An assessment of Pinus contorta seed production in British Columbia: Geographic variation and dynamically-downscaled climate correlates from the Canadian Regional Climate Model

(1)

Citation for this paper:

Lew, A., von Aderkas, P., Berland, A., Curry, C.L., Lacourse, T., Tencer, B. &

Weaver, A. (2017). An assessment of Pinus contorta seed production in British

Columbia: Geographic variation and dynamically-downscaled climate correlates

from the Canadian Regional Climate Model. Agricultural and Forest Meteorology,

236(April), 194-210.

https://doi.org/10.1016/j.agrformet.2016.12.013

UVicSPACE: Research & Learning Repository

_____________________________________________________________

Faculty of Science

Faculty Publications

_____________________________________________________________

An assessment of Pinus contorta seed production in British Columbia: Geographic

variation and dynamically-downscaled climate correlates from the Canadian

Regional Climate Model

Alicia Lew, Patrick von Aderkas, Anne Berland, Charles L. Curry, Terri Lacourse,

Bárbara Tencer, Andrew Weaver

April 2017

© 2016 The Author(s). Published by Elsevier B.V. This is an open access article under the

CC BY license (

http://creativecommons.org/licenses/by/4.0/

).

This article was originally published at:

https://doi.org/10.1016/j.agrformet.2016.12.013

(2)

ContentslistsavailableatScienceDirect

Agricultural

and

Forest

Meteorology

jou rn al h om ep a ge :w w w . e l s e v i e r . c o m / l o c a t e / a g r f o r m e t

An

assessment

of

Pinus

contorta

seed

production

in

British

Columbia:

Geographic

variation

and

dynamically-downscaled

climate

correlates

from

the

Canadian

Regional

Climate

Model

Alicia

Lew

a

_,

_Patrick

_von

_Aderkas

b,∗

_,

_Anne

_Berland

a

_,

_Charles

_L.

_Curry

a

_,

_Terri

_Lacourse

b

_,

Bárbara

Tencer

a

_,

_Andrew

_Weaver

a

a_School_of_Earth_and_Ocean_Sciences,_University_of_Victoria,₃₈₀₀_Finnerty_Road,_Victoria,_BC,_V8P_5C2,_Canada b_Department_of_Biology,_University_of_Victoria,₃₈₀₀_Finnerty_Road,_Victoria,_BC,_V8P_5C2,_Canada

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received16August2016

Receivedinrevisedform5December2016 Accepted18December2016

Availableonline26January2017 Keywords:

Lodgepolepine Seedyield Climate BritishColumbia Seedplanningzones Naturalpopulations

a

b

s

t

r

a

c

t

Theecologicalandeconomicimportanceoflodgepolepine(PinuscontortaDouglasexLouden)inBritish Columbia(BC)hasheightenedinterestintheadaptabilityandeffectivemanagementofthespecies, especiallyasclimatechanges.Therelationshipbetweenclimateandtheseedproductionofnatural pop-ulationsisakeymanagementissuethathasyettobeassessed.Thepurposeofthisstudyistodetermine ifvariationinP.contortaseedyieldisrelatedtotheclimateofBC.

Regionaldifferencesintheseedproductionoflodgepolepinewereexaminedusing1924archived seedlotcollectionsacross18differentnaturalstandseedplanningzones(SPZs)inBC.Therelationship betweenclimatevariationandtheseedproductionofP.contortawasthenevaluatedusing dynamically-downscaledoutputfromtheCanadianRegionalClimateModel(CRCM).

SeedproductionisrelativelyconsistentacrossSPZsspanningawiderangeofclimateregimes,with theexceptionofNassSkeenaTransition(NST)whereseedyieldisanorderofmagnitudehigherthan elsewhere.Significanttemporalcorrelationsbetweenoveralltrendsinseedproductionandboth tem-peratureandprecipitationwerefoundusingtheCRCMoutput.However,onlythreeofthe18SPZsshowed asignificantoveralltrendinmeanannualseedyieldbasedonconecollectionsmadebetween1963and 2013,suggestingthatthereproductivecapacityofnaturalpopulationsiswelladaptedtodecadal-scale climatechange.Tolerancetosignificantvariationinclimatelikelyplaysanimportantroleinexplaining theabilityofthisspeciestothrivewelloutsideitsnaturalrange.

1. Introduction

Lodgepole pine (Pinus contorta Douglas ex Louden), known for its exceptional latitudinal range from 31◦N in Baja Califor-niato 64◦N in Yukon (Koch, 1996), has theability to grow in awide range of ecosystems(Richardson, 2000).Threedifferent varieties—P.contorta var. latifolia Engelm., P. contorta var. con-tortaDoug.exLoud.,andP.contortavar.murrayana(Balf.)Engelm. (Kral,1993)—occupyabroadelevationalrangefrom0to3900m (WheelerandCritchﬁeld,1985)andareadaptedtomaritime,

conti-Abbreviations:CRCM,CanadianRegionalClimateModel;GCM,General Circula-tionModel;NST,NassSkeenaTransition;SPZ,seedplanningzone.

∗ Correspondingauthor.

E-mailaddresses:[email protected](A.Lew),[email protected](P.vonAderkas),

[email protected](A.Berland),[email protected](C.L.Curry),[email protected]

(T.Lacourse),[email protected](B.Tencer),[email protected](A.Weaver).

nental,andsubalpineconditions(Rehfeldtetal.,1999;Richardson, 2000).

Inareaswithfrequentforestfires,P.contortaregenerationhas furtheradaptedthroughthedevelopmentofclosed,orserotinous, cones(Elfvingetal.,2001),whichdonotopenuponmaturity,but remainsealeduntiltheyaresubjectedtosufficientlyhigh tem-peratures.Consequently,lodgepolepineplaysanimportantrole incolonizingpost-firelandscapes(Lotan,1970;PerryandLotan, 1977).The pioneercharacteristicsofthespecies, alongwithits edaphicandclimatictolerance,givetheprerequisitesforawide ecologicamplitude(CaseandPeterson,2007;Richardson,2000).

Thecommercialimportanceofinteriorlodgepolepine(P. con-torta var. latifolia)in BC hasresulted in heightened interest in itsadaptationsandgrowthpotentialundervarious environmen-talconditions(Wangetal.,2004).Extensivereforestationandtree improvementprogramshavebeendevelopedtoensuresustainable forestrypracticesforthespecies(YingandYanchuk,2006).Tohelp

http://dx.doi.org/10.1016/j.agrformet.2016.12.013

(3)

meettheincreasingreforestationdemandsoftheprovince,seed orchardshavebeenestablishedtoprovidehighquality,genetically selectedseed(Owensetal.,2005);however,theseorchards con-tinuetobeunabletomeetthehighreforestationdemands(Anon., 2015).Themajorityoflodgepolepineseedlingsareproducedfrom seedobtainedfromwildstands(Hadleyetal.,2001).

DespiteitsimportancetoreforestationinBC,geographic varia-tionintheseedproductionofnaturallodgepolepinepopulationsis notfullyunderstood.Earlierresearchonconeproductionandseed yieldinnaturalstandsindicatedthatseedproductioncouldvary considerablybetweenregionsandyears(Critchﬁeld,1980).Bates (1930)reportedanannualmeanof180,000germinableseedsper hectareforasouthernWyomingstand.Overa10-yearperiodof col-lections,seedproductionrangedbetween0and336,000seedper hectare.Duringthesamedecade,theannualmeanseedproduction foraColoradostandwasreportedas790,000germinableseedsper hectare,witharangeof74,000to2,042,000(Bates,1930).Despite theseapparentdifferences,moreextensiveresearchonthespatial variationinwildstandlodgepolepinereproductionhasbeen prob-lematicduetothisspecies’widedistribution,longlifespan,and relativelyslowreproductiveturnover(Mátyás,1996). Understand-inghowreproductionmayvarygeographicallyinnaturalstands providesavaluablebaselinetopotentiallyimprovereforestation stockandpractices,particularlywhenfacedwithfuture environ-mentalchallenges.

Climatechangewill likely havesevere implicationsfor both nativeandplantationpopulations.ForBC,futuresummersare gen-erallyexpectedtobewarmeranddrier,whilewinterswilllikelybe warmerandwetterthanpresent;however,thechangesinclimate areexpectedtovaryacrosstheprovince(Werner,2011).In addi-tion,climatechangeisexpectedtoincreasetheseverityofforest ﬁresinBC(Flanniganetal.,2005)aswellastheincidenceoffoliar disease(Woodsetal.,2005)andpests(Bentzetal.,2010). Infor-mationabouthowforestsrespondtoclimaticvariabilitywillallow managerstobetteranticipateandplanforfuturechangesto ecosys-temdynamics.Althoughitispossibleforvegetationtonaturally remainoutofbalancewithclimateforhundredsoreventhousands ofyears,it isnotclearhowforestswillrespondtothe acceler-atedpaceofanthropogenicclimatechange(Adams,2007).InBC, concernsoverthemaladaptationofforestspeciesbasedontheir currentdistributionsandslowreproductiveturnoverhasinspired researchonmoreactiveforestryresponses,includinghuman inter-ventionthroughassisted migration (O’Neillet al.,2008b;Leech etal.,2011).

Muchoftheresearchtodatehasfocusedonclimateresponses pertainingtogrowth.Thegrowthandsurvivalofindividualtrees appearstoberelatedtotheclimaticconditionsofthelocal environ-ment(Rehfeldtetal.,1999;Wuetal.,2005;O’Neilletal.,2008a). UsingtestsitesfromtheIllingworthprovenancetrials(Illingworth, 1978),Berland(2013)foundthemeannumberofseedspercone appearedtoberelativelystableacrossclimateregimes,incontrast totreegrowthandsurvival,whichvariedwidelyacrossthesesame regimes.Furtherresearchfromnaturalstandsisneededto under-standthechallengesthatforestsmightfaceinachangingclimate. Understandinghistoricalclimatetrendsisessentialinordertoput futureprojections incontext. GlobalCirculationModels(GCMs) arecapableofsimulatinghistoricalandfuturechangesinclimate regimes,typicallyonscalesofhundredsofkilometers.However, theseanalysesareoftenoflittleusetodecisionmakers,whorequire informationonsmallerscalestoevaluaterisksanddevelop adapta-tionstrategies(Mearnsetal.,2003).Theissueofmodelscalecanbe addressedbydownscalinginformationusingeitherastatisticalor dynamicaltechnique.Todate,studiesexaminingpopulation-scale responsesofP.contortatoclimatehavebeenprimarilyconducted usingstatistically-downscaledclimateoutputfromClimateWNA (Rehfeldtetal.,1999;Wangetal.,2006;Berland,2013).Although

therearesomeadvantagestothistechnique,themethodological constraintsofstatisticaldownscalingcannotaccountforchanges inclimatevariabilityovertime(Fowleretal.,2007).Analternative technique,whichprovidessimilaroutcomes,utilizes dynamically-downscaledclimateoutputgeneratedbyRegionalClimateModels (RCMs).

Here, we characterize the spatial and temporal pattern of seed yield in 18 natural stand seed planning zones in BC, in order to establish a baseline estimate of how the reproduc-tive capabilities of P. contorta may vary geographically. Then, dynamically-downscaledoutputfromtheCanadianRegional Cli-mateModel(CRCM)version4.2.4(Sushamaetal.,2010)isused toexaminetherelationshipbetweenclimatevariationandnatural standP.contortaseedproductioninBC.

2. Materialsandmethods

2.1. Seedyieldofnaturalstands

Historical cone collectiondata wereobtainedfrom archived recordsof1948seedlotsin22differentnaturalstandseed plan-ningzones(SPZs):BigBar(BB),Bulkley(BLK),Bush(BSH),Chilcotin (CHL),CentralPlateau(CP),CaribooTransition(CT),DeaseKlappan (DK), East Kootenay (EK), Finlay (FIN), Fort Nelson (FN), Hud-sonHope(HH),McGregor(MGR),Mica(MIC),Mt.Robson(MRB), Nechako(NCH),NassSkeenaTransition(NST),QuesnelLakes(QL), ShuswapAdams(SA),Submaritime(SM),ThompsonOkanaganArid (TOA),ThompsonOkanaganDry(TOD)andWestKootenay(WK) (Fig.1).These22naturalstandSPZswereestablishedbytheBC MinistryofForestsbasedonprovenanceperformanceinfieldtests (Anon.,1987).SPZsinuseinBCwereoriginallycreatedby overlay-ingclinaladaptivegeneticvariationdeterminedfromthesefield tests onto an ecological classificationof forestlands (Ying and Yanchuk,2006).TheboundariesofthesenaturalstandSPZs cor-respondtobiogeoclimaticsubzones(Anon.,1987)andrepresent regionsthataremoreorlessenvironmentallyuniform(Yingand Yanchuk,2006).AllnaturalstandlodgepolepineSPZsusedinthis study,withtheexception of thecoastalSubmaritime zone,are categorizedasinteriorregions.

ConeswerecollectedaccordingtoguidelinesbyLavenderetal. (1990).Sampleswereprocessedintheyearofcollection; standard-izedmethodsforhandlingconesandextractingseedwereused (Koloteloetal.,2001).SeedlotswerecollectedforeachSPZbetween 1963and2013andaminimumconecollectionfrom10treeswas requiredforeachseedlot.Eachofthe1948seedlotswascollected onasingleoccasion.Thefreshweightoftheseeds(inkilograms,kg), alongwiththevolumeofcones(inhectolitres,hL),wasdetermined foreachseedlotcollection.Freshseedmasseswerestandardizedto acommonmoisturecontentof7%inordertominimizevariation associatedwiththedryingprocess.Seedyield(kgfreshseed/hL cone)foreachseedlotwasthendeterminedbydividingthe stan-dardizedweightoffreshseedbythevolumeofconescollected. SeedlotswerearrangedintotheirrespectiveSPZandupdatesfrom theBCMinistryofForestsSEEDMAPGISsoftwareensuredthatall seedlotsreceivedtheappropriateSPZdesignationbasedontheir geographiccoordinates.

2.2. AnalysisofregionalvariationinPinuscontortaseed production

Seedplanningzoneswithfewerthan10totalseedlot collec-tionsoverthefivedecades—SM,DK,MICandFN—wereexcluded fromstatisticalanalysesduetotheirsmallsamplesizeand insuf-ficienttemporalcoverage.Consequently,thefinalsamplesizewas 1924seedlots.TheBrown-Forsythetest(␣significancelevel=0.05),

(4)

Fig.1. Naturalstandlodgepolepine(PinuscontortaDouglasexLouden)seedplanningzonesinBritishColumbiaandthelocationsof1948seedlotssampledbetween1963 and2013.SM,DK,MICandFNwereexcludedfromsubsequentstatisticalanalysesduetosmallsamplesizes.

basedonabsolutedeviationsfromthemedian,wasperformedon untransformeddatatotestforhomogeneityofvariance.Inorder toaccountforheterogeneityofvarianceandunequalsamplesize, Welch’sanalysisofvariance(ANOVA)andpost-hocGames-Howell tests(␣=0.05)wereconductedonlog-transformedseedyieldsto testfordifferencesbetweenthemeanseedyieldofeachzone.Log transformationwasusedtonormalizethedataandreduceskew. AllstatisticalanalyseswerecarriedoutinRversion3.1.0(RCore Team,2015).

2.3. Climateoutput

RegionalClimateModeloutputis quitetemporallyrealitistic whenreanalysesareinputastheboundaryconditions,primarily becausereanalysisintegratesmeteorologicalobservations(Kalnay etal.,1996).However,thelimitedavailabilityofthesereanalyses restrictsthenumberofrealizationsoftheRCMthatcanbe gen-erated.Whenimplementedasboundaryconditions,GCMsallow formanymorerealizationsusingtheRCM(Pierceetal.,2009),but introducetemporalambiguitytotheoutput(Kendonetal.,2010). Model output fromCRCM version 4.2.4 (Sushamaet al., 2010) servedastheclimateinformationforthisstudy.Outputwas uti-lizedfromtworunswithNCEP-DOEReanalysis2boundaryforcing atbotha15km(CRCM15-NREA)and45km(CRCM45-NREA) res-olutionandfromeight45kmresolutionrunswithGCMboundary forcing.BoundaryforcingfortheGCM-drivenrunswereprovided bythreerealizationsoftheECHAM5coupledGCM(Roeckneretal., 2003)andﬁverealizationsoftheCGCM3coupledGCM(McFarlane etal.,2005).Theseeightensemblemembersimulationsare sub-sequentlyreferredtoasCRCM45-GCM.Thereanalysis-drivenruns ofCRCMextendedfrom1979to2004,whiletheGCM-drivenruns

wereanalyzedfrom1958to2000.TheGCM-drivenCRCM realiza-tionsweretruncatedattheyear2000,becausethatdaterepresents theendofthemodel’shistoricalemissionsscenario.Thetwo cli-matevariablesofinterestwere2-msurfaceairtemperature(◦C) and total precipitation (rain & snow,mm),henceforth referred toassurface temperatureand precipitation,respectively. Mean annual,winter(December–February),andsummer(June–August) temperatureandprecipitationwerecalculatedusingeachmodel simulation.

ThereasonwhyRCMmodeloutputwasusedratherthana grid-dedobservationproduct,likeClimateWNA,wasbecausethehigh resolutionofthishistoricaloutputisachievedthroughuseofPRISM (Dalyetal.,2002),whichemploysaclimate-elevationinterpolation scheme.Assuch,thereisnoapriorirequirementfordynamicand thermodynamicconsistencyofthedownscaledproduct.In addi-tion,ourgoalwastoeventuallyexplore thepotentialeffects of futureclimatechangeonlodgepolepineseedproduction.Assuch, itwasimportantthatweretainedthesamemodellingframework throughoutouranalysis

2.4. Lodgepolepineserotiny

Anygivennaturalstandmayincludebothserotinoustreesand treeswithconesthatopenatmaturity(Tinkeretal.,1994). How-ever,serotinycancomplicateassessmentsofconeage,asthree yearsaretypicallyrequiredforconestoreachmaturity,andclosed conesandviableseedsmayremainonthetreesforseveralyears fol-lowingmaturation(Lotan,1970).Inthepresentstudy,theagerange ofconesrepresentedineachseedlotcollectionwasunclear.Thus, theclimatecorrespondingtoeachcollectionhadtobeassigned moreconservatively usinga meanvalue.Amoving window,or

(5)

trailing-mean,of3,5and10yearswascalculatedforeachclimate variableusingeachCRCMsimulationinordertoaccountforthe serotinyoflodgepolepine.Inthecaseofannualmeans,the previ-ous3,5or10yearsofoutputwereaveragedandassignedtotheﬁnal yearofthewindow.Forthewinter(summer)mean,theprevious 3,5or10winter(summer)meanswereaveragedandassignedto theﬁnalwinter(summer)ofthewindow.Werefertoeachunique combinationofmodeloutput,trailing-meanandseasonalorannual averageasaclimatescheme.

2.5. Climateanalyses

Foreachclimatescheme,the1924seedlotswereindividually assignedanappropriatevalueofeachclimatevariablebasedon twocriteria:theyearofcollectionandtheSPZ.First,the collec-tionyearisolatedthecorrecttimesliceoftheclimateoutputfor thatseedlot.Next,theapplicableSPZdesignationforthat seed-lotfurtherisolatedgridcellsofaparticularlongitudeandlatitude. Thesegridcellswerethenweightedwithrespecttotheirarea cov-erageandthespatially-weightedmeanoftheclimatevariablewas assignedtotheseedlotofinterest.Thisprocedurewasrepeatedfor eachseedlotundereachclimatescheme.

Oncetheappropriateclimatevariablevalueswereassignedto eachseedlot,theseedlotsweregroupedbySPZ.Theseedyieldsof seedlotscollectedinthesameSPZduringthesameyearwere aver-agedtodetermineameanannualseedyieldforthatregion.Within individualSPZs,theanomalyfromthemeanwasthencalculated forboththemeanannualseedyieldandtheclimatevariableof interestusingeachsetofclimateoutput.Climateanomalies calcu-latedfromreanalysis-drivenCRCMoutputwererepresentedunder eachofthethreetrailing-means,i.e.3,5and10years.Anomalies ofclimatemodeloutputwereindependentlycalculatedforeach oftheeightCRCM45-GCMensemblemembersbeforeaveraging themtogethertogiveamulti-modelensemblemeananomalyand a_±2standarddeviationenvelopebasedontheensemblespread. TheCRCM45-GCMoutputwasalsorepresentedundereachofthe threetrailing-means.

ClimatevariableanomaliescalculatedfromCRCM45-NREA out-putwereinvestigatedﬁrst.Thetemporalcorrelationbetweeneach climatevariableanomalyandthemeanannualseedyieldanomaly ineachSPZwascalculatedusingSpearman’srankcorrelation, pri-marilyduetoconcernssurroundingbothsmallsamplesizesand non-normality.Correlationcoefﬁcients(rs)andp-values(␣=0.05)

fromindependenttestsforeachSPZwerecomparedgeographically usingmapsforeachclimatevariable.SPZswithfewerthan10 dis-creteyearsofseedyielddatafortheaffiliatedclimateschemewere excludedfromthestatisticalanalysisduetotheirsmallsamplesize andinsufficienttemporalcoverage.Globalhypothesesof signifi-cantcorrelation(␣global=0.05)weretestedusingtheMonteCarlo

approachdevelopedbyLivezeyandChen(1983).Thisapproach accountsforthepossibilitythatSPZsinclosespatialproximityto eachothermaybecorrelated.Specifically,itallowsonetoassess theminimumfractionalareaofthedomainoverwhichlocal sta-tisticalsignificanceatagivenlevel,say95%,needstobeachieved inorderthatthesetoflocallysignificantcorrelationsnotbedueto chancealone.

ClimatevariableanomaliescalculatedfromCRCM45-GCM sim-ulations were then analyzed. Instead of temporal correlations, overalltrendsinclimatevariablesweredeterminedineach SPZ asameansofdealingwiththeasynchronousnatureofthese real-izations.SPZswithfewerthan10discreteyearsofseedyielddata fortheafﬁliatedclimateschemewereonceagainexcludedfromthe analysisduetotheirsmallsamplesize.Thetrendwasdetermined byﬁttingalinearregressiontoeachclimatevariableanomaly.The slope(␤)oftheensemblemeananomalywasofprimaryinterestin eachSPZ,buttheslopesofindividualensemblememberanomalies

Table1

Summaryofthenumberofseedlotsanddescriptivestatistics(mean,standarderror andvariance)ofnaturalstandlodgepolepine(PinuscontortaDouglasexLouden) seedyield(kgseed/hLcone)for18seedplanningzones(SPZ)inBritishColumbia, Canada.

SPZ NumberofSeedlots MeanSeedYield

(kg/hL)±S.E. Variance NST 43 0.378±0.0030 0.0199 HH 51 0.296±0.0002 0.0012 WK 184 0.285±0.0003 0.0045 NCH 160 0.283±0.0001 0.0011 EK 117 0.278±0.0004 0.0044 CHL 128 0.275±0.0001 0.0014 CT 62 0.274±0.0003 0.0021 MGR 62 0.271±0.0002 0.0014 TOA 250 0.270±0.0002 0.0026 MRB 17 0.267±0.0012 0.0048 CP 115 0.263±0.0002 0.0019 SA 148 0.262±0.0004 0.0047 TOD 170 0.262±0.0002 0.0031 BSH 30 0.261±0.0005 0.0027 FIN 99 0.261±0.0002 0.0022 BLK 163 0.255±0.0001 0.0017 QL 29 0.254±0.0001 0.0007 BB 96 0.242±0.0002 0.0021

werealsocalculatedasawayofgaugingthevariabilityinslopes. Forcomparison,alinearregressionwasalsofittothemeanannual seedyieldanomalyoverthesameinterval.Eachslopewasthen testedtoseeifitwassignificantlydifferentfromzeroatthe5% level.Multi-modelagreementwasdeclarediffiveormore indi-vidualensemblemembersshowedasignificanttrendthatagreed withthesignofthesignificantensemblemeantrend(Tebaldietal., 2011).Thesignandsignificanceoftheensemblemeanandmean annualseedyieldlinearregressionswerecomparedgeographically betweenzonesusingmapsforeachclimatescheme.InSPZswhere themajorityofGCM-drivenensemblemembersshoweda signifi-canttrendand,additionally,asignificanttemporalcorrelationwas previouslyfoundwithCRCM45-NREAoutput,linearregressions werefitandtestedforCRCM45-NREAoutputoverthesametime periodasCRCM45-GCM.

3. Resultsanddiscussion

3.1. RegionalvariationinPinuscontortaseedproduction

The Brown-Forsythe test indicated unequal variances (F

(17,1906)=16.61,p<0.0001)betweentheseedyieldsofthe18SPZs

includedinthestatisticalanalysis.Inparticular,thevarianceinseed yieldforNassSkeenaTransition(NST)wasanorderofmagnitude higherthanthatofotherSPZs(Table1;Fig.2).Inaddition,themean seedyieldofNST(0.378kgseed/hLcone)wasconsiderablyhigher thanthemeanseedyieldofthe17otherzones(0.268kgseed/hL cone)(Table1).Welch’sANOVAindicatedsigniﬁcantdifferences inmeanseedyieldbetweenseedplanningzones(F(17,398)=11.2,

p<0.0001).Post-hocanalysesusingpairwiseGames-Howelltests revealedthatmeanseedyieldwassigniﬁcantlyhigherinNST com-paredtoallotherzones(Table2)exceptHH(t(49)=3.3,p=0.1260).

Whenthe sameanalysis wasperformed ontheuntransformed data,NSTseedyieldwassigniﬁcantlyhigherthanallotherzones (F(17,398)=10.2,p<0.0001),includingHH,thoughthedifference

betweenNSTandHHwasbarelysigniﬁcant(t(46)=3.7,p=0.0461).

SinceNSTisnearthecoastandHHisintheBCInterior,theseregions representverydifferentconditions.

Asaresultofitsgeography,NSThasaheterogeneous, transi-tionalclimatewithbothmaritimeandcontinentalcharacteristics (Bennuah et al., 2004). Consequently, its steep environmen-talgradientsmakes NSTawell-known hotspotfor interspeciﬁc

(6)

Fig.2. Probabilitydensityfunctionsofnaturalstandlodgepolepine(PinuscontortaDouglasexLouden)seedyield(kgseed/hLcone)for18seedplanningzonesinBritish Columbia,Canada.

hybridizationandintrogressionbetweencoastalandinterior conif-erousspecies(Hamiltonetal.,2013).Inparticular,introgressive hybridization is known to occur between Sitka spruce (Picea sitchensisBong.) andwhitespruce(Picea glauca(Moench) Voss) inthisregion(Roche,1969), resultinginincreasedgenetic vari-ability(Hamiltonand Aitken2013),pestresistance(Ying,1991; KingandAlfaro,2009)andhighertoleranceofenvironmentalstress (Fanetal.,1997).Althoughlesswell-studiedinNSTthanspruce, lodgepolepineisrepresentedbytwovarieties—shorepine(Pinus contortavar.contorta)andinteriorlodgepolepine(P.contortavar. latifolia)(Koch,1996).Analysisofgeographicalvariationof lodge-polepinemonoterpenesshowedsubstantialchemicaldifferences betweencoastalandinteriorvarieties,withpopulationsofunique intermediatecharacterin areassuchastheSkeenaRiverregion (Forrest, 1980). Concentrations of foliar secondary metabolites, knowntobeactiveinregulatingabioticandbioticstress,arealso

consistentlyhigherinstandsofP.contortavar.latifoliafoundin theInteriorCedar-Hemlockbiogeoclimaticzone(Meidingerand Pojar,1991)ofNSTcomparedtootherinteriorstands(Wallisetal., 2011).Theextendedgrowingseasonprovidedbythemild,moist, transitionalclimateofthisregionmayreduceabioticstresssuchas drought,therebyallowingthetreestoallocateadditionalresources toreproduction(HermsandMattson,1992).

AlthoughNSTisconsideredaninteriornaturalstandSPZ,gene ﬂowbetweencoastaland interior varietiesoflodgepolepine is assumedtooccur(Koch,1996).Forinstance,geneticinﬂuencefrom thesmaller-sizedcoastalvarietywassuspectedtocontributeto thelowerheightsattainedbyinteriorlodgepolepinetrees near thecoast-interiortransitionzoneinwesternCanada(XieandYing, 1995).Themaritimepopulationsoflodgepolepinearegenetically morevariable(Wheelerand Guries,1982a)and tendtoexhibit higherseedyieldsthantheinteriorvariety(Anon.,2010).These

(7)

Fig.3.Naturalstandlodgepolepine(PinuscontortaDouglasexLouden)seedyield(kgseed/hLcone)for18seedplanningzonesinBritishColumbia,Canada,withcollections between1963and2013.

twovarietiesoflodgepolepineintergradealongthecoastalrange ofBC(ArnoandHammerly,1978),soitispossiblethatinﬂuence fromthecoastalvarietymayberesponsiblefortheexceptionalseed yieldsintheinteriorNSTzone(Figs.2and3).

Inadditiontovarietaldifferences,thelackofserotinyincones inNSTcomparedtothoseininteriorSPZsmayalsoexplainthe dif-ferencesinseedproduction.Fire-inducedserotinyisecologically importantforlodgepolepine,ensuringthatitremainsanaggressive pioneerspecies(Muir,1993).Serotinyiscommoninlodgepolepine intheBCInterior.Incontrast,serotinyisrarelyfoundinthe mar-itimeregion(Fowells,1965)wherefireislessprevalent(Lertzman etal.,2002).Individualtreesmayneedtoproducefewerseedsin ordertohavethesamereforestationimpactinareaswhereserotiny isrelativelyubiquitousandfiredominatesthelandscape.Onthe PacificNorthwestCoast—aregionwithoutfrequentfire—lodgepole pinetendstobereplacedbymoreshade-tolerantspeciessuchas Sitkaspruce(Piceasitchensis(Bong.)Carr.)(ReebandShaw,2015) andmayrequirealargerseedsettoremaincompetitive.

HudsonHope(HH)SPZissituatedinanareainwhichextensive interspeciﬁchybridizationoccursbetweenlodgepolepineandjack pine,PinusbanksianaLamb.(WheelerandGuries,1987).Although naturalhybridzonesareknowntobeimportantsourcesofnovel geneticvariationinwhichrecombinantsmaybecomesuperiorly adapted(Hamiltonetal.,2013),itisnotknownwhetherlodgepole pineseedproductionisinﬂuenced.Enhancedseedproductionin

HHandNSTmayalsobeattributedtophenotypicplasticity,i.e.the abilityofagivengenotypetoprovidearangeofphysiologicalor morphologicalphenotypesinresponsetodifferentenvironmental conditions(Voltasetal.,2008).Itisdifﬁculttodistinguishbetween geneticadaptationandphenotypicplasticity,butbothare impor-tantmechanismsbywhichlong-livedorganismsareboundmore stronglytotheirenvironment(Mátyás,1996).

SeedyieldsinHHandNSTcontrastedwiththoseinotherSPZs. Bothenvironmentalandgeneticfactorsmayhelptoexplainthe geographicdifferences.Inapreviousstudy,correlationsbetween climateandgrowthwerefoundtovaryregionally(McLaneetal., 2011).Whengrownincommontestsites,i.e.theIllingworth lodge-polepineprovenancetrial,populationsfromwarmregionswere affectedmorestronglybysummeraridity,whereaspopulations originatingfromcoldregionswereaffectedmorebyannual tem-perature. According toMcLane et al. (2011),the differences in sensitivityamongpopulationsgrownundersimilar environmen-talconditionsprovideameasureoftheinﬂuenceofbothgenetics andsiteclimate.Incontrast,astudyofreproductivetraitsof lodge-polepineinIllingworthprovenancetestsitesfoundthatthemean numberofseedsperconeremainedrelativelystableregardlessof provenanceorclimate(Berland,2013).Inourstudy,asimilarlackof variabilityinseedyieldwasgenerallyobservedwithinandbetween naturalstandSPZs(Tables1and2;Fig.A.1inAppendixA). Compar-isontoBerland’sstudyislimited,becauseourstudyinvestigated

(8)

Table2

Schemesofclimatevariableanomaliesshowingastatisticallysigniﬁcant(␣=0.05) Spearman’srankcorrelation(rs)withthemeanannualPinuscontortaseedyield

anomaly(kg/hL)ofthecorrespondingseedplanningzone(SPZ)usingprecipitation (PCP)andsurfacetemperature(ST)climateoutputfromCRCM45-NREA. Trailing-meansindicatethenumberofprecedingyears,summersorwintersthatwere averagedtogetherforeachclimateoutputanalysis.

SPZ ClimateVariable Anomaly Trailing-mean(years) rs p-value df HH PCP Annual 5 −0.547 0.0306 14 10 −0.650 0.0259 10 Summer 3 −0.638 0.0094 14 5 −0.765 0.0009 14 10 −0.755 0.0066 10 ST Summer 3 −0.600 0.0160 14 5 −0.624 0.0116 14 10 −0.671 0.0204 10 CP PCP Annual 3 0.626 0.0253 11 5 0.790 0.0036 10 ST Annual 3 0.654 0.0183 11 Summer 3 0.857 0.0003 11 Winter 5 0.748 0.0074 10 FIN PCP Annual 5 0.560 0.0401 12 10 0.560 0.0499 11 Summer 10 −0.632 0.0237 11 Winter 5 0.609 0.0237 12 BLK PCP Annual 10 −0.582 0.0403 11 Summer 5 −0.600 0.0099 16 ST Summer 10 −0.599 0.0340 11 EK ST Winter 5 −0.511 0.0321 16 SA PCP Summer 5 0.579 0.0264 13 TOA PCP Winter 10 −0.679 0.0049 15 WK ST Winter 5 −0.454 0.0458 15

naturalstandsratherthanprovenancetrials,andwecannot dis-tinguishbetweengeneticandenvironmentaleffects.Sinceseedlot collectionswerenottemporallyconsistentbetweenzones(Fig.3), itisalsodifﬁculttomakeinferencesaboutthedifferencesbetween SPZsfromthis studyalone,particularlygiventhat regionaland globalclimatemayhave hadvariable impactsover the50-year timespanofthecollections.

3.2. Seedyield-climaterelations

Localcorrelationsbetweeneachclimatevariableanomalyand meanannualseedyieldanomalyweresignificant(␣=0.05)in24 uniquecombinationsofclimateschemesandSPZusing CRCM45-NREA output(Table 2; Fig. 4, Fig. A.2 in Appendix A). Climate schemes that produced a sufficient number of statistically sig-nificantindividualresultstowarrantrejectionoftheglobalnull hypothesis(␣global=0.05)arehighlightedwithredbordersinFig.4

andFigA.2(inAppendixA).Inthesescenarios,therewasa proba-bilitynohigherthan0.05thatanequivalentorgreaternumberof significantlocalcorrelationswouldhavebeenobservedbychance. Themajorityofsignificantlocalcorrelationsfoundusing CRCM45-NREAoutput werein more northern SPZs suchasHH, CP, FIN andBLK(Table2).Onlyfourdifferentclimateschemesshowed globalfieldsignificance(␣global=0.05):5-yearmeanannual

precip-itation,10-yearmeanannualprecipitation,5-yearmeansummer precipitation(Fig.A.2 in AppendixA), and5-year mean winter surfacetemperature(Fig.4).Anomaliesandcorrelationsinvolving 15km×15kmCRCMoutputwithNCEP-DOEReanalysis2 bound-aryforcing(i.e.,CRCM15-NREA)wereindistinguishablefromthe lowerresolutionanalysesdisplayedhereandarehenceforthnot discussedindetail.

Ingeneral,CRCM45-GCMoutputcollectivelycaptureda warm-ingsignalinsurfacetemperaturesthroughouttheprovince(Fig.5). Althoughwe calculated the overall trend of each climate vari-able anomaly and the mean annual seed yield anomaly was calculated using CRCM45-GCM output, only the overall trends

areillustratedforsurfacetemperatureanomalies(Fig.5)and,in theappendix, for precipitationanomalies(Fig.A.3 in Appendix A).Threeof theeighteenSPZs—EK, TOAand SA(Figs. 5,A.3 in AppendixA;Table3)—showedatrendinthemeanannualseed yieldanomalythatwassignificantlydifferent(␣=0.05)fromzero (dataforremaining15SPZsnotshown).AcrossthesethreeSPZs, 23 climateschemes showedclimate anomalytrendsthat were significantlydifferentfromzero(␣=0.05)amongsttheensemble membersusingCRCM45-GCMoutput(Table3).Significant over-alltrendsfoundusingtheensemblemeanofGCM-drivenCRCM outputthatweresupportedbythemajorityofensemblemembers werecapturedmoreofteninsurfacetemperatureanomalies(Fig.5) thaninprecipitationanomalies(Fig.A.3inAppendixA).

SPZswhereCRCM45-NREAoutputshowedstrongcorrelations withthemeanannualseedyieldanomaly(Table2)werethe nat-uralstartingpointforfurtheranalysis.Meteorologicalreanalysis representedourbestestimateofobservedclimateonlargescales. CRCMdynamicallydownscaledtheseboundaryconditionsto pro-duceasynchronizedsimulationofactualclimateoverBC.However, sinceonlyonerealizationofthereanalysis-drivenCRCMwasused, eightadditionalGCM-drivenensemblerunsofCRCMwere nec-essarytoincreaseconfidenceintheoveralltrendsoftheclimatic variables.ZoneswheretheCRCM45-NREAoutputwassignificantly correlatedwiththemeanannualseedyieldanomalyand, further-more,wherethemajorityofensemblemembersagreedwiththe significanttrendoftheensemblemeanCRCM45-GCMoutput,were highlightedasthebestopportunitiesforinvestigatingrelationships betweenclimateandseedproduction.

InmostSPZs,significanttrendsinCRCM45-GCMand CRCM45-NREA output were unfortunately coupled with non-significant trendsorcontradictoryresultsintheseedyieldanomalies, restrict-ing further analysis. In Hudson Hope (HH), for instance, the ensemblemeanofGCM-drivenoutputof5-yearmeansummer sur-facetemperatureanomaliesshowedasignificantpositiveincrease (_␤ =0.021◦Cyear−1, p<0.0001) from 1971 to 2000 (Fig. 7a). CRCM45-NREAoutputof5-yearmeansummersurface tempera-tureanomalieswerenegativelycorrelatedwiththemeanannual seedyieldanomaly(rs(14)=−0.624,p=0.0116)from1984to2004

(Fig. 6a). Over the period of overlap with CRCM45-NREA, i.e. 1984–2000,theresultis similar:␤ =0.021◦Cyear−1,p=0.0003. Thetrend inCRCM45-NREA5-year meansummersurface tem-peratureanomaliesover this periodwasalsosigniﬁcantand of thesamesign:␤=0.037◦Cyear−1,p=0.0174.Consequently,there wasnoreasontosuspectthatthesinglerealizationof CRCM45-NREAoutputwasunreasonablydifferentfromtheCRCM45-GCM output.However,sincetheslopeofthemeanannualseedyield anomalywasnotsigniﬁcantlydifferentfromzeroduringboththe 1971–2000(␤ =-0.001kghL−1year−1,p=0.4843)and1984–2000 (␤=-0.003kghL−1year−1,p=0.0583)intervals,theresultsforthis SPZwereinconclusive.

In the East Kootenay (EK) SPZ, where a sound comparison betweenallelementsof thestudywaspossible,CRCM45-NREA outputof5-yearmeanwintersurfacetemperatureanomalieswas negativelycorrelated withthemeanannualseedyieldanomaly (rs(16)=−0.511,p=0.0321)from1985to2004(Fig.6b).The

ensem-ble meanof GCM-drivenoutputof 5-yearmeanwinter surface temperature anomalies showed a significant positive increase (␤=0.038◦Cyear−1,p<0.0001)from1969to2000(Fig.7b).Both theGCMandreanalysis-drivenCRCMoutputcapturedsignificant relationshipsinthisSPZunderthisclimatescheme(Tables2and3). In addition, the slope of the mean annual seed yield anomaly wassignificantly differentfrom zeroin EK from1969to 2000, ␤ =0.002kghL−1year−1,p=0.0066(Fig.7b).However,whenthe overall trends of these components were investigated by fit-ting linear regressions for the overlapping period from 1985 to2000,the slopesof theensemble meanGCM-driven climate

(9)

Fig.4.Temporalcorrelationsforsurfacetemperature.Spearman’srankcorrelation(rs)betweenthemeanannualPinuscontortaseedyieldanomaly(kg/hL)ofthe

corre-spondingseedplanningzone(SPZ)anddifferentclimateschemesusingoutputfromCRCM45-NREA:Meanannual(a),winter(b)andsummer(c)surfacetemperaturewith a3-yeartrailing-mean;meanannual(d),winter(e)andsummer(f)surfacetemperaturewitha5-yeartrailing-mean;andmeanannual(g),winter(h)andsummer(i) surfacetemperaturewitha10-yeartrailing-mean.Localsignificance(␣=0.05)isidentifiedbyhatching,andredbordersaroundapanelindicateglobalfieldsignificance (␣global=0.05).SPZwithn<10discreteyearsofseedyielddatafortheaffiliatedclimateschemewereexcluded.

anomaly(_␤=0.014◦Cyear−1,p=0.0552),reanalysis-drivenclimate anomaly(␤=0.013◦Cyear−1,p=0.5219)andseedyieldanomaly(␤ =−0.001kghL−1year−1,p=0.5865)werenotsigniﬁcantlydifferent fromzero.

While investigations using the reanalysis and GCM-driven CRCMoutputindependentlyproducedsigniﬁcantresults,the pre-dictiveskillofeach analysiswashinderedfordifferentreasons. Thereanalysis-basedtemporalcorrelationswerecalculatedusinga singlerealizationofCRCM,limitingthereliabilityoftheclimate sig-nal.Incontrast,averagingeightindividualCRCM45-GCMmembers madetheclimatesignalmuchclearer,butthetemporalambiguity ofthisoutputonlyallowedforoveralltrendstobeinvestigated. Thus,therelationshipbetweenclimateandseedproductioncould notbedeterminedwithahighdegreeofcertainty.

The complexities of biological and climate systems makes exploringtherelationshipbetweenannual seedproductionand climatedifficult,particularlyat largespatialscales.Research to datehasfocusedontheimpactofclimateontreeseedproduction intheformofmastseedingevents—periodswheresomespecies produceexceptionallylargeseedcrops.NortonandKelly(1988) hypothesizedthatvariationsinannualclimatemaycauseperiodsof decreasedbiologicalstressandallowtreestoproducemoreseeds. Inparticular,meansummertemperatureswerefoundtohavea weak—butsignificant—positivecorrelationwiththeannualseed productionofrimu(DacrydiumcupressinumLamb.)inNewZealand, butnosignificantcorrelationswerefoundbetweentotalseed pro-ductionandrainfall(NortonandKelly,1988).Asimilarrelationship hasbeennotedinDouglas-fir(PseudotsugamenziesiiMirb.(Franco)) withmeantemperaturesinthepreviousJunepositivelycorrelated

(10)

Fig.5.Signsofsigniﬁcant(␣=0.05)linearregressionslopes(␤)forensemblemeansurfacetemperatureanomaliesofthecorrespondingseedplanningzone(SPZ)under differentclimateschemesusingCRCM45-GCMoutput:meanannual(a),winter(b)andsummer(c)surfacetemperaturewitha3-yeartrailing-mean;meanannual(d),winter (e)andsummer(f)surfacetemperaturewitha5-yeartrailing-mean;andmeanannual(g),winter(h)andsummer(i)surfacetemperaturewitha10-yeartrailing-mean.SPZs showingasigniﬁcantpositiveslopeforthemeanannualPinuscontortaseedyieldanomaly(kghL−1_year−1₎_are_indicated_by_hatching._White_SPZs_indicate_that_the_slope

oftheensemblemeansurfacetemperatureanomaly(◦Cyear−1)wasnotsignificantlydifferentfromzero(p>0.05).DarkredordarkbluecolorationidentifiesSPZswherea strictmajorityof5ormoreindividualensemblemembersagreewiththesignificantslopeoftheensemblemean.SPZswithn<10discreteyearsofseedyielddataforthe affiliatedclimateschemewereexcluded.

withconeproduction(Eis,1973).Otherresearchfoundthat vari-abilityinannualseedproductioninvariousmastingtreespeciesis correlatedwithannualrainfall(Sorketal.,1993;Woodwardetal., 1994).Incontrast,analysesconductedbyKoenigandKnops(2000) foundthatmastseedingdidnotcorrelatewithpatternsof vari-abilityineitherannualrainfallormeantemperature.Lodgepole pineconeproductionvariesfromyear-to-year,thoughnotasmuch asotherconifers(Elliott,1974;Herreraetal.,1998).Thepresent studyfoundsigniﬁcant,moderatetostrongcorrelationsbetween themeanannualseedyieldoflodgepolepineandclimatevariables fromareanalysis-drivenrunofCRCM(Table2),aswellas signiﬁ-cantoveralltrendsusinganensembleofGCM-drivenCRCMoutput (Table3).

Our study investigated the correlation between and overall trend in seed production and climate, and cannot conclusively attribute the causation of such relationships without further research.However,itisinterestingtorecallthatlodgepolepinehas exceptionallybroadecologicamplitude(CaseandPeterson,2007) andiscapableofadaptingtodiverse,andoftensevere, environ-mentalconditions(Rehfeldtetal.,1999).Yellowcedar(Callitropsis nootkatensisD.Don)—ageneralistspecies—isalsoadaptedtowide climatic gradientsand, in terms of its productivity, appearsto beinsensitivetodifferencesinmoisture(RussellandKrakowski, 2012).

A portion of lodgepole pine’s adaptability can be explained intermsofgeneticvariability,growthandsurvival. Populations

(11)

Fig.6.TemporalvariationofmeanannualPinuscontortaseedyieldanomaly(kg/hL)alongwith:a)5-yearmeansummersurfacetemperature(ST)anomalyinHudsonHope (HH),andb)5-yearmeanwinterSTanomalyinEastKootenay(EK)calculatedusingoutputfromCRCM45-NREA.

throughoutBC aresigniﬁcantlydifferentfromone another(Xie andYing,1995;Rehfeldtetal.,1999).Highgeneticvariationinthis speciesismaintainedthroughlongdistancedispersal(Critchﬁeld, 1980),highoutcrossingrates(Liewlaksaneeyanawin,2006), and anintoleranceofnaturalself-fertilization(YehandLayton,1979). However,WheelerandGuries(1982b)foundthatthegenetic vari-abilitybetweenpopulationswaslessthanthatcontainedwithin populations.Berland(2013)hypothesizedthatthisgenetic varia-tionwithinpopulationsmayberesponsiblefortherelativestability of reproductive traits—namely, the mean number of seeds per

cone—acrossclimateregimesinBC.Interestingly,ourworkfound asimilarlackofvariabilityinseedyieldbetween18differentseed planningzonesinBC(Figs.2and3),despitetheseregionshaving verydifferentclimates.Inaddition,onlythreeofthe18seed plan-ningzones—EK,SAandTOA—showedasigniﬁcantnon-zerotrend intheirmeanannualseedyieldanomalybasedonconecollections madeoverthepast50years(Figs.5,A.3inAppendixA,Table3). Despitenaturalyear-to-yearvariability,theremainingSPZsseem tobemaintainingarelativelyconstantoverallreproductiveoutput overthepastfewdecades.

(12)

Fig.7.TemporalvariationofmeanannualPinuscontortaseedyieldanomaly(kg/hL)alongwith:a)Fittedlinearregressionof5-yearmeansummersurfacetemperature(ST) anomalyinHudsonHope(HH),andb)Fittedlinearregressionof5-yearmeanwinterSTanomalyinEastKootenay(EK)calculatedfromtheeight-memberensemblemean ofCRCM45-GCMoutput.EnsemblemeanSTlinearregressionshaveunitsof◦_C_year−1_._The_ﬁtted_linear_regression_for_the_mean_annual_seed_yield_anomaly_(kg_hL−1_year−1₎

foreachzoneisindicatedinsolidblack.

Therearesomepracticalimplicationsofourwork,inparticular howtochoosesitesforseedorchardswiththeaimofoptimizing seedyield.Thisisbecauseseedyieldisnotalwaysstable.Studies byBates(1930)andCritchﬁeld(1980)showedthatinWyoming andColoradoseedproductionwashighlyvariablefromoneyearto thenext.Repeatedpoorseedproductioncontinuestoberecorded fromlodgepolepineseedorchardsplantedoutofitsnatural cli-maticrange.InthesepublicandprivateseedorchardsintheNorth OkanaganofBritishColumbia,coneproductionisgood,butseed yieldisverypoor(Anon.,2002;Owensetal.,2005).Inthisregion,

whichispartoftheTODSPZ,theorchardsareplantedoutofthe zoneofoperationalforestry,i.e.atlowerelevationswhere lodge-polepinedoesnotnormallyoccur.Althoughtheremaybemany reasonsforfailure,includingpestsanddiseases,anotablefeature ofthesesitesareextremeJulyandAugusttemperatures.Ourstudy wouldsuggestthatplantingoffutureorchardswithinthenatural rangeoflodgepolepineoughttorestoreseedyieldtolevelsseen insurroundingtrees.Indeed,seedorchardswithinrangehave nor-malseedproduction,e.g.PrinceGeorgeTreeImprovementStation, whichisplantedinthetransitionalzonebetweenthePrinceGeorge

(13)

Table3

Schemesshowingalineartrendlineslope(␤)thatissigniﬁcantlydifferentfromzero(␣=0.05)forboththemeanannualPinuscontortaseedyieldanomaly(kghL−1_year−1₎

ofthecorrespondingseedplanningzone(SPZ)andtheensemblemeanofeithertheCRCM45-GCMprecipitation(PCP,mmyear−1)orsurfacetemperature(ST,◦Cyear−1) anomalies.Inallcases,5ormoreindividualensemblemembersagreewiththeslopeoftheensemblemean.Trailing-meansindicatethenumberofprecedingyears,summers orwintersthatwereaveragedtogetherforeachclimateoutputanalysis.

SPZ ␤SeedYieldAnomaly p-value ClimateVariable

Anomaly

Trailing-mean (years)

␤EnsembleMean p-value Num.ofEnsemble

MembersinAgreement Degreesof Freedom EK 0.002 0.0066 ST Annual 3 0.026 <0.0001 7 22 5 0.025 <0.0001 6 22 10 0.023 <0.0001 7 22 Winter 3 0.040 <0.0001 6 22 5 0.038 <0.0001 5 22 10 0.030 <0.0001 6 22 Summer 3 0.026 <0.0001 5 22 5 0.025 <0.0001 7 22 10 0.023 <0.0001 8 22 TOA 0.004 0.0006 PCP Annual 10 0.092 <0.0001 6 18 ST Annual 3 0.025 <0.0001 5 18 5 0.024 <0.0001 6 18 10 0.023 <0.0001 7 18 Winter 10 0.031 <0.0001 5 18 Summer 3 0.023 0.0001 5 18 5 0.024 <0.0001 5 18 10 0.024 <0.0001 7 18 SA 0.003 0.0074 ST Annual 3 0.021 <0.0001 5 19 5 0.023 <0.0001 6 19

andCentralPlateauSPZ(Anon.,2002).Stableseedproductioncan alsobeaproblem,allowinglodgepolepineplantedinnew envi-ronmentstobecomeinvasive.ExoticplantationsinNewZealand producedabundantseed,which,becauseofitslightweightnature spreadquicklyoverlargeareas.Furthermore,theserotinouscones provedwell-adaptedtoﬁre-regulatednativehabitats,suchas tus-sockgrasscommunities(Richardsonetal.,1994).Thishasledto ecologicalshiftsthatareundesired,butdifﬁculttopreventonce thetreeisestablished.

InCanada,lodgepolepineisundergoingrangeexpansionatits northerndistributionlimitswithnoevidenceofstrongclimatic restrictionsonpopulationgrowth,indicatingthatthespecieshas yettoreachequilibriumwithcurrentclimaticconditionsinthis region(JohnstoneandChapin,2003).Theseﬁndingssuggestthat lodgepolepine mayhaveaneven broaderecologicalamplitude thanpreviouslyrecognized.

4. Conclusions

WefoundthatlodgepolepineinNSThadasigniﬁcantlyhigher meanseedyieldcomparedtoallotherzonesinBritishColumbia, withtheexceptionofHH.ThevarianceinseedyieldforNSTwas anorderofmagnitudehigherthanthatofotherSPZs,indicating thatseedproductionisexceptionallyvariableinthisregion.Inthe future,transitionalecosystemssuchasNSTpresentaunique oppor-tunitytoinvestigatehowenvironmentalvariabilityandselection pressuremayleadtoadaptiveevolution.

Analyses conducted using reanalysis and GCM-driven CRCM outputindependentlyfoundsigniﬁcantcorrelationsbetween cli-mateschemesandmeanannualseedyield,butindifferentSPZs. Thesedifferencesmaybeattributedtoambiguousagerangesfor eachconecollectionandtemporalrestrictionsoftheseedcollection dataandreanalysis-drivenclimateoutput.

WiththeexceptionofNST,seedproductionwasfoundtobe rel-ativelystableacrosstheclimategradientsrepresentedbytheSPZs. Onlythreeofthe18SPZsshowedasigniﬁcantlynon-zerotrend intheirmeanannualseedyieldbasedonconecollectionsmade overthepast50years—namely,EK,SAand TOA.The extraordi-naryadaptabilityoflodgepolepineandthehighgeneticvariationin naturalpopulationsmaybecontributingtotheconstancyofthese reproductivecharacteristics,therebyconcealinganypotential rela-tionshipswithclimateandpossiblyallowingthisspeciestotolerate decadal-scalechangesinclimatechange.

Acknowledgements

DaveKolotelooftheBCMinistryofForests,LandsandNatural ResourceOperationsisgratefullyacknowledgedforprovidingthe archivedconecollectiondatathatmadethisworkpossible.We sin-cerelythankGregO’Neill,MichaelCarlsonandNicholasUkrainetz fortheirpatientandtremendouslyconstructivediscussionsofthis work,EdWiebefortechnicalsupport,andCaitieFrenkelfor edi-torialassistance.WewouldalsoliketothankSusanZedel,Jared Lew,SpencerReitenbach,MarkusSchnorbusandLeslieMcAuleyfor theirassistancewiththeacquisitionandre-projectionofthespatial shapeﬁles.CRCMsimulationswereconductedbyMichelGiguère atOuranosConsortiumandattheUniversityofVictoriaaspartof theNaturalSciencesandEngineeringResearchCouncilofCanada’s (NSERC)−CRD project, DynamicalDownscalingof Westernand EasternCanadianHydroclimate,withinfrastructurefundingfrom theCanadianFoundationforInnovationandtheBritishColumbia KnowledgeDevelopmentFund.Theauthorsaregratefulforfunding obtainedthroughtheNSERC’sCREATEPrograminInterdisciplinary ClimateScience,DiscoveryGrantProgramandPGSProgram.

(14)

Fig.A.1.Boxplotsofnaturalstandlodgepolepine(PinuscontortaDouglasexLouden)seedyield(kgseed/hLcone)for18seedplanningzonesinBritishColumbia,Canada, organizedbyincreasingmedianvalue.Outliers(opencircles)fallmorethan1.5timestheinterquartilerangeawayfromthe1stand3rdquartiles.

(15)

Fig.A.2.Temporalcorrelationsforprecipitation.Spearman’srankcorrelation(rs)betweenthemeanannualPinuscontortaseedyieldanomaly(kg/hL)ofthecorresponding

seedplanningzone(SPZ)anddifferentclimateschemesusingoutputfromCRCM45-NREA:Meanannual(a),winter(b)andsummer(c)precipitationwitha3-year trailing-mean;meanannual(d),winter(e)andsummer(f)precipitationwitha5-yeartrailing-mean;andmeanannual(g),winter(h)andsummer(i)precipitationwitha10-year trailing-mean.Localsignificance(␣=0.05)isidentifiedbyhatching,andredbordersaroundapanelindicateglobalfieldsignificance(␣global=0.05).SPZwithn<10discrete

yearsofseedyielddatafortheafﬁliatedclimateschemewereexcluded.(Forinterpretationofthereferencestocolourinthisﬁgurelegend,thereaderisreferredtotheweb versionofthisarticle.)

(16)

Fig.A.3.Signsofsigniﬁcant(␣=0.05)lineartrendlineslopes(␤)forensemblemeanprecipitationanomaliesofthecorrespondingseedplanningzone(SPZ)underdifferent climateschemesusingCRCM45-GCMoutput:meanannual(a),winter(b)andsummer(c)precipitationwitha3-yeartrailing-mean;meanannual(d),winter(e)andsummer (f)precipitationwitha5-yeartrailing-mean;andmeanannual(g),winter(h)andsummer(i)precipitationwitha10-yeartrailing-mean.SPZsshowingasigniﬁcantpositive slopeforthemeanannualPinuscontortaseedyieldanomaly(kghL−1_year−1₎_are_indicated_by_hatching._White_SPZs_indicate_that_the_slope_of_the_ensemble_mean_{precipitation}

anomaly(mmyear−1₎_was_not_{signiﬁcantly}_different_from_zero_(p_>_0.05)._Dark_red_or_dark_blue_coloration_identiﬁes_SPZs_where_a_strict_majority_of₅_or_more_individual

ensemblemembersagreewiththesignificantslopeoftheensemblemean.SPZswithn<10discreteyearsofseedyielddatafortheaffiliatedclimateschemewereexcluded. (Forinterpretationofthereferencestocolourinthisfigurelegend,thereaderisreferredtothewebversionofthisarticle.)

References

Adams,J.M.,2007.Vegetation-climateInteraction:HowVegetationMakesthe

GlobalEnvironment.Springer,NewYork,NewYork.

Anon,1987.SeedPlanningZonesandTransferGuidelinesforInteriorSpruceand

LodgepolePine.BCMinistryofForests,SilvicultureBranch,Victoria,BC,

MiscellaneousReportMr006.

Anon,2002.FinalReportoftheLodgepolePineSeedSetTaskGroup.Interior

TechnicalAdvisoryCommittee,ForestGeneticsCouncilofBritishColumbia,

http://www.fgcouncil.bc.ca/PliSeedSetRept.pdf(Accessed09July2015).

Anon,2010.TreeSeedProcessingandTestingResultSummariesforFall

2008-spring2010.BCMinistryofForests,TreeImprovementBranch,Victoria,

BC,FileNo.18120-20.http://www.for.gov.bc.ca/hti/treeseedcentre/tsc/ﬁles/

ProtestCoverLetterandAttachmentsTextAugust2010.pdf(Accessed20

August2014).

Anon,2015.2013/14ForestGeneticsCouncilofBCTreeImprovementProject

Report.ForestGeneticsCouncilofBritishColumbia,Victoria,BC,

http://www.fgcouncil.bc.ca/TIPR2013-14April2015Opt.pdf(Accessed8

November2016).

Arno,S.F.,Hammerly,R.P.,1978.NorthwestTrees:IdentifyingandUnderstanding

theRegion’sNativeTrees.MountaineersBooks,Seattle,Washington.

Bates,C.G.,1930.TheProductionExtractionGerminationofLodgepolePineSeed.

USDATechBullTB191,Washington,D.C.

Bennuah,S.Y.,Wang,T.,Aitken,S.N.,2004.GeneticanalysisofthePicea

sitchensis×glaucaintrogressionzoneinBritishColumbia.For.Ecol.Manage.

197,65–77.

Bentz,B.J.,Régnière,J.,Fettig,C.J.,Hansen,E.M.,Hayes,J.L.,Hicke,J.A.,Kelsey,R.G., Negrón,J.F.,Seybold,S.J.,2010.Climatechangeandbarkbeetlesofthewestern

UnitedStatesandCanada:directandindirecteffects.BioSci60,602–613.

Berland,A.,2013.VariationinReproductiveCharacteristicsofLodgepolePine

(PinuscontortaVar.latifolia)inBritishColumbia.UniversityofVictoria,

Victoria,BC(Dissertation).

Case,M.J.,Peterson,D.L.,2007.NorthCascadesNationalPark,Washington.

NorthwestSci.81,62–75.

Critchﬁeld,W.B.,1980.GeneticsofLodgepolePine.USDAForestService,

(17)

Daly,C.,Gibson,W.P.,Taylor,G.H.,Johnson,G.L.,Pasteris,P.,2002.A

knowledge-basedapproachtothestatisticalmappingofclimate.Clim.Res.22,

99–113.

Eis,S.,1973.ConeproductionofDouglas-ﬁrandgrandﬁranditsclimatic

requirements.Can.J.For.Res.3,61–70.

Elfving,B.,Ericsson,T.,Rosvall,O.,2001.Theintroductionoflodgepolepinefor

woodproductioninSweden—areview.For.Ecol.Manage.141,15–29.

Elliott,P.F.,1974.Evolutionaryresponsesofplantstoseed-eaters:pinesquirrel

predationonlodgepolepine.Evolution28,221–231.

Fan,S.,Grossnickle,S.C.,Sutton,B.C.,1997.Relationshipsbetweengasexchange

adaptationofSitka×interiorsprucegenotypesandribosomalDNAmarkers.

TreePhysiol.17,115–123.

Flannigan,M.D.,Logan,K.A.,Amiro,B.D.,Skinner,W.R.,Stocks,B.J.,2005.Future

areaburnedinCanada.Clim.Change72,1–16.

Forrest,G.I.,1980.GeographicalvariationinthemonoterpenesofPinuscontorta

oleoresin.Biochem.Syst.Ecol.8,343–359.

Fowells,H.A.,1965.SilvicsofForestTreesoftheUnitedStates.USDAForest

Service,Washington,DC.

Fowler,H.J.,Blenkinsop,S.,Tebaldi,C.,2007.Linkingclimatechangemodellingto

impactsstudies:recentadvancesindownscalingtechniquesforhydrological

modeling.Int.J.Climatol.27,1547–1578.

Hadley,M.J.,Tanz,J.S.,Fraser,J.,2001.Biotechnology:PotentialApplicationsin

TreeImprovement.BCForestGeneticsCouncil,Victoria,BC(ExtensionNote2).

Hamilton,J.A.,Aitken,S.N.,2013.Geneticandmorphologicalstructureofaspruce

hybrid(Piceasitchensis×P.glauca)zonealongaclimaticgradient.Am.J.Bot.

100,1651–1662.

Hamilton,J.A.,Lexer,C.,Aitken,S.N.,2013.Genomicandphenotypicarchitectureof

asprucehybridzone(Piceasitchensis×P.glauca).Mol.Ecol.22,827–841.

Herms,D.A.,Mattson,W.J.,1992.Thedilemmaofplants—togrowordefend.Q.

Rev.Biol.67,283–335.

Herrera,C.M.,Jordano,P.,Guitián,J.,Traveset,A.,1998.Annualvariabilityinseed productionbywoodyplantsandthemastingconcept:reassessmentof principlesandrelationshiptopollinationandseeddispersal.Am.Naturalist 152,576–594,http://dx.doi.org/10.1086/286191469.

Illingworth,K.,1978.Studyoflodgepolepinegenotype-environmentinteractionin

BC.In:ProceedingsInternationalUnionofForestryResearchOrganizations

(IUFRO)JointMeetingofWorkingParties:Douglas-ﬁrProvenances,Lodgepole

PineProvenances,SitkaSpruceProvenancesandAbiesProvenances,

Vancouver,BC,pp.151–158.

Johnstone,J.F.,Chapin,F.S.,2003.Non-equilibriumsuccessiondynamicsindicate

continuednorthernmigrationoflodgepolepine.GlobalChangeBiol.9,

1401–1409.

Kalnay,E.,Kanamitsu,M.,Kistler,R.,Collins,W.,Deaven,D.,Gandin,L.,etal.,1996.

TheNCEP/NCAR40-yearreanalysisproject.Bull.Am.Meteorol.Soc.77,

437–471.

Kendon,E.J.,Jones,R.G.,Kjellström,E.,Murphy,J.M.,2010.Usinganddesigning

GCM-RCMensembleregionalclimateprojections.J.Clim.23,6485–6503.

King,J.N.,Alfaro,R.I.,2009.DevelopingSitkaSprucePopulationsforResistanceto

theWhitePineWeevil:SummaryofResearchandBreedingProgram.BC

MinistryofForests,ResearchBranch,Victoria,BC(TechRep050).

Koch,P.,1996.LodgepolePineinNorthAmerica.ForestProductsSociety,Madison, Wisconsin.

Koenig,W.D.,Knops,J.M.,2000.Patternsofannualseedproductionbynorthern

hemispheretrees:aglobalperspective.Am.Nat.155,59–69.

Kolotelo,D.,VanSteenis,E.,Peterson,M.,Bennett,R.,Trotter,D.,Dennis,J.J.,2001.

SeedHandlingGuidebook.BCMinistryofForests,TreeImprovementBranch,

Victoria,BC,http://www.for.gov.bc.ca/hti/publications/misc/seedhandling

guidebookhi.pdf(Accessed25August2014).

Kral,R.,1993.Pinus.FloraofNorthAmerica,vol.2.OxfordUniversityPress,New

York,NewYork,pp.373–398.

Lavender,D.P.,Parish,R.,Johnson,C.M.,Montgomery,G.,Vyse,A.H.,Willis,R.A., Winston,D.,1990.RegeneratingBritishColumbia’sForests.BCMinistryof

Forests,ResearchBranch,Victoria,BC(Misc.Rep.Mr063).

Leech,S.M.,LaraAlmuedo,P.,O’Neill,G.,2011.Assistedmigration:adaptingforest

managementtoachangingclimate.J.Ecosyst.Manage.12,18–34.

Lertzman,K.P.,Gavin,D.G.,Hallett,D.J.,Brubaker,L.B.,Lepofsky,D.S.,Mathewes, R.W.,2002.Long-termﬁreregimefromsoilcharcoalincoastaltemperate

rainforests.Conserv.Ecol.6,5.

Liewlaksaneeyanawin,C.,2006.GeneticEvaluationofNaturalandDomesticated

LodgepolePinePopulationsUsingMolecularMarkers.UniversityofBritish

Columbia(Dissertation).

Livezey,R.E.,Chen,W.Y.,1983.Statisticalﬁeldsigniﬁcanceanditsdetermination

byMonteCarlotechniques.Mon.WeatherRev.111,46–59.

Lotan,J.E.,1970.ConeSerotinyinPinuscontorta.UniversityofMichigan (Dissertation).

Mátyás,C.,1996.Climaticadaptationoftrees:rediscoveringprovenancetests.

Euphytica92,45–54.

McFarlane,N.A.,Scinocca,J.F.,Lazare,M.,Harvey,R.,Verseghy,D.,Li,J.,2005.The CCCmathirdgenerationatmosphericgeneralcirculationmodel.Rapport interneduCentreCanadiendelaModélisationetdel’AnalyseClimatique,pp.

25.http://www.cccma.ec.gc.ca/papers/jscinocca/AGCM3report.pdf(Accessed

27April2015).

McLane,S.C.,Daniels,L.D.,Aitken,S.N.,2011.Climateimpactsonlodgepolepine

(Pinuscontorta)radialgrowthinaprovenanceexperiment.For.Ecol.Manage.

262,115–123.

Mearns,L.O.,Giorgi,F.,McDaniel,L.,Shields,C.,2003.Climatescenariosforthe

southeasternU.S.basedonGCMandregionalmodelsimulations.Clim.Change

6,7–35.

Meidinger,D.,Pojar,J.,1991.EcosystemsofBritishColumbia.BCMinistryof

Forests,ResearchBranch,Victoria,BC(SpecialReportSeriesSrs06).

Muir,P.S.,1993.Disturbanceeffectsonstructureandtreespeciescompositionof

PinuscontortaforestsinwesternMontana.Can.J.For.Res.23,1617–1625.

Norton,D.A.,Kelly,D.,1988.Mastseedingover33yearsbyDacrydiumcupressinum

Lamb.(rimu)(Podocarpaceae)inNewZealand:theimportanceofeconomies

ofscale.Funct.Ecol.2,399–408.

O’Neill,G.A.,Hamann,A.,Wang,T.,2008a.Accountingforpopulationvariation

improvesestimatesoftheimpactofclimatechangeonspecies’growthand

distribution.J.Appl.Ecol.45,1040–1049.

O’Neill,G.A.,Ukrainetz,N.K.,Carlson,M.,Cartwright,C.V.,Jaquish,B.C.,King,J.N., Krakowski,J.,Russell,J.H.,Stoehr,M.U.,Xie,C.-Y.,etal.,2008b.Assisted

MigrationtoAddressClimateChangeinBritishColumbia−Recommendations

forInterimSeedTransferStandards.BCMinistryofForests,ForestScience

Program,Victoria,BC(TechRep048).

Owens,J.N.,Bennett,J.,L’Hirondelle,S.,2005.Pollinationandconemorphology

affectconeandseedproductioninlodgepolepineseedorchards.Can.J.For.

Res.35,383–400.

Perry,D.A.,Lotan,J.E.,1977.OpeningTemperaturesinSerotinousConesof

LodgepolePine.USDAFor.Serv.Res(NoteINT-228).

Pierce,D.W.,Barnett,T.P.,Santer,B.D.,Gleckler,P.J.,2009.Selectingglobalclimate

modelsforregionalclimatechangestudies.PNAS106,8441–8446.

RCoreTeam,2015.R:ALanguageandEnvironmentforStatisticalComputing.R

FoundationforStatisticalComputing,Vienna,Austria

(http://www.R-project.org/).

Reeb,J.,Shaw,D.,2015.CommonInsectPestsandDiseasesofSitkaSpruceonthe

OregonCoast.OregonStateUniversityExtensionService(EM9105).

Rehfeldt,G.E.,Ying,C.C.,Spittlehouse,D.L.,HamiltonJr.,D.A.,1999.Genetic

responsestoclimateinPinuscontorta:nichebreadth,climatechange,and

reforestation.Ecol.Monogr.69,375–407.

Richardson,D.M.,Williams,P.A.,Hobbs,R.J.,1994.Pineinvasionsinthesouthern

hemisphere:determinantsofspreadandinvadability.J.Biogeogr.21,511–527.

Richardson,D.M.,2000.EcologyandBiogeographyofPinus.CambridgeUniversity

Press,NewYork,NY.

Roche,L.,1969.AgenecologicalstudyofthegenusPiceainBritishColumbia.New

Phytol.68,505–554.

Roeckner,E.,Bäuml,G.,Bonaventura,L.,Brokopf,R.,Esch,M.,Giorgetta,M., Hagemann,S.,Kirchner,I.,Kornblueh,L.,Manzini,E.,Rhodin,A.,Schlese,U., Schulzweida,U.,Tompkins,A.,2003.TheAtmosphericGeneralCirculation

ModelECHAM5.PartI:ModelDescription,vol.349.MaxPlanckInstitutefor

MeteorologyRep,pp.127.

Russell,J.H.,Krakowski,J.,2012.Geographicvariationandadaptationtocurrent

andfutureclimatesofCallitropsisnootkatensispopulations.Can.J.For.Res.42,

2118–2129.

Sork,V.L.,Bramble,J.,Sexton,O.,1993.Ecologyofmast-fruitinginthreespeciesof

NorthAmericandeciduousoaks.Ecology74,528–541.

Sushama,L.,Khaliq,M.N.,Laprise,R.,2010.DryspellcharacteristicsoverCanadain

achangingclimateassimulatedbytheCanadianRCM.GlobalPlanet.Change

74,1–14.

Tebaldi,C.,Arblaster,J.M.,Knutti,R.,2011.Mappingmodelagreementonfuture

climateprojections.Geophys.Res.Lett.38,L23701.

Tinker,D.B.,Romme,W.H.,Hargrove,W.W.,Gardner,R.H.,Turner,M.G.,1994.

Landscape-scaleheterogeneityinlodgepolepineserotiny.Can.J.For.Res.24,

897–903.

Voltas,J.,Chambel,M.R.,Prada,M.A.,Ferrio,J.P.,2008.Climate-relatedvariability

incarbonandoxygenstableisotopesamongpopulationsofAleppopinegrown

incommon-gardentests.Trees22,759–769.

Wallis,C.M.,Huber,D.P.,Lewis,K.J.,2011.Ecosystem,location,andclimateeffects

onfoliarsecondarymetabolitesoflodgepolepinepopulationsfromcentral

BritishColumbia.J.Chem.Ecol.37,607–621.

Wang,G.,Huang,S.,Monserud,R.A.,Klos,R.J.,2004.Lodgepolepinesiteindexin

relationtosynopticmeasuresofclimate,soilmoistureandsoilnutrients.For.

Chron.80,678–686.

Wang,T.,Hamann,A.,Yanchuk,A.,O’Neill,G.A.,Aitken,S.N.,2006.Useofresponse

functionsinselectinglodgepolepinepopulationsforfutureclimates.Global

ChangeBiol.12,2404–2416.

Werner,A.T.,2011.BCSDDownscaledTransientClimateProjectionsforEight

SelectGCMsoverBritishColumbia,Canada.PaciﬁcClimateImpacts

ConsortiumRep.UniversityofVictoria,Victoria,BC(pp63).

Wheeler,N.C.,Critchﬁeld,W.B.,1985.Thedistributionandbotanical

characteristicsoflodgepolepine.In:Baumgartner,etal.(Eds.),Symposium

Proceedings,LodgepolePine:theSpeciesandItsManagement.Spokane,WA

andVancouver,BC,8–16May1984.WashingtonStateUniversity,Cooperative

ExtensionService,Pullman,Washington,pp.1–14.

Wheeler,N.C.,Guries,R.P.,1982a.Biogeographyoflodgepolepine.Can.J.Bot.60, 1805–1814.

Wheeler,N.C.,Guries,R.P.,1982b.Populationstructure,genicdiversity,and

morphologicalvariationinPinuscontortaDougl.Can.J.For.Res.12,595–606.

Wheeler,N.C.,Guries,R.P.,1987.Aquantitativemeasureofintrogressionbetween

lodgepoleandjackpines.Can.J.Bot.65,1876–1885.

Woods,A.,Coates,K.D.,Hamann,A.,2005.IsanunprecedentedDothistroma

(18)

Woodward,A.,Silsbee,D.G.,Schreiner,E.G.,Means,J.E.,1994.Inﬂuenceofclimate

onradialgrowthandconeproductioninsubalpineﬁr(Abieslasiocarpa)and

mountainhemlock(Tsugamertensiana).Can.J.For.Res.24,1133–1143.

Wu,H.X.,Ying,C.C.,Ju,H.,2005.Predictingsiteproductivityandpesthazardin

lodgepolepineusingbiogeoclimaticsystemandgeographicvariablesinBritish

Columbia.Ann.For.Sci.62,31–42.

Xie,C.Y.,Ying,C.C.,1995.Geneticarchitectureandadaptivelandscapeofinterior

lodgepolepine(Pinuscontortassp.latifolia)inCanada.Can.J.For.Res.25,

2010–2021.

Yeh,F.C.,Layton,C.,1979.Theorganizationofgeneticvariabilityincentraland

marginalpopulationsoflodgepolepine(Pinuscontortassp.latifolia).Can.J.

Genet.Cytol.21,487–503.

Ying,C.C.,Yanchuk,A.D.,2006.ThedevelopmentofBritishColumbia’streeseed

transferguidelines:purpose,concept,methodology,andimplementation.For.

Ecol.Manage.227,1–13.

Ying,C.C.,1991.GeneticResistancetotheWhitePineWeevilinSitkaSpruce.BC