecipitation individually (Table 3) Texas ecotypes had halfthe number of flowering tillers at the end of season comparedwith Kansas and Oklahoma ecotypes (Table 4 Fig 4) Texasecotypes also initiated flowering a month later than the otherecotypes and flowered at a slower rate (Fig 4) Floweringphenology varied significantly with precipitation treatmentsIndividuals in the increased treatment had 40 and 38 moreflowering tillers at the end of the season compared with theaverage and decreased treatments respectively (Table 4Fig 4) Precipitation treatments affected flowering rates withthe increased treatment flowering at a greater rate compared withthe average and decreased treatment (Fig 4)

Discussion

The primary objective of this study was to characterise theecotype-specific physiological responses and growth ofP virgatum to changes in forecasted precipitation amount forthe Central Plains region of the United States by the year 2100Our results show that the precipitation treatments significantlyaffected soil moisture and the corresponding responses ofplant growth and physiology Significant effects of ecotype

gs (mol H2O mndash2 sndash1)

00 01 02 03 04 05

Fv

Fm

072

074

076

078

080

082

KansasOklahoma Texas

y = y0 + aln(abs(x))R-adj = 02442

Fig 2 Representative FvFmndashgs curves for all ecotypes Curves were fitfor all data points for each ecotype The R-adjusted value is for the overall fitof all ecotype data points

Early Mid Late Early Mid Late

Am

ax (

μmol

mndash2

sndash1

)E

(m

mol

mndash2

sndash1

)

0

10

20

30

40DecreasedAverageIncreased

000

005

010

015

020

025

Summer sampling period

0

2

4

6

8

10

Summer sampling period

WU

E (

μmol

mm

olndash1

)

0

1

2

3

4

5

6

a ab

a a

b

a a a

a a

b

a abb

a aa

a abb

aab

b

aab b

a a aa a

ba

aa

(a)

Early Mid LateEarly Mid Late

g s (

mol

H2O

mndash2

sndash1

)

(b)

(c) (d )

Fig 3 Gas exchange responses by treatment during the earlymid-season and late periods of the growing season (a)Carbon assimilationat ambient CO2 (b) stomatal conductance (c) transpiration (d) water use efficiency (Amax E

ndash1) Bars are means (1 se) and significantdifferences are indicated by letters within each sampling period

Responses of switchgrass to altered precipitation Functional Plant Biology 131

and precipitation were present but there were fewecotype precipitation interactions In general physiologicaldifferences reflected ecotype origin whereas differences inaboveground biomass largely reflected differences in soilwater availability from the precipitation treatments

Ecotype effects

Differences among means in the physiological variablesmeasured largely reflected differences by ecotype andcorresponding adaptation to the local environment of originWithin each ecotype similar changes in Amax gs and E reflectthe coupled gas exchange relationships documented previouslyfor other dominant prairie grasses (Polley et al 1992) It hasalso been shown that carbon assimilation in switchgrass iscontrolled mostly by population responses to localenvironmental conditions (Wullschleger et al 1996) In this

study the Kansas ecotypes had the lowest carbon assimilationrates providing support for the local environmental conditionresponse (Wullschleger et al 1996)

The FvFm ndash gs relationship developed (Fig 2) shows astatistical relationship between responses of leafphotochemistry to changes in stomatal conductance that variesby ecotypeReducedFvFmvalues indicate a decreased efficiencyof non-photochemical quenching and are reflective of thefunctioning of the light reactions of photosynthesis (Maxwelland Johnson 2000) Changes in FvFm can be especially useful instudies where physiological performance is strongly decreasedfrom longperiods of drought (Resco et al 2008) These data showthat nearly one-quarter of the variation themaximumefficiencyofPSII can be statistically attributed to changes in stomatalconductance (Fig 2) From first principles one would assumethat a reduction in stomatal water loss would result in increased

Table 3 Mixed-effects model ANOVA of precipitation treatments and ecotype (fixed precipitation ecotype random cell) effects on abovegroundbiomass variables

F- and P-values are provided with significant responses P lt 005 Plt 010

Response Precipitation Ecotype Ecotype precipitationF P F P F P

Tillers (per individual) 371 00326 338 00429 11 035Biomass (g per individual) 948 00004 21 012 15 021Biomass per tiller (g per individual) 449 00168 1109 00001 431 00050 allocated to reproduction (per individual) 483 00127 599 00050 191 012Flowering tillers (per individual) 754 00015 848 00008 05 068Flowering tiller biomass (g per individual) 1146 lt00001 08 044 15 019Biomass per flowering tiller (g per individual) 286 00681 1333 lt00001 351 00142Flowering tiller height (cm per individual) 512 00100 18 016 16 017Leaves per flowering tiller (per individual) 547 00075 22 011 11 035Non-flowering tillers (per individual) 325 00481 291 00651 287 00337Non-flowering tiller biomass (g per individual) 11 032 484 00126 05 067Biomass per non-flowering tiller (g per individual) 007 093 1585 lt00001 17 015Non-flowering tiller height (cm per individual) 09 038 305 00576 11 036Leaves per non-flowering tiller (per individual) 02 080 670 00029 18 013

Table 4 Mean responses (1 se) of various aboveground biomass measurements for ecotype and precipitation treatment effectsWithin the ecotype or treatment category bolded numbers show significance at the P 005 level and italicised numbers show marginal significance at theP 010 level Superscripts display significant differences between precipitation treatments or ecotypes within a category NS signifies no significance between

the ecotypes or the precipitation treatments

Variables Precipitation EcotypeDecreased Average Increased Kansas Oklahoma Texas

Tillers (per individual) 56 plusmn 5ab 47 plusmn 6b 71 plusmn 6a 69 plusmn 5a 58 plusmn 6ab 48 plusmn 6b

Biomass (g per individual) 202plusmn 28b 181 plusmn 30b 352 plusmn 30a NS NS NSBiomass per tiller (g per individual) 36 plusmn 03b 40 plusmn 03ab 50 plusmn 03a 30 plusmn 03b 44 plusmn 03a 52 plusmn 03a

Flowering tillers (per individual) 35 plusmn 4b 36 plusmn 4b 58 plusmn 4a 53 plusmn 4a 48 plusmn 5a 27 plusmn 5b

Flowering tiller biomass (g per individual) 157plusmn 27b 157 plusmn 28b 325 plusmn 28a NS NS NSBiomass per flowering tiller (g per individual) 41plusmn 04b 49plusmn 07ab 56plusmn 04a 32 plusmn 04b 50 plusmn 04a 64 plusmn 04a

Flowering tiller height (cm per individual) 94 plusmn 7b 101 plusmn 8b 129 plusmn 8a NS NS NSLeaves per flowering tiller (per individual) 47 plusmn 01b 51 plusmn 02ab 56 plusmn 02a NS NS NSNon-flowering tillers (per individual) 21 plusmn 2a 11 plusmn 3b 12 plusmn 3ab 15plusmn 2ab 9plusmn 3b 20plusmn 3a

Non-flowering tiller biomass (g per individual) NS NS NS 20 plusmn 9b 14 plusmn 11b 59 plusmn 11a

Biomass per non-flowering tiller (g per individual) NS NS NS 07 plusmn 01b 08 plusmn 01b 20 plusmn 01a

Non-flowering tiller height (cm per individual) NS NS NS 27plusmn 3ab 24plusmn 4b 38plusmn 4a

Leaves per non-flowering tiller (per individual) NS NS NS 32 plusmn 02b 30 plusmn 03b 44 plusmn 03a

allocated to reproduction (per individual) 77 plusmn 3b 85 plusmn 3ab 94 plusmn 3a 87 plusmn 3a 94 plusmn 4a 75 plusmn 4b

132 Functional Plant Biology J C Hartman et al

leaf temperature with the potential for corresponding decreasedphotochemistry during hot or dry summer conditions For allecotypesFvFm decreased nonlinearly with stomatal closure buttherewere clear differences in this relationship between ecotypes

with a general trend of increasing FvFm values for a given rate ofstomatal conductance from Kansas to Oklahoma to Texasecotypes Texas ecotypes had higher maximum FvFm valuescompared with Kansas ecotypes with values near the acceptedmaximal value of 083 (Krause and Weis 1991) As gs valuesdecreased Texas and Oklahoma ecotypes were able to maintainhigher FvFm values compared with the Kansas ecotypesindicating a greater decrease in the photochemical efficiencyof PSII in Kansas ecotypes Interpretation of the relationshipbetween FvFm and gs requires further investigation but thesedata show promise for linking changes in stomatal conductancewith themaintenance of leaf photochemistry as the ability to coolthe leaf through latent heat exchange becomes very limited withlow gs (Nippert et al 2009) and photochemistry is subsequentlyreducedby increased leaf temperatures (Albert et al 2011)Basedon theseFvFmndashgs response curves (Fig 2) Oklahoma and Texasecotypes are potentially more tolerant of higher leaf temperaturewith higher FvFm at low rates of gas exchange compared withKansas ecotypes

P virgatum is a highly photoperiod sensitive species(Benedict 1940) and differences in flowering dates have beendocumented for various ecotype and cultivars (McMillan 1965VanEsbroeck et al 2003Casler et al 2004) Sanderson andWolf(1995) recorded similar flowering dates for Alamo and Cave-in-Rock cultivars compared with the ecotypes in this study Alamocultivars flowered around the end of September and Cave-in-Rock cultivars flowered around the first week of July Theflowering time and rates seen in this study correspond withknown ecotype responses to changes in latitude

Many of the aboveground biomass responses with significantecotypic differences were a result of the Texas ecotypesproducing fewer tillers but with an increased biomass pertiller and number of leaves per tiller while the Oklahoma andKansas ecotypes had higher tiller density (Table 4) Typically thebest estimates to predict changes in biomass yield for this speciesare increased biomass per tiller and number of leaves per tiller(seen in Texas ecotypes) or higher tiller density (noted in the

Table 5 Mean responses (1 se) of various aboveground biomass measurements for interactions between ecotypes and precipitation treatmentsBolded numbers show significance at theP 005 level Superscripts display significant differences between ecotype precipitation combinations NS signifies

no significant difference between the means

Variables Ecotype precipitationDecreased Average Increased

Kansas Oklahoma Texas Kansas Oklahoma Texas Kansas Oklahoma Texas

Tillers (per individual) NS NS NS NS NS NS NS NS NSBiomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per tiller (g per individual) 28 plusmn 05b 42 plusmn 05b 38 plusmn 05b 29 plusmn 05b 50 plusmn 06ab 42 plusmn 06b 34 plusmn 05b 39 plusmn 06b 76 plusmn 06a

Flowering tillers (per individual) NS NS NS NS NS NS NS NS NSFlowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per flowering tiller (g per individual) 28 plusmn 07bc 44 plusmn 07bc 51 plusmn 07bc 31 plusmn 07bc 65 plusmn 08ab 52 plusmn 08abc 36 plusmn 07bc 42 plusmn 08bc 90 plusmn 08a

Flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per flowering tiller (per individual) NS NS NS NS NS NS NS NS NSNon-flowering tillers (per individual) 31 plusmn 4a 5 plusmn 5b 28 plusmn 5ab 8 plusmn 4b 7 plusmn 5ab 18 plusmn 5ab 7 plusmn 4b 15 plusmn 5ab 15 plusmn 5ab

Non-flowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per non-flowering tiller (g per individual) NS NS NS NS NS NS NS NS NSNon-flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per non-flowering tiller (per individual) NS NS NS NS NS NS NS NS NS allocated to reproduction (per individual) NS NS NS NS NS NS NS NS NS

Nov

Till

ers

flow

ered

(in

divndash1

)T

iller

s flo

wer

ed (

indi

vndash1)

0

10

20

30

40

50

60IncreasedAverageDecreased

0

10

20

30

40

50

60 Kansas Oklahoma Texas

(a)

(b)

Jul Aug Sep Oct

Fig 4 Total tillers flowered per individual over the course of the growingseason (a) Flowering tillers by ecotype (b) flowering tillers by precipitationtreatment Data points are means 1se Asterisks indicate statisticalsignificance (Plt 005) between groups within each sampling date

Responses of switchgrass to altered precipitation Functional Plant Biology 133

Oklahoma and Kansas ecotypes Boe and Beck 2008) Althoughtotal biomass was not significantly different among ecotypes theTexas ecotypes contain components (higher biomass and leavesper tiller) that if heritablemay lead to greater biomass inmultiplegeneration studies

Precipitation effects

The altered environmental conditions resulting from theprecipitation treatments significantly altered responses inaboveground biomass measurements with statisticallysignificant responses in the physiological variables largelyreflecting changes across the sampling periods (Fig 3)Flowering tiller heights were greatest in the increasedprecipitation treatments reaching mean heights of 129mThese heights are reflective of those observed in prairieremnant populations although they are lower than those foundin P virgatum cultivars (Das et al 2004 Casler 2005Alexopoulou et al 2008) but it should be noted that plantheights are variable from year to year (Alexopoulou et al2008) The average number of leaves per flowering tiller waslower than those found in agronomic P virgatum cultivars (VanEsbroeck et al 1997) however these cultivars have been bred tooptimise yield These data illustrate individual biomass responsesto precipitation of a dominant species in the tallgrass prairieecosystem further supporting the conclusions of others thatfuture changes in precipitation are likely to significantly alterecosystem productivity (Knapp 1984 Fay et al 2003 Fay et al2008 Zhou et al 2009)

Previous attempts to link leaf-level to whole-plant responsesfor grassland species have shown positive relationships betweenphotosynthesis and plant success (McAllister et al 1998) butleaf-level to whole-plant relationships in grasslands arecommonly weak (Nippert et al 2007) In this study no clearrelationships between leaf-level photosynthetic responses andtotal aboveground biomass were present for these ecotypesAlthough the increases in leaf-level photosynthesis did notcorrespond to total biomass increases other measurements ofbiomass increased from the increased precipitation partiallysupporting the positive relationship between higher rates ofgas exchange and plant success previously noted byMcAllister et al (1998)

The only physiological responses affected by precipitationtreatments were Ymid and WUE with WUE being lowest inthe increased treatments andYmid highest in increased treatmentsAlthough ANOVA results showed significant precipitationsample period interactions (Table 2) many of the effects wereonly seen in the early season period (Fig 3) Later in the seasonincreasing summer temperatures negated the positive impacts ofthe added water Decreased WUE across the growing seasoncorresponding to higher air temperatures has been previouslynoted for C4 species where WUE can be reduced by 40 indrought conditions (Taylor et al 2011) As water availabilitydecreasedYmid decreased to amean of ndash22MPa The decreasedprecipitation treatment did not lower soil water availability tothe point where the critical water potential (leaf pressurepotential when stomatal closure is less than 5 of themaximum) was reached for switchgrass (Ycrit = ndash32MPaTucker et al 2011) Indeed previous studies have shown that

switchgrass can maintain growth and functioning at lower waterpotentials than those recorded here (Knapp 1984 Stroup et al2003) With more severe decreases in water availability YmidAmax gs E and WUE should continue to decline with thepotential for increased cuticular conductance when drought-stressed (Manzoni et al 2011)

Ecotype precipitation interactions

Surprisingly few ecotype precipitation interactions werepresent for the variablesmeasured in this study (Tables 2 3) withthe primary exception being greater biomass per tiller in theincreased treatment for Texas ecotypes These differences largelyreflect the southern Texas ecotype being adapted to a longergrowing season (Newell 1968) Growth of the Texas ecotype in alocation with a shorter growing season delayed reproductivematurity and although the overall biomass yield was notsignificantly greater for Texas ecotypes compared withOklahoma and Kansas under greater precipitation differencesthe trends observed in the current study may become statisticallysignificant Because Texas ecotypes had a longer period in whichthey grew vegetatively theywere better able to take advantage ofthe increased precipitation treatment to produce higherbiomass per tiller greater biomass per flowering tiller and highnon-flowering tiller counts in all precipitation treatments Theseresults from a southern Texas ecotype growing in amore northernlocation are similar to other studies that show increases inbiomass longer photosynthetic activity and delayed floweringfor southern ecotypes grown in more northern latitudes (Quinn1969 Lemus et al 2002 Casler et al 2004 Berdahl et al 2005Casler et al 2007)

Conclusion

The physiological responses of Amax gs E FvFm and Ymid inP virgatum were explained by ecotype differences howeverrobust responses to altered precipitationwere seen inWUEYmidflowering times and many aboveground biomass variablesEcotype differences were also seen in several abovegroundbiomass variables and most strikingly in flowering initiationand rates In this study there were few interactions betweenecotype and precipitation amount suggesting that althoughprecipitation is a strong driver of biomass production localadaptations in the ecotypes used in the current study have lessimpact on these responses than initially expected It was expectedthat the environmental differences among the origins of theecotypes mainly the effects of higher temperature coupledwith similar mean annual precipitation (860ndash878mm) wouldlead to significant ecotype precipitation interactions in theparameters measured with the Texas ecotypes being the mosttolerant of the changes The lack of interactions between geneticbackground and precipitation also has broader evolutionaryimplications gene environment interactions are a primaryindicator of the potential for natural selection in a given trait(Schlichting 1989) The broader implications of the fewecotype precipitation interactions for this widely-distributedspecies suggest that it is not likely toundergo selection in responseto changes in precipitation As many have correctly suggested itis necessary to examine the role of genetic variation in theresponse of plant species to future changes in climate

134 Functional Plant Biology J C Hartman et al

(Callaway et al 2003 Ward and Kelly 2004) however asindicated by this study it may not always be necessary toconsider localised adaptation when working to project theimpacts of climate change on the performance of a species likeswitchgrass across its geographical distribution

Acknowledgements

We thankDr JohnBlair andDr TedMorgan for their advice and comments onthemanuscriptWealso thank theNatureConservancy for allowingcollectionof switchgrass rhizomes from the Tallgrass Prairie Preserve inOklahoma andlandholdingsnear Temple TXGracieOrozcoRebeccaOrozco JacobCarterand Teall Culbertson provided assistance with fieldwork and the Division ofBiology at Kansas State University Konza Prairie Biological Station KonzaPrairie LTER (DEB-0823341) and the US Department of Energy(EEE000314) provided funding and access to facilities

References

Albert KR Mikkelsen TN Michelsen A Ro-Poulsen H van der Linden L(2011) Interactive effects of drought elevated CO2 and warming onphotosynthetic capacity and photosystem performance in temperateheath plants Journal of Plant Physiology 168 1550ndash1561 doi101016jjplph201102011

Alexopoulou E Sharma N Papatheohari Y Christou M Piscioneri IPanoutsou D Pignatelli V (2008) Biomass yields for upland andlowland switchgrass varieties grown in the Mediterranean regionBiomass and Bioenergy 32 926ndash933 doi101016jbiombioe200801015

AlleyRBBerntsen T Bindoff NL ChenZ ChidthaisongA et al (2007) Hephysical science basis summary for policy makersrsquo (IPCC SecretariatGeneva)

Barney JN Mann JJ Kyser GB Blumwald B Deynze AV DiTomaso JM(2009)Toleranceof switchgrass to extremesoilmoisture stress ecologicalimplications Plant Science 177 724ndash732 doi101016jplantsci200909003

Benedict HM (1940) Effect of day length and temperature on the floweringand growth of four species of grasses Journal of Agricultural Research61 661ndash671

Berdahl JD Frank AB Krupinsky JM Carr PM Hanson JD Johnson HA(2005) Biomass yield phenology and survival of diverse switchgrasscultivars and experimental strains in western North Dakota AgronomyJournal 97 549ndash555 doi102134agronj20050549

Berg CC (1971) Forage yield of switchgrass (Panicum virgatum) inPennsylvania Agronomy Journal 63 785ndash786 doi102134agronj197100021962006300050038x

Boe A Beck DL (2008) Yield components of biomass in switchgrass CropScience 48 1306ndash1311 doi102135cropsci2007080482

Callaway RM Pennings SC Richards CL (2003) Phenotypic plasticity andinteractions among plants Ecology 84 1115ndash1128 doi1018900012-9658(2003)084[1115PPAIAP]20CO2

Casler MD (2005) Ecotypic variation among switchgrass populations fromthe northern USA Crop Science 45 388ndash398 doi102135cropsci20050388

Casler MD Vogel KP Taliaferro CM Wynia RL (2004) Latitudinaladaptation of switchgrass populations Crop Science 44 293ndash303

Casler MD Vogel KP Taliaferro CM Ehlke NJ Berdahl JD Brummer ECKallenbach RL West CP Mitchell RB (2007) Latitudinal andlongitudinal adaptation of switchgrass populations Crop Science 472249ndash2260 doi102135cropsci2006120780

Christensen JH Hewitson B Busuioc A Chen A Gao X Held I Jones RKolli RK KwonWT Laprise R Magantildea Rueda VMearns L MeneacutendezCG Raumlisaumlnen J Rinke A Sarr A (2007) Regional climate projectionsIn lsquoClimate change 2007 the physical science basis Contributionof Working Group I to the Fourth Assessment Report of the

Intergovernmental Panel on Climate Change (eds S Solomon D QinM Manning Z Chen M Marquis KB Averyt M Tignor HL Miller)pp 887ndash892 (Cambridge University Press Cambridge UK)

Das MK Fuentes RG Taliaferro CM (2004) Genetic variability and traitrelationships in switchgrass Crop Science 44 443ndash448

Easterling DR Meehl GA Parmesan C Changnon SA Karl TR Mearns LO(2000) Climate extremes observations modeling and impacts Science289 2068ndash2074 doi101126science28954872068

Emery SM Gross KL (2007) Dominant species identity not communityevenness regulates invasion in experimental grassland communitiesEcology 88 954ndash964 doi10189006-0568

Evers GW Parsons MJ (2003) Soil type and moisture level influence onAlamo switchgrass emergence and seedling growth Crop Science 43288ndash294 doi102135cropsci20030288

Fay PA Carlisle JD Knapp AK Blair JM Collins SL (2003) Productivityresponses to altered rainfall patterns in a C4-dominated grasslandOecologia 137 245ndash251 doi101007s00442-003-1331-3

FayPAKaufmanDMNippert JBCarlisle JDHarperCW(2008)Changes ingrassland ecosystem function due to extreme rainfall events implicationsfor responses to climate changeGlobal Change Biology 14 1600ndash1608doi101111j1365-2486200801605x

Fay PA Blair JM Smith MD Nippert JB Carlisle JD Knapp AK (2011)Relative effects of precipitation variability and warming on tallgrassprairie ecosystem function Biogeosciences 8 3053ndash3068 doi105194bg-8-3053-2011

Grime JP (1998) Benefits of plant diversity to ecosystems immediatefilter and founder effects Journal of Ecology 86 902ndash910doi101046j1365-2745199800306x

Hartman JC Nippert JB OrozcoRA Springer CJ (2011) Potential ecologicalimpacts of switchgrass (Panicum virgatum L) biofuel cultivation in theCentral Great Plains USA Biomass and Bioenergy 35 3415ndash3421doi101016jbiombioe201104055

Heaton E Voigt T Long SP (2004) A quantitative review comparing theyields of two candidate C4 perennial biomass crops in relation to nitrogentemperature and water Biomass and Bioenergy 27 21ndash30 doi101016jbiombioe200310005

HillebrandHBennettDMCadotteMW(2008)Consequences of dominancea review of evenness effects on local and regional ecosystem processesEcology 89 1510ndash1520 doi10189007-10531

Hughes AR Inouye BD Johnson MTJ Underwood N Vellend M (2008)Ecological consequences of genetic diversity Ecology Letters 11609ndash623 doi101111j1461-0248200801179x

Jump AS Pentildeuelas J (2005) Running to stand still adaptation and responseof plants to rapid climate change Ecology Letters 8 1010ndash1020doi101111j1461-0248200500796x

KnappAK (1984)Water relations and growth of three grasses duringwet anddrought years in a tallgrass prairie Oecologia 65 35ndash43 doi101007BF00384460

Knapp AK (1985) Effect of fire and drought on the ecophysiology ofAndropogon gerardii and Panicum virgatum in a tallgrass prairieEcology 66 1309ndash1320 doi1023071939184

KnappAKBriggs JMHartnett DCCollins SC (1998) lsquoGrassland dynamicslong-term ecological research in tallgrass prairiersquo (Oxford UniversityPress New York)

KnappAKFayPABlair JMCollinsSLSmithMDCarlisle JDHarperCWDanner BT Lett MS McCarron JK (2002) Rainfall variability carboncycling and plant species diversity in mesic grassland Science 2982202ndash2205 doi101126science1076347

Krause GH Weis E (1991) Chlorophyll fluorescence and photosynthesisthe basics Annual Review of Plant Physiology and Plant MolecularBiology 42 313ndash349 doi101146annurevpp42060191001525

Lemus R Brummer EC Moore KJ Molstad ME Burras CE Barker MF(2002) Biomass yield and quality of 20 switchgrass populations insouthern Iowa USA Biomass and Bioenergy 23 433ndash442doi101016S0961-9534(02)00073-9

Responses of switchgrass to altered precipitation Functional Plant Biology 135

Manzoni S Vico G Katul G Fay PA Polley W Palmroth S Porporato A(2011) Optimizing stomatal conductance for maximum carbon gainunder water stress a meta-analysis across plant functional types andclimates Functional Ecology 25 456ndash467 doi101111j1365-2435201001822x

Maxwell K JohnsonGN (2000) Chlorophyll fluorescence ndash a practical guideJournal of Experimental Botany 51 659ndash668 doi101093jexbot51345659

McAllister CA Knapp AK Maragni LA (1998) Is leaf-level photosynthesisrelated to plant success in a highly productive grasslandOecologia 11740ndash46 doi101007s004420050629

McLaughlin SB Kszos LN (2005) Development of switchgrass (Panicumvirgatum) as a bioenergy feedstock in the United States Biomass andBioenergy 28 515ndash535 doi101016jbiombioe200405006

McMillan C (1965) Ecotypic differentiation within four NorthAmerican prairie grasses II Behavioral variation within transplantedcommunity fractions American Journal of Botany 52 55ndash65doi1023072439975

McNaughton SJ Wolf LL (1970) Dominance and the niche in ecologicalsystems Science 167 131ndash139 doi101126science1673915131

Meehl GA Arblaster JM Tebaldi C (2005) Understanding future patterns ofincreased precipitation intensity in climate model simulationsGeophysical Research Letters 32 L18719 doi1010292005GL023680

Muir JP Sanderson MA Ocumpaugh WR Jones RM Reed RL (2001)Biomass production of lsquoAlamorsquo switchgrass in response to nitrogenphosphorus and row spacing Agronomy Journal 93 896ndash901doi102134agronj2001934896x

Newell LC (1968) Effects of strain source and management practice onforage yields of two warm-season prairie grasses Crop Science 8205ndash210 doi102135cropsci19680011183X000800020022x

Nippert JBKnappAKBriggs JM (2006) Intra-annual rainfall variability andgrassland productivity can the past predict the futurePlant Ecology 18465ndash74 doi101007s11258-005-9052-9

Nippert JB Fay PA Knapp AK (2007) Photosynthetic traits in C3 and C4

grassland species in mesocosm and field environments Environmentaland Experimental Botany 60 412ndash420 doi101016jenvexpbot200612012

Nippert JB Fay PA Carlisle JD Knapp AK Smith MD (2009)Ecophysiological responses of two dominant grasses to alteredtemperature and precipitation regimes Acta Oecologica 35 400ndash408doi101016jactao200901010

Norberg J Swaney DP Dushoff J Lin J Casagrandi R Levin SA (2001)Phenotypic diversity and ecosystem functioning in changingenvironments a theoretical framework Proceedings of the NationalAcademy of Sciences of the United States of America 9811376ndash11381 doi101073pnas171315998

Parrish DJ Fike JH (2005) The biology and agronomy of switchgrass forbiofuels Critical Reviews in Plant Sciences 24 423ndash459 doi10108007352680500316433

Polley HW Norman JM Arkebauer TJ Walter-Shea EA Greegor DH JrBramer B (1992) Leaf gas exchange of Andropogon gerardii VitmanPanicum virgatum L and Sorghastrum nutans (L) Nash in a tallgrassprairie Journal of Geophysical Research 97 18 837ndash18 844

Quinn JA (1969) Variability among High Plains populations of Panicumvirgatum Bulletin of the Torrey Botanical Club 96 20ndash41 doi1023072484005

Resco V IgnanceDD SunWHuxman TEWeltzin JFWilliams DG (2008)Chlorophyll fluorescence predawn water potential and photosynthesisin pulse-drive ecosystems ndash implications for ecological studiesFunctional Ecology 22 479ndash483 doi101111j1365-2435200801396x

Sanderson MA (1992) Morphological development of switchgrass andkleingrass Agronomy Journal 84 415ndash419 doi102134agronj199200021962008400030013x

Sanderson MA Reed RL (2000) Switchgrass growth and developmentwater nitrogen and plant density effects Journal of RangeManagement 53 221ndash227 doi1023074003287

SandersonMAWolf DD (1995)Morphological development of switchgrassin diverse environments Agronomy Journal 87 908ndash915 doi102134agronj199500021962008700050022x

Schlichting CD (1989) Phenotypic integration and environmental changeBioscience 39 460ndash464 doi1023071311138

SillettiAMKnappAK(2001)Responsesof the codominantgrassland speciesAndropogon gerardii and Sorghastrum nutans to long-termmanipulations of nitrogen and water American Midland Naturalist145 159ndash167 doi1016740003-0031(2001)145[0159ROTCGS]20CO2

SmithMDKnappAK (2003)Dominant speciesmaintain ecosystem functionwith non-random species loss Ecology Letters 6 509ndash517 doi101046j1461-0248200300454x

Sokal RRRohlf FJ (1995) lsquoBiometry the principles and practices of statisticsin biological research 3rd ednrsquo (WH Freeman New York)

StoutWL (1992)Water-use efficiency of grasses as affected by soil nitrogenand temperature Soil Science Society of America Journal 56 897ndash902doi102136sssaj199203615995005600030036x

Stout WL Jung GA Shaffer JA (1988) Effects of soil and nitrogen on wateruse efficiency of tall fescue and switchgrass under humid conditions SoilScience Society of America Journal 52 429ndash434 doi102136sssaj198803615995005200020023x

Stroup JA Sanderson MA Muir JP McFarland MJ Reed RL (2003)Comparison of growth and performance in upland and lowlandswitchgrass types to water and nitrogen stress Bioresource Technology86 65ndash72 doi101016S0960-8524(02)00102-5

Taylor SH Ripley BSWoodward FI Osborne CP (2011) Drought limitationofphotosynthesis differs betweenC3andC4grass species in a comparativeexperiment Plant Cell amp Environment 34 65ndash75 doi101111j1365-3040201002226x

Tucker SS Craine JM Nippert JB (2011) Physiological drought toleranceand the structuring of tallgrass prairie assemblages Ecosphere 2 1ndash19doi101890ES11-000231

Van Esbroeck GA Hussey MA Sanderson MA (1997) Leaf appearancerate and final leaf number of switchgrass cultivars Crop Science 37864ndash870 doi102135cropsci19970011183X003700030028x

Van Esbroeck GA Hussey MA Sanderson MA (2003) Variation betweenAlamo and Cave-in- Rock switchgrass in response to photoperiodextension Crop Science 43 639ndash643

WangDLeBauerDSDietzeMC(2010)Aquantitative reviewcomparing theyield of switchgrass in monocultures and mixtures in relation to climateand management factors GCB Bioenergy 2 16ndash25 doi101111j1757-1707201001035x

Ward JK Kelly JK (2004) Scaling up evolutionary responses to CO2 lessonsfrom Arabidopsis Ecology Letters 7 427ndash440 doi101111j1461-0248200400589x

Wullschleger SD SandersonMAMcLaughlin SB Biradar DP Rayburn AL(1996) Photosynthetic rates and ploidy levels among populations ofswitchgrassCrop Science 36 306ndash312 doi102135cropsci19960011183X003600020016x

Zhou X Talley M Luo Y (2009) Biomass litter and soil respiration along aprecipitation gradient in southern Great Plains USA Ecosystems(New York NY) 12 1369ndash1380 doi101007s10021-009-9296-7

136 Functional Plant Biology J C Hartman et al

wwwpublishcsiroaujournalsfpb

Page 5: Ecotypic responses of switchgrass to altered precipitation

response to the precipitation treatments (Table 4) Individuals inthe increased treatment had on average 15 more tillers anddouble the total aboveground and flowering tiller biomassThose individuals in the increased treatment also displayed a22 and 28 increase in mean tiller height compared with theaverage and decreased treatments respectively The increasedtreatment contained individuals with a greater number of leavesper tiller and that allocated over 94 of their biomass toreproductive tillers (Table 4)

Ecotype was a significant predictor of some abovegroundbiomass responses (Table 3) with differences in the total number

of tillers biomass per non-flowering tiller non-flowering tillerbiomass leaves per non-flowering tiller and percent biomassallocated to reproduction (Table 4) Texas ecotypes had amean ofless than 50 tillers per individual and over double the non-flowering tiller biomass and biomass per non-flowering tillercompared with Kansas and Oklahoma ecotypes Texas ecotypesalso displayed significantly less biomass allocation toreproductive tillers (75) compared with Kansas (87) andOklahoma ecotypes (94)

There were few ecotype precipitation interactions(Table 3) but the total biomass per tiller biomass per

EcotypeKansas Oklahoma Texas

Kansas Oklahoma Texas

Am

ax (

μ mol

mndash2

sndash1

)

0

5

10

15

20

25

(a)

Ecotype

g s (

mol

H2O

mndash2

sndash1

)E

(m

mol

mndash2

sndash1

)

WU

E (

μmol

mm

olndash1

)

000

002

004

006

008

010

012

014

Ecotype

0

1

2

3

4

5

6

Treatment

Decreased Average Increased0

1

2

3

4

5

Ecotype

FvF

m

073

074

075

076

077

078

079

080

081

Decre

ased

Avera

ge

Incr

ease

d

Kansa

s

Oklaho

ma

Texas

ψm

id (

MP

a)

ndash25

ndash20

ndash15

ndash10

ndash05

00

a

bb

a

b

b

a

bb

a

b b

b b

a

ab

c

ab b

Kansas Oklahoma Texas

Kansas Oklahoma Texas

(b)

(c) (d )

(e) (f )

Fig 1 Physiological responses for Panicum virgatum ecotypes across precipitation treatments Each point is the overall meanresponse (1 se) of the ecotype within the precipitation treatment (a) carbon assimilation at ambient CO2 (b) dark-adaptedchlorophyllfluorescence (c) stomatal conductance (d) water use efficiency (e) transpiration ( f )midday leaf-levelwater potentialIn (a) (b) (c) and (e) statistically significant (Plt 005) ecotype effects were present (varying letters among ecotypes denotesPlt 005) Significant precipitation treatment effects are present in (d) In ( f ) statistically significant ecotype and treatment effectswere present and significant differences are denoted by varying letters within each section of the panel

130 Functional Plant Biology J C Hartman et al

flowering tiller and the number of non-flowering tillers showedinteractions among ecotypes and precipitation treatments(Table 5) Texas ecotypes had significantly greater biomassper tiller in the increased treatments with the same results forbiomass per flowering tiller Kansas and Texas ecotypesdecreased the amount of non-flowering tillers as the soilmoisture increased Oklahoma ecotypes displayed the oppositetrend Oklahoma ecotypes had significantly fewer non-flowering

tillers in the decreased treatment compared with the other twoecotypes (Table 5)

Flowering

Flowering phenology was affected by both ecotype andprecipitation individually (Table 3) Texas ecotypes had halfthe number of flowering tillers at the end of season comparedwith Kansas and Oklahoma ecotypes (Table 4 Fig 4) Texasecotypes also initiated flowering a month later than the otherecotypes and flowered at a slower rate (Fig 4) Floweringphenology varied significantly with precipitation treatmentsIndividuals in the increased treatment had 40 and 38 moreflowering tillers at the end of the season compared with theaverage and decreased treatments respectively (Table 4Fig 4) Precipitation treatments affected flowering rates withthe increased treatment flowering at a greater rate compared withthe average and decreased treatment (Fig 4)

Discussion

The primary objective of this study was to characterise theecotype-specific physiological responses and growth ofP virgatum to changes in forecasted precipitation amount forthe Central Plains region of the United States by the year 2100Our results show that the precipitation treatments significantlyaffected soil moisture and the corresponding responses ofplant growth and physiology Significant effects of ecotype

gs (mol H2O mndash2 sndash1)

00 01 02 03 04 05

Fv

Fm

072

074

076

078

080

082

KansasOklahoma Texas

y = y0 + aln(abs(x))R-adj = 02442

Fig 2 Representative FvFmndashgs curves for all ecotypes Curves were fitfor all data points for each ecotype The R-adjusted value is for the overall fitof all ecotype data points

Early Mid Late Early Mid Late

Am

ax (

μmol

mndash2

sndash1

)E

(m

mol

mndash2

sndash1

)

0

10

20

30

40DecreasedAverageIncreased

000

005

010

015

020

025

Summer sampling period

0

2

4

6

8

10

Summer sampling period

WU

E (

μmol

mm

olndash1

)

0

1

2

3

4

5

6

a ab

a a

b

a a a

a a

b

a abb

a aa

a abb

aab

b

aab b

a a aa a

ba

aa

(a)

Early Mid LateEarly Mid Late

g s (

mol

H2O

mndash2

sndash1

)

(b)

(c) (d )

Fig 3 Gas exchange responses by treatment during the earlymid-season and late periods of the growing season (a)Carbon assimilationat ambient CO2 (b) stomatal conductance (c) transpiration (d) water use efficiency (Amax E

ndash1) Bars are means (1 se) and significantdifferences are indicated by letters within each sampling period

Responses of switchgrass to altered precipitation Functional Plant Biology 131

and precipitation were present but there were fewecotype precipitation interactions In general physiologicaldifferences reflected ecotype origin whereas differences inaboveground biomass largely reflected differences in soilwater availability from the precipitation treatments

Ecotype effects

Differences among means in the physiological variablesmeasured largely reflected differences by ecotype andcorresponding adaptation to the local environment of originWithin each ecotype similar changes in Amax gs and E reflectthe coupled gas exchange relationships documented previouslyfor other dominant prairie grasses (Polley et al 1992) It hasalso been shown that carbon assimilation in switchgrass iscontrolled mostly by population responses to localenvironmental conditions (Wullschleger et al 1996) In this

study the Kansas ecotypes had the lowest carbon assimilationrates providing support for the local environmental conditionresponse (Wullschleger et al 1996)

The FvFm ndash gs relationship developed (Fig 2) shows astatistical relationship between responses of leafphotochemistry to changes in stomatal conductance that variesby ecotypeReducedFvFmvalues indicate a decreased efficiencyof non-photochemical quenching and are reflective of thefunctioning of the light reactions of photosynthesis (Maxwelland Johnson 2000) Changes in FvFm can be especially useful instudies where physiological performance is strongly decreasedfrom longperiods of drought (Resco et al 2008) These data showthat nearly one-quarter of the variation themaximumefficiencyofPSII can be statistically attributed to changes in stomatalconductance (Fig 2) From first principles one would assumethat a reduction in stomatal water loss would result in increased

Table 3 Mixed-effects model ANOVA of precipitation treatments and ecotype (fixed precipitation ecotype random cell) effects on abovegroundbiomass variables

F- and P-values are provided with significant responses P lt 005 Plt 010

Response Precipitation Ecotype Ecotype precipitationF P F P F P

Tillers (per individual) 371 00326 338 00429 11 035Biomass (g per individual) 948 00004 21 012 15 021Biomass per tiller (g per individual) 449 00168 1109 00001 431 00050 allocated to reproduction (per individual) 483 00127 599 00050 191 012Flowering tillers (per individual) 754 00015 848 00008 05 068Flowering tiller biomass (g per individual) 1146 lt00001 08 044 15 019Biomass per flowering tiller (g per individual) 286 00681 1333 lt00001 351 00142Flowering tiller height (cm per individual) 512 00100 18 016 16 017Leaves per flowering tiller (per individual) 547 00075 22 011 11 035Non-flowering tillers (per individual) 325 00481 291 00651 287 00337Non-flowering tiller biomass (g per individual) 11 032 484 00126 05 067Biomass per non-flowering tiller (g per individual) 007 093 1585 lt00001 17 015Non-flowering tiller height (cm per individual) 09 038 305 00576 11 036Leaves per non-flowering tiller (per individual) 02 080 670 00029 18 013

Table 4 Mean responses (1 se) of various aboveground biomass measurements for ecotype and precipitation treatment effectsWithin the ecotype or treatment category bolded numbers show significance at the P 005 level and italicised numbers show marginal significance at theP 010 level Superscripts display significant differences between precipitation treatments or ecotypes within a category NS signifies no significance between

the ecotypes or the precipitation treatments

Variables Precipitation EcotypeDecreased Average Increased Kansas Oklahoma Texas

Tillers (per individual) 56 plusmn 5ab 47 plusmn 6b 71 plusmn 6a 69 plusmn 5a 58 plusmn 6ab 48 plusmn 6b

Biomass (g per individual) 202plusmn 28b 181 plusmn 30b 352 plusmn 30a NS NS NSBiomass per tiller (g per individual) 36 plusmn 03b 40 plusmn 03ab 50 plusmn 03a 30 plusmn 03b 44 plusmn 03a 52 plusmn 03a

Flowering tillers (per individual) 35 plusmn 4b 36 plusmn 4b 58 plusmn 4a 53 plusmn 4a 48 plusmn 5a 27 plusmn 5b

Flowering tiller biomass (g per individual) 157plusmn 27b 157 plusmn 28b 325 plusmn 28a NS NS NSBiomass per flowering tiller (g per individual) 41plusmn 04b 49plusmn 07ab 56plusmn 04a 32 plusmn 04b 50 plusmn 04a 64 plusmn 04a

Flowering tiller height (cm per individual) 94 plusmn 7b 101 plusmn 8b 129 plusmn 8a NS NS NSLeaves per flowering tiller (per individual) 47 plusmn 01b 51 plusmn 02ab 56 plusmn 02a NS NS NSNon-flowering tillers (per individual) 21 plusmn 2a 11 plusmn 3b 12 plusmn 3ab 15plusmn 2ab 9plusmn 3b 20plusmn 3a

Non-flowering tiller biomass (g per individual) NS NS NS 20 plusmn 9b 14 plusmn 11b 59 plusmn 11a

Biomass per non-flowering tiller (g per individual) NS NS NS 07 plusmn 01b 08 plusmn 01b 20 plusmn 01a

Non-flowering tiller height (cm per individual) NS NS NS 27plusmn 3ab 24plusmn 4b 38plusmn 4a

Leaves per non-flowering tiller (per individual) NS NS NS 32 plusmn 02b 30 plusmn 03b 44 plusmn 03a

allocated to reproduction (per individual) 77 plusmn 3b 85 plusmn 3ab 94 plusmn 3a 87 plusmn 3a 94 plusmn 4a 75 plusmn 4b

132 Functional Plant Biology J C Hartman et al

leaf temperature with the potential for corresponding decreasedphotochemistry during hot or dry summer conditions For allecotypesFvFm decreased nonlinearly with stomatal closure buttherewere clear differences in this relationship between ecotypes

with a general trend of increasing FvFm values for a given rate ofstomatal conductance from Kansas to Oklahoma to Texasecotypes Texas ecotypes had higher maximum FvFm valuescompared with Kansas ecotypes with values near the acceptedmaximal value of 083 (Krause and Weis 1991) As gs valuesdecreased Texas and Oklahoma ecotypes were able to maintainhigher FvFm values compared with the Kansas ecotypesindicating a greater decrease in the photochemical efficiencyof PSII in Kansas ecotypes Interpretation of the relationshipbetween FvFm and gs requires further investigation but thesedata show promise for linking changes in stomatal conductancewith themaintenance of leaf photochemistry as the ability to coolthe leaf through latent heat exchange becomes very limited withlow gs (Nippert et al 2009) and photochemistry is subsequentlyreducedby increased leaf temperatures (Albert et al 2011)Basedon theseFvFmndashgs response curves (Fig 2) Oklahoma and Texasecotypes are potentially more tolerant of higher leaf temperaturewith higher FvFm at low rates of gas exchange compared withKansas ecotypes

P virgatum is a highly photoperiod sensitive species(Benedict 1940) and differences in flowering dates have beendocumented for various ecotype and cultivars (McMillan 1965VanEsbroeck et al 2003Casler et al 2004) Sanderson andWolf(1995) recorded similar flowering dates for Alamo and Cave-in-Rock cultivars compared with the ecotypes in this study Alamocultivars flowered around the end of September and Cave-in-Rock cultivars flowered around the first week of July Theflowering time and rates seen in this study correspond withknown ecotype responses to changes in latitude

Many of the aboveground biomass responses with significantecotypic differences were a result of the Texas ecotypesproducing fewer tillers but with an increased biomass pertiller and number of leaves per tiller while the Oklahoma andKansas ecotypes had higher tiller density (Table 4) Typically thebest estimates to predict changes in biomass yield for this speciesare increased biomass per tiller and number of leaves per tiller(seen in Texas ecotypes) or higher tiller density (noted in the

Table 5 Mean responses (1 se) of various aboveground biomass measurements for interactions between ecotypes and precipitation treatmentsBolded numbers show significance at theP 005 level Superscripts display significant differences between ecotype precipitation combinations NS signifies

no significant difference between the means

Variables Ecotype precipitationDecreased Average Increased

Kansas Oklahoma Texas Kansas Oklahoma Texas Kansas Oklahoma Texas

Tillers (per individual) NS NS NS NS NS NS NS NS NSBiomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per tiller (g per individual) 28 plusmn 05b 42 plusmn 05b 38 plusmn 05b 29 plusmn 05b 50 plusmn 06ab 42 plusmn 06b 34 plusmn 05b 39 plusmn 06b 76 plusmn 06a

Flowering tillers (per individual) NS NS NS NS NS NS NS NS NSFlowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per flowering tiller (g per individual) 28 plusmn 07bc 44 plusmn 07bc 51 plusmn 07bc 31 plusmn 07bc 65 plusmn 08ab 52 plusmn 08abc 36 plusmn 07bc 42 plusmn 08bc 90 plusmn 08a

Flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per flowering tiller (per individual) NS NS NS NS NS NS NS NS NSNon-flowering tillers (per individual) 31 plusmn 4a 5 plusmn 5b 28 plusmn 5ab 8 plusmn 4b 7 plusmn 5ab 18 plusmn 5ab 7 plusmn 4b 15 plusmn 5ab 15 plusmn 5ab

Non-flowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per non-flowering tiller (g per individual) NS NS NS NS NS NS NS NS NSNon-flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per non-flowering tiller (per individual) NS NS NS NS NS NS NS NS NS allocated to reproduction (per individual) NS NS NS NS NS NS NS NS NS

Nov

Till

ers

flow

ered

(in

divndash1

)T

iller

s flo

wer

ed (

indi

vndash1)

0

10

20

30

40

50

60IncreasedAverageDecreased

0

10

20

30

40

50

60 Kansas Oklahoma Texas

(a)

(b)

Jul Aug Sep Oct

Fig 4 Total tillers flowered per individual over the course of the growingseason (a) Flowering tillers by ecotype (b) flowering tillers by precipitationtreatment Data points are means 1se Asterisks indicate statisticalsignificance (Plt 005) between groups within each sampling date

Responses of switchgrass to altered precipitation Functional Plant Biology 133

Oklahoma and Kansas ecotypes Boe and Beck 2008) Althoughtotal biomass was not significantly different among ecotypes theTexas ecotypes contain components (higher biomass and leavesper tiller) that if heritablemay lead to greater biomass inmultiplegeneration studies

Precipitation effects

The altered environmental conditions resulting from theprecipitation treatments significantly altered responses inaboveground biomass measurements with statisticallysignificant responses in the physiological variables largelyreflecting changes across the sampling periods (Fig 3)Flowering tiller heights were greatest in the increasedprecipitation treatments reaching mean heights of 129mThese heights are reflective of those observed in prairieremnant populations although they are lower than those foundin P virgatum cultivars (Das et al 2004 Casler 2005Alexopoulou et al 2008) but it should be noted that plantheights are variable from year to year (Alexopoulou et al2008) The average number of leaves per flowering tiller waslower than those found in agronomic P virgatum cultivars (VanEsbroeck et al 1997) however these cultivars have been bred tooptimise yield These data illustrate individual biomass responsesto precipitation of a dominant species in the tallgrass prairieecosystem further supporting the conclusions of others thatfuture changes in precipitation are likely to significantly alterecosystem productivity (Knapp 1984 Fay et al 2003 Fay et al2008 Zhou et al 2009)

Previous attempts to link leaf-level to whole-plant responsesfor grassland species have shown positive relationships betweenphotosynthesis and plant success (McAllister et al 1998) butleaf-level to whole-plant relationships in grasslands arecommonly weak (Nippert et al 2007) In this study no clearrelationships between leaf-level photosynthetic responses andtotal aboveground biomass were present for these ecotypesAlthough the increases in leaf-level photosynthesis did notcorrespond to total biomass increases other measurements ofbiomass increased from the increased precipitation partiallysupporting the positive relationship between higher rates ofgas exchange and plant success previously noted byMcAllister et al (1998)

The only physiological responses affected by precipitationtreatments were Ymid and WUE with WUE being lowest inthe increased treatments andYmid highest in increased treatmentsAlthough ANOVA results showed significant precipitationsample period interactions (Table 2) many of the effects wereonly seen in the early season period (Fig 3) Later in the seasonincreasing summer temperatures negated the positive impacts ofthe added water Decreased WUE across the growing seasoncorresponding to higher air temperatures has been previouslynoted for C4 species where WUE can be reduced by 40 indrought conditions (Taylor et al 2011) As water availabilitydecreasedYmid decreased to amean of ndash22MPa The decreasedprecipitation treatment did not lower soil water availability tothe point where the critical water potential (leaf pressurepotential when stomatal closure is less than 5 of themaximum) was reached for switchgrass (Ycrit = ndash32MPaTucker et al 2011) Indeed previous studies have shown that

switchgrass can maintain growth and functioning at lower waterpotentials than those recorded here (Knapp 1984 Stroup et al2003) With more severe decreases in water availability YmidAmax gs E and WUE should continue to decline with thepotential for increased cuticular conductance when drought-stressed (Manzoni et al 2011)

Ecotype precipitation interactions

Surprisingly few ecotype precipitation interactions werepresent for the variablesmeasured in this study (Tables 2 3) withthe primary exception being greater biomass per tiller in theincreased treatment for Texas ecotypes These differences largelyreflect the southern Texas ecotype being adapted to a longergrowing season (Newell 1968) Growth of the Texas ecotype in alocation with a shorter growing season delayed reproductivematurity and although the overall biomass yield was notsignificantly greater for Texas ecotypes compared withOklahoma and Kansas under greater precipitation differencesthe trends observed in the current study may become statisticallysignificant Because Texas ecotypes had a longer period in whichthey grew vegetatively theywere better able to take advantage ofthe increased precipitation treatment to produce higherbiomass per tiller greater biomass per flowering tiller and highnon-flowering tiller counts in all precipitation treatments Theseresults from a southern Texas ecotype growing in amore northernlocation are similar to other studies that show increases inbiomass longer photosynthetic activity and delayed floweringfor southern ecotypes grown in more northern latitudes (Quinn1969 Lemus et al 2002 Casler et al 2004 Berdahl et al 2005Casler et al 2007)

Conclusion

The physiological responses of Amax gs E FvFm and Ymid inP virgatum were explained by ecotype differences howeverrobust responses to altered precipitationwere seen inWUEYmidflowering times and many aboveground biomass variablesEcotype differences were also seen in several abovegroundbiomass variables and most strikingly in flowering initiationand rates In this study there were few interactions betweenecotype and precipitation amount suggesting that althoughprecipitation is a strong driver of biomass production localadaptations in the ecotypes used in the current study have lessimpact on these responses than initially expected It was expectedthat the environmental differences among the origins of theecotypes mainly the effects of higher temperature coupledwith similar mean annual precipitation (860ndash878mm) wouldlead to significant ecotype precipitation interactions in theparameters measured with the Texas ecotypes being the mosttolerant of the changes The lack of interactions between geneticbackground and precipitation also has broader evolutionaryimplications gene environment interactions are a primaryindicator of the potential for natural selection in a given trait(Schlichting 1989) The broader implications of the fewecotype precipitation interactions for this widely-distributedspecies suggest that it is not likely toundergo selection in responseto changes in precipitation As many have correctly suggested itis necessary to examine the role of genetic variation in theresponse of plant species to future changes in climate

134 Functional Plant Biology J C Hartman et al

(Callaway et al 2003 Ward and Kelly 2004) however asindicated by this study it may not always be necessary toconsider localised adaptation when working to project theimpacts of climate change on the performance of a species likeswitchgrass across its geographical distribution

Acknowledgements

We thankDr JohnBlair andDr TedMorgan for their advice and comments onthemanuscriptWealso thank theNatureConservancy for allowingcollectionof switchgrass rhizomes from the Tallgrass Prairie Preserve inOklahoma andlandholdingsnear Temple TXGracieOrozcoRebeccaOrozco JacobCarterand Teall Culbertson provided assistance with fieldwork and the Division ofBiology at Kansas State University Konza Prairie Biological Station KonzaPrairie LTER (DEB-0823341) and the US Department of Energy(EEE000314) provided funding and access to facilities

References

Albert KR Mikkelsen TN Michelsen A Ro-Poulsen H van der Linden L(2011) Interactive effects of drought elevated CO2 and warming onphotosynthetic capacity and photosystem performance in temperateheath plants Journal of Plant Physiology 168 1550ndash1561 doi101016jjplph201102011

Alexopoulou E Sharma N Papatheohari Y Christou M Piscioneri IPanoutsou D Pignatelli V (2008) Biomass yields for upland andlowland switchgrass varieties grown in the Mediterranean regionBiomass and Bioenergy 32 926ndash933 doi101016jbiombioe200801015

AlleyRBBerntsen T Bindoff NL ChenZ ChidthaisongA et al (2007) Hephysical science basis summary for policy makersrsquo (IPCC SecretariatGeneva)

Barney JN Mann JJ Kyser GB Blumwald B Deynze AV DiTomaso JM(2009)Toleranceof switchgrass to extremesoilmoisture stress ecologicalimplications Plant Science 177 724ndash732 doi101016jplantsci200909003

Benedict HM (1940) Effect of day length and temperature on the floweringand growth of four species of grasses Journal of Agricultural Research61 661ndash671

Berdahl JD Frank AB Krupinsky JM Carr PM Hanson JD Johnson HA(2005) Biomass yield phenology and survival of diverse switchgrasscultivars and experimental strains in western North Dakota AgronomyJournal 97 549ndash555 doi102134agronj20050549

Berg CC (1971) Forage yield of switchgrass (Panicum virgatum) inPennsylvania Agronomy Journal 63 785ndash786 doi102134agronj197100021962006300050038x

Boe A Beck DL (2008) Yield components of biomass in switchgrass CropScience 48 1306ndash1311 doi102135cropsci2007080482

Callaway RM Pennings SC Richards CL (2003) Phenotypic plasticity andinteractions among plants Ecology 84 1115ndash1128 doi1018900012-9658(2003)084[1115PPAIAP]20CO2

Casler MD (2005) Ecotypic variation among switchgrass populations fromthe northern USA Crop Science 45 388ndash398 doi102135cropsci20050388

Casler MD Vogel KP Taliaferro CM Wynia RL (2004) Latitudinaladaptation of switchgrass populations Crop Science 44 293ndash303

Casler MD Vogel KP Taliaferro CM Ehlke NJ Berdahl JD Brummer ECKallenbach RL West CP Mitchell RB (2007) Latitudinal andlongitudinal adaptation of switchgrass populations Crop Science 472249ndash2260 doi102135cropsci2006120780

Christensen JH Hewitson B Busuioc A Chen A Gao X Held I Jones RKolli RK KwonWT Laprise R Magantildea Rueda VMearns L MeneacutendezCG Raumlisaumlnen J Rinke A Sarr A (2007) Regional climate projectionsIn lsquoClimate change 2007 the physical science basis Contributionof Working Group I to the Fourth Assessment Report of the

Intergovernmental Panel on Climate Change (eds S Solomon D QinM Manning Z Chen M Marquis KB Averyt M Tignor HL Miller)pp 887ndash892 (Cambridge University Press Cambridge UK)

Das MK Fuentes RG Taliaferro CM (2004) Genetic variability and traitrelationships in switchgrass Crop Science 44 443ndash448

Easterling DR Meehl GA Parmesan C Changnon SA Karl TR Mearns LO(2000) Climate extremes observations modeling and impacts Science289 2068ndash2074 doi101126science28954872068

Emery SM Gross KL (2007) Dominant species identity not communityevenness regulates invasion in experimental grassland communitiesEcology 88 954ndash964 doi10189006-0568

Evers GW Parsons MJ (2003) Soil type and moisture level influence onAlamo switchgrass emergence and seedling growth Crop Science 43288ndash294 doi102135cropsci20030288

Fay PA Carlisle JD Knapp AK Blair JM Collins SL (2003) Productivityresponses to altered rainfall patterns in a C4-dominated grasslandOecologia 137 245ndash251 doi101007s00442-003-1331-3

FayPAKaufmanDMNippert JBCarlisle JDHarperCW(2008)Changes ingrassland ecosystem function due to extreme rainfall events implicationsfor responses to climate changeGlobal Change Biology 14 1600ndash1608doi101111j1365-2486200801605x

Fay PA Blair JM Smith MD Nippert JB Carlisle JD Knapp AK (2011)Relative effects of precipitation variability and warming on tallgrassprairie ecosystem function Biogeosciences 8 3053ndash3068 doi105194bg-8-3053-2011

Grime JP (1998) Benefits of plant diversity to ecosystems immediatefilter and founder effects Journal of Ecology 86 902ndash910doi101046j1365-2745199800306x

Hartman JC Nippert JB OrozcoRA Springer CJ (2011) Potential ecologicalimpacts of switchgrass (Panicum virgatum L) biofuel cultivation in theCentral Great Plains USA Biomass and Bioenergy 35 3415ndash3421doi101016jbiombioe201104055

Heaton E Voigt T Long SP (2004) A quantitative review comparing theyields of two candidate C4 perennial biomass crops in relation to nitrogentemperature and water Biomass and Bioenergy 27 21ndash30 doi101016jbiombioe200310005

HillebrandHBennettDMCadotteMW(2008)Consequences of dominancea review of evenness effects on local and regional ecosystem processesEcology 89 1510ndash1520 doi10189007-10531

Hughes AR Inouye BD Johnson MTJ Underwood N Vellend M (2008)Ecological consequences of genetic diversity Ecology Letters 11609ndash623 doi101111j1461-0248200801179x

Jump AS Pentildeuelas J (2005) Running to stand still adaptation and responseof plants to rapid climate change Ecology Letters 8 1010ndash1020doi101111j1461-0248200500796x

KnappAK (1984)Water relations and growth of three grasses duringwet anddrought years in a tallgrass prairie Oecologia 65 35ndash43 doi101007BF00384460

Knapp AK (1985) Effect of fire and drought on the ecophysiology ofAndropogon gerardii and Panicum virgatum in a tallgrass prairieEcology 66 1309ndash1320 doi1023071939184

KnappAKBriggs JMHartnett DCCollins SC (1998) lsquoGrassland dynamicslong-term ecological research in tallgrass prairiersquo (Oxford UniversityPress New York)

KnappAKFayPABlair JMCollinsSLSmithMDCarlisle JDHarperCWDanner BT Lett MS McCarron JK (2002) Rainfall variability carboncycling and plant species diversity in mesic grassland Science 2982202ndash2205 doi101126science1076347

Krause GH Weis E (1991) Chlorophyll fluorescence and photosynthesisthe basics Annual Review of Plant Physiology and Plant MolecularBiology 42 313ndash349 doi101146annurevpp42060191001525

Lemus R Brummer EC Moore KJ Molstad ME Burras CE Barker MF(2002) Biomass yield and quality of 20 switchgrass populations insouthern Iowa USA Biomass and Bioenergy 23 433ndash442doi101016S0961-9534(02)00073-9

Responses of switchgrass to altered precipitation Functional Plant Biology 135

Manzoni S Vico G Katul G Fay PA Polley W Palmroth S Porporato A(2011) Optimizing stomatal conductance for maximum carbon gainunder water stress a meta-analysis across plant functional types andclimates Functional Ecology 25 456ndash467 doi101111j1365-2435201001822x

Maxwell K JohnsonGN (2000) Chlorophyll fluorescence ndash a practical guideJournal of Experimental Botany 51 659ndash668 doi101093jexbot51345659

McAllister CA Knapp AK Maragni LA (1998) Is leaf-level photosynthesisrelated to plant success in a highly productive grasslandOecologia 11740ndash46 doi101007s004420050629

McLaughlin SB Kszos LN (2005) Development of switchgrass (Panicumvirgatum) as a bioenergy feedstock in the United States Biomass andBioenergy 28 515ndash535 doi101016jbiombioe200405006

McMillan C (1965) Ecotypic differentiation within four NorthAmerican prairie grasses II Behavioral variation within transplantedcommunity fractions American Journal of Botany 52 55ndash65doi1023072439975

McNaughton SJ Wolf LL (1970) Dominance and the niche in ecologicalsystems Science 167 131ndash139 doi101126science1673915131

Meehl GA Arblaster JM Tebaldi C (2005) Understanding future patterns ofincreased precipitation intensity in climate model simulationsGeophysical Research Letters 32 L18719 doi1010292005GL023680

Muir JP Sanderson MA Ocumpaugh WR Jones RM Reed RL (2001)Biomass production of lsquoAlamorsquo switchgrass in response to nitrogenphosphorus and row spacing Agronomy Journal 93 896ndash901doi102134agronj2001934896x

Newell LC (1968) Effects of strain source and management practice onforage yields of two warm-season prairie grasses Crop Science 8205ndash210 doi102135cropsci19680011183X000800020022x

Nippert JBKnappAKBriggs JM (2006) Intra-annual rainfall variability andgrassland productivity can the past predict the futurePlant Ecology 18465ndash74 doi101007s11258-005-9052-9

Nippert JB Fay PA Knapp AK (2007) Photosynthetic traits in C3 and C4

grassland species in mesocosm and field environments Environmentaland Experimental Botany 60 412ndash420 doi101016jenvexpbot200612012

Nippert JB Fay PA Carlisle JD Knapp AK Smith MD (2009)Ecophysiological responses of two dominant grasses to alteredtemperature and precipitation regimes Acta Oecologica 35 400ndash408doi101016jactao200901010

Norberg J Swaney DP Dushoff J Lin J Casagrandi R Levin SA (2001)Phenotypic diversity and ecosystem functioning in changingenvironments a theoretical framework Proceedings of the NationalAcademy of Sciences of the United States of America 9811376ndash11381 doi101073pnas171315998

Parrish DJ Fike JH (2005) The biology and agronomy of switchgrass forbiofuels Critical Reviews in Plant Sciences 24 423ndash459 doi10108007352680500316433

Polley HW Norman JM Arkebauer TJ Walter-Shea EA Greegor DH JrBramer B (1992) Leaf gas exchange of Andropogon gerardii VitmanPanicum virgatum L and Sorghastrum nutans (L) Nash in a tallgrassprairie Journal of Geophysical Research 97 18 837ndash18 844

Quinn JA (1969) Variability among High Plains populations of Panicumvirgatum Bulletin of the Torrey Botanical Club 96 20ndash41 doi1023072484005

Resco V IgnanceDD SunWHuxman TEWeltzin JFWilliams DG (2008)Chlorophyll fluorescence predawn water potential and photosynthesisin pulse-drive ecosystems ndash implications for ecological studiesFunctional Ecology 22 479ndash483 doi101111j1365-2435200801396x

Sanderson MA (1992) Morphological development of switchgrass andkleingrass Agronomy Journal 84 415ndash419 doi102134agronj199200021962008400030013x

Sanderson MA Reed RL (2000) Switchgrass growth and developmentwater nitrogen and plant density effects Journal of RangeManagement 53 221ndash227 doi1023074003287

SandersonMAWolf DD (1995)Morphological development of switchgrassin diverse environments Agronomy Journal 87 908ndash915 doi102134agronj199500021962008700050022x

Schlichting CD (1989) Phenotypic integration and environmental changeBioscience 39 460ndash464 doi1023071311138

SillettiAMKnappAK(2001)Responsesof the codominantgrassland speciesAndropogon gerardii and Sorghastrum nutans to long-termmanipulations of nitrogen and water American Midland Naturalist145 159ndash167 doi1016740003-0031(2001)145[0159ROTCGS]20CO2

SmithMDKnappAK (2003)Dominant speciesmaintain ecosystem functionwith non-random species loss Ecology Letters 6 509ndash517 doi101046j1461-0248200300454x

Sokal RRRohlf FJ (1995) lsquoBiometry the principles and practices of statisticsin biological research 3rd ednrsquo (WH Freeman New York)

StoutWL (1992)Water-use efficiency of grasses as affected by soil nitrogenand temperature Soil Science Society of America Journal 56 897ndash902doi102136sssaj199203615995005600030036x

Stout WL Jung GA Shaffer JA (1988) Effects of soil and nitrogen on wateruse efficiency of tall fescue and switchgrass under humid conditions SoilScience Society of America Journal 52 429ndash434 doi102136sssaj198803615995005200020023x

Stroup JA Sanderson MA Muir JP McFarland MJ Reed RL (2003)Comparison of growth and performance in upland and lowlandswitchgrass types to water and nitrogen stress Bioresource Technology86 65ndash72 doi101016S0960-8524(02)00102-5

Taylor SH Ripley BSWoodward FI Osborne CP (2011) Drought limitationofphotosynthesis differs betweenC3andC4grass species in a comparativeexperiment Plant Cell amp Environment 34 65ndash75 doi101111j1365-3040201002226x

Tucker SS Craine JM Nippert JB (2011) Physiological drought toleranceand the structuring of tallgrass prairie assemblages Ecosphere 2 1ndash19doi101890ES11-000231

Van Esbroeck GA Hussey MA Sanderson MA (1997) Leaf appearancerate and final leaf number of switchgrass cultivars Crop Science 37864ndash870 doi102135cropsci19970011183X003700030028x

Van Esbroeck GA Hussey MA Sanderson MA (2003) Variation betweenAlamo and Cave-in- Rock switchgrass in response to photoperiodextension Crop Science 43 639ndash643

WangDLeBauerDSDietzeMC(2010)Aquantitative reviewcomparing theyield of switchgrass in monocultures and mixtures in relation to climateand management factors GCB Bioenergy 2 16ndash25 doi101111j1757-1707201001035x

Ward JK Kelly JK (2004) Scaling up evolutionary responses to CO2 lessonsfrom Arabidopsis Ecology Letters 7 427ndash440 doi101111j1461-0248200400589x

Wullschleger SD SandersonMAMcLaughlin SB Biradar DP Rayburn AL(1996) Photosynthetic rates and ploidy levels among populations ofswitchgrassCrop Science 36 306ndash312 doi102135cropsci19960011183X003600020016x

Zhou X Talley M Luo Y (2009) Biomass litter and soil respiration along aprecipitation gradient in southern Great Plains USA Ecosystems(New York NY) 12 1369ndash1380 doi101007s10021-009-9296-7

136 Functional Plant Biology J C Hartman et al

wwwpublishcsiroaujournalsfpb

Page 6: Ecotypic responses of switchgrass to altered precipitation

flowering tiller and the number of non-flowering tillers showedinteractions among ecotypes and precipitation treatments(Table 5) Texas ecotypes had significantly greater biomassper tiller in the increased treatments with the same results forbiomass per flowering tiller Kansas and Texas ecotypesdecreased the amount of non-flowering tillers as the soilmoisture increased Oklahoma ecotypes displayed the oppositetrend Oklahoma ecotypes had significantly fewer non-flowering

tillers in the decreased treatment compared with the other twoecotypes (Table 5)

Flowering

Flowering phenology was affected by both ecotype andprecipitation individually (Table 3) Texas ecotypes had halfthe number of flowering tillers at the end of season comparedwith Kansas and Oklahoma ecotypes (Table 4 Fig 4) Texasecotypes also initiated flowering a month later than the otherecotypes and flowered at a slower rate (Fig 4) Floweringphenology varied significantly with precipitation treatmentsIndividuals in the increased treatment had 40 and 38 moreflowering tillers at the end of the season compared with theaverage and decreased treatments respectively (Table 4Fig 4) Precipitation treatments affected flowering rates withthe increased treatment flowering at a greater rate compared withthe average and decreased treatment (Fig 4)

Discussion

The primary objective of this study was to characterise theecotype-specific physiological responses and growth ofP virgatum to changes in forecasted precipitation amount forthe Central Plains region of the United States by the year 2100Our results show that the precipitation treatments significantlyaffected soil moisture and the corresponding responses ofplant growth and physiology Significant effects of ecotype

gs (mol H2O mndash2 sndash1)

00 01 02 03 04 05

Fv

Fm

072

074

076

078

080

082

KansasOklahoma Texas

y = y0 + aln(abs(x))R-adj = 02442

Fig 2 Representative FvFmndashgs curves for all ecotypes Curves were fitfor all data points for each ecotype The R-adjusted value is for the overall fitof all ecotype data points

Early Mid Late Early Mid Late

Am

ax (

μmol

mndash2

sndash1

)E

(m

mol

mndash2

sndash1

)

0

10

20

30

40DecreasedAverageIncreased

000

005

010

015

020

025

Summer sampling period

0

2

4

6

8

10

Summer sampling period

WU

E (

μmol

mm

olndash1

)

0

1

2

3

4

5

6

a ab

a a

b

a a a

a a

b

a abb

a aa

a abb

aab

b

aab b

a a aa a

ba

aa

(a)

Early Mid LateEarly Mid Late

g s (

mol

H2O

mndash2

sndash1

)

(b)

(c) (d )

Fig 3 Gas exchange responses by treatment during the earlymid-season and late periods of the growing season (a)Carbon assimilationat ambient CO2 (b) stomatal conductance (c) transpiration (d) water use efficiency (Amax E

ndash1) Bars are means (1 se) and significantdifferences are indicated by letters within each sampling period

Responses of switchgrass to altered precipitation Functional Plant Biology 131

and precipitation were present but there were fewecotype precipitation interactions In general physiologicaldifferences reflected ecotype origin whereas differences inaboveground biomass largely reflected differences in soilwater availability from the precipitation treatments

Ecotype effects

Differences among means in the physiological variablesmeasured largely reflected differences by ecotype andcorresponding adaptation to the local environment of originWithin each ecotype similar changes in Amax gs and E reflectthe coupled gas exchange relationships documented previouslyfor other dominant prairie grasses (Polley et al 1992) It hasalso been shown that carbon assimilation in switchgrass iscontrolled mostly by population responses to localenvironmental conditions (Wullschleger et al 1996) In this

study the Kansas ecotypes had the lowest carbon assimilationrates providing support for the local environmental conditionresponse (Wullschleger et al 1996)

The FvFm ndash gs relationship developed (Fig 2) shows astatistical relationship between responses of leafphotochemistry to changes in stomatal conductance that variesby ecotypeReducedFvFmvalues indicate a decreased efficiencyof non-photochemical quenching and are reflective of thefunctioning of the light reactions of photosynthesis (Maxwelland Johnson 2000) Changes in FvFm can be especially useful instudies where physiological performance is strongly decreasedfrom longperiods of drought (Resco et al 2008) These data showthat nearly one-quarter of the variation themaximumefficiencyofPSII can be statistically attributed to changes in stomatalconductance (Fig 2) From first principles one would assumethat a reduction in stomatal water loss would result in increased

Table 3 Mixed-effects model ANOVA of precipitation treatments and ecotype (fixed precipitation ecotype random cell) effects on abovegroundbiomass variables

F- and P-values are provided with significant responses P lt 005 Plt 010

Response Precipitation Ecotype Ecotype precipitationF P F P F P

Tillers (per individual) 371 00326 338 00429 11 035Biomass (g per individual) 948 00004 21 012 15 021Biomass per tiller (g per individual) 449 00168 1109 00001 431 00050 allocated to reproduction (per individual) 483 00127 599 00050 191 012Flowering tillers (per individual) 754 00015 848 00008 05 068Flowering tiller biomass (g per individual) 1146 lt00001 08 044 15 019Biomass per flowering tiller (g per individual) 286 00681 1333 lt00001 351 00142Flowering tiller height (cm per individual) 512 00100 18 016 16 017Leaves per flowering tiller (per individual) 547 00075 22 011 11 035Non-flowering tillers (per individual) 325 00481 291 00651 287 00337Non-flowering tiller biomass (g per individual) 11 032 484 00126 05 067Biomass per non-flowering tiller (g per individual) 007 093 1585 lt00001 17 015Non-flowering tiller height (cm per individual) 09 038 305 00576 11 036Leaves per non-flowering tiller (per individual) 02 080 670 00029 18 013

Table 4 Mean responses (1 se) of various aboveground biomass measurements for ecotype and precipitation treatment effectsWithin the ecotype or treatment category bolded numbers show significance at the P 005 level and italicised numbers show marginal significance at theP 010 level Superscripts display significant differences between precipitation treatments or ecotypes within a category NS signifies no significance between

the ecotypes or the precipitation treatments

Variables Precipitation EcotypeDecreased Average Increased Kansas Oklahoma Texas

Tillers (per individual) 56 plusmn 5ab 47 plusmn 6b 71 plusmn 6a 69 plusmn 5a 58 plusmn 6ab 48 plusmn 6b

Biomass (g per individual) 202plusmn 28b 181 plusmn 30b 352 plusmn 30a NS NS NSBiomass per tiller (g per individual) 36 plusmn 03b 40 plusmn 03ab 50 plusmn 03a 30 plusmn 03b 44 plusmn 03a 52 plusmn 03a

Flowering tillers (per individual) 35 plusmn 4b 36 plusmn 4b 58 plusmn 4a 53 plusmn 4a 48 plusmn 5a 27 plusmn 5b

Flowering tiller biomass (g per individual) 157plusmn 27b 157 plusmn 28b 325 plusmn 28a NS NS NSBiomass per flowering tiller (g per individual) 41plusmn 04b 49plusmn 07ab 56plusmn 04a 32 plusmn 04b 50 plusmn 04a 64 plusmn 04a

Flowering tiller height (cm per individual) 94 plusmn 7b 101 plusmn 8b 129 plusmn 8a NS NS NSLeaves per flowering tiller (per individual) 47 plusmn 01b 51 plusmn 02ab 56 plusmn 02a NS NS NSNon-flowering tillers (per individual) 21 plusmn 2a 11 plusmn 3b 12 plusmn 3ab 15plusmn 2ab 9plusmn 3b 20plusmn 3a

Non-flowering tiller biomass (g per individual) NS NS NS 20 plusmn 9b 14 plusmn 11b 59 plusmn 11a

Biomass per non-flowering tiller (g per individual) NS NS NS 07 plusmn 01b 08 plusmn 01b 20 plusmn 01a

Non-flowering tiller height (cm per individual) NS NS NS 27plusmn 3ab 24plusmn 4b 38plusmn 4a

Leaves per non-flowering tiller (per individual) NS NS NS 32 plusmn 02b 30 plusmn 03b 44 plusmn 03a

allocated to reproduction (per individual) 77 plusmn 3b 85 plusmn 3ab 94 plusmn 3a 87 plusmn 3a 94 plusmn 4a 75 plusmn 4b

132 Functional Plant Biology J C Hartman et al

leaf temperature with the potential for corresponding decreasedphotochemistry during hot or dry summer conditions For allecotypesFvFm decreased nonlinearly with stomatal closure buttherewere clear differences in this relationship between ecotypes

with a general trend of increasing FvFm values for a given rate ofstomatal conductance from Kansas to Oklahoma to Texasecotypes Texas ecotypes had higher maximum FvFm valuescompared with Kansas ecotypes with values near the acceptedmaximal value of 083 (Krause and Weis 1991) As gs valuesdecreased Texas and Oklahoma ecotypes were able to maintainhigher FvFm values compared with the Kansas ecotypesindicating a greater decrease in the photochemical efficiencyof PSII in Kansas ecotypes Interpretation of the relationshipbetween FvFm and gs requires further investigation but thesedata show promise for linking changes in stomatal conductancewith themaintenance of leaf photochemistry as the ability to coolthe leaf through latent heat exchange becomes very limited withlow gs (Nippert et al 2009) and photochemistry is subsequentlyreducedby increased leaf temperatures (Albert et al 2011)Basedon theseFvFmndashgs response curves (Fig 2) Oklahoma and Texasecotypes are potentially more tolerant of higher leaf temperaturewith higher FvFm at low rates of gas exchange compared withKansas ecotypes

P virgatum is a highly photoperiod sensitive species(Benedict 1940) and differences in flowering dates have beendocumented for various ecotype and cultivars (McMillan 1965VanEsbroeck et al 2003Casler et al 2004) Sanderson andWolf(1995) recorded similar flowering dates for Alamo and Cave-in-Rock cultivars compared with the ecotypes in this study Alamocultivars flowered around the end of September and Cave-in-Rock cultivars flowered around the first week of July Theflowering time and rates seen in this study correspond withknown ecotype responses to changes in latitude

Many of the aboveground biomass responses with significantecotypic differences were a result of the Texas ecotypesproducing fewer tillers but with an increased biomass pertiller and number of leaves per tiller while the Oklahoma andKansas ecotypes had higher tiller density (Table 4) Typically thebest estimates to predict changes in biomass yield for this speciesare increased biomass per tiller and number of leaves per tiller(seen in Texas ecotypes) or higher tiller density (noted in the

Table 5 Mean responses (1 se) of various aboveground biomass measurements for interactions between ecotypes and precipitation treatmentsBolded numbers show significance at theP 005 level Superscripts display significant differences between ecotype precipitation combinations NS signifies

no significant difference between the means

Variables Ecotype precipitationDecreased Average Increased

Kansas Oklahoma Texas Kansas Oklahoma Texas Kansas Oklahoma Texas

Tillers (per individual) NS NS NS NS NS NS NS NS NSBiomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per tiller (g per individual) 28 plusmn 05b 42 plusmn 05b 38 plusmn 05b 29 plusmn 05b 50 plusmn 06ab 42 plusmn 06b 34 plusmn 05b 39 plusmn 06b 76 plusmn 06a

Flowering tillers (per individual) NS NS NS NS NS NS NS NS NSFlowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per flowering tiller (g per individual) 28 plusmn 07bc 44 plusmn 07bc 51 plusmn 07bc 31 plusmn 07bc 65 plusmn 08ab 52 plusmn 08abc 36 plusmn 07bc 42 plusmn 08bc 90 plusmn 08a

Flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per flowering tiller (per individual) NS NS NS NS NS NS NS NS NSNon-flowering tillers (per individual) 31 plusmn 4a 5 plusmn 5b 28 plusmn 5ab 8 plusmn 4b 7 plusmn 5ab 18 plusmn 5ab 7 plusmn 4b 15 plusmn 5ab 15 plusmn 5ab

Non-flowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per non-flowering tiller (g per individual) NS NS NS NS NS NS NS NS NSNon-flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per non-flowering tiller (per individual) NS NS NS NS NS NS NS NS NS allocated to reproduction (per individual) NS NS NS NS NS NS NS NS NS

Nov

Till

ers

flow

ered

(in

divndash1

)T

iller

s flo

wer

ed (

indi

vndash1)

0

10

20

30

40

50

60IncreasedAverageDecreased

0

10

20

30

40

50

60 Kansas Oklahoma Texas

(a)

(b)

Jul Aug Sep Oct

Fig 4 Total tillers flowered per individual over the course of the growingseason (a) Flowering tillers by ecotype (b) flowering tillers by precipitationtreatment Data points are means 1se Asterisks indicate statisticalsignificance (Plt 005) between groups within each sampling date

Responses of switchgrass to altered precipitation Functional Plant Biology 133

Oklahoma and Kansas ecotypes Boe and Beck 2008) Althoughtotal biomass was not significantly different among ecotypes theTexas ecotypes contain components (higher biomass and leavesper tiller) that if heritablemay lead to greater biomass inmultiplegeneration studies

Precipitation effects

The altered environmental conditions resulting from theprecipitation treatments significantly altered responses inaboveground biomass measurements with statisticallysignificant responses in the physiological variables largelyreflecting changes across the sampling periods (Fig 3)Flowering tiller heights were greatest in the increasedprecipitation treatments reaching mean heights of 129mThese heights are reflective of those observed in prairieremnant populations although they are lower than those foundin P virgatum cultivars (Das et al 2004 Casler 2005Alexopoulou et al 2008) but it should be noted that plantheights are variable from year to year (Alexopoulou et al2008) The average number of leaves per flowering tiller waslower than those found in agronomic P virgatum cultivars (VanEsbroeck et al 1997) however these cultivars have been bred tooptimise yield These data illustrate individual biomass responsesto precipitation of a dominant species in the tallgrass prairieecosystem further supporting the conclusions of others thatfuture changes in precipitation are likely to significantly alterecosystem productivity (Knapp 1984 Fay et al 2003 Fay et al2008 Zhou et al 2009)

Previous attempts to link leaf-level to whole-plant responsesfor grassland species have shown positive relationships betweenphotosynthesis and plant success (McAllister et al 1998) butleaf-level to whole-plant relationships in grasslands arecommonly weak (Nippert et al 2007) In this study no clearrelationships between leaf-level photosynthetic responses andtotal aboveground biomass were present for these ecotypesAlthough the increases in leaf-level photosynthesis did notcorrespond to total biomass increases other measurements ofbiomass increased from the increased precipitation partiallysupporting the positive relationship between higher rates ofgas exchange and plant success previously noted byMcAllister et al (1998)

The only physiological responses affected by precipitationtreatments were Ymid and WUE with WUE being lowest inthe increased treatments andYmid highest in increased treatmentsAlthough ANOVA results showed significant precipitationsample period interactions (Table 2) many of the effects wereonly seen in the early season period (Fig 3) Later in the seasonincreasing summer temperatures negated the positive impacts ofthe added water Decreased WUE across the growing seasoncorresponding to higher air temperatures has been previouslynoted for C4 species where WUE can be reduced by 40 indrought conditions (Taylor et al 2011) As water availabilitydecreasedYmid decreased to amean of ndash22MPa The decreasedprecipitation treatment did not lower soil water availability tothe point where the critical water potential (leaf pressurepotential when stomatal closure is less than 5 of themaximum) was reached for switchgrass (Ycrit = ndash32MPaTucker et al 2011) Indeed previous studies have shown that

switchgrass can maintain growth and functioning at lower waterpotentials than those recorded here (Knapp 1984 Stroup et al2003) With more severe decreases in water availability YmidAmax gs E and WUE should continue to decline with thepotential for increased cuticular conductance when drought-stressed (Manzoni et al 2011)

Ecotype precipitation interactions

Surprisingly few ecotype precipitation interactions werepresent for the variablesmeasured in this study (Tables 2 3) withthe primary exception being greater biomass per tiller in theincreased treatment for Texas ecotypes These differences largelyreflect the southern Texas ecotype being adapted to a longergrowing season (Newell 1968) Growth of the Texas ecotype in alocation with a shorter growing season delayed reproductivematurity and although the overall biomass yield was notsignificantly greater for Texas ecotypes compared withOklahoma and Kansas under greater precipitation differencesthe trends observed in the current study may become statisticallysignificant Because Texas ecotypes had a longer period in whichthey grew vegetatively theywere better able to take advantage ofthe increased precipitation treatment to produce higherbiomass per tiller greater biomass per flowering tiller and highnon-flowering tiller counts in all precipitation treatments Theseresults from a southern Texas ecotype growing in amore northernlocation are similar to other studies that show increases inbiomass longer photosynthetic activity and delayed floweringfor southern ecotypes grown in more northern latitudes (Quinn1969 Lemus et al 2002 Casler et al 2004 Berdahl et al 2005Casler et al 2007)

Conclusion

The physiological responses of Amax gs E FvFm and Ymid inP virgatum were explained by ecotype differences howeverrobust responses to altered precipitationwere seen inWUEYmidflowering times and many aboveground biomass variablesEcotype differences were also seen in several abovegroundbiomass variables and most strikingly in flowering initiationand rates In this study there were few interactions betweenecotype and precipitation amount suggesting that althoughprecipitation is a strong driver of biomass production localadaptations in the ecotypes used in the current study have lessimpact on these responses than initially expected It was expectedthat the environmental differences among the origins of theecotypes mainly the effects of higher temperature coupledwith similar mean annual precipitation (860ndash878mm) wouldlead to significant ecotype precipitation interactions in theparameters measured with the Texas ecotypes being the mosttolerant of the changes The lack of interactions between geneticbackground and precipitation also has broader evolutionaryimplications gene environment interactions are a primaryindicator of the potential for natural selection in a given trait(Schlichting 1989) The broader implications of the fewecotype precipitation interactions for this widely-distributedspecies suggest that it is not likely toundergo selection in responseto changes in precipitation As many have correctly suggested itis necessary to examine the role of genetic variation in theresponse of plant species to future changes in climate

134 Functional Plant Biology J C Hartman et al

(Callaway et al 2003 Ward and Kelly 2004) however asindicated by this study it may not always be necessary toconsider localised adaptation when working to project theimpacts of climate change on the performance of a species likeswitchgrass across its geographical distribution

Acknowledgements

We thankDr JohnBlair andDr TedMorgan for their advice and comments onthemanuscriptWealso thank theNatureConservancy for allowingcollectionof switchgrass rhizomes from the Tallgrass Prairie Preserve inOklahoma andlandholdingsnear Temple TXGracieOrozcoRebeccaOrozco JacobCarterand Teall Culbertson provided assistance with fieldwork and the Division ofBiology at Kansas State University Konza Prairie Biological Station KonzaPrairie LTER (DEB-0823341) and the US Department of Energy(EEE000314) provided funding and access to facilities

References

Albert KR Mikkelsen TN Michelsen A Ro-Poulsen H van der Linden L(2011) Interactive effects of drought elevated CO2 and warming onphotosynthetic capacity and photosystem performance in temperateheath plants Journal of Plant Physiology 168 1550ndash1561 doi101016jjplph201102011

Alexopoulou E Sharma N Papatheohari Y Christou M Piscioneri IPanoutsou D Pignatelli V (2008) Biomass yields for upland andlowland switchgrass varieties grown in the Mediterranean regionBiomass and Bioenergy 32 926ndash933 doi101016jbiombioe200801015

AlleyRBBerntsen T Bindoff NL ChenZ ChidthaisongA et al (2007) Hephysical science basis summary for policy makersrsquo (IPCC SecretariatGeneva)

Barney JN Mann JJ Kyser GB Blumwald B Deynze AV DiTomaso JM(2009)Toleranceof switchgrass to extremesoilmoisture stress ecologicalimplications Plant Science 177 724ndash732 doi101016jplantsci200909003

Benedict HM (1940) Effect of day length and temperature on the floweringand growth of four species of grasses Journal of Agricultural Research61 661ndash671

Berdahl JD Frank AB Krupinsky JM Carr PM Hanson JD Johnson HA(2005) Biomass yield phenology and survival of diverse switchgrasscultivars and experimental strains in western North Dakota AgronomyJournal 97 549ndash555 doi102134agronj20050549

Berg CC (1971) Forage yield of switchgrass (Panicum virgatum) inPennsylvania Agronomy Journal 63 785ndash786 doi102134agronj197100021962006300050038x

Boe A Beck DL (2008) Yield components of biomass in switchgrass CropScience 48 1306ndash1311 doi102135cropsci2007080482

Callaway RM Pennings SC Richards CL (2003) Phenotypic plasticity andinteractions among plants Ecology 84 1115ndash1128 doi1018900012-9658(2003)084[1115PPAIAP]20CO2

Casler MD (2005) Ecotypic variation among switchgrass populations fromthe northern USA Crop Science 45 388ndash398 doi102135cropsci20050388

Casler MD Vogel KP Taliaferro CM Wynia RL (2004) Latitudinaladaptation of switchgrass populations Crop Science 44 293ndash303

Casler MD Vogel KP Taliaferro CM Ehlke NJ Berdahl JD Brummer ECKallenbach RL West CP Mitchell RB (2007) Latitudinal andlongitudinal adaptation of switchgrass populations Crop Science 472249ndash2260 doi102135cropsci2006120780

Christensen JH Hewitson B Busuioc A Chen A Gao X Held I Jones RKolli RK KwonWT Laprise R Magantildea Rueda VMearns L MeneacutendezCG Raumlisaumlnen J Rinke A Sarr A (2007) Regional climate projectionsIn lsquoClimate change 2007 the physical science basis Contributionof Working Group I to the Fourth Assessment Report of the

Intergovernmental Panel on Climate Change (eds S Solomon D QinM Manning Z Chen M Marquis KB Averyt M Tignor HL Miller)pp 887ndash892 (Cambridge University Press Cambridge UK)

Das MK Fuentes RG Taliaferro CM (2004) Genetic variability and traitrelationships in switchgrass Crop Science 44 443ndash448

Easterling DR Meehl GA Parmesan C Changnon SA Karl TR Mearns LO(2000) Climate extremes observations modeling and impacts Science289 2068ndash2074 doi101126science28954872068

Emery SM Gross KL (2007) Dominant species identity not communityevenness regulates invasion in experimental grassland communitiesEcology 88 954ndash964 doi10189006-0568

Evers GW Parsons MJ (2003) Soil type and moisture level influence onAlamo switchgrass emergence and seedling growth Crop Science 43288ndash294 doi102135cropsci20030288

Fay PA Carlisle JD Knapp AK Blair JM Collins SL (2003) Productivityresponses to altered rainfall patterns in a C4-dominated grasslandOecologia 137 245ndash251 doi101007s00442-003-1331-3

FayPAKaufmanDMNippert JBCarlisle JDHarperCW(2008)Changes ingrassland ecosystem function due to extreme rainfall events implicationsfor responses to climate changeGlobal Change Biology 14 1600ndash1608doi101111j1365-2486200801605x

Fay PA Blair JM Smith MD Nippert JB Carlisle JD Knapp AK (2011)Relative effects of precipitation variability and warming on tallgrassprairie ecosystem function Biogeosciences 8 3053ndash3068 doi105194bg-8-3053-2011

Grime JP (1998) Benefits of plant diversity to ecosystems immediatefilter and founder effects Journal of Ecology 86 902ndash910doi101046j1365-2745199800306x

Hartman JC Nippert JB OrozcoRA Springer CJ (2011) Potential ecologicalimpacts of switchgrass (Panicum virgatum L) biofuel cultivation in theCentral Great Plains USA Biomass and Bioenergy 35 3415ndash3421doi101016jbiombioe201104055

Heaton E Voigt T Long SP (2004) A quantitative review comparing theyields of two candidate C4 perennial biomass crops in relation to nitrogentemperature and water Biomass and Bioenergy 27 21ndash30 doi101016jbiombioe200310005

HillebrandHBennettDMCadotteMW(2008)Consequences of dominancea review of evenness effects on local and regional ecosystem processesEcology 89 1510ndash1520 doi10189007-10531

Hughes AR Inouye BD Johnson MTJ Underwood N Vellend M (2008)Ecological consequences of genetic diversity Ecology Letters 11609ndash623 doi101111j1461-0248200801179x

Jump AS Pentildeuelas J (2005) Running to stand still adaptation and responseof plants to rapid climate change Ecology Letters 8 1010ndash1020doi101111j1461-0248200500796x

KnappAK (1984)Water relations and growth of three grasses duringwet anddrought years in a tallgrass prairie Oecologia 65 35ndash43 doi101007BF00384460

Knapp AK (1985) Effect of fire and drought on the ecophysiology ofAndropogon gerardii and Panicum virgatum in a tallgrass prairieEcology 66 1309ndash1320 doi1023071939184

KnappAKBriggs JMHartnett DCCollins SC (1998) lsquoGrassland dynamicslong-term ecological research in tallgrass prairiersquo (Oxford UniversityPress New York)

KnappAKFayPABlair JMCollinsSLSmithMDCarlisle JDHarperCWDanner BT Lett MS McCarron JK (2002) Rainfall variability carboncycling and plant species diversity in mesic grassland Science 2982202ndash2205 doi101126science1076347

Krause GH Weis E (1991) Chlorophyll fluorescence and photosynthesisthe basics Annual Review of Plant Physiology and Plant MolecularBiology 42 313ndash349 doi101146annurevpp42060191001525

Lemus R Brummer EC Moore KJ Molstad ME Burras CE Barker MF(2002) Biomass yield and quality of 20 switchgrass populations insouthern Iowa USA Biomass and Bioenergy 23 433ndash442doi101016S0961-9534(02)00073-9

Responses of switchgrass to altered precipitation Functional Plant Biology 135

Manzoni S Vico G Katul G Fay PA Polley W Palmroth S Porporato A(2011) Optimizing stomatal conductance for maximum carbon gainunder water stress a meta-analysis across plant functional types andclimates Functional Ecology 25 456ndash467 doi101111j1365-2435201001822x

Maxwell K JohnsonGN (2000) Chlorophyll fluorescence ndash a practical guideJournal of Experimental Botany 51 659ndash668 doi101093jexbot51345659

McAllister CA Knapp AK Maragni LA (1998) Is leaf-level photosynthesisrelated to plant success in a highly productive grasslandOecologia 11740ndash46 doi101007s004420050629

McLaughlin SB Kszos LN (2005) Development of switchgrass (Panicumvirgatum) as a bioenergy feedstock in the United States Biomass andBioenergy 28 515ndash535 doi101016jbiombioe200405006

McMillan C (1965) Ecotypic differentiation within four NorthAmerican prairie grasses II Behavioral variation within transplantedcommunity fractions American Journal of Botany 52 55ndash65doi1023072439975

McNaughton SJ Wolf LL (1970) Dominance and the niche in ecologicalsystems Science 167 131ndash139 doi101126science1673915131

Meehl GA Arblaster JM Tebaldi C (2005) Understanding future patterns ofincreased precipitation intensity in climate model simulationsGeophysical Research Letters 32 L18719 doi1010292005GL023680

Muir JP Sanderson MA Ocumpaugh WR Jones RM Reed RL (2001)Biomass production of lsquoAlamorsquo switchgrass in response to nitrogenphosphorus and row spacing Agronomy Journal 93 896ndash901doi102134agronj2001934896x

Newell LC (1968) Effects of strain source and management practice onforage yields of two warm-season prairie grasses Crop Science 8205ndash210 doi102135cropsci19680011183X000800020022x

Nippert JBKnappAKBriggs JM (2006) Intra-annual rainfall variability andgrassland productivity can the past predict the futurePlant Ecology 18465ndash74 doi101007s11258-005-9052-9

Nippert JB Fay PA Knapp AK (2007) Photosynthetic traits in C3 and C4

grassland species in mesocosm and field environments Environmentaland Experimental Botany 60 412ndash420 doi101016jenvexpbot200612012

Nippert JB Fay PA Carlisle JD Knapp AK Smith MD (2009)Ecophysiological responses of two dominant grasses to alteredtemperature and precipitation regimes Acta Oecologica 35 400ndash408doi101016jactao200901010

Norberg J Swaney DP Dushoff J Lin J Casagrandi R Levin SA (2001)Phenotypic diversity and ecosystem functioning in changingenvironments a theoretical framework Proceedings of the NationalAcademy of Sciences of the United States of America 9811376ndash11381 doi101073pnas171315998

Parrish DJ Fike JH (2005) The biology and agronomy of switchgrass forbiofuels Critical Reviews in Plant Sciences 24 423ndash459 doi10108007352680500316433

Polley HW Norman JM Arkebauer TJ Walter-Shea EA Greegor DH JrBramer B (1992) Leaf gas exchange of Andropogon gerardii VitmanPanicum virgatum L and Sorghastrum nutans (L) Nash in a tallgrassprairie Journal of Geophysical Research 97 18 837ndash18 844

Quinn JA (1969) Variability among High Plains populations of Panicumvirgatum Bulletin of the Torrey Botanical Club 96 20ndash41 doi1023072484005

Resco V IgnanceDD SunWHuxman TEWeltzin JFWilliams DG (2008)Chlorophyll fluorescence predawn water potential and photosynthesisin pulse-drive ecosystems ndash implications for ecological studiesFunctional Ecology 22 479ndash483 doi101111j1365-2435200801396x

Sanderson MA (1992) Morphological development of switchgrass andkleingrass Agronomy Journal 84 415ndash419 doi102134agronj199200021962008400030013x

Sanderson MA Reed RL (2000) Switchgrass growth and developmentwater nitrogen and plant density effects Journal of RangeManagement 53 221ndash227 doi1023074003287

SandersonMAWolf DD (1995)Morphological development of switchgrassin diverse environments Agronomy Journal 87 908ndash915 doi102134agronj199500021962008700050022x

Schlichting CD (1989) Phenotypic integration and environmental changeBioscience 39 460ndash464 doi1023071311138

SillettiAMKnappAK(2001)Responsesof the codominantgrassland speciesAndropogon gerardii and Sorghastrum nutans to long-termmanipulations of nitrogen and water American Midland Naturalist145 159ndash167 doi1016740003-0031(2001)145[0159ROTCGS]20CO2

SmithMDKnappAK (2003)Dominant speciesmaintain ecosystem functionwith non-random species loss Ecology Letters 6 509ndash517 doi101046j1461-0248200300454x

Sokal RRRohlf FJ (1995) lsquoBiometry the principles and practices of statisticsin biological research 3rd ednrsquo (WH Freeman New York)

StoutWL (1992)Water-use efficiency of grasses as affected by soil nitrogenand temperature Soil Science Society of America Journal 56 897ndash902doi102136sssaj199203615995005600030036x

Stout WL Jung GA Shaffer JA (1988) Effects of soil and nitrogen on wateruse efficiency of tall fescue and switchgrass under humid conditions SoilScience Society of America Journal 52 429ndash434 doi102136sssaj198803615995005200020023x

Stroup JA Sanderson MA Muir JP McFarland MJ Reed RL (2003)Comparison of growth and performance in upland and lowlandswitchgrass types to water and nitrogen stress Bioresource Technology86 65ndash72 doi101016S0960-8524(02)00102-5

Taylor SH Ripley BSWoodward FI Osborne CP (2011) Drought limitationofphotosynthesis differs betweenC3andC4grass species in a comparativeexperiment Plant Cell amp Environment 34 65ndash75 doi101111j1365-3040201002226x

Tucker SS Craine JM Nippert JB (2011) Physiological drought toleranceand the structuring of tallgrass prairie assemblages Ecosphere 2 1ndash19doi101890ES11-000231

Van Esbroeck GA Hussey MA Sanderson MA (1997) Leaf appearancerate and final leaf number of switchgrass cultivars Crop Science 37864ndash870 doi102135cropsci19970011183X003700030028x

Van Esbroeck GA Hussey MA Sanderson MA (2003) Variation betweenAlamo and Cave-in- Rock switchgrass in response to photoperiodextension Crop Science 43 639ndash643

WangDLeBauerDSDietzeMC(2010)Aquantitative reviewcomparing theyield of switchgrass in monocultures and mixtures in relation to climateand management factors GCB Bioenergy 2 16ndash25 doi101111j1757-1707201001035x

Ward JK Kelly JK (2004) Scaling up evolutionary responses to CO2 lessonsfrom Arabidopsis Ecology Letters 7 427ndash440 doi101111j1461-0248200400589x

Wullschleger SD SandersonMAMcLaughlin SB Biradar DP Rayburn AL(1996) Photosynthetic rates and ploidy levels among populations ofswitchgrassCrop Science 36 306ndash312 doi102135cropsci19960011183X003600020016x

Zhou X Talley M Luo Y (2009) Biomass litter and soil respiration along aprecipitation gradient in southern Great Plains USA Ecosystems(New York NY) 12 1369ndash1380 doi101007s10021-009-9296-7

136 Functional Plant Biology J C Hartman et al

wwwpublishcsiroaujournalsfpb

Page 7: Ecotypic responses of switchgrass to altered precipitation

and precipitation were present but there were fewecotype precipitation interactions In general physiologicaldifferences reflected ecotype origin whereas differences inaboveground biomass largely reflected differences in soilwater availability from the precipitation treatments

Ecotype effects

Differences among means in the physiological variablesmeasured largely reflected differences by ecotype andcorresponding adaptation to the local environment of originWithin each ecotype similar changes in Amax gs and E reflectthe coupled gas exchange relationships documented previouslyfor other dominant prairie grasses (Polley et al 1992) It hasalso been shown that carbon assimilation in switchgrass iscontrolled mostly by population responses to localenvironmental conditions (Wullschleger et al 1996) In this

study the Kansas ecotypes had the lowest carbon assimilationrates providing support for the local environmental conditionresponse (Wullschleger et al 1996)

The FvFm ndash gs relationship developed (Fig 2) shows astatistical relationship between responses of leafphotochemistry to changes in stomatal conductance that variesby ecotypeReducedFvFmvalues indicate a decreased efficiencyof non-photochemical quenching and are reflective of thefunctioning of the light reactions of photosynthesis (Maxwelland Johnson 2000) Changes in FvFm can be especially useful instudies where physiological performance is strongly decreasedfrom longperiods of drought (Resco et al 2008) These data showthat nearly one-quarter of the variation themaximumefficiencyofPSII can be statistically attributed to changes in stomatalconductance (Fig 2) From first principles one would assumethat a reduction in stomatal water loss would result in increased

Table 3 Mixed-effects model ANOVA of precipitation treatments and ecotype (fixed precipitation ecotype random cell) effects on abovegroundbiomass variables

F- and P-values are provided with significant responses P lt 005 Plt 010

Response Precipitation Ecotype Ecotype precipitationF P F P F P

Tillers (per individual) 371 00326 338 00429 11 035Biomass (g per individual) 948 00004 21 012 15 021Biomass per tiller (g per individual) 449 00168 1109 00001 431 00050 allocated to reproduction (per individual) 483 00127 599 00050 191 012Flowering tillers (per individual) 754 00015 848 00008 05 068Flowering tiller biomass (g per individual) 1146 lt00001 08 044 15 019Biomass per flowering tiller (g per individual) 286 00681 1333 lt00001 351 00142Flowering tiller height (cm per individual) 512 00100 18 016 16 017Leaves per flowering tiller (per individual) 547 00075 22 011 11 035Non-flowering tillers (per individual) 325 00481 291 00651 287 00337Non-flowering tiller biomass (g per individual) 11 032 484 00126 05 067Biomass per non-flowering tiller (g per individual) 007 093 1585 lt00001 17 015Non-flowering tiller height (cm per individual) 09 038 305 00576 11 036Leaves per non-flowering tiller (per individual) 02 080 670 00029 18 013

Table 4 Mean responses (1 se) of various aboveground biomass measurements for ecotype and precipitation treatment effectsWithin the ecotype or treatment category bolded numbers show significance at the P 005 level and italicised numbers show marginal significance at theP 010 level Superscripts display significant differences between precipitation treatments or ecotypes within a category NS signifies no significance between

the ecotypes or the precipitation treatments

Variables Precipitation EcotypeDecreased Average Increased Kansas Oklahoma Texas

Tillers (per individual) 56 plusmn 5ab 47 plusmn 6b 71 plusmn 6a 69 plusmn 5a 58 plusmn 6ab 48 plusmn 6b

Biomass (g per individual) 202plusmn 28b 181 plusmn 30b 352 plusmn 30a NS NS NSBiomass per tiller (g per individual) 36 plusmn 03b 40 plusmn 03ab 50 plusmn 03a 30 plusmn 03b 44 plusmn 03a 52 plusmn 03a

Flowering tillers (per individual) 35 plusmn 4b 36 plusmn 4b 58 plusmn 4a 53 plusmn 4a 48 plusmn 5a 27 plusmn 5b

Flowering tiller biomass (g per individual) 157plusmn 27b 157 plusmn 28b 325 plusmn 28a NS NS NSBiomass per flowering tiller (g per individual) 41plusmn 04b 49plusmn 07ab 56plusmn 04a 32 plusmn 04b 50 plusmn 04a 64 plusmn 04a

Flowering tiller height (cm per individual) 94 plusmn 7b 101 plusmn 8b 129 plusmn 8a NS NS NSLeaves per flowering tiller (per individual) 47 plusmn 01b 51 plusmn 02ab 56 plusmn 02a NS NS NSNon-flowering tillers (per individual) 21 plusmn 2a 11 plusmn 3b 12 plusmn 3ab 15plusmn 2ab 9plusmn 3b 20plusmn 3a

Non-flowering tiller biomass (g per individual) NS NS NS 20 plusmn 9b 14 plusmn 11b 59 plusmn 11a

Biomass per non-flowering tiller (g per individual) NS NS NS 07 plusmn 01b 08 plusmn 01b 20 plusmn 01a

Non-flowering tiller height (cm per individual) NS NS NS 27plusmn 3ab 24plusmn 4b 38plusmn 4a

Leaves per non-flowering tiller (per individual) NS NS NS 32 plusmn 02b 30 plusmn 03b 44 plusmn 03a

allocated to reproduction (per individual) 77 plusmn 3b 85 plusmn 3ab 94 plusmn 3a 87 plusmn 3a 94 plusmn 4a 75 plusmn 4b

132 Functional Plant Biology J C Hartman et al

leaf temperature with the potential for corresponding decreasedphotochemistry during hot or dry summer conditions For allecotypesFvFm decreased nonlinearly with stomatal closure buttherewere clear differences in this relationship between ecotypes

with a general trend of increasing FvFm values for a given rate ofstomatal conductance from Kansas to Oklahoma to Texasecotypes Texas ecotypes had higher maximum FvFm valuescompared with Kansas ecotypes with values near the acceptedmaximal value of 083 (Krause and Weis 1991) As gs valuesdecreased Texas and Oklahoma ecotypes were able to maintainhigher FvFm values compared with the Kansas ecotypesindicating a greater decrease in the photochemical efficiencyof PSII in Kansas ecotypes Interpretation of the relationshipbetween FvFm and gs requires further investigation but thesedata show promise for linking changes in stomatal conductancewith themaintenance of leaf photochemistry as the ability to coolthe leaf through latent heat exchange becomes very limited withlow gs (Nippert et al 2009) and photochemistry is subsequentlyreducedby increased leaf temperatures (Albert et al 2011)Basedon theseFvFmndashgs response curves (Fig 2) Oklahoma and Texasecotypes are potentially more tolerant of higher leaf temperaturewith higher FvFm at low rates of gas exchange compared withKansas ecotypes

P virgatum is a highly photoperiod sensitive species(Benedict 1940) and differences in flowering dates have beendocumented for various ecotype and cultivars (McMillan 1965VanEsbroeck et al 2003Casler et al 2004) Sanderson andWolf(1995) recorded similar flowering dates for Alamo and Cave-in-Rock cultivars compared with the ecotypes in this study Alamocultivars flowered around the end of September and Cave-in-Rock cultivars flowered around the first week of July Theflowering time and rates seen in this study correspond withknown ecotype responses to changes in latitude

Many of the aboveground biomass responses with significantecotypic differences were a result of the Texas ecotypesproducing fewer tillers but with an increased biomass pertiller and number of leaves per tiller while the Oklahoma andKansas ecotypes had higher tiller density (Table 4) Typically thebest estimates to predict changes in biomass yield for this speciesare increased biomass per tiller and number of leaves per tiller(seen in Texas ecotypes) or higher tiller density (noted in the

Table 5 Mean responses (1 se) of various aboveground biomass measurements for interactions between ecotypes and precipitation treatmentsBolded numbers show significance at theP 005 level Superscripts display significant differences between ecotype precipitation combinations NS signifies

no significant difference between the means

Variables Ecotype precipitationDecreased Average Increased

Kansas Oklahoma Texas Kansas Oklahoma Texas Kansas Oklahoma Texas

Tillers (per individual) NS NS NS NS NS NS NS NS NSBiomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per tiller (g per individual) 28 plusmn 05b 42 plusmn 05b 38 plusmn 05b 29 plusmn 05b 50 plusmn 06ab 42 plusmn 06b 34 plusmn 05b 39 plusmn 06b 76 plusmn 06a

Flowering tillers (per individual) NS NS NS NS NS NS NS NS NSFlowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per flowering tiller (g per individual) 28 plusmn 07bc 44 plusmn 07bc 51 plusmn 07bc 31 plusmn 07bc 65 plusmn 08ab 52 plusmn 08abc 36 plusmn 07bc 42 plusmn 08bc 90 plusmn 08a

Flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per flowering tiller (per individual) NS NS NS NS NS NS NS NS NSNon-flowering tillers (per individual) 31 plusmn 4a 5 plusmn 5b 28 plusmn 5ab 8 plusmn 4b 7 plusmn 5ab 18 plusmn 5ab 7 plusmn 4b 15 plusmn 5ab 15 plusmn 5ab

Non-flowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per non-flowering tiller (g per individual) NS NS NS NS NS NS NS NS NSNon-flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per non-flowering tiller (per individual) NS NS NS NS NS NS NS NS NS allocated to reproduction (per individual) NS NS NS NS NS NS NS NS NS

Nov

Till

ers

flow

ered

(in

divndash1

)T

iller

s flo

wer

ed (

indi

vndash1)

0

10

20

30

40

50

60IncreasedAverageDecreased

0

10

20

30

40

50

60 Kansas Oklahoma Texas

(a)

(b)

Jul Aug Sep Oct

Fig 4 Total tillers flowered per individual over the course of the growingseason (a) Flowering tillers by ecotype (b) flowering tillers by precipitationtreatment Data points are means 1se Asterisks indicate statisticalsignificance (Plt 005) between groups within each sampling date

Responses of switchgrass to altered precipitation Functional Plant Biology 133

Oklahoma and Kansas ecotypes Boe and Beck 2008) Althoughtotal biomass was not significantly different among ecotypes theTexas ecotypes contain components (higher biomass and leavesper tiller) that if heritablemay lead to greater biomass inmultiplegeneration studies

Precipitation effects

The altered environmental conditions resulting from theprecipitation treatments significantly altered responses inaboveground biomass measurements with statisticallysignificant responses in the physiological variables largelyreflecting changes across the sampling periods (Fig 3)Flowering tiller heights were greatest in the increasedprecipitation treatments reaching mean heights of 129mThese heights are reflective of those observed in prairieremnant populations although they are lower than those foundin P virgatum cultivars (Das et al 2004 Casler 2005Alexopoulou et al 2008) but it should be noted that plantheights are variable from year to year (Alexopoulou et al2008) The average number of leaves per flowering tiller waslower than those found in agronomic P virgatum cultivars (VanEsbroeck et al 1997) however these cultivars have been bred tooptimise yield These data illustrate individual biomass responsesto precipitation of a dominant species in the tallgrass prairieecosystem further supporting the conclusions of others thatfuture changes in precipitation are likely to significantly alterecosystem productivity (Knapp 1984 Fay et al 2003 Fay et al2008 Zhou et al 2009)

Previous attempts to link leaf-level to whole-plant responsesfor grassland species have shown positive relationships betweenphotosynthesis and plant success (McAllister et al 1998) butleaf-level to whole-plant relationships in grasslands arecommonly weak (Nippert et al 2007) In this study no clearrelationships between leaf-level photosynthetic responses andtotal aboveground biomass were present for these ecotypesAlthough the increases in leaf-level photosynthesis did notcorrespond to total biomass increases other measurements ofbiomass increased from the increased precipitation partiallysupporting the positive relationship between higher rates ofgas exchange and plant success previously noted byMcAllister et al (1998)

The only physiological responses affected by precipitationtreatments were Ymid and WUE with WUE being lowest inthe increased treatments andYmid highest in increased treatmentsAlthough ANOVA results showed significant precipitationsample period interactions (Table 2) many of the effects wereonly seen in the early season period (Fig 3) Later in the seasonincreasing summer temperatures negated the positive impacts ofthe added water Decreased WUE across the growing seasoncorresponding to higher air temperatures has been previouslynoted for C4 species where WUE can be reduced by 40 indrought conditions (Taylor et al 2011) As water availabilitydecreasedYmid decreased to amean of ndash22MPa The decreasedprecipitation treatment did not lower soil water availability tothe point where the critical water potential (leaf pressurepotential when stomatal closure is less than 5 of themaximum) was reached for switchgrass (Ycrit = ndash32MPaTucker et al 2011) Indeed previous studies have shown that

switchgrass can maintain growth and functioning at lower waterpotentials than those recorded here (Knapp 1984 Stroup et al2003) With more severe decreases in water availability YmidAmax gs E and WUE should continue to decline with thepotential for increased cuticular conductance when drought-stressed (Manzoni et al 2011)

Ecotype precipitation interactions

Surprisingly few ecotype precipitation interactions werepresent for the variablesmeasured in this study (Tables 2 3) withthe primary exception being greater biomass per tiller in theincreased treatment for Texas ecotypes These differences largelyreflect the southern Texas ecotype being adapted to a longergrowing season (Newell 1968) Growth of the Texas ecotype in alocation with a shorter growing season delayed reproductivematurity and although the overall biomass yield was notsignificantly greater for Texas ecotypes compared withOklahoma and Kansas under greater precipitation differencesthe trends observed in the current study may become statisticallysignificant Because Texas ecotypes had a longer period in whichthey grew vegetatively theywere better able to take advantage ofthe increased precipitation treatment to produce higherbiomass per tiller greater biomass per flowering tiller and highnon-flowering tiller counts in all precipitation treatments Theseresults from a southern Texas ecotype growing in amore northernlocation are similar to other studies that show increases inbiomass longer photosynthetic activity and delayed floweringfor southern ecotypes grown in more northern latitudes (Quinn1969 Lemus et al 2002 Casler et al 2004 Berdahl et al 2005Casler et al 2007)

Conclusion

The physiological responses of Amax gs E FvFm and Ymid inP virgatum were explained by ecotype differences howeverrobust responses to altered precipitationwere seen inWUEYmidflowering times and many aboveground biomass variablesEcotype differences were also seen in several abovegroundbiomass variables and most strikingly in flowering initiationand rates In this study there were few interactions betweenecotype and precipitation amount suggesting that althoughprecipitation is a strong driver of biomass production localadaptations in the ecotypes used in the current study have lessimpact on these responses than initially expected It was expectedthat the environmental differences among the origins of theecotypes mainly the effects of higher temperature coupledwith similar mean annual precipitation (860ndash878mm) wouldlead to significant ecotype precipitation interactions in theparameters measured with the Texas ecotypes being the mosttolerant of the changes The lack of interactions between geneticbackground and precipitation also has broader evolutionaryimplications gene environment interactions are a primaryindicator of the potential for natural selection in a given trait(Schlichting 1989) The broader implications of the fewecotype precipitation interactions for this widely-distributedspecies suggest that it is not likely toundergo selection in responseto changes in precipitation As many have correctly suggested itis necessary to examine the role of genetic variation in theresponse of plant species to future changes in climate

134 Functional Plant Biology J C Hartman et al

(Callaway et al 2003 Ward and Kelly 2004) however asindicated by this study it may not always be necessary toconsider localised adaptation when working to project theimpacts of climate change on the performance of a species likeswitchgrass across its geographical distribution

Acknowledgements

We thankDr JohnBlair andDr TedMorgan for their advice and comments onthemanuscriptWealso thank theNatureConservancy for allowingcollectionof switchgrass rhizomes from the Tallgrass Prairie Preserve inOklahoma andlandholdingsnear Temple TXGracieOrozcoRebeccaOrozco JacobCarterand Teall Culbertson provided assistance with fieldwork and the Division ofBiology at Kansas State University Konza Prairie Biological Station KonzaPrairie LTER (DEB-0823341) and the US Department of Energy(EEE000314) provided funding and access to facilities

References

Albert KR Mikkelsen TN Michelsen A Ro-Poulsen H van der Linden L(2011) Interactive effects of drought elevated CO2 and warming onphotosynthetic capacity and photosystem performance in temperateheath plants Journal of Plant Physiology 168 1550ndash1561 doi101016jjplph201102011

Alexopoulou E Sharma N Papatheohari Y Christou M Piscioneri IPanoutsou D Pignatelli V (2008) Biomass yields for upland andlowland switchgrass varieties grown in the Mediterranean regionBiomass and Bioenergy 32 926ndash933 doi101016jbiombioe200801015

AlleyRBBerntsen T Bindoff NL ChenZ ChidthaisongA et al (2007) Hephysical science basis summary for policy makersrsquo (IPCC SecretariatGeneva)

Barney JN Mann JJ Kyser GB Blumwald B Deynze AV DiTomaso JM(2009)Toleranceof switchgrass to extremesoilmoisture stress ecologicalimplications Plant Science 177 724ndash732 doi101016jplantsci200909003

Benedict HM (1940) Effect of day length and temperature on the floweringand growth of four species of grasses Journal of Agricultural Research61 661ndash671

Berdahl JD Frank AB Krupinsky JM Carr PM Hanson JD Johnson HA(2005) Biomass yield phenology and survival of diverse switchgrasscultivars and experimental strains in western North Dakota AgronomyJournal 97 549ndash555 doi102134agronj20050549

Berg CC (1971) Forage yield of switchgrass (Panicum virgatum) inPennsylvania Agronomy Journal 63 785ndash786 doi102134agronj197100021962006300050038x

Boe A Beck DL (2008) Yield components of biomass in switchgrass CropScience 48 1306ndash1311 doi102135cropsci2007080482

Callaway RM Pennings SC Richards CL (2003) Phenotypic plasticity andinteractions among plants Ecology 84 1115ndash1128 doi1018900012-9658(2003)084[1115PPAIAP]20CO2

Casler MD (2005) Ecotypic variation among switchgrass populations fromthe northern USA Crop Science 45 388ndash398 doi102135cropsci20050388

Casler MD Vogel KP Taliaferro CM Wynia RL (2004) Latitudinaladaptation of switchgrass populations Crop Science 44 293ndash303

Casler MD Vogel KP Taliaferro CM Ehlke NJ Berdahl JD Brummer ECKallenbach RL West CP Mitchell RB (2007) Latitudinal andlongitudinal adaptation of switchgrass populations Crop Science 472249ndash2260 doi102135cropsci2006120780

Christensen JH Hewitson B Busuioc A Chen A Gao X Held I Jones RKolli RK KwonWT Laprise R Magantildea Rueda VMearns L MeneacutendezCG Raumlisaumlnen J Rinke A Sarr A (2007) Regional climate projectionsIn lsquoClimate change 2007 the physical science basis Contributionof Working Group I to the Fourth Assessment Report of the

Intergovernmental Panel on Climate Change (eds S Solomon D QinM Manning Z Chen M Marquis KB Averyt M Tignor HL Miller)pp 887ndash892 (Cambridge University Press Cambridge UK)

Das MK Fuentes RG Taliaferro CM (2004) Genetic variability and traitrelationships in switchgrass Crop Science 44 443ndash448

Easterling DR Meehl GA Parmesan C Changnon SA Karl TR Mearns LO(2000) Climate extremes observations modeling and impacts Science289 2068ndash2074 doi101126science28954872068

Emery SM Gross KL (2007) Dominant species identity not communityevenness regulates invasion in experimental grassland communitiesEcology 88 954ndash964 doi10189006-0568

Evers GW Parsons MJ (2003) Soil type and moisture level influence onAlamo switchgrass emergence and seedling growth Crop Science 43288ndash294 doi102135cropsci20030288

Fay PA Carlisle JD Knapp AK Blair JM Collins SL (2003) Productivityresponses to altered rainfall patterns in a C4-dominated grasslandOecologia 137 245ndash251 doi101007s00442-003-1331-3

FayPAKaufmanDMNippert JBCarlisle JDHarperCW(2008)Changes ingrassland ecosystem function due to extreme rainfall events implicationsfor responses to climate changeGlobal Change Biology 14 1600ndash1608doi101111j1365-2486200801605x

Fay PA Blair JM Smith MD Nippert JB Carlisle JD Knapp AK (2011)Relative effects of precipitation variability and warming on tallgrassprairie ecosystem function Biogeosciences 8 3053ndash3068 doi105194bg-8-3053-2011

Grime JP (1998) Benefits of plant diversity to ecosystems immediatefilter and founder effects Journal of Ecology 86 902ndash910doi101046j1365-2745199800306x

Hartman JC Nippert JB OrozcoRA Springer CJ (2011) Potential ecologicalimpacts of switchgrass (Panicum virgatum L) biofuel cultivation in theCentral Great Plains USA Biomass and Bioenergy 35 3415ndash3421doi101016jbiombioe201104055

Heaton E Voigt T Long SP (2004) A quantitative review comparing theyields of two candidate C4 perennial biomass crops in relation to nitrogentemperature and water Biomass and Bioenergy 27 21ndash30 doi101016jbiombioe200310005

HillebrandHBennettDMCadotteMW(2008)Consequences of dominancea review of evenness effects on local and regional ecosystem processesEcology 89 1510ndash1520 doi10189007-10531

Hughes AR Inouye BD Johnson MTJ Underwood N Vellend M (2008)Ecological consequences of genetic diversity Ecology Letters 11609ndash623 doi101111j1461-0248200801179x

Jump AS Pentildeuelas J (2005) Running to stand still adaptation and responseof plants to rapid climate change Ecology Letters 8 1010ndash1020doi101111j1461-0248200500796x

KnappAK (1984)Water relations and growth of three grasses duringwet anddrought years in a tallgrass prairie Oecologia 65 35ndash43 doi101007BF00384460

Knapp AK (1985) Effect of fire and drought on the ecophysiology ofAndropogon gerardii and Panicum virgatum in a tallgrass prairieEcology 66 1309ndash1320 doi1023071939184

KnappAKBriggs JMHartnett DCCollins SC (1998) lsquoGrassland dynamicslong-term ecological research in tallgrass prairiersquo (Oxford UniversityPress New York)

KnappAKFayPABlair JMCollinsSLSmithMDCarlisle JDHarperCWDanner BT Lett MS McCarron JK (2002) Rainfall variability carboncycling and plant species diversity in mesic grassland Science 2982202ndash2205 doi101126science1076347

Krause GH Weis E (1991) Chlorophyll fluorescence and photosynthesisthe basics Annual Review of Plant Physiology and Plant MolecularBiology 42 313ndash349 doi101146annurevpp42060191001525

Lemus R Brummer EC Moore KJ Molstad ME Burras CE Barker MF(2002) Biomass yield and quality of 20 switchgrass populations insouthern Iowa USA Biomass and Bioenergy 23 433ndash442doi101016S0961-9534(02)00073-9

Responses of switchgrass to altered precipitation Functional Plant Biology 135

Manzoni S Vico G Katul G Fay PA Polley W Palmroth S Porporato A(2011) Optimizing stomatal conductance for maximum carbon gainunder water stress a meta-analysis across plant functional types andclimates Functional Ecology 25 456ndash467 doi101111j1365-2435201001822x

Maxwell K JohnsonGN (2000) Chlorophyll fluorescence ndash a practical guideJournal of Experimental Botany 51 659ndash668 doi101093jexbot51345659

McAllister CA Knapp AK Maragni LA (1998) Is leaf-level photosynthesisrelated to plant success in a highly productive grasslandOecologia 11740ndash46 doi101007s004420050629

McLaughlin SB Kszos LN (2005) Development of switchgrass (Panicumvirgatum) as a bioenergy feedstock in the United States Biomass andBioenergy 28 515ndash535 doi101016jbiombioe200405006

McMillan C (1965) Ecotypic differentiation within four NorthAmerican prairie grasses II Behavioral variation within transplantedcommunity fractions American Journal of Botany 52 55ndash65doi1023072439975

McNaughton SJ Wolf LL (1970) Dominance and the niche in ecologicalsystems Science 167 131ndash139 doi101126science1673915131

Meehl GA Arblaster JM Tebaldi C (2005) Understanding future patterns ofincreased precipitation intensity in climate model simulationsGeophysical Research Letters 32 L18719 doi1010292005GL023680

Muir JP Sanderson MA Ocumpaugh WR Jones RM Reed RL (2001)Biomass production of lsquoAlamorsquo switchgrass in response to nitrogenphosphorus and row spacing Agronomy Journal 93 896ndash901doi102134agronj2001934896x

Newell LC (1968) Effects of strain source and management practice onforage yields of two warm-season prairie grasses Crop Science 8205ndash210 doi102135cropsci19680011183X000800020022x

Nippert JBKnappAKBriggs JM (2006) Intra-annual rainfall variability andgrassland productivity can the past predict the futurePlant Ecology 18465ndash74 doi101007s11258-005-9052-9

Nippert JB Fay PA Knapp AK (2007) Photosynthetic traits in C3 and C4

grassland species in mesocosm and field environments Environmentaland Experimental Botany 60 412ndash420 doi101016jenvexpbot200612012

Nippert JB Fay PA Carlisle JD Knapp AK Smith MD (2009)Ecophysiological responses of two dominant grasses to alteredtemperature and precipitation regimes Acta Oecologica 35 400ndash408doi101016jactao200901010

Norberg J Swaney DP Dushoff J Lin J Casagrandi R Levin SA (2001)Phenotypic diversity and ecosystem functioning in changingenvironments a theoretical framework Proceedings of the NationalAcademy of Sciences of the United States of America 9811376ndash11381 doi101073pnas171315998

Parrish DJ Fike JH (2005) The biology and agronomy of switchgrass forbiofuels Critical Reviews in Plant Sciences 24 423ndash459 doi10108007352680500316433

Polley HW Norman JM Arkebauer TJ Walter-Shea EA Greegor DH JrBramer B (1992) Leaf gas exchange of Andropogon gerardii VitmanPanicum virgatum L and Sorghastrum nutans (L) Nash in a tallgrassprairie Journal of Geophysical Research 97 18 837ndash18 844

Quinn JA (1969) Variability among High Plains populations of Panicumvirgatum Bulletin of the Torrey Botanical Club 96 20ndash41 doi1023072484005

Resco V IgnanceDD SunWHuxman TEWeltzin JFWilliams DG (2008)Chlorophyll fluorescence predawn water potential and photosynthesisin pulse-drive ecosystems ndash implications for ecological studiesFunctional Ecology 22 479ndash483 doi101111j1365-2435200801396x

Sanderson MA (1992) Morphological development of switchgrass andkleingrass Agronomy Journal 84 415ndash419 doi102134agronj199200021962008400030013x

Sanderson MA Reed RL (2000) Switchgrass growth and developmentwater nitrogen and plant density effects Journal of RangeManagement 53 221ndash227 doi1023074003287

SandersonMAWolf DD (1995)Morphological development of switchgrassin diverse environments Agronomy Journal 87 908ndash915 doi102134agronj199500021962008700050022x

Schlichting CD (1989) Phenotypic integration and environmental changeBioscience 39 460ndash464 doi1023071311138

SillettiAMKnappAK(2001)Responsesof the codominantgrassland speciesAndropogon gerardii and Sorghastrum nutans to long-termmanipulations of nitrogen and water American Midland Naturalist145 159ndash167 doi1016740003-0031(2001)145[0159ROTCGS]20CO2

SmithMDKnappAK (2003)Dominant speciesmaintain ecosystem functionwith non-random species loss Ecology Letters 6 509ndash517 doi101046j1461-0248200300454x

Sokal RRRohlf FJ (1995) lsquoBiometry the principles and practices of statisticsin biological research 3rd ednrsquo (WH Freeman New York)

StoutWL (1992)Water-use efficiency of grasses as affected by soil nitrogenand temperature Soil Science Society of America Journal 56 897ndash902doi102136sssaj199203615995005600030036x

Stout WL Jung GA Shaffer JA (1988) Effects of soil and nitrogen on wateruse efficiency of tall fescue and switchgrass under humid conditions SoilScience Society of America Journal 52 429ndash434 doi102136sssaj198803615995005200020023x

Stroup JA Sanderson MA Muir JP McFarland MJ Reed RL (2003)Comparison of growth and performance in upland and lowlandswitchgrass types to water and nitrogen stress Bioresource Technology86 65ndash72 doi101016S0960-8524(02)00102-5

Taylor SH Ripley BSWoodward FI Osborne CP (2011) Drought limitationofphotosynthesis differs betweenC3andC4grass species in a comparativeexperiment Plant Cell amp Environment 34 65ndash75 doi101111j1365-3040201002226x

Tucker SS Craine JM Nippert JB (2011) Physiological drought toleranceand the structuring of tallgrass prairie assemblages Ecosphere 2 1ndash19doi101890ES11-000231

Van Esbroeck GA Hussey MA Sanderson MA (1997) Leaf appearancerate and final leaf number of switchgrass cultivars Crop Science 37864ndash870 doi102135cropsci19970011183X003700030028x

Van Esbroeck GA Hussey MA Sanderson MA (2003) Variation betweenAlamo and Cave-in- Rock switchgrass in response to photoperiodextension Crop Science 43 639ndash643

WangDLeBauerDSDietzeMC(2010)Aquantitative reviewcomparing theyield of switchgrass in monocultures and mixtures in relation to climateand management factors GCB Bioenergy 2 16ndash25 doi101111j1757-1707201001035x

Ward JK Kelly JK (2004) Scaling up evolutionary responses to CO2 lessonsfrom Arabidopsis Ecology Letters 7 427ndash440 doi101111j1461-0248200400589x

Wullschleger SD SandersonMAMcLaughlin SB Biradar DP Rayburn AL(1996) Photosynthetic rates and ploidy levels among populations ofswitchgrassCrop Science 36 306ndash312 doi102135cropsci19960011183X003600020016x

Zhou X Talley M Luo Y (2009) Biomass litter and soil respiration along aprecipitation gradient in southern Great Plains USA Ecosystems(New York NY) 12 1369ndash1380 doi101007s10021-009-9296-7

136 Functional Plant Biology J C Hartman et al

wwwpublishcsiroaujournalsfpb

Page 8: Ecotypic responses of switchgrass to altered precipitation

leaf temperature with the potential for corresponding decreasedphotochemistry during hot or dry summer conditions For allecotypesFvFm decreased nonlinearly with stomatal closure buttherewere clear differences in this relationship between ecotypes

with a general trend of increasing FvFm values for a given rate ofstomatal conductance from Kansas to Oklahoma to Texasecotypes Texas ecotypes had higher maximum FvFm valuescompared with Kansas ecotypes with values near the acceptedmaximal value of 083 (Krause and Weis 1991) As gs valuesdecreased Texas and Oklahoma ecotypes were able to maintainhigher FvFm values compared with the Kansas ecotypesindicating a greater decrease in the photochemical efficiencyof PSII in Kansas ecotypes Interpretation of the relationshipbetween FvFm and gs requires further investigation but thesedata show promise for linking changes in stomatal conductancewith themaintenance of leaf photochemistry as the ability to coolthe leaf through latent heat exchange becomes very limited withlow gs (Nippert et al 2009) and photochemistry is subsequentlyreducedby increased leaf temperatures (Albert et al 2011)Basedon theseFvFmndashgs response curves (Fig 2) Oklahoma and Texasecotypes are potentially more tolerant of higher leaf temperaturewith higher FvFm at low rates of gas exchange compared withKansas ecotypes

P virgatum is a highly photoperiod sensitive species(Benedict 1940) and differences in flowering dates have beendocumented for various ecotype and cultivars (McMillan 1965VanEsbroeck et al 2003Casler et al 2004) Sanderson andWolf(1995) recorded similar flowering dates for Alamo and Cave-in-Rock cultivars compared with the ecotypes in this study Alamocultivars flowered around the end of September and Cave-in-Rock cultivars flowered around the first week of July Theflowering time and rates seen in this study correspond withknown ecotype responses to changes in latitude

Many of the aboveground biomass responses with significantecotypic differences were a result of the Texas ecotypesproducing fewer tillers but with an increased biomass pertiller and number of leaves per tiller while the Oklahoma andKansas ecotypes had higher tiller density (Table 4) Typically thebest estimates to predict changes in biomass yield for this speciesare increased biomass per tiller and number of leaves per tiller(seen in Texas ecotypes) or higher tiller density (noted in the

Table 5 Mean responses (1 se) of various aboveground biomass measurements for interactions between ecotypes and precipitation treatmentsBolded numbers show significance at theP 005 level Superscripts display significant differences between ecotype precipitation combinations NS signifies

no significant difference between the means

Variables Ecotype precipitationDecreased Average Increased

Kansas Oklahoma Texas Kansas Oklahoma Texas Kansas Oklahoma Texas

Tillers (per individual) NS NS NS NS NS NS NS NS NSBiomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per tiller (g per individual) 28 plusmn 05b 42 plusmn 05b 38 plusmn 05b 29 plusmn 05b 50 plusmn 06ab 42 plusmn 06b 34 plusmn 05b 39 plusmn 06b 76 plusmn 06a

Flowering tillers (per individual) NS NS NS NS NS NS NS NS NSFlowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per flowering tiller (g per individual) 28 plusmn 07bc 44 plusmn 07bc 51 plusmn 07bc 31 plusmn 07bc 65 plusmn 08ab 52 plusmn 08abc 36 plusmn 07bc 42 plusmn 08bc 90 plusmn 08a

Flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per flowering tiller (per individual) NS NS NS NS NS NS NS NS NSNon-flowering tillers (per individual) 31 plusmn 4a 5 plusmn 5b 28 plusmn 5ab 8 plusmn 4b 7 plusmn 5ab 18 plusmn 5ab 7 plusmn 4b 15 plusmn 5ab 15 plusmn 5ab

Non-flowering tiller biomass (g per individual) NS NS NS NS NS NS NS NS NSBiomass per non-flowering tiller (g per individual) NS NS NS NS NS NS NS NS NSNon-flowering tiller height (cm per individual) NS NS NS NS NS NS NS NS NSLeaves per non-flowering tiller (per individual) NS NS NS NS NS NS NS NS NS allocated to reproduction (per individual) NS NS NS NS NS NS NS NS NS

Nov

Till

ers

flow

ered

(in

divndash1

)T

iller

s flo

wer

ed (

indi

vndash1)

0

10

20

30

40

50

60IncreasedAverageDecreased

0

10

20

30

40

50

60 Kansas Oklahoma Texas

(a)

(b)

Jul Aug Sep Oct

Fig 4 Total tillers flowered per individual over the course of the growingseason (a) Flowering tillers by ecotype (b) flowering tillers by precipitationtreatment Data points are means 1se Asterisks indicate statisticalsignificance (Plt 005) between groups within each sampling date

Responses of switchgrass to altered precipitation Functional Plant Biology 133

Oklahoma and Kansas ecotypes Boe and Beck 2008) Althoughtotal biomass was not significantly different among ecotypes theTexas ecotypes contain components (higher biomass and leavesper tiller) that if heritablemay lead to greater biomass inmultiplegeneration studies

Precipitation effects

The altered environmental conditions resulting from theprecipitation treatments significantly altered responses inaboveground biomass measurements with statisticallysignificant responses in the physiological variables largelyreflecting changes across the sampling periods (Fig 3)Flowering tiller heights were greatest in the increasedprecipitation treatments reaching mean heights of 129mThese heights are reflective of those observed in prairieremnant populations although they are lower than those foundin P virgatum cultivars (Das et al 2004 Casler 2005Alexopoulou et al 2008) but it should be noted that plantheights are variable from year to year (Alexopoulou et al2008) The average number of leaves per flowering tiller waslower than those found in agronomic P virgatum cultivars (VanEsbroeck et al 1997) however these cultivars have been bred tooptimise yield These data illustrate individual biomass responsesto precipitation of a dominant species in the tallgrass prairieecosystem further supporting the conclusions of others thatfuture changes in precipitation are likely to significantly alterecosystem productivity (Knapp 1984 Fay et al 2003 Fay et al2008 Zhou et al 2009)

Previous attempts to link leaf-level to whole-plant responsesfor grassland species have shown positive relationships betweenphotosynthesis and plant success (McAllister et al 1998) butleaf-level to whole-plant relationships in grasslands arecommonly weak (Nippert et al 2007) In this study no clearrelationships between leaf-level photosynthetic responses andtotal aboveground biomass were present for these ecotypesAlthough the increases in leaf-level photosynthesis did notcorrespond to total biomass increases other measurements ofbiomass increased from the increased precipitation partiallysupporting the positive relationship between higher rates ofgas exchange and plant success previously noted byMcAllister et al (1998)

The only physiological responses affected by precipitationtreatments were Ymid and WUE with WUE being lowest inthe increased treatments andYmid highest in increased treatmentsAlthough ANOVA results showed significant precipitationsample period interactions (Table 2) many of the effects wereonly seen in the early season period (Fig 3) Later in the seasonincreasing summer temperatures negated the positive impacts ofthe added water Decreased WUE across the growing seasoncorresponding to higher air temperatures has been previouslynoted for C4 species where WUE can be reduced by 40 indrought conditions (Taylor et al 2011) As water availabilitydecreasedYmid decreased to amean of ndash22MPa The decreasedprecipitation treatment did not lower soil water availability tothe point where the critical water potential (leaf pressurepotential when stomatal closure is less than 5 of themaximum) was reached for switchgrass (Ycrit = ndash32MPaTucker et al 2011) Indeed previous studies have shown that

switchgrass can maintain growth and functioning at lower waterpotentials than those recorded here (Knapp 1984 Stroup et al2003) With more severe decreases in water availability YmidAmax gs E and WUE should continue to decline with thepotential for increased cuticular conductance when drought-stressed (Manzoni et al 2011)

Ecotype precipitation interactions

Surprisingly few ecotype precipitation interactions werepresent for the variablesmeasured in this study (Tables 2 3) withthe primary exception being greater biomass per tiller in theincreased treatment for Texas ecotypes These differences largelyreflect the southern Texas ecotype being adapted to a longergrowing season (Newell 1968) Growth of the Texas ecotype in alocation with a shorter growing season delayed reproductivematurity and although the overall biomass yield was notsignificantly greater for Texas ecotypes compared withOklahoma and Kansas under greater precipitation differencesthe trends observed in the current study may become statisticallysignificant Because Texas ecotypes had a longer period in whichthey grew vegetatively theywere better able to take advantage ofthe increased precipitation treatment to produce higherbiomass per tiller greater biomass per flowering tiller and highnon-flowering tiller counts in all precipitation treatments Theseresults from a southern Texas ecotype growing in amore northernlocation are similar to other studies that show increases inbiomass longer photosynthetic activity and delayed floweringfor southern ecotypes grown in more northern latitudes (Quinn1969 Lemus et al 2002 Casler et al 2004 Berdahl et al 2005Casler et al 2007)

Conclusion

The physiological responses of Amax gs E FvFm and Ymid inP virgatum were explained by ecotype differences howeverrobust responses to altered precipitationwere seen inWUEYmidflowering times and many aboveground biomass variablesEcotype differences were also seen in several abovegroundbiomass variables and most strikingly in flowering initiationand rates In this study there were few interactions betweenecotype and precipitation amount suggesting that althoughprecipitation is a strong driver of biomass production localadaptations in the ecotypes used in the current study have lessimpact on these responses than initially expected It was expectedthat the environmental differences among the origins of theecotypes mainly the effects of higher temperature coupledwith similar mean annual precipitation (860ndash878mm) wouldlead to significant ecotype precipitation interactions in theparameters measured with the Texas ecotypes being the mosttolerant of the changes The lack of interactions between geneticbackground and precipitation also has broader evolutionaryimplications gene environment interactions are a primaryindicator of the potential for natural selection in a given trait(Schlichting 1989) The broader implications of the fewecotype precipitation interactions for this widely-distributedspecies suggest that it is not likely toundergo selection in responseto changes in precipitation As many have correctly suggested itis necessary to examine the role of genetic variation in theresponse of plant species to future changes in climate

134 Functional Plant Biology J C Hartman et al

(Callaway et al 2003 Ward and Kelly 2004) however asindicated by this study it may not always be necessary toconsider localised adaptation when working to project theimpacts of climate change on the performance of a species likeswitchgrass across its geographical distribution

Acknowledgements

We thankDr JohnBlair andDr TedMorgan for their advice and comments onthemanuscriptWealso thank theNatureConservancy for allowingcollectionof switchgrass rhizomes from the Tallgrass Prairie Preserve inOklahoma andlandholdingsnear Temple TXGracieOrozcoRebeccaOrozco JacobCarterand Teall Culbertson provided assistance with fieldwork and the Division ofBiology at Kansas State University Konza Prairie Biological Station KonzaPrairie LTER (DEB-0823341) and the US Department of Energy(EEE000314) provided funding and access to facilities

References

Albert KR Mikkelsen TN Michelsen A Ro-Poulsen H van der Linden L(2011) Interactive effects of drought elevated CO2 and warming onphotosynthetic capacity and photosystem performance in temperateheath plants Journal of Plant Physiology 168 1550ndash1561 doi101016jjplph201102011

Alexopoulou E Sharma N Papatheohari Y Christou M Piscioneri IPanoutsou D Pignatelli V (2008) Biomass yields for upland andlowland switchgrass varieties grown in the Mediterranean regionBiomass and Bioenergy 32 926ndash933 doi101016jbiombioe200801015

AlleyRBBerntsen T Bindoff NL ChenZ ChidthaisongA et al (2007) Hephysical science basis summary for policy makersrsquo (IPCC SecretariatGeneva)

Barney JN Mann JJ Kyser GB Blumwald B Deynze AV DiTomaso JM(2009)Toleranceof switchgrass to extremesoilmoisture stress ecologicalimplications Plant Science 177 724ndash732 doi101016jplantsci200909003

Benedict HM (1940) Effect of day length and temperature on the floweringand growth of four species of grasses Journal of Agricultural Research61 661ndash671

Berdahl JD Frank AB Krupinsky JM Carr PM Hanson JD Johnson HA(2005) Biomass yield phenology and survival of diverse switchgrasscultivars and experimental strains in western North Dakota AgronomyJournal 97 549ndash555 doi102134agronj20050549

Berg CC (1971) Forage yield of switchgrass (Panicum virgatum) inPennsylvania Agronomy Journal 63 785ndash786 doi102134agronj197100021962006300050038x

Boe A Beck DL (2008) Yield components of biomass in switchgrass CropScience 48 1306ndash1311 doi102135cropsci2007080482

Callaway RM Pennings SC Richards CL (2003) Phenotypic plasticity andinteractions among plants Ecology 84 1115ndash1128 doi1018900012-9658(2003)084[1115PPAIAP]20CO2

Casler MD (2005) Ecotypic variation among switchgrass populations fromthe northern USA Crop Science 45 388ndash398 doi102135cropsci20050388

Casler MD Vogel KP Taliaferro CM Wynia RL (2004) Latitudinaladaptation of switchgrass populations Crop Science 44 293ndash303

Casler MD Vogel KP Taliaferro CM Ehlke NJ Berdahl JD Brummer ECKallenbach RL West CP Mitchell RB (2007) Latitudinal andlongitudinal adaptation of switchgrass populations Crop Science 472249ndash2260 doi102135cropsci2006120780

Christensen JH Hewitson B Busuioc A Chen A Gao X Held I Jones RKolli RK KwonWT Laprise R Magantildea Rueda VMearns L MeneacutendezCG Raumlisaumlnen J Rinke A Sarr A (2007) Regional climate projectionsIn lsquoClimate change 2007 the physical science basis Contributionof Working Group I to the Fourth Assessment Report of the

Intergovernmental Panel on Climate Change (eds S Solomon D QinM Manning Z Chen M Marquis KB Averyt M Tignor HL Miller)pp 887ndash892 (Cambridge University Press Cambridge UK)

Das MK Fuentes RG Taliaferro CM (2004) Genetic variability and traitrelationships in switchgrass Crop Science 44 443ndash448

Easterling DR Meehl GA Parmesan C Changnon SA Karl TR Mearns LO(2000) Climate extremes observations modeling and impacts Science289 2068ndash2074 doi101126science28954872068

Emery SM Gross KL (2007) Dominant species identity not communityevenness regulates invasion in experimental grassland communitiesEcology 88 954ndash964 doi10189006-0568

Evers GW Parsons MJ (2003) Soil type and moisture level influence onAlamo switchgrass emergence and seedling growth Crop Science 43288ndash294 doi102135cropsci20030288

Fay PA Carlisle JD Knapp AK Blair JM Collins SL (2003) Productivityresponses to altered rainfall patterns in a C4-dominated grasslandOecologia 137 245ndash251 doi101007s00442-003-1331-3

FayPAKaufmanDMNippert JBCarlisle JDHarperCW(2008)Changes ingrassland ecosystem function due to extreme rainfall events implicationsfor responses to climate changeGlobal Change Biology 14 1600ndash1608doi101111j1365-2486200801605x

Fay PA Blair JM Smith MD Nippert JB Carlisle JD Knapp AK (2011)Relative effects of precipitation variability and warming on tallgrassprairie ecosystem function Biogeosciences 8 3053ndash3068 doi105194bg-8-3053-2011

Grime JP (1998) Benefits of plant diversity to ecosystems immediatefilter and founder effects Journal of Ecology 86 902ndash910doi101046j1365-2745199800306x

Hartman JC Nippert JB OrozcoRA Springer CJ (2011) Potential ecologicalimpacts of switchgrass (Panicum virgatum L) biofuel cultivation in theCentral Great Plains USA Biomass and Bioenergy 35 3415ndash3421doi101016jbiombioe201104055

Heaton E Voigt T Long SP (2004) A quantitative review comparing theyields of two candidate C4 perennial biomass crops in relation to nitrogentemperature and water Biomass and Bioenergy 27 21ndash30 doi101016jbiombioe200310005

HillebrandHBennettDMCadotteMW(2008)Consequences of dominancea review of evenness effects on local and regional ecosystem processesEcology 89 1510ndash1520 doi10189007-10531

Hughes AR Inouye BD Johnson MTJ Underwood N Vellend M (2008)Ecological consequences of genetic diversity Ecology Letters 11609ndash623 doi101111j1461-0248200801179x

Jump AS Pentildeuelas J (2005) Running to stand still adaptation and responseof plants to rapid climate change Ecology Letters 8 1010ndash1020doi101111j1461-0248200500796x

KnappAK (1984)Water relations and growth of three grasses duringwet anddrought years in a tallgrass prairie Oecologia 65 35ndash43 doi101007BF00384460

Knapp AK (1985) Effect of fire and drought on the ecophysiology ofAndropogon gerardii and Panicum virgatum in a tallgrass prairieEcology 66 1309ndash1320 doi1023071939184

KnappAKBriggs JMHartnett DCCollins SC (1998) lsquoGrassland dynamicslong-term ecological research in tallgrass prairiersquo (Oxford UniversityPress New York)

KnappAKFayPABlair JMCollinsSLSmithMDCarlisle JDHarperCWDanner BT Lett MS McCarron JK (2002) Rainfall variability carboncycling and plant species diversity in mesic grassland Science 2982202ndash2205 doi101126science1076347

Krause GH Weis E (1991) Chlorophyll fluorescence and photosynthesisthe basics Annual Review of Plant Physiology and Plant MolecularBiology 42 313ndash349 doi101146annurevpp42060191001525

Lemus R Brummer EC Moore KJ Molstad ME Burras CE Barker MF(2002) Biomass yield and quality of 20 switchgrass populations insouthern Iowa USA Biomass and Bioenergy 23 433ndash442doi101016S0961-9534(02)00073-9

Responses of switchgrass to altered precipitation Functional Plant Biology 135

Manzoni S Vico G Katul G Fay PA Polley W Palmroth S Porporato A(2011) Optimizing stomatal conductance for maximum carbon gainunder water stress a meta-analysis across plant functional types andclimates Functional Ecology 25 456ndash467 doi101111j1365-2435201001822x

Maxwell K JohnsonGN (2000) Chlorophyll fluorescence ndash a practical guideJournal of Experimental Botany 51 659ndash668 doi101093jexbot51345659

McAllister CA Knapp AK Maragni LA (1998) Is leaf-level photosynthesisrelated to plant success in a highly productive grasslandOecologia 11740ndash46 doi101007s004420050629

McLaughlin SB Kszos LN (2005) Development of switchgrass (Panicumvirgatum) as a bioenergy feedstock in the United States Biomass andBioenergy 28 515ndash535 doi101016jbiombioe200405006

McMillan C (1965) Ecotypic differentiation within four NorthAmerican prairie grasses II Behavioral variation within transplantedcommunity fractions American Journal of Botany 52 55ndash65doi1023072439975

McNaughton SJ Wolf LL (1970) Dominance and the niche in ecologicalsystems Science 167 131ndash139 doi101126science1673915131

Meehl GA Arblaster JM Tebaldi C (2005) Understanding future patterns ofincreased precipitation intensity in climate model simulationsGeophysical Research Letters 32 L18719 doi1010292005GL023680

Muir JP Sanderson MA Ocumpaugh WR Jones RM Reed RL (2001)Biomass production of lsquoAlamorsquo switchgrass in response to nitrogenphosphorus and row spacing Agronomy Journal 93 896ndash901doi102134agronj2001934896x

Newell LC (1968) Effects of strain source and management practice onforage yields of two warm-season prairie grasses Crop Science 8205ndash210 doi102135cropsci19680011183X000800020022x

Nippert JBKnappAKBriggs JM (2006) Intra-annual rainfall variability andgrassland productivity can the past predict the futurePlant Ecology 18465ndash74 doi101007s11258-005-9052-9

Nippert JB Fay PA Knapp AK (2007) Photosynthetic traits in C3 and C4

grassland species in mesocosm and field environments Environmentaland Experimental Botany 60 412ndash420 doi101016jenvexpbot200612012

Nippert JB Fay PA Carlisle JD Knapp AK Smith MD (2009)Ecophysiological responses of two dominant grasses to alteredtemperature and precipitation regimes Acta Oecologica 35 400ndash408doi101016jactao200901010

Norberg J Swaney DP Dushoff J Lin J Casagrandi R Levin SA (2001)Phenotypic diversity and ecosystem functioning in changingenvironments a theoretical framework Proceedings of the NationalAcademy of Sciences of the United States of America 9811376ndash11381 doi101073pnas171315998

Parrish DJ Fike JH (2005) The biology and agronomy of switchgrass forbiofuels Critical Reviews in Plant Sciences 24 423ndash459 doi10108007352680500316433

Polley HW Norman JM Arkebauer TJ Walter-Shea EA Greegor DH JrBramer B (1992) Leaf gas exchange of Andropogon gerardii VitmanPanicum virgatum L and Sorghastrum nutans (L) Nash in a tallgrassprairie Journal of Geophysical Research 97 18 837ndash18 844

Quinn JA (1969) Variability among High Plains populations of Panicumvirgatum Bulletin of the Torrey Botanical Club 96 20ndash41 doi1023072484005

Resco V IgnanceDD SunWHuxman TEWeltzin JFWilliams DG (2008)Chlorophyll fluorescence predawn water potential and photosynthesisin pulse-drive ecosystems ndash implications for ecological studiesFunctional Ecology 22 479ndash483 doi101111j1365-2435200801396x

Sanderson MA (1992) Morphological development of switchgrass andkleingrass Agronomy Journal 84 415ndash419 doi102134agronj199200021962008400030013x

Sanderson MA Reed RL (2000) Switchgrass growth and developmentwater nitrogen and plant density effects Journal of RangeManagement 53 221ndash227 doi1023074003287

SandersonMAWolf DD (1995)Morphological development of switchgrassin diverse environments Agronomy Journal 87 908ndash915 doi102134agronj199500021962008700050022x

Schlichting CD (1989) Phenotypic integration and environmental changeBioscience 39 460ndash464 doi1023071311138

SillettiAMKnappAK(2001)Responsesof the codominantgrassland speciesAndropogon gerardii and Sorghastrum nutans to long-termmanipulations of nitrogen and water American Midland Naturalist145 159ndash167 doi1016740003-0031(2001)145[0159ROTCGS]20CO2

SmithMDKnappAK (2003)Dominant speciesmaintain ecosystem functionwith non-random species loss Ecology Letters 6 509ndash517 doi101046j1461-0248200300454x

Sokal RRRohlf FJ (1995) lsquoBiometry the principles and practices of statisticsin biological research 3rd ednrsquo (WH Freeman New York)

StoutWL (1992)Water-use efficiency of grasses as affected by soil nitrogenand temperature Soil Science Society of America Journal 56 897ndash902doi102136sssaj199203615995005600030036x

Stout WL Jung GA Shaffer JA (1988) Effects of soil and nitrogen on wateruse efficiency of tall fescue and switchgrass under humid conditions SoilScience Society of America Journal 52 429ndash434 doi102136sssaj198803615995005200020023x

Stroup JA Sanderson MA Muir JP McFarland MJ Reed RL (2003)Comparison of growth and performance in upland and lowlandswitchgrass types to water and nitrogen stress Bioresource Technology86 65ndash72 doi101016S0960-8524(02)00102-5

Taylor SH Ripley BSWoodward FI Osborne CP (2011) Drought limitationofphotosynthesis differs betweenC3andC4grass species in a comparativeexperiment Plant Cell amp Environment 34 65ndash75 doi101111j1365-3040201002226x

Tucker SS Craine JM Nippert JB (2011) Physiological drought toleranceand the structuring of tallgrass prairie assemblages Ecosphere 2 1ndash19doi101890ES11-000231

Van Esbroeck GA Hussey MA Sanderson MA (1997) Leaf appearancerate and final leaf number of switchgrass cultivars Crop Science 37864ndash870 doi102135cropsci19970011183X003700030028x

Van Esbroeck GA Hussey MA Sanderson MA (2003) Variation betweenAlamo and Cave-in- Rock switchgrass in response to photoperiodextension Crop Science 43 639ndash643

WangDLeBauerDSDietzeMC(2010)Aquantitative reviewcomparing theyield of switchgrass in monocultures and mixtures in relation to climateand management factors GCB Bioenergy 2 16ndash25 doi101111j1757-1707201001035x

Ward JK Kelly JK (2004) Scaling up evolutionary responses to CO2 lessonsfrom Arabidopsis Ecology Letters 7 427ndash440 doi101111j1461-0248200400589x

Wullschleger SD SandersonMAMcLaughlin SB Biradar DP Rayburn AL(1996) Photosynthetic rates and ploidy levels among populations ofswitchgrassCrop Science 36 306ndash312 doi102135cropsci19960011183X003600020016x

Zhou X Talley M Luo Y (2009) Biomass litter and soil respiration along aprecipitation gradient in southern Great Plains USA Ecosystems(New York NY) 12 1369ndash1380 doi101007s10021-009-9296-7

136 Functional Plant Biology J C Hartman et al

wwwpublishcsiroaujournalsfpb

Page 9: Ecotypic responses of switchgrass to altered precipitation

Oklahoma and Kansas ecotypes Boe and Beck 2008) Althoughtotal biomass was not significantly different among ecotypes theTexas ecotypes contain components (higher biomass and leavesper tiller) that if heritablemay lead to greater biomass inmultiplegeneration studies

Precipitation effects

The altered environmental conditions resulting from theprecipitation treatments significantly altered responses inaboveground biomass measurements with statisticallysignificant responses in the physiological variables largelyreflecting changes across the sampling periods (Fig 3)Flowering tiller heights were greatest in the increasedprecipitation treatments reaching mean heights of 129mThese heights are reflective of those observed in prairieremnant populations although they are lower than those foundin P virgatum cultivars (Das et al 2004 Casler 2005Alexopoulou et al 2008) but it should be noted that plantheights are variable from year to year (Alexopoulou et al2008) The average number of leaves per flowering tiller waslower than those found in agronomic P virgatum cultivars (VanEsbroeck et al 1997) however these cultivars have been bred tooptimise yield These data illustrate individual biomass responsesto precipitation of a dominant species in the tallgrass prairieecosystem further supporting the conclusions of others thatfuture changes in precipitation are likely to significantly alterecosystem productivity (Knapp 1984 Fay et al 2003 Fay et al2008 Zhou et al 2009)

Previous attempts to link leaf-level to whole-plant responsesfor grassland species have shown positive relationships betweenphotosynthesis and plant success (McAllister et al 1998) butleaf-level to whole-plant relationships in grasslands arecommonly weak (Nippert et al 2007) In this study no clearrelationships between leaf-level photosynthetic responses andtotal aboveground biomass were present for these ecotypesAlthough the increases in leaf-level photosynthesis did notcorrespond to total biomass increases other measurements ofbiomass increased from the increased precipitation partiallysupporting the positive relationship between higher rates ofgas exchange and plant success previously noted byMcAllister et al (1998)

The only physiological responses affected by precipitationtreatments were Ymid and WUE with WUE being lowest inthe increased treatments andYmid highest in increased treatmentsAlthough ANOVA results showed significant precipitationsample period interactions (Table 2) many of the effects wereonly seen in the early season period (Fig 3) Later in the seasonincreasing summer temperatures negated the positive impacts ofthe added water Decreased WUE across the growing seasoncorresponding to higher air temperatures has been previouslynoted for C4 species where WUE can be reduced by 40 indrought conditions (Taylor et al 2011) As water availabilitydecreasedYmid decreased to amean of ndash22MPa The decreasedprecipitation treatment did not lower soil water availability tothe point where the critical water potential (leaf pressurepotential when stomatal closure is less than 5 of themaximum) was reached for switchgrass (Ycrit = ndash32MPaTucker et al 2011) Indeed previous studies have shown that

switchgrass can maintain growth and functioning at lower waterpotentials than those recorded here (Knapp 1984 Stroup et al2003) With more severe decreases in water availability YmidAmax gs E and WUE should continue to decline with thepotential for increased cuticular conductance when drought-stressed (Manzoni et al 2011)

Ecotype precipitation interactions

Surprisingly few ecotype precipitation interactions werepresent for the variablesmeasured in this study (Tables 2 3) withthe primary exception being greater biomass per tiller in theincreased treatment for Texas ecotypes These differences largelyreflect the southern Texas ecotype being adapted to a longergrowing season (Newell 1968) Growth of the Texas ecotype in alocation with a shorter growing season delayed reproductivematurity and although the overall biomass yield was notsignificantly greater for Texas ecotypes compared withOklahoma and Kansas under greater precipitation differencesthe trends observed in the current study may become statisticallysignificant Because Texas ecotypes had a longer period in whichthey grew vegetatively theywere better able to take advantage ofthe increased precipitation treatment to produce higherbiomass per tiller greater biomass per flowering tiller and highnon-flowering tiller counts in all precipitation treatments Theseresults from a southern Texas ecotype growing in amore northernlocation are similar to other studies that show increases inbiomass longer photosynthetic activity and delayed floweringfor southern ecotypes grown in more northern latitudes (Quinn1969 Lemus et al 2002 Casler et al 2004 Berdahl et al 2005Casler et al 2007)

Conclusion

The physiological responses of Amax gs E FvFm and Ymid inP virgatum were explained by ecotype differences howeverrobust responses to altered precipitationwere seen inWUEYmidflowering times and many aboveground biomass variablesEcotype differences were also seen in several abovegroundbiomass variables and most strikingly in flowering initiationand rates In this study there were few interactions betweenecotype and precipitation amount suggesting that althoughprecipitation is a strong driver of biomass production localadaptations in the ecotypes used in the current study have lessimpact on these responses than initially expected It was expectedthat the environmental differences among the origins of theecotypes mainly the effects of higher temperature coupledwith similar mean annual precipitation (860ndash878mm) wouldlead to significant ecotype precipitation interactions in theparameters measured with the Texas ecotypes being the mosttolerant of the changes The lack of interactions between geneticbackground and precipitation also has broader evolutionaryimplications gene environment interactions are a primaryindicator of the potential for natural selection in a given trait(Schlichting 1989) The broader implications of the fewecotype precipitation interactions for this widely-distributedspecies suggest that it is not likely toundergo selection in responseto changes in precipitation As many have correctly suggested itis necessary to examine the role of genetic variation in theresponse of plant species to future changes in climate

134 Functional Plant Biology J C Hartman et al

(Callaway et al 2003 Ward and Kelly 2004) however asindicated by this study it may not always be necessary toconsider localised adaptation when working to project theimpacts of climate change on the performance of a species likeswitchgrass across its geographical distribution

Acknowledgements

We thankDr JohnBlair andDr TedMorgan for their advice and comments onthemanuscriptWealso thank theNatureConservancy for allowingcollectionof switchgrass rhizomes from the Tallgrass Prairie Preserve inOklahoma andlandholdingsnear Temple TXGracieOrozcoRebeccaOrozco JacobCarterand Teall Culbertson provided assistance with fieldwork and the Division ofBiology at Kansas State University Konza Prairie Biological Station KonzaPrairie LTER (DEB-0823341) and the US Department of Energy(EEE000314) provided funding and access to facilities

References

Albert KR Mikkelsen TN Michelsen A Ro-Poulsen H van der Linden L(2011) Interactive effects of drought elevated CO2 and warming onphotosynthetic capacity and photosystem performance in temperateheath plants Journal of Plant Physiology 168 1550ndash1561 doi101016jjplph201102011

Alexopoulou E Sharma N Papatheohari Y Christou M Piscioneri IPanoutsou D Pignatelli V (2008) Biomass yields for upland andlowland switchgrass varieties grown in the Mediterranean regionBiomass and Bioenergy 32 926ndash933 doi101016jbiombioe200801015

AlleyRBBerntsen T Bindoff NL ChenZ ChidthaisongA et al (2007) Hephysical science basis summary for policy makersrsquo (IPCC SecretariatGeneva)

Barney JN Mann JJ Kyser GB Blumwald B Deynze AV DiTomaso JM(2009)Toleranceof switchgrass to extremesoilmoisture stress ecologicalimplications Plant Science 177 724ndash732 doi101016jplantsci200909003

Benedict HM (1940) Effect of day length and temperature on the floweringand growth of four species of grasses Journal of Agricultural Research61 661ndash671

Berdahl JD Frank AB Krupinsky JM Carr PM Hanson JD Johnson HA(2005) Biomass yield phenology and survival of diverse switchgrasscultivars and experimental strains in western North Dakota AgronomyJournal 97 549ndash555 doi102134agronj20050549

Berg CC (1971) Forage yield of switchgrass (Panicum virgatum) inPennsylvania Agronomy Journal 63 785ndash786 doi102134agronj197100021962006300050038x

Boe A Beck DL (2008) Yield components of biomass in switchgrass CropScience 48 1306ndash1311 doi102135cropsci2007080482

Callaway RM Pennings SC Richards CL (2003) Phenotypic plasticity andinteractions among plants Ecology 84 1115ndash1128 doi1018900012-9658(2003)084[1115PPAIAP]20CO2

Casler MD (2005) Ecotypic variation among switchgrass populations fromthe northern USA Crop Science 45 388ndash398 doi102135cropsci20050388

Casler MD Vogel KP Taliaferro CM Wynia RL (2004) Latitudinaladaptation of switchgrass populations Crop Science 44 293ndash303

Casler MD Vogel KP Taliaferro CM Ehlke NJ Berdahl JD Brummer ECKallenbach RL West CP Mitchell RB (2007) Latitudinal andlongitudinal adaptation of switchgrass populations Crop Science 472249ndash2260 doi102135cropsci2006120780

Christensen JH Hewitson B Busuioc A Chen A Gao X Held I Jones RKolli RK KwonWT Laprise R Magantildea Rueda VMearns L MeneacutendezCG Raumlisaumlnen J Rinke A Sarr A (2007) Regional climate projectionsIn lsquoClimate change 2007 the physical science basis Contributionof Working Group I to the Fourth Assessment Report of the

Intergovernmental Panel on Climate Change (eds S Solomon D QinM Manning Z Chen M Marquis KB Averyt M Tignor HL Miller)pp 887ndash892 (Cambridge University Press Cambridge UK)

Das MK Fuentes RG Taliaferro CM (2004) Genetic variability and traitrelationships in switchgrass Crop Science 44 443ndash448

Easterling DR Meehl GA Parmesan C Changnon SA Karl TR Mearns LO(2000) Climate extremes observations modeling and impacts Science289 2068ndash2074 doi101126science28954872068

Emery SM Gross KL (2007) Dominant species identity not communityevenness regulates invasion in experimental grassland communitiesEcology 88 954ndash964 doi10189006-0568

Evers GW Parsons MJ (2003) Soil type and moisture level influence onAlamo switchgrass emergence and seedling growth Crop Science 43288ndash294 doi102135cropsci20030288

Fay PA Carlisle JD Knapp AK Blair JM Collins SL (2003) Productivityresponses to altered rainfall patterns in a C4-dominated grasslandOecologia 137 245ndash251 doi101007s00442-003-1331-3

FayPAKaufmanDMNippert JBCarlisle JDHarperCW(2008)Changes ingrassland ecosystem function due to extreme rainfall events implicationsfor responses to climate changeGlobal Change Biology 14 1600ndash1608doi101111j1365-2486200801605x

Fay PA Blair JM Smith MD Nippert JB Carlisle JD Knapp AK (2011)Relative effects of precipitation variability and warming on tallgrassprairie ecosystem function Biogeosciences 8 3053ndash3068 doi105194bg-8-3053-2011

Grime JP (1998) Benefits of plant diversity to ecosystems immediatefilter and founder effects Journal of Ecology 86 902ndash910doi101046j1365-2745199800306x

Hartman JC Nippert JB OrozcoRA Springer CJ (2011) Potential ecologicalimpacts of switchgrass (Panicum virgatum L) biofuel cultivation in theCentral Great Plains USA Biomass and Bioenergy 35 3415ndash3421doi101016jbiombioe201104055

Heaton E Voigt T Long SP (2004) A quantitative review comparing theyields of two candidate C4 perennial biomass crops in relation to nitrogentemperature and water Biomass and Bioenergy 27 21ndash30 doi101016jbiombioe200310005

HillebrandHBennettDMCadotteMW(2008)Consequences of dominancea review of evenness effects on local and regional ecosystem processesEcology 89 1510ndash1520 doi10189007-10531

Hughes AR Inouye BD Johnson MTJ Underwood N Vellend M (2008)Ecological consequences of genetic diversity Ecology Letters 11609ndash623 doi101111j1461-0248200801179x

Jump AS Pentildeuelas J (2005) Running to stand still adaptation and responseof plants to rapid climate change Ecology Letters 8 1010ndash1020doi101111j1461-0248200500796x

KnappAK (1984)Water relations and growth of three grasses duringwet anddrought years in a tallgrass prairie Oecologia 65 35ndash43 doi101007BF00384460

Knapp AK (1985) Effect of fire and drought on the ecophysiology ofAndropogon gerardii and Panicum virgatum in a tallgrass prairieEcology 66 1309ndash1320 doi1023071939184

KnappAKBriggs JMHartnett DCCollins SC (1998) lsquoGrassland dynamicslong-term ecological research in tallgrass prairiersquo (Oxford UniversityPress New York)

KnappAKFayPABlair JMCollinsSLSmithMDCarlisle JDHarperCWDanner BT Lett MS McCarron JK (2002) Rainfall variability carboncycling and plant species diversity in mesic grassland Science 2982202ndash2205 doi101126science1076347

Krause GH Weis E (1991) Chlorophyll fluorescence and photosynthesisthe basics Annual Review of Plant Physiology and Plant MolecularBiology 42 313ndash349 doi101146annurevpp42060191001525

Lemus R Brummer EC Moore KJ Molstad ME Burras CE Barker MF(2002) Biomass yield and quality of 20 switchgrass populations insouthern Iowa USA Biomass and Bioenergy 23 433ndash442doi101016S0961-9534(02)00073-9

Responses of switchgrass to altered precipitation Functional Plant Biology 135

Manzoni S Vico G Katul G Fay PA Polley W Palmroth S Porporato A(2011) Optimizing stomatal conductance for maximum carbon gainunder water stress a meta-analysis across plant functional types andclimates Functional Ecology 25 456ndash467 doi101111j1365-2435201001822x

Maxwell K JohnsonGN (2000) Chlorophyll fluorescence ndash a practical guideJournal of Experimental Botany 51 659ndash668 doi101093jexbot51345659

McAllister CA Knapp AK Maragni LA (1998) Is leaf-level photosynthesisrelated to plant success in a highly productive grasslandOecologia 11740ndash46 doi101007s004420050629

McLaughlin SB Kszos LN (2005) Development of switchgrass (Panicumvirgatum) as a bioenergy feedstock in the United States Biomass andBioenergy 28 515ndash535 doi101016jbiombioe200405006

McMillan C (1965) Ecotypic differentiation within four NorthAmerican prairie grasses II Behavioral variation within transplantedcommunity fractions American Journal of Botany 52 55ndash65doi1023072439975

McNaughton SJ Wolf LL (1970) Dominance and the niche in ecologicalsystems Science 167 131ndash139 doi101126science1673915131

Meehl GA Arblaster JM Tebaldi C (2005) Understanding future patterns ofincreased precipitation intensity in climate model simulationsGeophysical Research Letters 32 L18719 doi1010292005GL023680

Muir JP Sanderson MA Ocumpaugh WR Jones RM Reed RL (2001)Biomass production of lsquoAlamorsquo switchgrass in response to nitrogenphosphorus and row spacing Agronomy Journal 93 896ndash901doi102134agronj2001934896x

Newell LC (1968) Effects of strain source and management practice onforage yields of two warm-season prairie grasses Crop Science 8205ndash210 doi102135cropsci19680011183X000800020022x

Nippert JBKnappAKBriggs JM (2006) Intra-annual rainfall variability andgrassland productivity can the past predict the futurePlant Ecology 18465ndash74 doi101007s11258-005-9052-9

Nippert JB Fay PA Knapp AK (2007) Photosynthetic traits in C3 and C4

grassland species in mesocosm and field environments Environmentaland Experimental Botany 60 412ndash420 doi101016jenvexpbot200612012

Nippert JB Fay PA Carlisle JD Knapp AK Smith MD (2009)Ecophysiological responses of two dominant grasses to alteredtemperature and precipitation regimes Acta Oecologica 35 400ndash408doi101016jactao200901010

Norberg J Swaney DP Dushoff J Lin J Casagrandi R Levin SA (2001)Phenotypic diversity and ecosystem functioning in changingenvironments a theoretical framework Proceedings of the NationalAcademy of Sciences of the United States of America 9811376ndash11381 doi101073pnas171315998

Parrish DJ Fike JH (2005) The biology and agronomy of switchgrass forbiofuels Critical Reviews in Plant Sciences 24 423ndash459 doi10108007352680500316433

Polley HW Norman JM Arkebauer TJ Walter-Shea EA Greegor DH JrBramer B (1992) Leaf gas exchange of Andropogon gerardii VitmanPanicum virgatum L and Sorghastrum nutans (L) Nash in a tallgrassprairie Journal of Geophysical Research 97 18 837ndash18 844

Quinn JA (1969) Variability among High Plains populations of Panicumvirgatum Bulletin of the Torrey Botanical Club 96 20ndash41 doi1023072484005

Resco V IgnanceDD SunWHuxman TEWeltzin JFWilliams DG (2008)Chlorophyll fluorescence predawn water potential and photosynthesisin pulse-drive ecosystems ndash implications for ecological studiesFunctional Ecology 22 479ndash483 doi101111j1365-2435200801396x

Sanderson MA (1992) Morphological development of switchgrass andkleingrass Agronomy Journal 84 415ndash419 doi102134agronj199200021962008400030013x

Sanderson MA Reed RL (2000) Switchgrass growth and developmentwater nitrogen and plant density effects Journal of RangeManagement 53 221ndash227 doi1023074003287

SandersonMAWolf DD (1995)Morphological development of switchgrassin diverse environments Agronomy Journal 87 908ndash915 doi102134agronj199500021962008700050022x

Schlichting CD (1989) Phenotypic integration and environmental changeBioscience 39 460ndash464 doi1023071311138

SillettiAMKnappAK(2001)Responsesof the codominantgrassland speciesAndropogon gerardii and Sorghastrum nutans to long-termmanipulations of nitrogen and water American Midland Naturalist145 159ndash167 doi1016740003-0031(2001)145[0159ROTCGS]20CO2

SmithMDKnappAK (2003)Dominant speciesmaintain ecosystem functionwith non-random species loss Ecology Letters 6 509ndash517 doi101046j1461-0248200300454x

Sokal RRRohlf FJ (1995) lsquoBiometry the principles and practices of statisticsin biological research 3rd ednrsquo (WH Freeman New York)

StoutWL (1992)Water-use efficiency of grasses as affected by soil nitrogenand temperature Soil Science Society of America Journal 56 897ndash902doi102136sssaj199203615995005600030036x

Stout WL Jung GA Shaffer JA (1988) Effects of soil and nitrogen on wateruse efficiency of tall fescue and switchgrass under humid conditions SoilScience Society of America Journal 52 429ndash434 doi102136sssaj198803615995005200020023x

Stroup JA Sanderson MA Muir JP McFarland MJ Reed RL (2003)Comparison of growth and performance in upland and lowlandswitchgrass types to water and nitrogen stress Bioresource Technology86 65ndash72 doi101016S0960-8524(02)00102-5

Taylor SH Ripley BSWoodward FI Osborne CP (2011) Drought limitationofphotosynthesis differs betweenC3andC4grass species in a comparativeexperiment Plant Cell amp Environment 34 65ndash75 doi101111j1365-3040201002226x

Tucker SS Craine JM Nippert JB (2011) Physiological drought toleranceand the structuring of tallgrass prairie assemblages Ecosphere 2 1ndash19doi101890ES11-000231

Van Esbroeck GA Hussey MA Sanderson MA (1997) Leaf appearancerate and final leaf number of switchgrass cultivars Crop Science 37864ndash870 doi102135cropsci19970011183X003700030028x

Van Esbroeck GA Hussey MA Sanderson MA (2003) Variation betweenAlamo and Cave-in- Rock switchgrass in response to photoperiodextension Crop Science 43 639ndash643

WangDLeBauerDSDietzeMC(2010)Aquantitative reviewcomparing theyield of switchgrass in monocultures and mixtures in relation to climateand management factors GCB Bioenergy 2 16ndash25 doi101111j1757-1707201001035x

Ward JK Kelly JK (2004) Scaling up evolutionary responses to CO2 lessonsfrom Arabidopsis Ecology Letters 7 427ndash440 doi101111j1461-0248200400589x

Wullschleger SD SandersonMAMcLaughlin SB Biradar DP Rayburn AL(1996) Photosynthetic rates and ploidy levels among populations ofswitchgrassCrop Science 36 306ndash312 doi102135cropsci19960011183X003600020016x

Zhou X Talley M Luo Y (2009) Biomass litter and soil respiration along aprecipitation gradient in southern Great Plains USA Ecosystems(New York NY) 12 1369ndash1380 doi101007s10021-009-9296-7

136 Functional Plant Biology J C Hartman et al

wwwpublishcsiroaujournalsfpb

Page 10: Ecotypic responses of switchgrass to altered precipitation

(Callaway et al 2003 Ward and Kelly 2004) however asindicated by this study it may not always be necessary toconsider localised adaptation when working to project theimpacts of climate change on the performance of a species likeswitchgrass across its geographical distribution

Acknowledgements

We thankDr JohnBlair andDr TedMorgan for their advice and comments onthemanuscriptWealso thank theNatureConservancy for allowingcollectionof switchgrass rhizomes from the Tallgrass Prairie Preserve inOklahoma andlandholdingsnear Temple TXGracieOrozcoRebeccaOrozco JacobCarterand Teall Culbertson provided assistance with fieldwork and the Division ofBiology at Kansas State University Konza Prairie Biological Station KonzaPrairie LTER (DEB-0823341) and the US Department of Energy(EEE000314) provided funding and access to facilities

References

Albert KR Mikkelsen TN Michelsen A Ro-Poulsen H van der Linden L(2011) Interactive effects of drought elevated CO2 and warming onphotosynthetic capacity and photosystem performance in temperateheath plants Journal of Plant Physiology 168 1550ndash1561 doi101016jjplph201102011

Alexopoulou E Sharma N Papatheohari Y Christou M Piscioneri IPanoutsou D Pignatelli V (2008) Biomass yields for upland andlowland switchgrass varieties grown in the Mediterranean regionBiomass and Bioenergy 32 926ndash933 doi101016jbiombioe200801015

AlleyRBBerntsen T Bindoff NL ChenZ ChidthaisongA et al (2007) Hephysical science basis summary for policy makersrsquo (IPCC SecretariatGeneva)

Barney JN Mann JJ Kyser GB Blumwald B Deynze AV DiTomaso JM(2009)Toleranceof switchgrass to extremesoilmoisture stress ecologicalimplications Plant Science 177 724ndash732 doi101016jplantsci200909003

Benedict HM (1940) Effect of day length and temperature on the floweringand growth of four species of grasses Journal of Agricultural Research61 661ndash671

Berdahl JD Frank AB Krupinsky JM Carr PM Hanson JD Johnson HA(2005) Biomass yield phenology and survival of diverse switchgrasscultivars and experimental strains in western North Dakota AgronomyJournal 97 549ndash555 doi102134agronj20050549

Berg CC (1971) Forage yield of switchgrass (Panicum virgatum) inPennsylvania Agronomy Journal 63 785ndash786 doi102134agronj197100021962006300050038x

Boe A Beck DL (2008) Yield components of biomass in switchgrass CropScience 48 1306ndash1311 doi102135cropsci2007080482

Callaway RM Pennings SC Richards CL (2003) Phenotypic plasticity andinteractions among plants Ecology 84 1115ndash1128 doi1018900012-9658(2003)084[1115PPAIAP]20CO2

Casler MD (2005) Ecotypic variation among switchgrass populations fromthe northern USA Crop Science 45 388ndash398 doi102135cropsci20050388

Casler MD Vogel KP Taliaferro CM Wynia RL (2004) Latitudinaladaptation of switchgrass populations Crop Science 44 293ndash303

Casler MD Vogel KP Taliaferro CM Ehlke NJ Berdahl JD Brummer ECKallenbach RL West CP Mitchell RB (2007) Latitudinal andlongitudinal adaptation of switchgrass populations Crop Science 472249ndash2260 doi102135cropsci2006120780

Christensen JH Hewitson B Busuioc A Chen A Gao X Held I Jones RKolli RK KwonWT Laprise R Magantildea Rueda VMearns L MeneacutendezCG Raumlisaumlnen J Rinke A Sarr A (2007) Regional climate projectionsIn lsquoClimate change 2007 the physical science basis Contributionof Working Group I to the Fourth Assessment Report of the

Intergovernmental Panel on Climate Change (eds S Solomon D QinM Manning Z Chen M Marquis KB Averyt M Tignor HL Miller)pp 887ndash892 (Cambridge University Press Cambridge UK)

Das MK Fuentes RG Taliaferro CM (2004) Genetic variability and traitrelationships in switchgrass Crop Science 44 443ndash448

Easterling DR Meehl GA Parmesan C Changnon SA Karl TR Mearns LO(2000) Climate extremes observations modeling and impacts Science289 2068ndash2074 doi101126science28954872068

Emery SM Gross KL (2007) Dominant species identity not communityevenness regulates invasion in experimental grassland communitiesEcology 88 954ndash964 doi10189006-0568

Evers GW Parsons MJ (2003) Soil type and moisture level influence onAlamo switchgrass emergence and seedling growth Crop Science 43288ndash294 doi102135cropsci20030288

Fay PA Carlisle JD Knapp AK Blair JM Collins SL (2003) Productivityresponses to altered rainfall patterns in a C4-dominated grasslandOecologia 137 245ndash251 doi101007s00442-003-1331-3

FayPAKaufmanDMNippert JBCarlisle JDHarperCW(2008)Changes ingrassland ecosystem function due to extreme rainfall events implicationsfor responses to climate changeGlobal Change Biology 14 1600ndash1608doi101111j1365-2486200801605x

Fay PA Blair JM Smith MD Nippert JB Carlisle JD Knapp AK (2011)Relative effects of precipitation variability and warming on tallgrassprairie ecosystem function Biogeosciences 8 3053ndash3068 doi105194bg-8-3053-2011

Grime JP (1998) Benefits of plant diversity to ecosystems immediatefilter and founder effects Journal of Ecology 86 902ndash910doi101046j1365-2745199800306x

Hartman JC Nippert JB OrozcoRA Springer CJ (2011) Potential ecologicalimpacts of switchgrass (Panicum virgatum L) biofuel cultivation in theCentral Great Plains USA Biomass and Bioenergy 35 3415ndash3421doi101016jbiombioe201104055

Heaton E Voigt T Long SP (2004) A quantitative review comparing theyields of two candidate C4 perennial biomass crops in relation to nitrogentemperature and water Biomass and Bioenergy 27 21ndash30 doi101016jbiombioe200310005

HillebrandHBennettDMCadotteMW(2008)Consequences of dominancea review of evenness effects on local and regional ecosystem processesEcology 89 1510ndash1520 doi10189007-10531

Hughes AR Inouye BD Johnson MTJ Underwood N Vellend M (2008)Ecological consequences of genetic diversity Ecology Letters 11609ndash623 doi101111j1461-0248200801179x

Jump AS Pentildeuelas J (2005) Running to stand still adaptation and responseof plants to rapid climate change Ecology Letters 8 1010ndash1020doi101111j1461-0248200500796x

KnappAK (1984)Water relations and growth of three grasses duringwet anddrought years in a tallgrass prairie Oecologia 65 35ndash43 doi101007BF00384460

Knapp AK (1985) Effect of fire and drought on the ecophysiology ofAndropogon gerardii and Panicum virgatum in a tallgrass prairieEcology 66 1309ndash1320 doi1023071939184

KnappAKBriggs JMHartnett DCCollins SC (1998) lsquoGrassland dynamicslong-term ecological research in tallgrass prairiersquo (Oxford UniversityPress New York)

KnappAKFayPABlair JMCollinsSLSmithMDCarlisle JDHarperCWDanner BT Lett MS McCarron JK (2002) Rainfall variability carboncycling and plant species diversity in mesic grassland Science 2982202ndash2205 doi101126science1076347

Krause GH Weis E (1991) Chlorophyll fluorescence and photosynthesisthe basics Annual Review of Plant Physiology and Plant MolecularBiology 42 313ndash349 doi101146annurevpp42060191001525

Lemus R Brummer EC Moore KJ Molstad ME Burras CE Barker MF(2002) Biomass yield and quality of 20 switchgrass populations insouthern Iowa USA Biomass and Bioenergy 23 433ndash442doi101016S0961-9534(02)00073-9

Responses of switchgrass to altered precipitation Functional Plant Biology 135

Manzoni S Vico G Katul G Fay PA Polley W Palmroth S Porporato A(2011) Optimizing stomatal conductance for maximum carbon gainunder water stress a meta-analysis across plant functional types andclimates Functional Ecology 25 456ndash467 doi101111j1365-2435201001822x

Maxwell K JohnsonGN (2000) Chlorophyll fluorescence ndash a practical guideJournal of Experimental Botany 51 659ndash668 doi101093jexbot51345659

McAllister CA Knapp AK Maragni LA (1998) Is leaf-level photosynthesisrelated to plant success in a highly productive grasslandOecologia 11740ndash46 doi101007s004420050629

McLaughlin SB Kszos LN (2005) Development of switchgrass (Panicumvirgatum) as a bioenergy feedstock in the United States Biomass andBioenergy 28 515ndash535 doi101016jbiombioe200405006

McMillan C (1965) Ecotypic differentiation within four NorthAmerican prairie grasses II Behavioral variation within transplantedcommunity fractions American Journal of Botany 52 55ndash65doi1023072439975

McNaughton SJ Wolf LL (1970) Dominance and the niche in ecologicalsystems Science 167 131ndash139 doi101126science1673915131

Meehl GA Arblaster JM Tebaldi C (2005) Understanding future patterns ofincreased precipitation intensity in climate model simulationsGeophysical Research Letters 32 L18719 doi1010292005GL023680

Muir JP Sanderson MA Ocumpaugh WR Jones RM Reed RL (2001)Biomass production of lsquoAlamorsquo switchgrass in response to nitrogenphosphorus and row spacing Agronomy Journal 93 896ndash901doi102134agronj2001934896x

Newell LC (1968) Effects of strain source and management practice onforage yields of two warm-season prairie grasses Crop Science 8205ndash210 doi102135cropsci19680011183X000800020022x

Nippert JBKnappAKBriggs JM (2006) Intra-annual rainfall variability andgrassland productivity can the past predict the futurePlant Ecology 18465ndash74 doi101007s11258-005-9052-9

Nippert JB Fay PA Knapp AK (2007) Photosynthetic traits in C3 and C4

grassland species in mesocosm and field environments Environmentaland Experimental Botany 60 412ndash420 doi101016jenvexpbot200612012

Nippert JB Fay PA Carlisle JD Knapp AK Smith MD (2009)Ecophysiological responses of two dominant grasses to alteredtemperature and precipitation regimes Acta Oecologica 35 400ndash408doi101016jactao200901010

Norberg J Swaney DP Dushoff J Lin J Casagrandi R Levin SA (2001)Phenotypic diversity and ecosystem functioning in changingenvironments a theoretical framework Proceedings of the NationalAcademy of Sciences of the United States of America 9811376ndash11381 doi101073pnas171315998

Parrish DJ Fike JH (2005) The biology and agronomy of switchgrass forbiofuels Critical Reviews in Plant Sciences 24 423ndash459 doi10108007352680500316433

Polley HW Norman JM Arkebauer TJ Walter-Shea EA Greegor DH JrBramer B (1992) Leaf gas exchange of Andropogon gerardii VitmanPanicum virgatum L and Sorghastrum nutans (L) Nash in a tallgrassprairie Journal of Geophysical Research 97 18 837ndash18 844

Quinn JA (1969) Variability among High Plains populations of Panicumvirgatum Bulletin of the Torrey Botanical Club 96 20ndash41 doi1023072484005

Resco V IgnanceDD SunWHuxman TEWeltzin JFWilliams DG (2008)Chlorophyll fluorescence predawn water potential and photosynthesisin pulse-drive ecosystems ndash implications for ecological studiesFunctional Ecology 22 479ndash483 doi101111j1365-2435200801396x

Sanderson MA (1992) Morphological development of switchgrass andkleingrass Agronomy Journal 84 415ndash419 doi102134agronj199200021962008400030013x

Sanderson MA Reed RL (2000) Switchgrass growth and developmentwater nitrogen and plant density effects Journal of RangeManagement 53 221ndash227 doi1023074003287

SandersonMAWolf DD (1995)Morphological development of switchgrassin diverse environments Agronomy Journal 87 908ndash915 doi102134agronj199500021962008700050022x

Schlichting CD (1989) Phenotypic integration and environmental changeBioscience 39 460ndash464 doi1023071311138

SillettiAMKnappAK(2001)Responsesof the codominantgrassland speciesAndropogon gerardii and Sorghastrum nutans to long-termmanipulations of nitrogen and water American Midland Naturalist145 159ndash167 doi1016740003-0031(2001)145[0159ROTCGS]20CO2

SmithMDKnappAK (2003)Dominant speciesmaintain ecosystem functionwith non-random species loss Ecology Letters 6 509ndash517 doi101046j1461-0248200300454x

Sokal RRRohlf FJ (1995) lsquoBiometry the principles and practices of statisticsin biological research 3rd ednrsquo (WH Freeman New York)

StoutWL (1992)Water-use efficiency of grasses as affected by soil nitrogenand temperature Soil Science Society of America Journal 56 897ndash902doi102136sssaj199203615995005600030036x

Stout WL Jung GA Shaffer JA (1988) Effects of soil and nitrogen on wateruse efficiency of tall fescue and switchgrass under humid conditions SoilScience Society of America Journal 52 429ndash434 doi102136sssaj198803615995005200020023x

Stroup JA Sanderson MA Muir JP McFarland MJ Reed RL (2003)Comparison of growth and performance in upland and lowlandswitchgrass types to water and nitrogen stress Bioresource Technology86 65ndash72 doi101016S0960-8524(02)00102-5

Taylor SH Ripley BSWoodward FI Osborne CP (2011) Drought limitationofphotosynthesis differs betweenC3andC4grass species in a comparativeexperiment Plant Cell amp Environment 34 65ndash75 doi101111j1365-3040201002226x

Tucker SS Craine JM Nippert JB (2011) Physiological drought toleranceand the structuring of tallgrass prairie assemblages Ecosphere 2 1ndash19doi101890ES11-000231

Van Esbroeck GA Hussey MA Sanderson MA (1997) Leaf appearancerate and final leaf number of switchgrass cultivars Crop Science 37864ndash870 doi102135cropsci19970011183X003700030028x

Van Esbroeck GA Hussey MA Sanderson MA (2003) Variation betweenAlamo and Cave-in- Rock switchgrass in response to photoperiodextension Crop Science 43 639ndash643

WangDLeBauerDSDietzeMC(2010)Aquantitative reviewcomparing theyield of switchgrass in monocultures and mixtures in relation to climateand management factors GCB Bioenergy 2 16ndash25 doi101111j1757-1707201001035x

Ward JK Kelly JK (2004) Scaling up evolutionary responses to CO2 lessonsfrom Arabidopsis Ecology Letters 7 427ndash440 doi101111j1461-0248200400589x

Wullschleger SD SandersonMAMcLaughlin SB Biradar DP Rayburn AL(1996) Photosynthetic rates and ploidy levels among populations ofswitchgrassCrop Science 36 306ndash312 doi102135cropsci19960011183X003600020016x

Zhou X Talley M Luo Y (2009) Biomass litter and soil respiration along aprecipitation gradient in southern Great Plains USA Ecosystems(New York NY) 12 1369ndash1380 doi101007s10021-009-9296-7

136 Functional Plant Biology J C Hartman et al

wwwpublishcsiroaujournalsfpb

Page 11: Ecotypic responses of switchgrass to altered precipitation

Manzoni S Vico G Katul G Fay PA Polley W Palmroth S Porporato A(2011) Optimizing stomatal conductance for maximum carbon gainunder water stress a meta-analysis across plant functional types andclimates Functional Ecology 25 456ndash467 doi101111j1365-2435201001822x

Maxwell K JohnsonGN (2000) Chlorophyll fluorescence ndash a practical guideJournal of Experimental Botany 51 659ndash668 doi101093jexbot51345659

McAllister CA Knapp AK Maragni LA (1998) Is leaf-level photosynthesisrelated to plant success in a highly productive grasslandOecologia 11740ndash46 doi101007s004420050629

McLaughlin SB Kszos LN (2005) Development of switchgrass (Panicumvirgatum) as a bioenergy feedstock in the United States Biomass andBioenergy 28 515ndash535 doi101016jbiombioe200405006

McMillan C (1965) Ecotypic differentiation within four NorthAmerican prairie grasses II Behavioral variation within transplantedcommunity fractions American Journal of Botany 52 55ndash65doi1023072439975

McNaughton SJ Wolf LL (1970) Dominance and the niche in ecologicalsystems Science 167 131ndash139 doi101126science1673915131

Meehl GA Arblaster JM Tebaldi C (2005) Understanding future patterns ofincreased precipitation intensity in climate model simulationsGeophysical Research Letters 32 L18719 doi1010292005GL023680

Muir JP Sanderson MA Ocumpaugh WR Jones RM Reed RL (2001)Biomass production of lsquoAlamorsquo switchgrass in response to nitrogenphosphorus and row spacing Agronomy Journal 93 896ndash901doi102134agronj2001934896x

Newell LC (1968) Effects of strain source and management practice onforage yields of two warm-season prairie grasses Crop Science 8205ndash210 doi102135cropsci19680011183X000800020022x

Nippert JBKnappAKBriggs JM (2006) Intra-annual rainfall variability andgrassland productivity can the past predict the futurePlant Ecology 18465ndash74 doi101007s11258-005-9052-9

Nippert JB Fay PA Knapp AK (2007) Photosynthetic traits in C3 and C4

grassland species in mesocosm and field environments Environmentaland Experimental Botany 60 412ndash420 doi101016jenvexpbot200612012

Nippert JB Fay PA Carlisle JD Knapp AK Smith MD (2009)Ecophysiological responses of two dominant grasses to alteredtemperature and precipitation regimes Acta Oecologica 35 400ndash408doi101016jactao200901010

Norberg J Swaney DP Dushoff J Lin J Casagrandi R Levin SA (2001)Phenotypic diversity and ecosystem functioning in changingenvironments a theoretical framework Proceedings of the NationalAcademy of Sciences of the United States of America 9811376ndash11381 doi101073pnas171315998

Parrish DJ Fike JH (2005) The biology and agronomy of switchgrass forbiofuels Critical Reviews in Plant Sciences 24 423ndash459 doi10108007352680500316433

Polley HW Norman JM Arkebauer TJ Walter-Shea EA Greegor DH JrBramer B (1992) Leaf gas exchange of Andropogon gerardii VitmanPanicum virgatum L and Sorghastrum nutans (L) Nash in a tallgrassprairie Journal of Geophysical Research 97 18 837ndash18 844

Quinn JA (1969) Variability among High Plains populations of Panicumvirgatum Bulletin of the Torrey Botanical Club 96 20ndash41 doi1023072484005

Resco V IgnanceDD SunWHuxman TEWeltzin JFWilliams DG (2008)Chlorophyll fluorescence predawn water potential and photosynthesisin pulse-drive ecosystems ndash implications for ecological studiesFunctional Ecology 22 479ndash483 doi101111j1365-2435200801396x

Sanderson MA (1992) Morphological development of switchgrass andkleingrass Agronomy Journal 84 415ndash419 doi102134agronj199200021962008400030013x

Sanderson MA Reed RL (2000) Switchgrass growth and developmentwater nitrogen and plant density effects Journal of RangeManagement 53 221ndash227 doi1023074003287

SandersonMAWolf DD (1995)Morphological development of switchgrassin diverse environments Agronomy Journal 87 908ndash915 doi102134agronj199500021962008700050022x

Schlichting CD (1989) Phenotypic integration and environmental changeBioscience 39 460ndash464 doi1023071311138

SillettiAMKnappAK(2001)Responsesof the codominantgrassland speciesAndropogon gerardii and Sorghastrum nutans to long-termmanipulations of nitrogen and water American Midland Naturalist145 159ndash167 doi1016740003-0031(2001)145[0159ROTCGS]20CO2

SmithMDKnappAK (2003)Dominant speciesmaintain ecosystem functionwith non-random species loss Ecology Letters 6 509ndash517 doi101046j1461-0248200300454x

Sokal RRRohlf FJ (1995) lsquoBiometry the principles and practices of statisticsin biological research 3rd ednrsquo (WH Freeman New York)

StoutWL (1992)Water-use efficiency of grasses as affected by soil nitrogenand temperature Soil Science Society of America Journal 56 897ndash902doi102136sssaj199203615995005600030036x

Stout WL Jung GA Shaffer JA (1988) Effects of soil and nitrogen on wateruse efficiency of tall fescue and switchgrass under humid conditions SoilScience Society of America Journal 52 429ndash434 doi102136sssaj198803615995005200020023x

Stroup JA Sanderson MA Muir JP McFarland MJ Reed RL (2003)Comparison of growth and performance in upland and lowlandswitchgrass types to water and nitrogen stress Bioresource Technology86 65ndash72 doi101016S0960-8524(02)00102-5

Taylor SH Ripley BSWoodward FI Osborne CP (2011) Drought limitationofphotosynthesis differs betweenC3andC4grass species in a comparativeexperiment Plant Cell amp Environment 34 65ndash75 doi101111j1365-3040201002226x

Tucker SS Craine JM Nippert JB (2011) Physiological drought toleranceand the structuring of tallgrass prairie assemblages Ecosphere 2 1ndash19doi101890ES11-000231

Van Esbroeck GA Hussey MA Sanderson MA (1997) Leaf appearancerate and final leaf number of switchgrass cultivars Crop Science 37864ndash870 doi102135cropsci19970011183X003700030028x

Van Esbroeck GA Hussey MA Sanderson MA (2003) Variation betweenAlamo and Cave-in- Rock switchgrass in response to photoperiodextension Crop Science 43 639ndash643

WangDLeBauerDSDietzeMC(2010)Aquantitative reviewcomparing theyield of switchgrass in monocultures and mixtures in relation to climateand management factors GCB Bioenergy 2 16ndash25 doi101111j1757-1707201001035x

Ward JK Kelly JK (2004) Scaling up evolutionary responses to CO2 lessonsfrom Arabidopsis Ecology Letters 7 427ndash440 doi101111j1461-0248200400589x

Wullschleger SD SandersonMAMcLaughlin SB Biradar DP Rayburn AL(1996) Photosynthetic rates and ploidy levels among populations ofswitchgrassCrop Science 36 306ndash312 doi102135cropsci19960011183X003600020016x

Zhou X Talley M Luo Y (2009) Biomass litter and soil respiration along aprecipitation gradient in southern Great Plains USA Ecosystems(New York NY) 12 1369ndash1380 doi101007s10021-009-9296-7

136 Functional Plant Biology J C Hartman et al

wwwpublishcsiroaujournalsfpb


Identification of altered regulatory interactions in disease

Identification of altered regulatory interactions in disease

Contrasting ecotypic differentiation for growth and survival in Pinus canariensis

Contrasting ecotypic differentiation for growth and survival in Pinus canariensis

Environmental and economic analysis of switchgrass production for water quality improvement in northeast Kansas

Environmental and economic analysis of switchgrass production for water quality improvement in northeast Kansas

Altered metabolome and microbiome features provide clues in ...

Altered metabolome and microbiome features provide clues in ...

Altered Cingulostriatal Coupling in Obsessive–Compulsive Disorder

Altered Cingulostriatal Coupling in Obsessive–Compulsive Disorder

Metagenomes of tropical soil-derived anaerobic switchgrass-adapted consortia with and without iron

Metagenomes of tropical soil-derived anaerobic switchgrass-adapted consortia with and without iron

Altered motor nerve excitability in end-stage kidney disease

Altered motor nerve excitability in end-stage kidney disease

Altered cytokine (receptor) mRNA expression as a tool in immunotoxicology

Altered cytokine (receptor) mRNA expression as a tool in immunotoxicology

LOOSE-FILLED AND TAPPED DENSITIES OF CHOPPED SWITCHGRASS, CORN STOVER AND WHEAT STRAW

LOOSE-FILLED AND TAPPED DENSITIES OF CHOPPED SWITCHGRASS, CORN STOVER AND WHEAT STRAW

Altered Traits Science Reveals How Meditation Changes Your ...

Altered Traits Science Reveals How Meditation Changes Your ...

Quantum Resonance Magnetic Analyzer - Altered States *

Quantum Resonance Magnetic Analyzer - Altered States *

Serpentine–Hisingerite Solid Solution in Altered Ferroan ...

Serpentine–Hisingerite Solid Solution in Altered Ferroan ...

Altered lipid metabolism in gastroschisis: A novel hypothesis

Altered lipid metabolism in gastroschisis: A novel hypothesis

Fat Accumulation with Altered Inflammation and Regeneration in Skeletal

Fat Accumulation with Altered Inflammation and Regeneration in Skeletal

A Challenge to Analyze and Detect Altered Human Fingerprints

A Challenge to Analyze and Detect Altered Human Fingerprints

Altered HIV-1 Viral Copy Number and Gene Expression ...

Altered HIV-1 Viral Copy Number and Gene Expression ...

Altered Hippocampal Place Cell Representation and Theta ...

Altered Hippocampal Place Cell Representation and Theta ...

Altered resting-state activity in seasonal affective disorder

Altered resting-state activity in seasonal affective disorder

Mineralogy and Geochemistry of Hydrothermally altered ...

Mineralogy and Geochemistry of Hydrothermally altered ...

722 - Special Laws, Altered, Amended or Repealed by the ...

722 - Special Laws, Altered, Amended or Repealed by the ...