Electronic Supplementary Material Spatio-temporal … ·  · 2017-01-23Spatio-temporal environmental variation mediates geographic differences in ... (~ 80 animals per site) ...

1

ElectronicSupplementaryMaterial

Spatio-temporalenvironmentalvariationmediatesgeographicdifferencesin

phenotypicresponsestooceanacidification

JuanDiegoGaitán-Espitia1,2*,PaolaVillanueva3,JorgeLopez3,RodrigoTorres4,Jorge

Navarro3&LeonardoBacigalupe2

1CSIROOceans&Atmosphere,GPOBox1538,Hobart7001,TAS,Australia

2InstitutodeCienciasAmbientalesyEvolutivas,UniversidadAustraldeChile,

Casilla567Valdivia,Chile.

4InstitutodeCienciasMarinasyLimnologicas,UniversidadAustraldeChile,Casilla

567Valdivia,Chile.

5CentrodeInvestigaciónenEcosistemasdelaPatagonia,Coyhaique,Chile.

*Correspondingauthor:JuanDiegoGaitán-Espitia.Emailaddress:

[email protected]

2

Legendsofsupplementaryfiles

*Supplementarymethods:

a) Spawningandexperimentalset-up

b) Phenotypicanalyses

c) Metabolicrates

d) Developmentalprogressionandsurvival

*FigureS1.A)PopulationsofLoxechinusalbusdistributedalongalatitudinal

gradientinthePacificcoastofChilecharacterizedbynaturalfluctuationsin

pCO2/pHconditions.Populationsaregroupedinthreemainregionsaccordingtothe

intensityanddurationofpCO2variability.B)Spatialvariabilityinsurfaceseawater

pCO2alongthecoastofChilebetween1995and2009.Dotsrepresenttheair-surface

waterCO2(ΔpCO2),estimatedasthedifferenceintheCO2partialpressurebetween

surfacewaterandair.Colorsrepresentsamplinglocations(Modifiedfrom[1]).C)

Carbonatesystemtimeseriesinintertidalwatersofanupwellingarea(24S)and

non-upwellingarea(46S)ofChile.CO2variabilityishigherwithintheupwelling

area,reachingshortperiodsofpCO2values>1000uatmduringtheAustralspring

andsummer(UnpublisheddataRodrigoTorres).

*FigureS2.Modelfittingofdevelopmentalprogressionuntilthe4-armspluteus

stageinLoxechinusalbus.

*TableS1.Seawatercarbonatechemistry(mean±SE)usedinexperimental

treatmentsofL.albus

3

inexperimental *TableS2.Seawatercarbonatechemistry(mean±SE)used

treatmentsofL.albus

*PhenotypicdatafromdifferentpopulationsofL.albus.

4

SupplementaryMethods

a)Animalcollection,spawningandexperimentalset-up

Adultseaurchinswerecollectedfromninelocalities(~80animalspersite)

alongtheupwellingsystemandfjordsofthePacificcoastofChileduringthe

spawningseason(JulytoSeptember2014)reportedforthisspeciesalongits

geographicdistribution[2].Gameteswereobtainedbystandardmethodswithin2-3

daysaftercollection.Weinducedspawningviacoelomicinjectionof0.5MKCl.The

eggswerecollectedin0.35µmfiltered,UV-sterilizedseawateratambient

conditions(13°Cand390μatmpCO2).Thespermwascollecteddry,andkeptonice

untilfertilization.Fromeachlocality,20independentcrossesofasinglemalewith

pooledeggsofthreefemalesweredevelopedtoavoidmale-femaleincompatibility.

Eggswerefertilizedusingfiltered(0.1μm)seawateratambientconditions(13°C

and390μatmpCO2)inordertoeliminateconfoundingeffectsofhighpCO2on

spermmobility.Embryoswereequallyandrandomlydistributedbetweentwo

replicatebucketsintheexperimentalpCO2treatments(currentglobal:390μatm;

andprojectedOA:1200μatm)withcontrolledtemperature(13-14°C)ata

concentrationofapproximately0.7embryosml-1.Embryosweremaintainedunder

experimentalconditionsuntilcompletetransitiontolarvalstage(4-armspluteus,

~96h).Larvaewerenotfedduringtheexperimentalperiodtoavoidchangesin

developmentandphysiologyinducedbydiet[3],aswellaschangesinexperimental

conditionsproducedbyrespirationandphotosynthesisofmicroalgae.Asamodeof

separatingthermalenvironmentaleffectsfromlocaldifferentiationamong

populations,common-gardenexperimentsarethegold-standarddesign.Weapplied

5

thisapproachwithhomogeneouslaboratoryacclimationtoourninepopulationsof

L.albus.Hence,anyobservedphenotypicdifferencewillbeattributabletothe

geneticsofthesourcepopulations[4].

Seawatercarbonatechemistrywasmaintainedusingasemi-automaticflow-

throughCO2-mixingsystem.Thesystemwasmodifiedbypumpingtheappropriate

CO2–gasmixforthetreatmentdirectlyintotheheadspaceoftheculturevessels,

usingtheA-1experimentaldesign(Sensu[5]).Simultaneouswatersampleswere

collectedtodeterminepHvariationstroughthepotentiometricmethod.pHwas

measuredinaclosed25-mLcellthermo-staticallycontrolledat25.0°Cusinga

Metrohm713pHmeter(inputresistanceN1013Ohm,0.1mVsensitivity,and

nominalresolution0.001pHunits)andaglasscombineddoublejunctionAg/AgCl

electrode(Metrohmmodel6.0219.100)calibratedwith8.089Trisbuffer(25°C).pH

isreportedonthetotalhydrogenionscale.Totalalkalinity(TA)wasmeasured

everythreedayswhereastemperature,pHandsalinityweremonitoredeveryday.

TA,pH,phosphatesanddissolvedsilicateswereusedtocalculatetherestofthe

carbonatesystemparametersandthesaturationstageofOmega,Aragoniteand

CalciteusingCO2calc[6]withthedissociationconstantsofMehrbach,refittedby

DicksonandMillero(1987).

b)Phenotypicanalyses

Morphometrictraitssuchastotallarvallengthandpostoralarmlengthwere

analyzedatthe4-armspluteusstage.Fromeachcrossineachexperimental

condition,20larvaewerecollectedandimmediatelyfixedin2%formalinsaturated

6

withsodiumboratetopreventacidityanddissolutionoftheskeleton.Photographs

formorphometricanalysesweretakenwithintwoweeksoffixation,usingadigital

cameraattachedtoacompoundmicroscopeatamagnificationof×10,withlarvae

orientateddorsalsidedown.Totallarvallengthandpostoralarmlengthwere

estimatedfollowing[7].Morphometricvalueswereestimatedusingtheimage

analysissoftwareImage-ProPlus.DifferencesinTLandPOLwereanalyzedusing

lmerfunctioninthe‘lme4’packageofRv.3.3(RCoreTeam,2016),withCO2

treatmentandregionasfixedfactorsandlocalityasrandomfactor.Significance

testswereperformedwiththe‘lmerTest’package.Post-hoccomparisonsformixed

effectsmodelsweredonewiththe‘multcomp’and‘lsmeans’packages.

c)Metabolicrates

Ratesofoxygenconsumptionwereobtainedinearlypluteuslarvae

accordingto[7].Inbrief,differentdensitiesoflarvae(50–1000individuals)were

placedinrespiratorychambers(1000μlvials)andincubatedfor4–7hinorderto

generateastandardcurvefromwhichwecouldestimatetherateofoxygen

consumptionperindividual(expressedaspmolO2/hr/larva)underexperimental

conditions.Twocontrolvialscontainingonlyfreshseawaterwereincubated

simultaneouslytoaccountforbackgroundrespiration.Afterincubation,larvaewere

removedandO2measuredusinganopticfiberoxygen-meter(MicroxTX3,PreSens,

Germany),withdiametertipsof20–50um.Zerocalibrationwasperformedusinga

Na2O3Ssolution(0%saturation)and100%wascalibratedusingair-bubbled

seawater.Temperaturewasstabilizedusingatemperature-controlledwaterbath

7

(JioTech,Co).Inordertotestforthepotentialeffectoflarvalsizeonrespiration,we

performedalinearregressionanalysisusingtheaveragerespirationrateandthe

averagelarvalsizeforeachcrosswithineachtreatment.Duetothesignificant

correlationbetweenthesetwovariables(r=0.75;P<0.001),residualswereusedto

removetheeffectofsize.Differencesinrespirationrate(aproxyofmetabolicrate)

wereanalyzedusinglmerfunctioninthe‘lme4’packageofRv.3.3(RCoreTeam,

2016),withCO2treatmentandregionasfixedfactorsandlocalityasrandomfactor.

Significancetestswereperformedwiththe‘lmerTest’package.Post-hoc

comparisonsformixedeffectsmodelsweredonewiththe‘multcomp’and‘lsmeans’

packages.

d)Developmentalprogressionandlarvalsurvival

Developmentwastrackedbyrecordingtheproportionofembryosthat

reachedblastula,gastrula,prismandearlypluteusstagesundereachexperimental

condition.Culturesweremonitoredbyremoving50larvaeevery12handscoring

eachlarvalstage.Theexperimentwasterminatedwhen100%ofthecollected

larvaereachedthe4-armspluteusstage.Developmentalprogressioncurveswere

fittedandanalysedwithGraphPadPrismsoftware(GraphPad,SanDiego,CA).

Comparisonsoffits(Gompertz,logistic,exponentialandWeibull)wereperformed

usingAkaike’sinformativecriteria(AICc)andtheExtrasum-of-squaresFtest.Inall

ofthecases,theWeibullgrowthmodelshowedthebestfit.Thismodelisdescribed

bytheequation:𝑌! = 𝐿! − (𝐿! − 𝛽) ∗ 𝑒(!! !∗! !)(−1 ∗ ((𝑘 ∗ 𝑡)^𝑔))

8

WhereYisthelength(orweight,height,size)atatimet,𝛽isthelowerasymptote,L

istheupperasymptote,kisthegrowthrateandgisaparameterthatcontrolsthex-

ordinateforthepointofinflectioninthexaxis.Comparisonsamongpopulations

andexperimentaltreatmentswerebasedontheparameterkasaproxyoftheshape

ofthedevelopmentalprogression.

Larvalsurvivalwasestimateddailybyaspiringthebottomofthe

experimentalbuckets.Differencesinsurvivalat4-armspluteusstagewereanalyzed

usingglmerfunctionwithbinomialdistributioninthe‘lme4’packageofRv.3.3(R

CoreTeam,2016),withCO2treatmentandregionasfixedfactorsandlocalityas

randomfactor.Significancetestswereperformedwiththe‘lmerTest’package.Post-

hoccomparisonsformixedeffectsmodelsweredonewiththe‘multcomp’and

‘lsmeans’packages.

Seasonal CO2-supersaturated CO2-undersaturated

pH@

26C

Salin

ityTe

mpe

ratu

re (C

)pC

O2 (u

atm

) 800

600

400

200

0

Om

ega

Ara

goni

te

5

4

3

2

1 0

8.4

8.2

8

7.8

7.6

7.4

30

20

10

0

Nitr

ate

Silic

ate

50

40

30

20

10

0

16

14

12

10 34

32

30

28

pH@

26C

Tem

pera

ture

(C)

pCO

2 (u

atm

) 800

600

400

200

0

Om

ega

Ara

goni

te

5

4

3

2

1 0

8.4

8.2

8

7.8

7.6

7.4

Nitr

ate

Silic

ate

36

32 28 24

20

16

25

20

15

10

5

0

36

32

28

24

20

16

20

18

12

8

100 200 300 400 500 600 700

Fall$2011$

Winter$2011$

Spring$2011$

Summer$2012$

Fall$2012$

Winter$2012$

Spring$2012$

Summer$2013$

Spring'2013'

Summer'2013'

Fall'2014'

Winter'2014'

Spring'2014'

Summer'2014'

Fall'2015'

Winter'2015'

100 200 300 400 500 600 700

Lat

itude

(°S)

-18

-20

-22

-24

-26

-28

-30

-32 -34 -36 -38

-40

-42

-44

-46

-48

-50

-52

-54

-56

-200 0 200 400 >600 ΔpCO2(µatm)

Iquique

Antofagasta

San Antonio

Lenga

Los Molinos

Ancud

Arica

Puerto Aguirre

Punta Arenas

41°52’

45°15’

53°10’

Seas

onal

CO

2-su

pers

atur

ated

A

lmos

t yea

r-ro

und

C

O2-

supe

rsat

urat

ed

CO

2-un

ders

atur

ated

39°84’

36°45’

33°59’

23°46’

20°19’

18°30’

A B C

Days Days

0 50 1000

20

40

60

80

100

Hours after fertilization

Plut

eus

(%) Arica low pCO2

Iquique low pCO2

Arica high pCO2

Iquique high pCO2

Antofagasta high pCO2

St. Antonio high pCO2

Lenga high pCO2

Molinos high pCO2

Ancud high pCO2

P. Aguirre high pCO2

P. Arenas high pCO2

Antofagasta high pCO2

St. Antonio low pCO2

P. Aguirre low pCO2

Ancud low pCO2

Lenga low pCO2

Molinos low pCO2

P. Arenas low pCO2

TableS1.Seawatercarbonatechemistry(mean±SE)usedinexperimental

treatmentsofL.albus

Parameters ExperimentalCO2treatments

Low High

pCO2(μatm) 403.5±36.2 1175.2±53.8

pH 8.02±0.03 7.64±0.04

Temperature(°C) 13.4±0.6 13.6±0.7

Salinity(psu) 33.6±0.2 33.58±0.2

Totalalkalinity(μmolkg−1) 2156.4±6.8 2160.5±5.3

ΩCa 3.13±0.2 1.59±0.08

ΩAr 2.01±0.1 1.02±0.06

**ParametersofpCO2,ΩArandΩCawerecalculatedusingCO2calc[6]withthe

dissociationconstantsofMehrbach,refittedbyDicksonandMillero(1987).

References1. Torres,R.etal.2011Air-seaCO2fluxesalongthecoastofChile:FromCO

2outgassingincentralnorthernupwellingwaterstoCO2uptakein

southernPatagonianfjords.J.Geophys.Res.116,C09006.

(doi:10.1029/2010JC006344)

2. Vásquez,J.2007EcologyofLoxechinusalbus.InEdibleSeaUrchins:

BiologyandEcology(edJ.M.Lawrence),pp.227–241.Elsevier.

(doi:http://dx.doi.org/10.1016/S0167-9309(01)80012-0)

3. Stumpp,M.,Wren,J.,Melzner,F.,Thorndyke,M.&Dupont,S.2011CO2

inducedseawateracidificationimpactsseaurchinlarvaldevelopmentI:

elevatedmetabolicratesdecreasescopeforgrowthandinduce

developmentaldelay.Comp.Biochem.Physiol.PartAMol.Integr.Physiol.

160,331–40.(doi:10.1016/j.cbpa.2011.06.022)

4. Lindgren,B.&Laurila,A.2009Physiologicalvariationalongageographical

gradient:isgrowthratecorrelatedwithroutinemetabolicrateinRana

temporariatadpoles?Biol.J.Linn.Soc.98,217–224.(doi:10.1111/j.1095-

8312.2009.01255.x)

5. Cornwall,C.E.&Hurd,C.L.2015Experimentaldesigninocean

acidificationresearch:problemsandsolutions.ICESJ.Mar.Sci.2015,177–

202.

6. Robbins,L.,Hansen,M.,Kleypas,J.&Meylan,S.2010CO2calc—auser-

friendlyseawatercarboncalculatorforWindows,MaxOSX,andiOS

(iPhone).

7. Padilla-Gamiño,J.,Kelly,M.,Evans,T.&Hofmann,G.2013Temperature

andCO2additivelyregulatephysiology,morphologyandgenomic

responsesoflarvalseaurchins,Strongylocentrotuspurpuratus.Proc.R.

Soc.BBiol.Sci.280,20130155.