Spatial and temporal patterns of pesticide use on California almonds and associated risks to the ...

13 Pages·2013·4.6 MB·English

by  Yu Zhan

ScienceoftheTotalEnvironment472(2014)517–529
ContentslistsavailableatScienceDirect
Science of the Total Environment
journal homepage: www.elsevier.com/locate/scitotenv
Spatial and temporal patterns of pesticide use on California almonds and
associated risks to the surrounding environment
⁎
YuZhan,MinghuaZhang
DepartmentofLand,Air,andWaterResources,UniversityofCalifornia,Davis,CA95616,USA
H I G H L I G H T S
•Saptiotemporalpatternsofpesticideuse/riskinCaliforniaalmondswerestudied.
•Useintensitiesofinsecticides/fungicides/herbicidesshowedlatitudinalgradients.
•Overall,herbicideuseincreasedconsiderably,whilefungicideusedecreased.
•Theriskstosurfacewater,groundwater,andsoildecreasedinmanyareas.
•Riskpatternsweremainlyassociatedwithusepatternsofhigh-riskpesticides.
a r t i c l e i n f o a b s t r a c t
Articlehistory: VariousstakeholdersofCaliforniaalmondshavebeeninvestingeffortsintomitigatingpesticideimpactson
Received14September2013 humanandecosystemhealth.Thisstudyisthefirstcomprehensiveevaluationthatexaminesthespatialandtem-
Receivedinrevisedform2November2013 poralpatternsofpesticideuseandassociatedenvironmentalrisks.Thepesticideusedatafrom1996to2010were
Accepted3November2013
obtainedfromthePesticideUseReportingdatabase.ThePesticideUseRiskEvaluationindicatorwasemployedto
Availableonlinexxxx
evaluatethepesticideenvironmentalrisksbasedonthepesticidepropertiesandlocalenvironmentalconditions.
Analysesshowedthattheuseintensities(UI)ofinsecticides(oilsaccountedfor86%ofthetotalinsecticideUI)
Keywords:
andherbicidesbothincreasedfromnorthtosouth;fungicidesshowedtheoppositespatialpattern;andfumi-
Pesticide
Environmentalrisk gantswereusedmostintensivelyinthemiddleregion.TheUIoffungicidesandherbicidessignificantlydecreased
Pestmanagement andincreased,respectively,throughoutthestudyarea.TheinsecticideUIsignificantlydecreasedinthenorthbut
Regulation increasedinmanyareasinthesouth.Inparticular,theorganophosphateUIsignificantlydecreasedacrossthe
Almond studyarea,whilethepyrethroidUIsignificantlyincreasedinthesouth.ThefumigantUIdidnotshowatrend.
California Theregionalriskintensitiesofsurfacewater(RI ),soil(RI),andair(RI )allincreasedfromnorthtosouth,
W S A
whilethegroundwaterregionalriskintensity(RI )decreasedfromnorthtosouth.ThemaintrendsofRI ,RI ,
G W G
andRI weredecreasing,whiletheRI didnotshowatrendinanyregion.It'snoticeablethatalthoughtheher-
S A
bicideUIsignificantlyincreased,theUIofhigh-leachingherbicidessignificantlydecreased,whichledtothesig-
nificantdecreaseofRI .Insummary,thetemporaltrendsofthepesticideuseandrisksindicatethattheCalifornia
G
almondgrowersaremakingconsiderableprogresstowardssustainablepestmanagementviaintegratedpest
management,butstillrequiremoreeffortstocurbthefastincreaseofherbicideuse.
©2013ElsevierB.V.Allrightsreserved.
1.Introduction orangeworm (Amyelois transitella), San Jose scale (Quadraspidiotus
perniciosus),peachtwigborer(Anarsialineatella),web-spinningspider
AlmondsareoneofthemostimportantspecialtycropsinCalifornia, mites,andants(CEPA,2011).Inthedormantseason,oilsprayalone
USA,whichproducedabout80%oftheglobalalmondsupplyandgener- cancontrollowtomoderatepopulationsofSanJosescaleandmites.
ated$3.87billioninrevenuein2012(AlmondBoardofCalifornia,2012). Whenpopulationsofpeachtwigborer(alsotargetedduringbloom)
AlmostalltheCaliforniaalmondorchards(3080km2in2012)arelocated andSanJosescalearehigh,oilsarelikelysprayedwithotherinsecticides.
intheCentralValley(58,000km2),whichhasamildclimate,fertilesoil, Inthegrowingseason,insecticidetreatments(mainlyinJulyandAugust)
andabundantsunshine.TheCentralValleyisoneofthemostproductive mostlycontrolnavelorangeworm.Diseasesduringwintersandearly
agricultural areas in the world. Key pests in almond are navel springs, such as anthracnose (pathogen: Colletotrichum acutatum),
brown rot blossom blight (pathogen: Monilinia laxa), and scab
(pathogen:Cladosporiumcarpophilum)arecontrolledbyvariousfungi-
⁎ Correspondingauthor at. DepartmentofLand, Air,andWaterResources, 131
VeihmeyerHall,UniversityofCalifornia,Davis,CA95616,USA.Tel.:[phone]; cides,e.g.,captan,copper,orziram(UCIPM,2012).Weeds,suchas
fax:[phone]. bermudagrass(Cynodondactylon),dallisgrass(Paspalumdilatatum),and
0048-9697/$–seefrontmatter©2013ElsevierB.V.Allrightsreserved.
http://dx.doi.org/10.1016/j.scitotenv.2013.11.022
518 Y.Zhan,M.Zhang/ScienceoftheTotalEnvironment472(2014)517–529
hairyfleabane(Conyzabonariensis),aretreatedwithpre-emergenceor resultsandconclusionsareexpectedtoreflecttheoutcomeofCalifornia
post-emergenceherbicides.Tominimizetheyieldlosscausedbypests, almondstakeholders'effortstowardssustainablepestmanagementand
pathogens,andweeds,Californiaalmondgrowersapplyalargeamount toprovidesuggestionsforprioritizingpestmanagementpractices.
ofpesticides;9.3milliontonsofpesticideactiveingredientswereapplied
in2010(CEPA,2012).However,theappliedpesticidesthreatentheenvi- 2.Materialsandmethods
ronmentandhumanhealth,asevidencedbypesticidedetectionsin
groundwater(Kolpinetal.,2000)andsurfacewater(Guoetal.,2007; 2.1.Studyarea
Hladiketal.,2009).
Variousstakeholdershavemadeeffortstoreduceoreliminatetheir TheCentralValley,wherealmostallofthealmondsinCalifornia
usesofthepesticidesthatareknowntoharmhumanhealthordegrade werecultivated,wasselectedasthestudyarea(Fig.1a).Thestudy
environmentalquality.TheUnitedStatesEnvironmentalProtection
Agency(USEPA)regulatespesticideuseundertheFederalInsecticide,
Fungicide,andRodenticideAct(FIFRA)andtheFederalFood,Drug,
and Cosmetic Act (FFDCA). Both of these acts were significantly
amendedbytheFoodQualityProtectionActof1996(FQPA),which
settoughersafetystandards,includingmandatorypesticidereregistra-
tion(USEPA,2012).Inaddition,integratedpestmanagement(IPM)
practiceshavebeenpromotedtoachievethegoalofsustainablepest
management. Growers monitor pest pressure and apply pesticides
only when necessary, and high-risk pesticides tend to be replaced
with reduced-risk pesticides. For instance, organophosphates that
werefoundtodeterioratesurfacewaterqualitywerepartiallyreplaced
withoilsorBacillusthuringiensis(Bt),andhencethemajorityofinsecti-
cides(intermsofmass)appliedonalmondsinrecentyearswereoils
(Epsteinetal.,2001;Zhangetal.,2005).Topresentanoveralland
more recent picture of the shift of pest management practices for
Californiaalmonds,itisimportanttoevaluateallthepesticidesthat
areused,whichhasnotbeendoneinpreviousstudies.
Analyzingthedataforpesticideusealoneisinsufficientforevaluat-
ing environmental consequences of pest management practices
(Barnardetal.,1997),thusnumerouspesticideriskindicatorsconsider-
ingpesticideeffectsandexposurehavebeendevelopedaroundthe
world (Bockstaller et al., 2009), including PRoMPT (Whelan et al.,
2007),SPIDER(Renaudetal.,2008;RenaudandBrown,2008),EPRIP
(Trevisan et al., 2009), and I-Phy (Lindahl and Bockstaller, 2012).
Theseindicatorsvaryinmethodologies,inputdatarequirements,indi-
catoroutputs,andapplicablescales.Severalindicatorcomparisonstud-
ies have been carried out to identify ideal indicators for different
purposes(Maudetal.,2001;Reusetal.,2002;Stenrodetal.,2008),
but they have failed to reach clear agreements. In recent years,
alongwiththeadvancementoftheGeographicInformationSystem
(GIS)softwaretechniquesandaccumulationofenvironmentaldata,
pesticideriskindicatorshavebecomecloselyintegratedwithGISfor
preparingsite-specificenvironmentalconditiondataandpresenting
riskmaps(e.g.,Centofantietal.,2008;Salaetal.,2010;Schrieverand
Liess,2007;Vajetal.,2011).
Yet,twoobstaclesexistinappliedpesticideriskevaluation:(1)the
shortageofrealpesticideapplicationdata;and(2)thelackofasuitable
pesticideriskindicatorequippedwithextensivedataofpesticideproper-
tiesandenvironmentalconditions.Thisstudyovercamethesetwoobsta-
cleswiththePesticideUseReporting(PUR)database(CEPA,2012)and
thePesticideUseRiskEvaluation(PURE)indicator(ZhanandZhang,
2012).ThePURdatabasehascomprehensivelyrecordedtemporaland
spatialdataforagriculturalpesticideuseinCalifornia,USAsince1990.
ThePUREindicatorwasspecificallydevelopedforCaliforniaagricultural
pesticideuse,andevaluatespesticide'sriskstosurfacewater,groundwa-
ter,soil,andair,byconsideringpesticidepropertiesandon-siteenviron-
mentalconditions.ThePUREindicatorwasvalidatedagainstsurface
water monitoring data (Zhan and Zhang, 2012) and was evaluated
withasensitivityanalysis(ZhanandZhang,2013).
Thisstudyprovidesthefirstcomprehensiveanalysisofoverallpesti-
cideuseforacrop,alongwithriskevaluationbyapesticideriskindica-
tor.Thegoalistoevaluatethepastperformanceofpestmanagementin
Californiaalmonds.Thespecificobjectivesare:(1)tocharacterizethe
Fig.1.Spatialandtemporalpatternsofthealmondplantedareasfrom1996to2010in
spatialandtemporalpatternsofpesticideuse;and(2)toanalyzethe
theCentralValley,California,USA.(a)Theaverageannualplantedareasattownship
spatialandtemporalpatternsofpesticideenvironmentalrisks.The (~9.7×9.7km2)level,and(b)theannualplantedareasforeachregion.
Y.Zhan,M.Zhang/ScienceoftheTotalEnvironment472(2014)517–529 519
area was divided according to convention into three regions tosurfacewatertothemaximumacuteandchronicpesticidetoxicities,
fromnorthtosouthinthestate:theSacramentoValley(SAC),the respectively,totheaquaticorganisms(includingfish,Daphnia,and
SanJoaquinValley(SJQ),andtheTulareBasin(TUL).Theaverageannu- algae).PEC wasdeterminedbythepesticidedriftprocessmodeled
WS
al planted area of almonds in California from 1996 to 2010 was withtheDriftCalculator(FOCUS,2001)andthepesticiderunoffprocess
277,000ha,whichwascomposedof50,000hainSAC,128,000hain usingtheSCScurvenumbermethod(SCS,1972).PEC wastheaverage
WL
SJQ,and99,000hainTUL.Almondsarethemostdenselycultivatedin concentrationduringthe21days(thetypicalperiodformeasuringthe
centralSJQ,southTUL,andnorthSAC(Fig.1a).Theplantedareasin- chronictoxicity)afterapplication.
creasedinallthreeregionsfrom1996to2010,withasharpincrease Secondly,R wastheratioofthepredictedpesticideconcentration
G
from2003to2007(Fig.1b).Thethreeregionshavesomewhatdifferent leachingtogroundwater(PEC )toADI.Theadaptedattenuationfactor
G
climaticconditions.Fromnorthtosouth,temperatureincreaseswhile methodoriginallyproposedbyRaoetal.(1985)wasusedtocalculate
humiditydecreases,resultingindifferentpestpatternsandpestman- PEC ,wherepesticidedegradation,convection,anddispersionwere
G
agementpractices. takenintoaccount.
Theenvironmentalconditions,includingtheclimaticconditions,soil Thirdly,R,similartoR ,wasthemaximumoftheshort-termand
S W
properties,andotherdataforenvironmentalfactorswerecompiled long-termriskscoresforsoil,whichweretheratiosofthepredicted
fromvariouspublicdatasources.Theclimaticconditionswereobtained short-term(PEC )andlong-term(PEC )pesticideconcentrationsin
SS SL
fromtheCaliforniaIrrigationManagementInformationSystem(CIMIS) topsoiltotheacuteandchronicpesticidetoxicitiestoearthworms,
(CDWR,2010).ThesoilpropertieswereextractedfromtheSoilSurvey respectively.PEC wasdeterminedbytheamountofpesticidereaching
SS
Geographic (SSURGO) and the State Soil Geographic (STATSGO) thegroundrightafterthepesticideapplication,andhencePEC wasthe
SL
databases(NRCS,2008).Thegroundslopewascalculatedfromadigital averageconcentrationintopsoilduringthe21daysafterapplication.
elevationmodel(DEM)database(NRCS,2008).Thegroundwaterdepth Finally,R wastheproductofthepesticideapplicationrate(RATE),
A
wasinterpolatedfromtheUSGSgroundwatermonitoringdata(USGS, theEP,andtheapplicationmethodadjustmentfactor(AMAF).Fora
2010),andthefarmlanddistancetosurfacewaterwascalculatedfrom pesticideproductcontainingmultipleAIs,theproduct-levelR was
A
theCal-Atlasstreammap(Cal-Atlas,2008). assignedtoeachAIinproportiontotheirmasspercentagesinthat
product.
2.2.Pesticideusedataandpesticideproperties Asthefourtypesofriskvalueswerecalculatedfordifferentenviron-
mental compartments, they cannot be compared with each other.
ThepesticideusedataforCaliforniaalmondsfrom1996to2010were SimilartoUI,theannualfield-levelpesticideriskintensities(RI;unit:
queriedfromthePURdatabasemaintainedbyCDPR(CEPA,2012).Nearly R/ha)werealsosummarizedbyAI,pesticidecategories,andallpesti-
2millionpesticideapplicationrecordswereretrieved,eachincludingthe cides.RI =Σ(pesticideriskvalues)/fieldarea,wherei=W,G,S,
i
applicationdateandspatialsection(~1.6×1.6km2)(USDI,2009).This or A, which represent surface water, groundwater, soil, and air,
studytookallpossiblepesticidesintoaccount,withafocusonthemain respectively. Then the field-level RIs were aggregated to township
pesticidecategoriesofinsecticides,fungicides,herbicides,andfumigants, (~9.66×9.66km2)(USDI,2009),region,andstatelevels.
whichrepresented139,76,76,and8pesticideAIs(activeingredients)
(TableA1),respectively.ElevenAIs(e.g.,sulfur)wereclassifiedasinsecti- 2.4.Trendanalysisandspatialmapping
cidesaswellasfungicides.Furthermore,twohighlyconcernedchemical
groupsofinsecticides–organophosphateandpyrethroids(TableA2)– ThetrendanalysisandspatialmappingofUIandRIwereperformed
werealsoanalyzedaspesticidecategories.Theannualfield-levelpesticide inR(RDevelopmentCoreTeam,2013),whichisafreeandversatile
useintensity(UI=Σ(pesticideuseamount)/fieldarea;unit:kg/ha) computationplatform.TrendsweredetectedwiththeMann–Kendall
wassummarizedbyindividualAIs,pesticidecategories,andallpesticides. trendtest(Mann,1945)implementedinpackageKendall(McLeod,
Thenthefield-levelUIswereaggregatedtotownship(~9.66×9.66km2) 2011),andslopeswerecalculatedbytheTheil–Senestimator(Sen,
(USDI,2009),region,andstatelevelsusingthearea-weighted-mean 1968)implementedinpackagezyp(BronaughandWerner,2013).
approach. ThecombinationoftheMann–KendallmethodandtheTheil–Senesti-
Theproduct-andAI-levelpesticidepropertieswereobtainedfrom matorisrobustandwidelyusedforanalyzingenvironmentaltime-
severaldatasources.Theproduct-levelproperties,includingtheemis- seriesdata(HelselandHirsch,2002).Significancewasconsideredas
sionpotential(EP)andpercentageofAI,werefromthepesticideprod- pb0.1. In addition, the UI and RI were mapped at township level
uct/label database maintained by CDPR (CEPA, 2010). The AI-level (~9.7×9.7km2)(USDI,2009)byusingpackagesrgdal(Bivandetal.,
propertiesincludechemical,physical,andtoxicityproperties.Specifi- 2013)andsp(PebesmaandBivand,2005).
cally,thesorptioncoefficient(K ),theHenry'slawconstant(K ),the
OC H
aerobic(DT )andanaerobic(DT )half-lifeinsoil,thehalf-lifein 3.Results
SO SA
water(DT ),themaximumacute(LEC )andchronic(NOEC )toxicity
W A A
toaquaticorganisms(includingfish,Daphnia,andalgae),theacute 3.1.Pesticideuseintensity(UI)
(LC )andchronic(NOEC )toxicitytoearthworms,andtheacceptable
W W
dailyintake(ADI)wereobtainedfromtheChemPestdatabase(CEPA, Between1996and2010,thestateaverageannualUIsofinsecticides,
2009),thePesticidePropertiesDatabase(PPDB)(PPDB,2012),andthe fungicides,herbicides,andfumigantswere17.00kg/ha,4.05kg/ha,
Pesticide Action Network (PAN) (Kegley et al., 2011) in order of 3.21kg/ha,and1.09kg/ha,respectively(Table1);andtheaverage
preference. annualUIsoforganophosphatesandpyrethroidswere0.98kg/haand
0.06kg/ha,respectively(TableA3).Attheregionallevel,theaverage
2.3.Pesticideriskindicator annualUIofinsecticidesandherbicidesbothincreasedfromnorthto
south,thefungicideUIdecreasedfromnorthtosouth,andthefumigant
Onthebasisofthepesticidepropertiesandlocalenvironmental UIwasthehighestinthemiddleregion.Thesamelatitudinalpatterns
conditions,thePUREindicator(ZhanandZhang,2012)wasusedto wereobservedonthetownshipscale,withsmoothspatialtransition
evaluatetheriskvaluesofeachpesticideapplicationtosurfacewater (Fig.2).Furthermore,theregionalaverageannualUIoforganophos-
(R ),groundwater(R ),soil(R),andair(R ). phatesinTULwasaboutthreetimesasthatinSACorSJQ,whilethe
W G S A
Firstly,R wasthemaximumoftheshort-termandlong-termrisk regionalaverageannualUIofpyrethroidsinSACwaslessthanahalf
W
valuesforsurfacewater,whichweretheratiosofthepredictedshort- ofthatinSJQorTUL(TableA3).Thespatialmaps(Fig.A1aandA1b)
term(PEC )andlong-term(PEC )pesticideconcentrationsloaded confirmtheregionalUIpatternsoforganophosphatesandpyrethroids.
WS WL
520 Y.Zhan,M.Zhang/ScienceoftheTotalEnvironment472(2014)517–529
Table1
Useintensities(UI)bypesticideusecategoryforCaliforniaalmondsfrom1996to2010,withstatewidetop-fivepesticidesineachusecategory.
Usecategory/pesticide State SAC SJQ TUL
Mean Slope Mean Slope Mean Slope Mean Slope
Insecticides 17.00 −0.22 7.63 −0.83** 13.28 −0.28· 26.54 0.26
Petroleumoil,unclassified 10.03 −0.22* 2.81 −0.43** 6.79 −0.27* 17.98 −0.01
Mineraloil 4.59 0.20* 2.60 −0.37** 4.70 0.13 5.28 0.43**
Sulfur 0.47 0.00 0.96 0.11* 0.44 −0.04** 0.25 0.01
Propargite 0.46 −0.06** 0.24 −0.01· 0.37 −0.05** 0.73 −0.12**
Chlorpyrifos 0.45 −0.01 0.23 0.01* 0.37 −0.02** 0.67 −0.01
Fungicides 4.05 −0.28** 5.34 −0.05 4.46 −0.41** 2.83 −0.22**
Ziram 0.95 −0.08** 2.06 −0.04 0.69 −0.09** 0.73 −0.08**
Copperhydroxide 0.81 −0.08** 0.29 −0.02** 1.15 −0.11** 0.63 −0.04*
Captan 0.48 −0.08** 0.64 −0.05* 0.58 −0.10** 0.25 −0.05**
Sulfur 0.47 0.00 0.96 0.11* 0.44 −0.04** 0.25 0.01
Maneb 0.35 −0.06** 0.54 −0.05* 0.37 −0.07** 0.22 −0.04**
Herbicides 3.21 0.17** 2.94 0.13** 3.10 0.14** 3.45 0.22**
Glyphosate,isopropylaminesalt 1.24 0.00 1.40 −0.01 1.21 0.02 1.21 −0.01
Paraquatdichloride 0.46 0.03* 0.31 0.05** 0.37 0.02· 0.65 0.04*
Glyphosate,potassiumsalt 0.26 0.04** 0.16 0.02* 0.20 0.04** 0.36 0.05**
Oryzalin 0.23 −0.00 0.34 0.01 0.24 −0.01 0.18 −0.01
Oxyfluorfen 0.22 0.01* 0.15 0.01** 0.21 0.01* 0.26 0.01*
Fumigants 1.09 −0.02 0.11 2E-04 1.51 −0.01 1.06 −0.02
1,3-Dichloropropene 0.77 0.07 0.05 0.00 1.06 0.09· 0.76 0.06*
Methylbromide 0.26 −0.04** 0.04 −0.01** 0.33 −0.06** 0.30 −0.04**
Sodiumtetrathiocarbonate 0.03 0.00 3E-3 0.00 0.06 0.00 7E-5 0.00
Metam-sodium 0.02 −1E-3** 7E-6 0.00 0.03 −3E-3** 0.01 0.00
Chloropicrin 0.01 −0.00 0.01 3E-4 0.01 −0.00 4E-3 −0.00
SAC:theSacramentoValley;SJQ:theSanJoaquinValley;TUL:theTulareBasin.Mean:averageannualuseintensity(kg/ha).Slope:theTheil–Senslope(kg/ha/year)withsignificancelevel
calculatedbytheMann–Kendalltrendtest.**pb0.01;*pb0.05;·pb0.1.
ThestatewideUIoffungicidesandherbicidessignificantlydecreased 3.2.Pesticideriskintensity(RI)
andincreased,respectively,whilethestatewideUIofinsecticidesand
fumigantsshowednotrends(Table1).AlltheUItrendsatregional Between 1996 and 2010, the state average annual RI , RI , RI ,
W G S
level were consistent with the trends at state level, except for the andRI were81R/ha,98R/ha,182R/ha,and90R/ha,respectively
A
increasedtrendsoftheinsecticideUIinSACandSJQ,andthelackof (Table2).Organophosphatescontributed45%,9%,17%,and16%ofthe
trendsobservedforthefungicideUIinSAC.Amonginsecticides,the totalRI ,RI ,RI,andRI ,respectively,whilepyrethroidscontributed
W G S A
organophosphateUIsignificantlydecreasedinallregions,whilethe 11%,0%,5%,and2%,respectively(TableA3).Atregionallevel,theaver-
pyrethroidUIsignificantlyincreasedonlyinTUL(TableA3).Figs.3 ageannualRI ,RI,andRI increasedfromnorthtosouth,whileRI de-
W S A G
andA1showspecificareaswithsignificantUIchanges.Increaseofher- creasedfromnorthtosouth.ThespatialgradientsofRIonthetownship
bicideUIoccurredacrossthewholeCentralValley,whiledecreaseof levelwerelessclearthanthoseofUI,andthehigh-RIareasweremore
fungicideUIspreadoverSJQandTUL.DecreaseoforganophosphateUI clustered(Fig.5).NorthernSACandsouthernTULhadclusteredareas
occurredoverthewholeCentralValley,whileincreaseofpyrethroid ofbothhighRI andhighRI ,withafewhigh-RI areasscattered
W G W
UImainlytookplaceinTUL.Figs.4andA2showtheyearlyUIbypesti- inmiddleSJQ.Incontrast,high-RI andhigh-RI areaswerelocatedin
S A
cideusecategoriesatregionallevel.ThedecreaseofinsecticideUIinSAC middleSJQandnortheasternTUL.Moreover,theriskmapsoforgano-
andSJQmainlyhappenedfrom1996to2000;andtheorganophosphate phosphates(Fig.A3)aresimilartothoseofallpesticides(Fig.5)toa
UIcontinuouslydecreasedinallregions,whilethepyrethroidUIkept certainextent.Forpyrethroids,RI andRI weretheonlyconcerns:
W S
steadyinmostyearsbutincreasedalotinSJQandTULin2010.InSJQ thehigh-RI areasscatteredacrossthewholeCentralValley,while
W
and TUL the fungicide UI decreased consistently, while in SAC it thehigh-RI areasclusteredinTUL(Fig.A6).
S
decreasedinitiallyuntil2001,followedbyaperiodofrapidincrease. ThestatewideRI andRI significantlydecreased,whilethestate-
G S
TheherbicideUIinallthreeregionsslightlydecreasedfrom1996to wideRI andRI didn'tshowtrends(Table2).RegionallyinSAC,
W A
2001andthenincreaseddramatically.ThefumigantUIkeptsteadyin noneoftherisktypesshowedanytrends.InSJQ,allrisktypessignif-
SACbutvariedwidelyinSJQandTUL. icantlydecreasedexceptRI ,whichdidnothavetrendsinanyre-
A
Thestatewidetop-five-UIpesticidesbyusecategoryaccountedfor gion.InTUL,bothRI andRI significantlydecreased.Fig.6shows
G S
94%,76%,75%,and99.6%oftheUIofinsecticides,fungicides,herbicides, specificareaswithsignificantRIchanges.InSAC,theareaswhere
andfumigants,respectively(Table1).Forinsecticides,“petroleumoil, RIsignificantlyincreased/decreasedwerescattered.InSJQRI ,RI
W G
unclassified”andmineraloilaccountedforthemajorityoftheinsecti- andRI significantlydecreasedacrosslargeareas.InTUL,bothRI and
S W
cideUIandthetotalpesticideUI.Mostofthetop-fiveinsecticideseither RI significantlydecreasedinaclusteredarealocatedinthesouth,and
S
significantlydecreasedorshowednotrendintheirUI.Forfungicides, RI significantlyincreasedatthewestedge.Moretemporallyspecific,
A
SACheavilyreliedonziramandsulfur,SJQusedcopperhydroxidethe theRI ofTULlargelydecreasedfrom1996to2002andthenbounced
W
mostintensively,andTULtendedtohaveevenapplicationsofziram back in 2006, while the RI of SAC and SJQ were relatively steady
W
andcopperhydroxide.MostoftheUIofthetop-fivefungicidesde- (Fig. 7a). The RI of SJQ and TUL showed two stages, separated in
G
creasedsignificantly.Forherbicides,“glyphosate,isopropylaminesalt” 2004and2000,respectively,whiletheRI ofSACdidnotshowaclear
G
wasthedominantherbicideinallregionsandshowednotrends.The trend(Fig.7b).TheRI ofallregionsdecreasedsharplyfrom1996to
S
UIoftheothertopherbicidesexceptoryzalinincreasedsignificantlyin 2000or2001,andthenroseslowlytill2006,followedbyashortde-
allregions.Forfumigants,1,3-dichloropropeneandmethylbromide creasingperiod(Fig.7c).TheRI ofallthreeregionsshowedsimilar
A
accountedfor94%ofstatewidefumigantuses.Theformerincreasedsig- temporalpatternsasRI (i.e.,decrease–increase–decrease)(Fig.7d).
S
nificantlyinSJQandTUL,whilethelattersignificantlydecreasedinall ThedifferenceisthatRI recoveredtotheinitiallevelattheendofthe
A
regions. increasestagewhileRI onlypartiallyrecovered.
S
Y.Zhan,M.Zhang/ScienceoftheTotalEnvironment472(2014)517–529 521
Fig.2.Spatialpatternsoftheaverageannualuseintensities(UI;kg/ha)of(a)insecticides,(b)fungicides,(c)herbicides,and(d)fumigantsforCaliforniaalmondsfrom1996to2010.
522 Y.Zhan,M.Zhang/ScienceoftheTotalEnvironment472(2014)517–529
Fig.3.Temporaltrendsoftheannualuseintensities(UI;kg/ha)of(a)insecticides,(b)fungicides,(c)herbicides,and(d)fumigantsforCaliforniaalmondsfrom1996to2010.
Y.Zhan,M.Zhang/ScienceoftheTotalEnvironment472(2014)517–529 523
Fig.4.Regionalannualuseintensities(UI;kg/ha)of(a)insecticides,(b)fungicides,(c)herbicides,and(d)fumigantsforCaliforniaalmondsfrom1996to2010.
ThetemporalpatternsofRIwerequitedifferentbetweenorgano- significantlyincreasedinSJQandTUL,butonlyaccountedfor3%ofRI
A
phosphatesandpyrethroids.Fororganophosphates,RI significantly inSACanddidnotshowatrend.
W
decreasedonlyinSJQ,RI andRI significantlydecreasedinallregions,
G S
and RI significantly decreased in SJQ and TUL (Table A3). Fig. A4 4.Discussion
A
showstheRItrendsfororganophosphatesattownshiplevel.RI signif-
W
icantlydecreasedinlargeareasofSJQandinasmallportionofSACand 4.1.Spatialpatternsofpesticideuseandrisk
TUL,whichwasreflectedinthebasin-leveltrendofRI .RI hadasim-
W A
ilarspatialpatternwithRI ,butsignificantlydecreasedinlargerareas 4.1.1.Pesticideuseintensity(UI)
W
inTUL.ThetemporaltrendsofRI andRI showedasimilarspatialpat- ThespatialpatternsofUI,mainlycausedbyspatiallydifferentpest
G S
tern.Fig.A5showstheyearlychangeofRIatregionallevel.Thedecrease pressures,werehighlyassociatedwithclimateconditionsandfarming
ofRI wasapparentinallregions,onepeakforRI andRI ofTULoc- activities.Inthestudyarea(i.e.,theCentralValley,California),thetem-
S W A
curredin2006,andonepeakforRI ofSACappearedin1999.Forpyre- peratureincreasesfromnorthtosouthwhilethehumiditydecreases
G
throids(RI andRI werenegligible),RI significantlyincreasedonlyin fromnorthtosouth.Insouthernareas,moreinsecticides(including
G A W
TUL,whileRI significantlyincreasedinSJQandTUL.RI significantly moreorganophosphates,pyrethroids,andotherinsecticides)andherbi-
S W
decreasedinsmallareasofnorthernSACandmiddleSJQ,andsignifi- cideswereapplied,indicatingthathighertemperatureswithsufficient
cantlyincreasedinlargeareasofTUL(Fig.A7a).RI significantlyin- watersupplyviairrigationfavoredthegrowthofinsectsandweeds.
S
creased across the whole Central Valley (Fig. A7c). The RI of TUL Incontrast,fungiprefercoolandmoistenvironments,whichresulted
W
increasedcontinuouslyandrapidlyfrom2005to2010(Fig.A8a).The inmoreintensivefungicideuseinnorthernareas.Inaddition,thespatial
RI of TUL increased slowly but steadily from the beginning of the patternoffumigantUIwasmainlyduetofarmingactivities.Fumigants
S
studyperiod,andincreasedmuchfasterfrom2005(Fig.A8c). weremainlyusedtotreatsoil-bornediseaseswhenalmondfieldswere
Thestatewidetop-fivepesticidesbyrisktypeaccountedfor80%, newlycultivatedorreplanted(CEPA,2008),aswellasforpost-harvest
86%,44%,and66%ofRI ,RI ,RI ,andRI ,respectively(Table2).For pests.TherapidexpansionofalmondfieldsinSJQandTULresultedin
W G S A
RI ,ziram,copperhydroxide,andchlorpyrifoswerethetopcontribu- higheraverageannualfumigantUIinthesetworegions(Fig.2d).
W
torsinSAC,SJQ,andTUL,respectively.TheRI fromziramsignificantly InadditiontothegeneralspatialpatternsoftheUIs,theexistenceof
W
decreasedinSJQonly.TheRI fromcopperhydroxidesignificantlyde- clusteredhigh-UIareasdemonstrateslocationspecificity,whichwas
W
creasedinallregions.Incontrast,theRI fromchlorpyrifosdidn'tshow likelyassociatedwithlocalpestpressure(includinginsects,pathogens,
W
atrendinanyregion.ForRI ,oxyfluorfenandsimazinewerethemain andweeds).Theareaswithdenseralmondfields(Fig.1a)seemedto
G
contributorsinallregions.TheRI ofallthetop-fivepesticidessignifi- sufferhigherpestpressuresreflectedinhigherpesticideUI.Innorthern
G
cantlydecreasedinallregions,exceptfortheRI ofoxyfluorfenwhich SACmoreintensefungicidesandherbicideswereapplied.IncentralSJQ
G
significantlyincreasedinSACandshowednotrendintheothertwore- higherUIsoffungicidesandfumigantswereobserved.InsouthernTUL
gions.ForRI,mostofthetop-fivecontributorssignificantlydecreased insecticidesandherbicideswereusedmoreintensively.Thespatial
S
inallregions,exceptfortheRI of1,3-dichloropropenethatsignificantly correlation between the cultivation density and the pest pressure
S
increasedinSJQandTUL,andtheRI ofmineraloilthatsignificantlyin- mightbeinducedbypestdispersionranges.Thatis,closerdistances
S
creasedinTUL.ForRI ,1,3-dichloropronewasthetopcontributorand amongalmondfieldsfacilitatedthedispersionandsubsequentburst
A
524 Y.Zhan,M.Zhang/ScienceoftheTotalEnvironment472(2014)517–529
Table2
Riskintensities(RI)forCaliforniaalmondsfrom1996to2010,withstatewidetop-fivepesticidesforeachrisktype.
Risk/pesticide State SAC SJQ TUL
Mean Slope Mean Slope Mean Slope Mean Slope
Surfacewater 81 −0.8 57 −2.0 67 −2.4· 111 −1.1
Chlorpyrifos 31 1.2 9 −0.03 21 −0.7 54 2.3
Copperhydroxide 18 −1.7** 10 −0.4** 22 −1.6* 18 −1.5
Ziram 7 −0.7* 19 −0.8 5 −0.7** 4 −0.3
Permethrin 4 −0.3· 2 −0.01 4 −0.3* 6 −0.2
Chloropicrin 4 −0.2 9 0.2* 5 −0.6** 1 −0.04
Groundwater 98 −4.1** 185 −2.5 90 −4.5** 64 −3.0·
Oxyfluorfen 37 1.0 83 6.8** 18 0.3 39 −0.5
Simazine 29 −1.9** 26 −1.6** 48 −2.8* 6 −0.5**
Diazinon 7 −0.7** 32 −3.3** 1 −0.1* 1 −0.02**
Norflurazon 6 −0.9** 7 −0.8** 8 −0.9** 3 −0.3**
Propargite 5 −0.7** 12 −0.7 3 −0.4** 5 −0.9**
Soil 182 −9.8* 145 −7.1 183 −9.4* 199 −9.7*
Copperhydroxide 22 −2.3** 8 −0.6** 31 −3.1** 17 −1.1*
1,3-Dichloropropene 19 1.4* 1 0.03 25 2.0· 19 1.7*
Ziram 17 −1.5** 39 −1.5 12 −1.6** 12 −1.6**
Methidathion 15 −2.8** 11 −0.9** 9 −1.4** 25 −5.6**
Mineraloil 9 0.1 6 −0.9** 9 0.03 10 0.7**
Air 90 1.7 42 0.2 92 0.6 112 3.0
1,3-Dichloropropene 20 1.8 1 0.1 27 2.2 20 1.7*
Oxyfluorfen 12 0.5 9 0.7* 12 0.3 15 0.5
Chlorpyrifos 12 −0.5 6 0.4 10 −0.6* 17 −0.7
Petroleumoil,unclassified 9 1.0** 1 −0.1* 6 0.6** 17 1.6**
Methylbromide 7 −1.0** 1 −0.1** 8 −1.4** 7 −0.9**
SAC:theSacramentoValley;SJQ:theSanJoaquinValley;TUL:theTulareBasin.Mean:averageannualriskintensity(R/ha).Slope:theTheil–Senslope(R/ha/year)withsignificancelevel
calculatedbytheMann–Kendalltrendtest.**pb0.01;*pb0.05;·pb0.1.
ofpests.Ontheotherhand,innortheasternTULwherethealmond animportantroleinRI andtheemissionpotentialswereimportantto
S
fieldswererelativelysparser,theUIsofinsecticides,fungicides,andfu- RI .Thefindingswereconsistentwiththesensitivityanalysisonthe
A
migantswerealsohigh.Itindicatedthatsomeotherfactors(e.g.,farm PUREindicator(ZhanandZhang,2013).Thereexistedhigh-RI and
S
managementpractices)influencedthelocalpesticideUIorpestpres- high-RI areasincentralSJQandnortheasternTUL,whichwasmainly
A
sure,whichrequiresmoreinvestigationinthefuture. causedbyhighfumigantUIinthoseareas.Inaddition,thehigh-RI
S
areasinnorthernSACwerelargelyduetotheintenseuseoffungicides.
4.1.2.Pesticideriskintensity(RI)
ComparedwithUI,thespatialpatternsofRIwereaffectedbymore 4.2.Temporalpatternsofpesticideuseandrisk
factors,includingenvironmentalconditions(e.g.,surfacewaterdis-
tanceandgroundwaterdepth)andpesticideproperties.Firstly,the ThetemporalpatternsofUIandRIweretheresultsoftheshiftofpest
high-RI areaswereallclosetosurfacewaterandusedpesticideshigh- managementpracticesledbygovernmentalregulations,theintegrated
W
lytoxictoaquaticorganisms(e.g.,chlorpyrifos).Therisktosurface pestmanagement(IPM)promotion,availabilitiesofnewpesticides,and
waterwasthemainenvironmentalconcernoftreatinginsectswithor- phasing-outofpesticidesknowntoposehighlyadverseimpactson
ganophosphatesandpyrethroids,whicharegenerallyhighlytoxicto humanhealthandenvironment.
aquaticorganisms,moderatelypersistent,andsolubleinwaterforor-
ganophosphatesorboundtosedimentforpyrethroids.Thehigh-RI 4.2.1.Insecticides
W
areasinsouthTULwereneartheKernRiverandthePosoCreek,with Insecticideusewasunderstringentregulation,whichlargelyaffect-
intensiveapplicationsoforganophosphatesandpyrethroids.Similarly, ed the temporal patterns of insecticide UI and the associated RI.
thehigh-RI areasinnorthSACwereclosetotheSacramentoRiver, Althoughpropargiteandchlorpyrifoswereusedmuchlessthanthe
W
andfungicideswereappliedintensivelyintheseareas.Additionally, top-three-UIinsecticides(i.e.,“petroleumoil,unclassified”,mineraloil,
thehigh-RI areasscatteredincentralSJQwereneartheSanJoaquin andsulfur),theirriskstohumanandecosystemhealthweremuch
W
River,withintensiveuseofinsecticidesandfungicides. higher. Propargite is known to cause human health problems as a
Secondly,high-RI areasweremainlycausedbythecombinedef- carcinogenandreproductivetoxicant(e.g.,MillsandYang,2007),as
G
fectsofhighherbicideuseandshallowgroundwaterlevel.Herbicides wellasenvironmentalproblems(e.g.,Bradfordetal.,2010).Therefore,
areusuallymobileinsoilasindicatedbylowsoilsorptioncoefficients theuseofpropargitehasbeenrestrictedbyregulations,resultingin
(K ).Inanationalgroundwatersurvey,mostofthedetectedpesticides thesignificantdecreaseinitsuseinallthreeregions.Inaddition,chlor-
OC
wereherbicidesinareaswithshallowgroundwaterlevel(Kolpinetal., pyrifos has been frequently detected in surface water in California
2000).BothnorthSACandsouthwesternTULhadhigh-RI areas.Asex- (CEPA,2007)andishighlytoxictoaquaticorganisms.Ithasbeenon
G
pected,intheseareasthegroundwaterlevelwasshallowandherbicide theCleanWaterAct303(d)listofimpairedwaterwayssince1998in
UIwashigh.Contrarily,intheareasneartheboundarybetweenSJQand theTotalMaximumDailyLoad(TMDL)program(CEPA,2013).Inthis
TULwherethegroundwaterlevelwasdeep,althoughtheherbicideUI riskevaluation,chlorpyrifoswasthetopRI contributorandoneof
W
wasalsohigh,theRI wasnotashighasthatinnorthSACandsouth- themainRI contributors(Table2).InTULtheelevatedRI after2005
G A W
westernTUL. wascausedbytheincreaseduseofchlorpyrifos(Fig.7a).Animportant
Finally,thespatialpatternsofRI andRI bothwerelargelyaffected concernisthesignificantlyincreaseduseofchlorpyrifosinSAC(though
S A
bytotalpesticideUI,whilethepesticidetoxicitiestoearthwormsplayed stilllowerthantheothertworegions),whichmightbeduetoelevated
Fig.5.Spatialpatternsoftheannualriskintensities(R/ha)of(a)surfacewaterrisk(RIW),(b)groundwaterrisk(RIG),(c)soilrisk(RIS),and(d)airrisk(RIA)forCaliforniaalmondsfrom@
1996to2010.
Y.Zhan,M.Zhang/ScienceoftheTotalEnvironment472(2014)517–529 525
526 Y.Zhan,M.Zhang/ScienceoftheTotalEnvironment472(2014)517–529