The long shadow of our chemical past - High DDT concentrations in fish near a former ...

12 Pages·2016·1.82 MB·English

by  Monika D. Jürgens

Chemosphere162(2016)333e344
ContentslistsavailableatScienceDirect
Chemosphere
journal homepage: www.elsevier.com/locate/chemosphere
The long shadow of our chemical past e High DDT concentrations in
fish near a former agrochemicals factory in England
Monika D. Jürgens a,*, John Crosse b, Patrick B. Hamilton c, Andrew C. Johnson a,
Kevin C. Jones b
aCentreforEcologyandHydrology,WallingfordOX108BB,UK
bLancasterUniversity,LancasterEnvironmentCentre,LA14YQ,UK
cUniversityofExeter,Biosciences,ExeterEX44QD,UK
h i g h l i g h t s g r a p h i c a l a b s t r a c t
(cid:1)Roach (Rutilus rutilus) from 13 UK
river sites were analysed for
pesticides.
(cid:1)FPish from one site had much higher
DDTcontent(DDTþDDEþDDD)
thanothers.
(cid:1)The explanation was found in a
formerpesticidefactorynearby.
(cid:1)A review found some sPimilar hot-
spotsinrecentEuropean DDTdata
infish.
(cid:1)Some fish contained levels of DDT
harmfultothemortheirpredators.
a r t i c l e i n f o a b s t r a c t
Articlehistory: Atotalof81roach(Rutilusrutilus)collectedfrom13southernEnglishriversitesbetween2007and2012,
Received2June2016 were analysed for organochlorine pesticides, PCBs, PBDEs and some metals. UnexpectePdly high con-
Receivedinrevisedform centrationsofthebannedinsecticideDDTanditsdegradationproductsDDEandDDD( DDTs)were
19July2016 foundinthe10fishfromtheriverLee(orLea)whichaveraged88±70(standarddeviation)mg/kgww,
Accepted24July2016 almost20timeshigherthantheaveragefortheremainingsites(4.8±3.1mg/kg).AllfishfromtPhatsite
exceededtheCanadianTissueResidueGuideline(environmentalqualitystandard)of14mg/kg DDTs.
HandlingEditor:MyrtoPetreas Concentrations of the insecticides chlordane and lindane as well as copper, which is often used as a
fungicide, were also elevated in fish from the Lee, though not as much as those of DDTs. A likely
Keywords: explanationfortheseobservationswasfoundinanearbyformerpesticidefactory,whichhadstopped
DDT productionaboutthreedecadesearlier.
Pesticides AnextensivereviewofrecentliteraturedataonDDTinwiPldEuropeanfishfoundthat,whilelevelsare
Freshwater nowgenerallylow,therewereseveralotherhotspotswith DDTslevelsthatmaystillbeofconcern.
Fish ©2016TheAuthors.PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCCBYlicense
Chlordane
(http://creativecommons.org/licenses/by/4.0/).
Copper
Abbreviations: DCM, dichloromethane; DDT, dichlorodiphenyltrichloroethane; DDE, dichlorodiphenyldichloroethylene (degradation product of DDT); DDD, dichlor-
odiphenyldichloroethanes,alsoknownasTDE,tetrachlorodiphenylethane(degradationproductofDDT);d.s.,downstream;EQS,environmentalqualitystandard;HCB,
Hexachlorobenzene;PCB,polychlorinatedbiphenyls(industrialinsulatingfluids);PBDE,polybrominateddiphenylethers(flameretardants);POPs,persistentorganicpol-
lutants;u.s.,upstream;ww,wetweight;±,denotesstandarddeviation(basedonasample).
* Correspondingauthor.
E-mailaddresses:[email protected](M.D.Jürgens),[email protected](J.Crosse),[email protected](P.B.Hamilton),[email protected](A.C.Johnson),k.c.jones@
lancaster.ac.uk(K.C.Jones).
http://dx.doi.org/10.1016/j.chemosphere.2016.07.078
0045-6535/©2016TheAuthors.PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).
334 M.D.Jürgensetal./Chemosphere162(2016)333e344
1. Introduction stability, reducing half-life in soil to approximately 1 month
comparedtoabout10yearsforDDT(Aronsonetal.,2006)or>20
Much of the concern about chemical hazards in the aquatic yearsforitsdegradationproductDDE(Mackayetal.,2006).Ifthe
environmentinrecentyearshasfocusedonpharmaceuticalsand conversion from DDT to dicofol is incomplete, some DDT may
personal care products (PPCPs) such as endocrine active com- remainintheproduct.Impuritiesindicofolhavebeenidentifiedas
poundsandmorerecentlynanoparticles,butastheirnameimplies, a recent source of DDT and related compounds mainly in Asia
persistent organic pollutants (POPs) have the potential to cause (RickingandSchwarzbauer,2012).IntheEuropeanUnion,theuse
harm many decades after their production and use was stopped. ofdicofolcontaining>0.1%residuesofDDTandrelatedcompounds
WecontinuetohaveconcernsoverPCBsandmorerecentlyPBDEs, or<78%pp'dicofolhasbeenbannedsince1991(EuropeanUnion,
butsomeoftheolderPOPsproducedjustafterWorldWarII,such 1990) and all dicofol use was banned in 2010 (European Union,
astheinsecticideDDT,nowreceiveconsiderablylessattention. 2008). Therefore, environmental DDTcontaminationfromdicofol
useeatleastafter1991eshouldbenegligibleinEurope.Dueto
1.1. Projectbackground theuseofDDTasanintermediate,localDDTcontaminationrelated
todicofolproductionhashoweverbeensignificantinsomecases
Therearefewrecentdataonlevelsortrendsoflegacychemicals (de la Cal et al., 2008). The Stockholm convention (UNEP, 2001)
inUKfreshwaterfishandwhethertheyposeathreattofishpop- originallymadeprovisionsfortheuseofDDTindicofolproduction,
ulationsand/ortheirpredators.Toaddressthisknowledgegap,we butby2009noneofthe180signatories(179statesþtheEuropean
have an ongoing project, the National Fish Tissue Archive (see Union)hadappliedtoproduceoruseDDTforthispurposeandthat
Jürgensetal.,2013andhttp://www.ceh.ac.uk/our-science/projects/ optionwasautomaticallywithdrawn.
national-fish-tissue-archive), which was started in 2007 and in
which fish samples (mainly roach, Rutilus rutilus) are regularly
1.3. CongenersofDDTanditsdegradationproducts
collectedfromanumberofriversitesinEnglandandstoredfrozen
forfutureanalyses.Whilethearchivewassetupprimarilyforthe
TechnicalDDT(themixturetypicallyused)consistsofabout85%
purposeofmonitoringchemicalpollution,otherapplications-for
pp’DDT, the active insecticidal ingredient, and about 15% op’DDT
example genetics of the fish and/or their associated
withtraceamountsofDDEsandDDDs(ATSDR,2002)(seeFig.S1
microorganisms-arealsopossible.Comparingstoredsampleswith
formolecular structures and degradationpathways). Underaero-
morerecentoneswillallowthedetectionoftemporaltrends,while
bicconditions,DDTmainlydegradestoDDEandunderanaerobic
aspreadofsamplingsitesgoessomewayofcoveringspatialdif-
conditions to DDD, which are both more stable than the parent
ferences. Sub-samples of a proportion of the fish collected have
compound(RickingandSchwarzbauer,2012).Duringthesetrans-
already been analysed for a number of persistent chemicals,
formations the pp’ or op’ configuration is maintained, but differ-
including organochlorine pesticides, PCBs, PBDEs, and a suite of
encesinwater-solubility,vapourpressureandstabilitybetweenthe
metals.ThefollowinPgdiscussionfocusesonDDT,becausesurpris- pp’andop’congeners(seeTableS2)tendtoleadtoshiftstowards
inglyhighlevelsof DDT(DDTanditsmaindegradationproducts
higher pp’/op’ ratios from DDT to DDD and DDE (Ricking and
DDEandDDD)werefoundinfishfromonesite(seeresultssection),
Schwarzbauer, 2012). The compound typically found in the high-
butsomeoftheotherresultsarePalsobrieflymentionedtohelpput est concentrations in the (aerobic) environment is therefore
the data into context. The DDT concentrations were also
pp’DDE. Environmental concentrations are often reported as the
compared to DDT data from other recent European fish studies
(Tables 2 and 3) and effect concentrations from eco-toxicology sumofthesixpp’andop’congenPersofDDT,DDEandDDD,which
arecollectivelyknownasDDTs( DDTs).
studies(Table1).
1.2. HistoryandsourcesofDDT 1.4. DDTintheaquaticenvironment
DDTwasfirstsynthesizedin1873anditwasdevelopedasan DDThasrarelybeendetectedinwater,especiallyafterthere-
insecticide in 1939 (Mellanby,1992). Its high efficiency, speed of strictions and bans on use took effect. In England and Wales the
action,relativelylowtoxicitytohumans,lowcostandpersistence EnvironmentAgencyhasbeenmonitoringaquaticpollutantssince
intheenvironmentappearedtomakeitperfectbothforcombat- the 1970s, but DDTs (sum of pp’DDT, pp’DDE, pp’DDD) were only
ting insect-borne diseases and protecting crops and it earned its detectedin2%ofover60,000riverwatersamplesaccordingtoa
inventorPaulMüllertheNobelPrizeformedicineandphysiologyin publicly available dataset for 1974e2013 (Environment Agency,
1948 (Mellanby,1992). Due to limited availability, it was mainly 2014). While water concentrations of POPs reduce fairly quickly
usedfordiseasecontrolduringthe2ndWorldWar,butitsusein aftertheyarerestrictedorbanned,concentrationsinsedimentscan
agriculture increased dramatically in the 1950s and 60s. Enthu- persist for decades and contamination stored in deep sediments
siasm for DDT waned in the 1960s when it was linked toa large canre-emergeinthefoodchain,especiallywhenthesesediments
decline in raptor populations across severalcontinentsand other aredisturbed.ThiswasobservedinGermany,whereDDTconcen-
negativeeffectsonnon-targetspecies(Mellanby,1992).IntheUK trations in eels increased sharply following dredging activities a
itsusewasrestrictedfromtheearly1970s(HansardReport,1969) decade after a nearby factory had stopped processing DDT
andagriculturaluseswerebannedthroughouttheEuropeanUnion (Heinischetal.,2007)andrecentincreasesofDDTinzooplankton
in1981(EEC,1978).In2001DDTwasincludedinANNEXBinthe andfishrelatedtoalongclosedDDTfactorywePrealsoobservedin
Stockholm convention of POPs to be eliminated or restricted Italy(Bettinettietal.,2012).Infishtissue,the DDTsconcentra-
worldwide.Thisallowsonlylimitedproductionandusefordisease tionsareusuallyevenhigherthaninsedimentseareviewfound
vector control, mainly to combat the spread of malaria (UNEP, fish-sedimentaccumulationfactorsforthemaincongenerpp’DDE
2001). tobetypicallybetween30and75onadryweightbasis(vander
Another source for DDT and related compounds is the more Oost et al., 2003) and a more recent study (Niewiadowska et al.,
recentlyusedpesticidedicofol,whichwasproducedfromDDTby 2014) measured factors of 2e59 in roach and 5 to 125 in bream.
replacingoneoftheprotonswithanOH-group(seeFigs.S1andS2 Therefore, fish are very suitable for monitoring DDTs and other
insupplementaryinformation).Thissmallmodificationdecreases persistentpollutantsintheaquaticenvironment.
M.D.Jürgensetal./Chemosphere162(2016)333e344 335
Table1
Tissue-burdenbasedrisksofDDTanditsdegradationproductstofishandtheirpredators(moredetailsaregiveninTableS3).
P
Species Effect Fishtissueconc.(wholebody DDTs)Reference
[mg/kg]
ObservedLOECsforfishexposedinearlylife P
Japanesemedakaa intersexandalteredgeneexpression 58( op’DDTs) Sunetal.(2016)
Atlanticcroaker behaviour 70ineggs Faulketal.(1999)
Japanesemedakaa intersex,increasedplasmaestradiol,alteredgene 271(pp’DDE) Sunetal.(2016)
expression
9Studies,7species survival 890-2400 Beckvaretal.(2005)
cohosalmon behaviour 1100 Beckvaretal.(2005)
Summerflounder,rainbowincreasedvitellogeninandphysiologicalchanges (cid:5)60,000b Millsetal.(2001)
trout (cid:5)135,000b DonohoeandCurtis(1996)
ObservedLOECsforfishexposedasadults
6Species survival 290-1130 Beckvaretal.(2005)
Goldfish behaviour 1650 Beckvaretal.(2005)
Brooktrout reproduction 7600 Beckvaretal.(2005)
Brooktrout growth 11,200 Beckvaretal.(2005)
Tilapia osmoregulation ca.20,000 Millsetal.(2001)
Publishedestimatedsafetissueconcentrationsforfishthemselves
Adultfish provisionalduetoinsufficientsub-lethaldata 600 Beckvaretal.(2005)
Fishearlylifestages provisionalduetoinsufficientsub-lethaldata 700 Beckvaretal.(2005)
Publishedestimatedsafetissueconcentrationsforconsumersoffish
Birds basedoneggshellthinninginducks 14 CanadianCouncilofMinistersofthe
Environment(1999)
Kingfisher basedoneggshellthinninginpelicans(USEPA, 20 Lazorchaketal.(2003)
1995)
Humans riskbasedthresholdforcancer(from4fishmeals/69 Stahletal.(2009)
month)
Osprey dietaryeffectsconcentration 90e190(pp’DDE) Hincketal.(2009a)
Mammals basedoneffectsonratsandfoodintakeratefor 94 CanadianCouncilofMinistersofthe
mink Environment(1999)
Baldeagle dietaryeffectsconcentration 130(pp’DDE) Hincketal.(2009a)
Pelicans reducedfledglingratewhenfishconc.werestill <150(LOEC) USEPA(1995)
0.15mg/g
Baldeagle noeffectshazardconcentration 270 Hincketal.(2009b)
Mink basedoneffectsonratsandfoodintakeratefor 360 Lazorchaketal.(2003)
mink
Otter basedoneffectsonratsandfoodintakeratefor 490 Lazorchaketal.(2003)
otter
Humans foodstandardformeat(noneexistsforfish) 1000 EuropeanUnion(2005)
a Transgenicmedakainwhichoocytesinmaletestesareeasiertodetect,butsensitivityshouldbeequaltothewild-type.Dataisformales.
b Totaldoseinjected.
1.5. ToxiceffectsandsafelevelsofDDTinfish 2. Materialsandmethods
DDThasanumberoftoxicologicaleffectsonfishthemselvesas The current study examined the presence of DDT and some
wellasontheir predators. Inthe1960s,studiesfound thataccu- otherpollutantsinroach(Rutilusrutilus)fromanumberofsitesin
mulationofDDTupthefoodchainhadresultedinegg-shellthin- the Thames and Anglian regions of England. Roach were chosen
ninginbirdstoapointwheretheeggsfrequentlybroke(Ratcliffe, because they are relatively common throughout the UK and not
1967).InfishDDTshavebeenlinkedtovariouseffectsonfertility much sought afterbyanglers. Although theycan migrate several
and development (Baatrup and Junge, 2001; Bayley et al., 2002; kmifsuitablespawningareasarenotavailablenearby,therangeis
Okoumassounetal.,2002;Uchidaetal.,2010)duetotheirstruc- much smaller than that of truly migratory species (Baade and
tural similarity to steroid hormones, and osmoregulation (Janicki Fredrich,1998). In the rivers sampled, their movement is further
and Kinter,1971; Kinter et al.,1972; Riou et al., 2012) or thyroid restrictedbyobstacles,suchaslocksandweirsandtheseparation
function(Braretal.,2010)mayalsobeaffected. of populations within the Thames Catchment was confirmed by
Beckvar et al. (2005) reviewed available toxicological data analysisofpopulation-geneticstructureusingDNAmicrosatellites
(reproducedinTable1andTableS3alongwithsomemorerecent (Fig.S4andHamiltonetal.,2014).
data)andestimatedthatamaximalbodyburdPenof600e700mg/kg
wetweight(ww)wouldlikelybeaprotective DDTsconcentration
forthefishthemselves,butdeclaredthisvalueprovisionalduetothe 2.1. Fishcollection
scarcity of data, especially on non-lethal endpoints. A number of
RoachwerecaughtbyEnvironmentAgencystaffinconnection
other studies estimated safe levels for human and wildlife con-
sumersoffish(Table1).OfthoseonlytheCanadianTissueResidue with regular fish population surveys using electrofishing tech-
niques. They were killed with an overdose of the anaesthetic 2-
GuiPdeline(i.e.environmentalqualitystandard,biotaEQS)of14mg/
kg DDTs,whichisbasedontherisktofish-eatingbirds(Canadian phenoxyethanol(ca.4mlina10lbucket)packedinplasticbags
(polyethylene/polyacryl or fluoro-ethylene-propylene) and then
Council of Ministers of the Environment,1999) has a legal status
frozenonsiteusingaliquidnitrogen-cooleddryshipper.Onreturn
withregardstotheenvironment.TheEuropeanUniondoesnothave
to the laboratory the frozen fish were transferred to a (cid:3)80 (cid:4)C
a25bniogt/alfEoQrSPfoDrDDTsDaTns,db1u0tnthg/elrfeorispopn’DeDfTor(EsuurrofapceeanwUatneiro,nw,2h0ic1h3)i.s freezer. For the current study a total of 81 roach (9e21 cm, esti-
matedabout2e8yearsold)werecollectedbetween2007and2012
336 M.D.Jürgensetal./Chemosphere162(2016)333e344
Table2
ConcentrationsofDDTanditsmetabolitesinfishfromconfirmedorsuspectedindustriallycontaminatedsitesa.Dataforsomecommonlymonitoredfreshwaterorbrackish
waterfishspeciescaughtinthelast20years(1995e2014).Eels,whichoftenhavehigherPOPconcentrationsthanotherspeciesbecauseoftheirhighfatcontentandas-
sociationwithsediments,havealreadybeenreviewedforpp’DDEinJürgensetal.(2015).AgriculturaluseofDDTceasedinthe1970sor1980sinallthecountriesmentionedin
thistable.
P
Country n n Sites Year(s)SamplebDDTs Site Overall Overall Ref. Comments
sitessamples averages average range
[mg/kg] [mg/kg] [mg/kg]
Barbel(cyprinid,bottomfeeder)
Italy 1 4 PoRiverd.s.of 1995 m S2 174d,e n.a. n.a. Vigano(cid:1)etal.(2000) Lambroisveryinfluencedbychemicalindustry
riverLambro DDTsc
Spain 2 2 Cincariverjust 1999 m S6 ca.8000, n.a. 1188 delaCaletal.
d.s.ofoperating DDTsf ca.2000 e10,431(2008)
dicofolfactory
and30kmd.s.
Spain 2 11 Cincariverjust 2002 m S6 997,562 n.a. 211- delaCaletal. decreasedsince1999(seeabove)
d.s.ofoperating DDTsf 2098 (2008)
dicofolfactory
and30kmds.
Spain 1 1 Cincariverjust 2003 m S6 616e n.a. n.a. Lacorteetal.(2006)slightfurtherdecreasesince2002
pooled d.s.ofoperating DDTsf
dicofolfactory.
Spain 1 5 lowerJarama 2006 m S3 20 n.a. 8.2e51 Nicolaetal.(2014) Highindustrialandurbanimpact.DDT/DDE>1
river,Iberia e09 DDTsg /recentcontamination.
Bream(cyprinid,omnivorous,bottomfeeder)
Czech 5 26 Elberiver 2008/ m S6 14e405 156 6.4e791Hr(cid:1)adkov(cid:1)aetal. Elberiverhasnumerousindustries
republic 09 DDTsf (2012)
Czech 3 3 2reservoirs,1 2006 m S6 97e263 178 n.a. Cervenyetal. Elberiverhasnumerousindustries
republic pooled riverElbesite e10 DDTsf (2014)
Germany 2 10 Elbetributaries 2009 m S2 151,173 162 61e249 Umweltbundesamt 1995e2015Saalepossibleupwardstrend,
SaaleandMulde e13h DDTsc (2016) Muldenotrend
Germany 5 25 Elberiver 2009 m S2 72e125 95 53e178 Umweltbundesamt 1993e2013,slightdownwardstrends,butnot
e13h DDTsc (2016) clear.Elberiverhasnumerousindustries
Poland 1 15 riverVistula 2011/ m S4 654 n.a. 179- Niewiadowskaetal.
(Wisła)near 12 DDTsi 1921 (2014)
Cracow
Roach(cyprinid,omnivorous)
Southern 1 10 Leeat 2011 w S6 88 n.a. 35e270 currentstudy justd.s.formerDDTfactory,whichclosedin
England Wheathampstead DDTsf 1982
Czech 2 6 ElbeatUstinad 2008 m S6 10,24 17 6.7e26 Hr(cid:1)adkov(cid:1)aetal. includedinpollutedlist,becausevaluesfor
republic LabemandDecin DDTsf (2012) breamwerehighatthesesites
Italy 1 4times LakeMaggiore 2009 m S3 ca.35j n.a. n.a. Bettinettietal. DDTfactoryclosed1996,butnewreleaseofDDT
DDTsg (2012) in2009waslikely,showingfirstinzooplankton
andtheninfish
Italy 1 4times LakeMaggiore 2010 m S3 ca.85j n.a. n.a. Bettinettietal.
DDTsg (2012)
Poland 1 17 riverVistula 2011/ m S4 121 n.a. 28e414 Niewiadowskaetal.
(Wisła)near 12 DDTsi (2014)
Cracow
Shad(clupeid(belongingtotheherringfamily),pelagic,omnivorous)
Italy 1 5times LakeMaggiore 2002 m S2 ca.90- 125 n.a. Bettinettietal. InfluencedbyDDTfactorywhichclosedin1996
e04 DDTsk190j (2006)
Italy 1 3times LakeMaggiore 2008 m S3 28e56d 43d n.a. Bettinettietal. InfluencedbyDDTfactorywhichclosedin1996
DDTsg (2010)
Italy 1 4times LakeMaggiore 2009 m S3 n.a. ca.55j n.a. Bettinettietal. InfluencedbyDDTfactorywhichclosedin1996
DDTsg (2012)
Italy 1 4times LakeMaggiore 2010 m S3 n.a. ca.80j n.a. Bettinettietal. InfluencedbyDDTfactorywhichclosedin1996,
DDTsg (2012) newreleaselikelyin2009,whichshowedfirst
inzooplanktonandtheninfish
Italy 1 5-6per LakeComo,Como2005 m S3 54e120d 84d n.a. Bettinettietal. InZebramusselslwDDTconc.increased>100-
year branch e09 DDTsg (2008;2016) foldbetween2003and2005inlakeComoand
lakeIseo,butnotlakeMaggiore.InLakeComo
fish,thehighestvalueswerealsoin2005.
Meltingglaciersweresuggestedassecondary
DDTsource.
Italy 1 5-6per LakeComo,Lecco 2007 m S3 79,98d 98d n.a. Bettinettietal.
year branch e08 DDTsg (2016)
1 5or6 LakeIseo 2007 m S3 71d n.a. n.a. Bettinettietal.
DDTsg (2008)
Perch(piscivorous)
Italy 1 5times LakeMaggiore 2002 m S3 ca.14- 93j n.a. Bettinettietal. InfluencedbyDDTfactorywhichclosedin1996
e04 DDTsg250j (2006)
n.a.enotavailable.
a Wherepossiblesiteswereassigned“contaminated”or“background”statusbasedontheinformationinthepaper,butsometimesthechoicewasbasedonthereported
concentrations.
b w:wholebody,m:muscle,thefootnotesdescribeanydatatransformationsperformedwhereconcentrationswerenotpresentedonawetweightbasis.
M.D.Jürgensetal./Chemosphere162(2016)333e344 337
P
c pp’DDEþpp’DDD.Inthecurrentstudythisprovidedonaverage85%of DDTsatthecontaminatedsiteand92%attheothersites.
d Convertedfromlipidweight(lw)towwusing(average)lipidcontentsgiveninthereference.
e Convertedfromdwtoww,assuming26%drymatter,whichisthevalueforaEU00standard”fish(EuropeanUnion,2014)andalsoapproximatelytheaveragefortheroach
inthisstudy. P
f op’DDT,pp’DDT,op’DDE,pp’DDE,op’DDD,pp’DDD( DDTs). P
g All3pp'congeners.Inthecurrentstudythisprovidedonaverage90%of DDTsatthecontaminatedsiteand93%attheothersites.
h Onecompositesampleperyearandsite.ThelastfiveavailableyearswereselectedP.
i All3pp'congenersþop’DDT.Inthecurrentstudythisprovidedonaverage98%of DDTsatthecontaminatedsiteand94%attheothersites.
j ConvertedfromlwtowwusingtheEU“standard”fishlipidcontentof5P%(EuropeanUnion,2014)toestimateafreshweightbasedconcentration.
k pp’DDEþpp’DDT.Inthecurrentstudythisprovidedonaverage82%of DDTsatthecontaminatedsiteand75%attheothersites.
fromseveralsitesalongtheRiversGlen,NeneandThamesaswell procedure without the addition of fish homogenate (i.e. extract-
as three Thames tributaries, the Kennet, Lee (also commonly ingonlysodiumsulphate).Theinstrumentlimitofdetection(LOD),
spelled“Lea”)andStort(Fig.1).ThesiteswerechosenfromEnvi- defined as the lowest observable standard was between 1 and
ronment Agency fish population monitoring sites. They varied in 6.25 pg/ml for the analysed chemicals, which is equivalent to
impactfromtreatedsewagebetween1%oftheflowonaveragein 5e31 ng/kg for a 5 g sample. To measure metal concentrations
theGlenand43%intheStort(TableS1)butwereotherwisethought 1e2.5 g wet weight was weighed into a PTFE vial, digested in a
torepresentfairlytypicallevelsofchemical inputsfor theirarea, microwavedigester(MARSXpress,CEM)with10mLultrapureni-
rather than deliberately choosing known pollution hotspots. De- tric acid (Baker, Ultrex II), then made up with ultrapure water
scriptionsoftheriversLeeanditstributaryStortarefoundinSnook (>18MU/cm)to25mLbeforeanalyzingusingaPerkinElmerElan
andWhitehead(2004)andoftheothersampledriversinJürgens DRCIIinductivelycoupledplasmamassspectrometer(ICPMS)in-
et al. (2013) and detailed catchment information is in Table S1. strument.Certifiedreferencematerials(DORM-3andadditionally
The catchments above most sampling sites were dominated by DOLT-4 for later batches, both from National Research Council,
agriculturallandexceptforthesiteontheriverLeewherethetown Canada) were analysed alongside each batch and for copper
of Lutonwith its suburbs occupies a largeproportion of the area measuredat83e116%ofthepublishedvalues.TheLOQwasabout
(12%urbanand35%suburbanorruraldevelopments). 4mg/kgwwfora2gsample.Thecurrentpaperisfocusedmainlyon
pesticides,notmetals,sotheonlymetalincludedintheresultsand
2.2. Chemicalanalysis discussion is copper, because it is frequently used as a fungicide.
MercuryisdiscussedinJürgensetal.(2013)andallresultscanbe
Forallsamplesfrom2008onwardsthewholefrozenfishwere foundinJürgens(2015).
groundintoapowderwithoutdefrostingthemusingacryogrinder
(SPEX SamplePrep 6850). The resulting frozen fish powder was 2.3. Geneticanalysis
divided into pre-cooled 20 ml glass scintillation vials and stored
at(cid:3)80(cid:4)Cuntiluse.Forthenineroachcollectedin2007theprocess Toinvestigatetheoriginofroachusedforanalyticalchemistry
wasslightlydifferent,inthattheliverandgallbladderaswellasa fromtheLee,Kennet,Stort,andtheThamesatCastleEaton,allfish
blood sample were removed before freezing and the remaining fromthese sites weregenotyped using 14 DNA microsatellites as
carcass was divided dorsally with one half (or part of it for very described by Hamilton et al. (2014). Fifty roach were also geno-
largeindividuals)beingusedforPOPsanalysis.Theotherhalfwas typedfromCalvertonFishfarm,whichisthesourceoffishusedfor
cryo-groundandusedformetalsanalysis.ForPOPsanalysisca.5g restocking by the England and Wales Environment Agency.
of the cryo-ground homogenate was dried by mixing with 30 g Together with University of Exeter's existing dataset (Hamilton
sodium-sulphate(orinthecaseofthe2007fish,thesamplewas et al., 2014), genotypes from 1858 roach from 33 different river
groundwithsodium-sulphateinapestleandmortar).Afteraddi- stretches in England were available. The whole dataset was ana-
tionofrecoverystandards(13CPCBmix:28,52,101,138,153,180 lysed using the Bayesian clustering program STRUCTURE v2.3.3
and PBDE mix 51, 128, 190), DDTs and other persistent organic which assigns individuals to probable genetically differentiated
pollutantswereextractedovernightwithdichloromethane(DCM) clusters (K) using a Bayesian approach (Pritchard et al., 2000). In
inaSoxhletapparatus.Lipidcontentwasdeterminedgravimetri- STRUCTURE, analyses were conducted using the standard model
callyfroma subsample of the Soxhletextract and the restof the and a “locprior” model and used 125,000 repetitions (burn-
DCMwasevaporatedinavacuumrotaryevaporatorandreplaced in ¼ 25,000 iterations). The number of clusters K was set to 16
with 10 ml hexane, which was reduced to about 1 ml. This was whichwaspreviouslyshowntodifferentiateroachfrommanyof
added to a glass column with 11 g acidified silica (200 ml silica theriverstretches(Hamiltonetal.,2014).
bakedat450(cid:4)Candacidifiedwith25mlconcentratedsulfuricacid)
andelutedwithhexaneasafirstcleanupstep,whichremovesthe 3. Resultsanddiscussion
fats.Thenthesamplewaspassedthroughagelpermeationchro-
matographycolumnwith50:50hexane:DCMandonlythefraction 3.1. HighlevelsofDDTandsomeotherchemicalsinfishfromthe
from 17 to 51 ml collected as second clean up step to remove LeeatWheathampstead
moleculesoutsidethesizerangeofinterest.Thesolventwasthen P
againreplacedwithhexaneandthesampleaddedto25mlinternal All the chemicals in this Pstudy ( PDDTs, chlordane (aþg),
standards (PCB 30, 13C-PCB141, 13C-PCB208, BDE69, BDE181) in lindane(g-HCH),HCB,copper, 7PCBs, 6PBDEs)weredetect-
dodecane,beforeevaporatingthehexane,sothatthewholesample able in all fish with statistically significant differences (ANOVA,
wascontainedinthe25mldodecane.Theextractswereanalysedby a ¼ 0.05) between at least some of the sampling sites (see
gas chromatography e mass spectrometry, single ion monitoring TablesS4eS10fordetaPilsofthestatisticalcalculations).Ofpartic-
usinga50mVarianCP-SIL8CBPesticidecolumn(Varian-Chrom- ular note were the DDTs concentrations in fish from the
pack, Middelburg, The Netherlands) with electron Wheathampstead site, which were very significantly higher
impactþionisation.Standardsandblankswererunalongwiththe (p<0.001,TabPleS4)thanatanyoftheothersites(Figs.2aand3).
samples.Theinstrumentblankcontainedonlysolventandproce- The average DDTs concentration in roach from that site was
dural blanks went through the whole extraction and clean-up 88±70mg/kgww,comparedtotheoverallaverageof4.8±3.1mg/
338 M.D.Jürgensetal./Chemosphere162(2016)333e344
Table3
ConcentrationsofDDTanditsmetabolitesinfishfrom“background”sitesa.Dataforsomecommonlymonitoredfreshwaterorbrackishwaterfishspeciescaughtinthelast20
years(1995e2014).Eels,whichoftenhavehigherPOPconcentrationsthanotherspeciesbecauseoftheirhighfatcontentandassociationwithsediments,havealreadybeen
reviewedforpp’DDEinJürgensetal.(2015).AgriculturaluseofDDTceasedinthe1970sor1980sinallthecountriesmentionedinthistable.
P
Country n n Sites Year(s)SamplebDDTs Site Overall Overall Ref. Comments
sites samples averages average range[mg/
[mg/kg] [mg/kg] kg]
Barbel(cyprinid,bottomfeeder)
UK 1 1 RiverDon 2008 m pp’DDEc10 n.a. n.a. Roseetal.(2015)
Greece 2 2 RiverNestos 2004 m S3 0.25,0.47 n.a. n.a. Christoforidisetal.
DDTsd (2008)
Italy 1 4 PoRiver 1995 m S2 50f,g n.a. n.a. Vigano(cid:1)etal.(2000)
DDTse
Macedonia3 11(?) LakesOhrid,Prespa, ca. m S3 8.8,13,16 n.a. ca.9-17 (Veljanoska-Sarafiloska
Dorjan 2005? DDTsd etal.,2011,2012;2013)
Spain 5 5 Ebroriverbasin 2003 w S6 <LOD-19g 8.7g n.a. Lacorteetal.(2006)
pooled DDTsh
Spain 2 12 Cincariveru.s.ofdicofol 2002 m S6 51,11 24 8e64 delaCaletal.(2008)
factory DDTsh
Spain 1 1 Cincariverimmediately 1999 m S6 n.a. n.a. 29e104 quotedindelaCaletal.
upstreamofdicofolfactory DDTsh (2008)
Spain 2 10 upperþmiddleJarama 2006 m S3 0.92,2.7 1.8 0.24e3.66 Nicolaetal.(2014) DDT/DDE<0.5–>old
river e09 DDTsd contamination
Bream(cyprinid,omnivorous,bottomfeeder)
UK 3 3 RiverDon,2ponds 2008 m pp’DDEc<5,<5,10 n.a. n.a. Roseetal.(2015)
Czech 1 2 RiverOder(Odra)at 2008/ m S6 24 24 23e25 Hra(cid:1)dkov(cid:1)aetal.(2012)
republic Ostrava 09 DDTsh
Czech 24 24 variousriversand 2006 m S6 1e65 17 n.a. Cervenyetal.(2014)
republic pooled reservoirs e10 DDTsh
Germany 2 10 RiverSaar 2009 m S2 18,24 21 14e30 Umweltbundesamt onesitecleardecrease
e13i DDTse (2016) 1994e2013,63->19
Germany 1 3 LakeBelau(cleancontrol 2009 m S2 2.2 n.a. 1.4e3.8 Umweltbundesamt nocleartrend1997
site) e13i DDTse (2016) e2013
Germany 4 18 RiverRhine 2009 m S2 11e39 28 8e55 Umweltbundesamt nocleartrend1997
e13i DDTse (2016) e2013
Germany 3 15 RiverDanube 2009 m S2 10e13 12 4.4e20 Umweltbundesamt nocleartrend2004
e13i DDTse (2016) e2013
Italy 1 2 RiverPo 2014? liver S6 27f,g n.a. 15,39 Vigano(cid:1)etal.(2015) 3furtherspecieswere
DDTsh mostlylower
Moldova 1 3 RiverDniesternear 2001 m S6 19 n.a. n.a. Sapozhnikovaetal.
Dubassari DDTsh (2005)
Poland 8 74 varioussites 2011/ m S4 11e65 37 3.8e295 Niewiadowskaetal.
12 DDTsj (2014)
Poland 1 5(?) Szczecinlagoon 1995 m S3 65 n.a. n.a. CiereszkoandWitczak
DDTsd (2002)
Romania 2 4 EnisalaandCaraorman, 2001 m S5 5.9,32f 19f n.a. Covacietal.(2006)
DanubeDelta DDTsk
Roach(omnivorous)
Southern 12 71 Thamescatchment,Glen, 2007 w S6 1.8e12 4.8 0.6e14.3 currentstudy
England Nene e12 DDTsh
UK 2 2 2ponds 2008 m pp’DDEc<5,2 Roseetal.(2015)
Bulgaria 2 2 LakeBurgas,LakeMandra 2014 m S3 2.31,2.59 2.45 e Georgievaetal.(2015)
pooled DDTsd
Moldova 2 6 riverDniesternear 2001 m S6 8,19 14 n.a. Sapozhnikovaetal.
DubassariMoldova DDTsh (2005)
Poland 1 5(?) Szczecinlagoon 1995 m S3 23.7 n.a. n.a. CiereszkoandWitczak
DDTsd (2002)
Poland 5 25 Oderdelta 2003 m S3 n.a. 9.1 n.a. Tomza-Marciniakand
DDTsd Witczak(2009)
Poland 9 52 varioussites 2011/ m S4 4.5e44 15 3.0e61 Niewiadowskaetal.
12 DDTsj (2014)
Romania 1 2 Enisala,DanubeDelta 2001 m S5 11.4f n.a. n.a. Covacietal.(2006)
DDTsk
Spain 1 1 riverZadorra(Ebroriver 2003 w S6 17g n.a. n.a. Lacorteetal.(2006)
pooled basin) DDTsh
Shad(clupeid(belongingtotheherringfamily),pelagic,omnivorous)
France 2,2 69 Vilaineestuary 2004 m S3 2.0e4.7g 8.7g n.a. BocqueneandAbarnou
times e05 DDTsd (2013)
Perch(piscivorous)
UK 5 5 3ponds,riverDon, 2008 m pp’DDEc<5 <5 n.a. Roseetal.(2015)
Thames
Bulgaria 1 1 LakeMandra 2014 m S3 1.81 n.a. n.a. Georgievaetal.(2015)
pooled DDTsd
Latvia 9 9 6lakes,2rivers 1995 m pp’DDEc0.6e5.9f 2.0f n.a. (Valtersetal.,1998)
pooled
Latvia 4 43 4lakes 1996 m S2 0.5e5.9f 2.3f n.a. (Valtersetal.,1999a)
e97 DDTse
M.D.Jürgensetal./Chemosphere162(2016)333e344 339
Table3(continued)
P
Country n n Sites Year(s)SamplebDDTs Site Overall Overall Ref. Comments
sites samples averages average range[mg/
[mg/kg] [mg/kg] kg]
Latvia 6 60 riversDaugavaandLielupe1997 m S3 1.0e2.8f 1.8f n.a. (Valtersetal.,1999b)
DDTsd
Moldova 2 5 riverDniesternear 2001 m S6 6,10 8 n.a. Sapozhnikovaetal.
Dubassari DDTsh (2005)
Romania 1 2 Caraorman,DanubeDelta 2001 m S5 10f n.a. n.a. Covacietal.(2006)
DDTsf
n.a.enotavailable.
a Wherepossiblesiteswereassigned“contaminated”or“background”statusbasedontheinformationinthepaper,butsometimesthechoicewasbasedonthereported
concentrations.
b w:wholebody,m:muscle,thefootnotesdescribeanydaPtatransformationsperformedwhereconcentrationswerenotpresentedonawetweightbasis.
c Inthecurrentstudypp’DDEprovidedonaverage75%of DDTs. P
d All3pp'congeners.Inthecurrentstudythisprovidedonaverage90%ofPDDTsatthecontaminatedsiteand93%attheothersites.
e pp’DDEþpp’DDD.Inthecurrentstudythisprovidedonaverage85%of DDTsatthecontaminatedsiteand92%attheothersites.
f Convertedfromlipidweight(lw)towwusing(average)lipidcontentsgiveninthereference.
g Convertedfromdwtoww,assuming26%drymatter,whichisthevalueforaEU00standard”fish(EuropeanUnion,2014)andalsoapproximatelytheaveragefortheroach
inthisstudy. P
h op’DDT,pp’DDT,op’DDE,pp’DDE,op’DDD,pp’DDD( DDTs).
i Onecompositesampleperyearandsite.Thelastfiveavailableyearswereselected.P
j APll3pp’congenersþop’DDT.Inthecurrentstudythisprovidedonaverage9P8%of DDTsatthecontaminatedsiteand94%attheothersites.
k DDTsexceptop’DDT.Inthecurrentstudythisprovidedonaverage92%of DDTsatthecontaminatedsiteand99%attheothersites.
the fungicide HCB nor PCBs concentrations were elevated in
Wheathampsteadroach,beingsignificantlydifferenttoonly1/11or
3/12 other sites respectively (Tables S7 and S9). Although the
Wheathampstead values for PBDE flame retardants were among
thehighestmeasured,thedifferencewasonlysignificantcompared
tohalfoftheother12sites(TableS10),butcopperconcentrations
were also significantly higher at Wheathampstead than at most
(10/12) of the other sites (Table S8). In summary: compared to
others inPthe study, the fish from Wheathampstead had strongly
elevated DDT levels and slightly elevated levels of the in-
secticides chlordane and lindane and also of copper, which may
Fig.1. MapoftheUK,showingfishsamplingsites(outline©DanielDalet/d-maps. havebeen usedas afungicide, while thesitedid notparticularly
com). standoutforthefungicideHCB,PCBs,orPBDEs.
3.2. Alikelysourceofthecontamination
kgwwforallothersites.TheroachfromWheathampsteadalsohad
unusually high lipid contents (8.8%e13.6%, average 11.6%,
AninvestigationintopotentialsourcesofDDTandotherpesti-
comparedto1.4%e7.9%,average4.7%intheremainingroach),but
cides in Wheathampstead roach revealed evidence of DDT pro-
thataloneisinsufficienttoexplainthealmost20folddifferencein
duction at Murphy Chemical Company in an advertisement
DDTconPcentrationscomparedtotheremainingfish. (reproduced in Fig. S3) published in a 1940s book on pesticides
The DDTsconcentrationsattheothersitesweremoresimilar, (Massee, 1946), where they stated “[…] we are equally in the
butthetwomostupstreamsitesintheThamescatchment(New- forefront with insecticides containing D.D.T. […]” and the com-
buryontheKennetandCastleEatonontheThames)hadthelowest pany'saddresswas given as “Wheathampstead, Herts”. The Mur-
concentrations(1.8±0.5and2.4±0.9mg/kgww),eachsignificantly
phy Chemical Company was founded when Murphy and Son
(a < 0.05) lower than seven of the 12 other sites. The highest
BrewerySuppliesacquiredtheWheathampsteadsitein1928and
concentrations, apart from Wheathampstead, were found in the
branchedoutintoagriculturalchemicals.Theagro-chemicalsarm
fourroachmeasuredfromtheriverGlen(12.4±1.9mg/kg),andthe “MurphyChemicalCompanyLtd”inWheathampsteadwassoldto
difference was significant compared to six of the 12 other sites
Glaxo in 1956 (http://www.murphyandson.co.uk/heritage) and
(TableS4andFig.3).
thenchangedhandsafewmoretimes(Dalgety,Dow,Fisons).Both
Some, but not all, of the other chemical concentrations
production and development of pesticides took place in Wheat-
measuredwerealPsoelevatedinWheathampsteadfish,thoughnot hampstead between about 1931 and 1982 and test beds were
as much as for DDTs (Fig. 3): For the insecticide chlordane, locatedclosetotheriverLee.In1967amajorfishkillwascausedby
WheathampsteadcamesecondtotheSunburytoMoleseyreachon
aspillof thepesticideMecarbam(notmeasuredinthisstudy)at
the river Thames, both being significantlyhigher than nine of 11
Wheathampstead (Environment Agency, 2010). Since the factory
othersitesanalysed.Theinsecticidelindane(g-HCH)wasanalysed
closed, an attempt has been made to clean up the considerable
withthesamemethodasforDDT,whichforg-HCHisonlysemi-
pesticidecontaminationofthegroundbyremovingcontaminated
quantitative.Extractsoffishfromothersiteshadpreviouslybeen
soilandcleaninggroundwaterbytreatingitonsiteinreedbeds
analysed with a separate GCMS method for HCH (Jürgens, 2015;
whichwerecompletedin1998(http://www.oceans-esu.com/case-
Jürgens et al., 2015) but the results are noteasilycomparable, so
studies/)andappearedtobestilloperationalwhentheleadauthor
semi-quantitativedatafromonlyfivesitesareshownhere.Theg-
revisited the area in summer 2016. Comparisons of historic and
HCHconcentrationsinWheathampsteadwerehighestandsignif-
recent aerial photos revealed the approximate extent of the area
icantlydifferentfromallfourothers(TableS6).Bycontrast,neither
formerly occupied by the company, which has been redeveloped
340 M.D.Jürgensetal./Chemosphere162(2016)333e344
Fig.2. Concentration(a)andrelativecontributionofcongeners(b)ofDDTanditsdegradationandby-productsDDEandDDDinroach.Individualsateachsiteareorderedbyyear
andlength(cm).Sitesoneachriverareorderedbydistancefromthesource(river-km).Br.-Bov:Bray-Boveney203e209km.TheCanadianTissueResidueGuidelineforthe
protectionofwildlifeconsumers(CanadianCouncilofMinistersoftheEnvironment,1999)isalsoshown(thereiscurrentlynoequivalentEUguideline).
withdomesticandcommercialproperties(Fig.4).Thesiteappears pesticides in this population. Roach from the Stort and the
tohavestretchedfromtheriverbanktothenorthernedgeoftown, Kennet also showed high genetic similarity to fish previously
bordering the river for several hundred meters immediately up- sampledfromthesameriverstretches(Fig.S4).
streamofthefish-samplingsite.Withoutknowingaboutthehis-
tory of the site, we had collected fish right next to the former
3.3. Congenerdistribution
factory with its associated testfields. This is likelytoexplainnot
P
onlytheveryelevatedDDTconcentrations,butalsotherelatively Atallsites,the DDTsconcentrationwasdominatedbypp’DDE,
highconcentrationsofthepesticideschlordaneandlindane,andof the main aerobic degradation product of pp’DDT. However, at
copper.Copperismentionedintheadvertisement(Fig.S3)under
Wheathampsteadroachhadanunusualcongenerdistributionwith
the name “Bordeaux Powder”, a mixture of copper (II) sulphate relatively high levels of op’DDT, and lower proportions of both
(CuSO4)andslakedlime(Ca(OH)2),whichiscommonlyusedasa op’DDDandpp’DDD,thanotherfishinthestudy(Fig.2b).Ingen-
fungicide,soitselevatedlevelsarelikelytobealsorelatedtothe eral, a higher DDT/DDE ratio reflects more recent contamination,
formerpesticidefactory.Bycontrast,HCB,whichwasbannedfrom while a high op’/pp’ ratio may indicate that DDT contamination
agricultural use in the European Union in the early 1980s along
originated from impurities in dicofol (Ricking and Schwarzbauer,
withchlordaneandDDT(EEC,1978),doesnotappeartohavebeen
2012). The higher proportion of untransformed DDT may there-
releasedinWheathampsteadtoasignificantdegreeandtherela-
forepointtoarelativelyrecentcontamination,despitethefactthat
tivelyhighPBDEcontaminationthereislikelytobearesultofthe
productionceasedatthefactoryin1982.Thelowerproportionof
original reason the site was chosen, namely its relatively high
bothDDDcongenersinWheathampsteadcomparedtoothersites
sewageeffluentcontentoftheriverflow(SchrederandLaGuardia,
wouldindicatethatthetransformationprocessesoccurringthere
2014;Jürgens,2015).
were mainly aerobic, hence favouring DDE, while at other sites
Previous analysis of the roach population genetic structure anaerobic processes in sediments may have had more influence.
revealed the population at Wheathampstead to be genetically
Thereis,however,alsoapossibilitythatdifferingcongenerdistri-
distinctfrompopulationssampledabout5kmupstreamand5km
butions between Wheathampstead and other sites resulted from
downstream,andmigrationintoandoutoftheWheathampstead
the use of different formulations. Murphy's factory had a devel-
population to be minimal (Hamilton et al., 2014). Bayesian clus-
opmentdepartmentforimprovingpesticides,soitispossiblethat
tering analysis of the DNA microsatellite genotype dataset in the
someformulationsorvarietiesofformulationsthatwerenotornot
programme STRUCTURE (Pritchard et al., 2000) showed that the yetonthemarketweretestedonthefieldsclosetotheriverLee.In
Wheathampstead roach used for chemical analysis in this study
thecaseofDDT,thisseems,however,unlikelybecauseitsconsti-
hadhighgeneticsimilaritytoroachsampledpreviouslyfromthe
tutionvariedlittlebetweenmanufacturers.
same river stretch, but were distinct from those from the Lee There were no obvious differences between op’ and pp’ ratios
tributaryStort(Fig.S4).Importantly,theywerealsodistinctfrom
between the different sites for any of the DDT family, and the
roachfromCalvertonFishFarm,soarenotrecentlyrestockedfish. overall distribution showed a strong dominance of pp’ congeners
Thissuggeststhatthesefishhadbeenconfinedtothisriverstretch
which points to contamination from DDT itself rather than from
fortheirentirelives,helpingtoexplainthehighconcentrationsof
impurities in dicofol (Ricking and Schwarzbauer, 2012). A shift
M.D.Jürgensetal./Chemosphere162(2016)333e344 341
P P
Fig.3. DDTconcentrationsinroach(detailsinFig.2P)comparedtotheinsecticideschlordaneandlindane(g(cid:3)HCH),thefungicideHCB,copper,PCBs( indicatorcongeners28,
52,101,118,138,153,180)andPBDEflameretardants( indicatorcongeners28,47,99,100,153,154).Fishwerecaughtbetween2007and2012,rivercodesaregiveninFig.1,the
distanceinbracketsreferstothedistanceofthesamplingsitefromthesourceoftheriver.Boxplotsshow,25-,50-,and75-%ileswith10-and90-%ilewhiskers.Wheathampsteadon
theLeehasbeencolouredin.*Lindaneconcentrationsaresemi-quantitativebecausetheywereanalysedwiththestandardGC/MSmethod,whichislessaccurateforHCHsthanthe
methodusedinJürgensetal.(2015).Groupsthatdonotshareacommonletteraresignificantlydifferentfromeachotherata¼0.5(detailsinTablesS4eS10).
342 M.D.Jürgensetal./Chemosphere162(2016)333e344
clearly followed the expected exponential decline with R2 > 0.5
(R2 ¼ 0.86, p < 0.001), the concentrations at five more locations
showedalessclearbutsignificant(a<0.05)downwardtrendand
atfoursitesasignificantincreasewasobserved.IntheUSA,con-
centrationsinlargemouthbasscaughtnearaformerDDTfactory
thathadcPlosedin1963,remainedatveryhighlevels(averageabout
9 mg/kg DDTs) for three decades (Hinck et al., 2009a) and at
some locations in GermanPy (Heinisch et al., 2005) and Italy
(Bettinetti et al., 2012) DDTs concentrations in fish and
zooplankton increased many years after a nearby factory had
stopped processing it. This suggests a continuing (secondary)
sourceofDDTssuchascontaminatedsoilsorsediments.
3.5. Implications
ThereviewofrecentEuropeandata(Tables2and3)showedthat
Fig.4. SchematicmapofWheathampstead,withthefishsamplingsite,obstaclesto while the DDT levels in fish influenced only by diffuse pollution
fish movement, and the approximateextentof the areaformerlyoccupied bythe fromitsagriculturalusearenowlow,somefishfromindustrially
MurphyChemicalCompany.
impactedsiteshadbodyburdensassociatedwithnegativeeffects
onthefishthemselvesortheirpredators(Table1).Thisshowsthat
towardshigherpp’/op’ratiosforthetransformationproductswas even decades after its ban this chemical still has the potential to
howevervisible(Fig.2b).Sincethepp’orop’configurationremains causeharm.AlltenroachfromWheathampstead,butonlyoneof
the71othersinthecurrentstudyexceededtheCanadianEQSfor
duringtransformationofDDTtoDDEorDDD,onewouldexpectthe
ratiotobeabout85%pp’and15%op’asintheparentcompound theprotectionoffish-eatingbirdsof14mg/kg.Theconcentrations
(ATSDR, 2002). However, for the pp’ congeners pp’DDE was the insomeofthemwerehigherthanreportedeffectconcentrationsin
largestcontributorwithpp’DDDplayingaminorroleandtheun- early life stages of fish and approaching reported effect concen-
transformed pp’DDT being almost absent, whereas for the op’ trations for adults (see Table 1), while whole-life exposure or
multigenerationaleffectsarelargelyunknown.However,therewas
congeners, DDD was generally more prominent than DDE and
op’DDToccurred at similar concentrations to pp’DDT despite the noevidenceofactualharmtotheroachinthecurrentstudyasthe
verydifferentstartingratio.Thisconfirmsthatthefateofop’and populationof thecontaminated siteappeared tobein goodcon-
pp’ DDT congeners in the environment differs, reflecting their ditionwithlargerandfatterindividualsthanwerecaughtatother
sites.Furthermorethereisgeneticevidenceforapopulation,which
differentphysico-chemicalproperties(RickingandSchwarzbauer,
isself-sustainingratherthanconsistingofrecentimmigrantsand
2012)(TableS2).
whichhasrecoveredfromapopulationbottleneckeperhapsdue
3.4. DDTcontaminationinotherrecentEuropeanfreshwaterfish tothe1967fishkill(Hamiltonetal.,2014).Howeverhealtheffects
onindividualsdonotnecessarilyresultinpopulationleveleffects
studies
(Hamilton et al., 2015) and we cannot rule out adaptation that
P enables fish to tolerate the harmful effects of exposure. Whether
The results from the current study were compared to DDTs
contamination of commonly monitored fish species collected at fish-eating birds near this and other contaminated sites are
adverselyaffectedisnotknown,butisatleastapossibilityjudging
freshwatersitesorlagoonsinEuropebetween1995and2014,i.e.
well after agricultural use of DDT was banned in the EU in 1981 fromtheeffectlevelsrepPortedinTable1.
Theunexpectedhigh DDTsconcentrationsfoundinsomefish
(Table2listsvaluesfromknownorsuspectedpollutedsites,while
“background” values are found inTable 3). Concentrations in our inthisstudycanserveasanexamplehowmonitoringchemicalsin
fishtissuecanbeusefultospotpreviouslyunknownproblems.A
roachfromsitesotherthanWheathampsteadwerefairlytypicalfor
recentEuropeandatafrom“background”sites(Table3),probably spikeinatemporalorspatialserieswouldindicatethatsomething
reflectingthelegacyofhistoricagriculturaluseofDDTandpossibly unusual happened, which warrants further investigation. In this
case,aformerpesticidefactorywasidentifiedasthelikelycauseof
dicofol. However, in a number of places high concentrations of
DDTsinfishwereobservedlongafteritsuseandproductionceased thepollutioninfish.
(Table2).Someareasofhighcontaminationwerelinkedto(former)
factories producing DDT (Bettinetti et al., 2006, 2010, 2012;
Acknowledgements
Hra(cid:1)dkova(cid:1) et al., 2012) or dicofol (Lacorte et al., 2006; de la Cal
et al., 2008). In other cases DDT hotspots were reported without
TheauthorswishtothankDEFRA(projectnumberCB0447)and
identifyingasource(Niewiadowskaetal.,2014)orasinthecaseof
CEH for funding this research, the Environment Agency for
the river Elbe, the legacy of several large industrial areas (pre- providing the fish and Dave Hughes, Montse Auladell-Mestre,
sumably including pesticide manufacturing) were considered
Qiong Lu, Alan Lawlor and Hayley Guyatt for help with analyses.
responsibleforelevatedDDTlevelsinfishalonglargepartsofits
P. Hamilton also received support from Natural Environmental
length through two countries (Hra(cid:1)dkova(cid:1) et al., 2012;
Research Council (NERC), grant numbers NE/G019355/1 and NE/
Umweltbundesamt,20P16). K004263/1.
Inafewlocations, DDTcontaminationoffishwasmonitored
overseveralyears.Itmightbeexpectedthattheconcentrationsin
fishwoulddeclinefollowingtheendofproductionanduseofDDT. AppendixA. Supplementarydata
However,thiswasnotalwaysobserved.Forexample,inGermany
DDEþDDDweremonitoredinbreamfor10e21yearsat17sites Supplementarydatarelatedtothisarticlecanbefoundathttp://
(Umweltbundesamt,2016).OftheseonlyonesiteontheriverSaar dx.doi.org/10.1016/j.chemosphere.2016.07.078.