Efficiency of Livestock Carcass Detection Dogs

(1)

RALA-00283;NoofPages199

Research Article

Efﬁciency of livestock carcass detection dogs

By Inger Hansen and Erlend Winje

On theGround

• There is an increasing use of carcass detection dogsto findremainsofdeadlivestockinNorwe- gianrangelands.Buthoweffectivearethesedogs actually?

• Wecomparedtheefficiencyofapprovedcarcass detection dog equipages (CDEs, i.e., dog and man) withpeoplesearching for sheepcarcasses withoutdogs.

• CDEs found significantly more carcasses than peoplewithoutdogs,andkilometerstraveledand minutesspentpercarcassdetectionindicatedthat dogswere >3xas effectiveintheirsearch.How- ever,CDEsfoundonly1in4ofthecarcasseslaid outexperimentally.

• Thetrainingprogram forCDEsinNorway isnow adjustedtoimprovethequalityoftheequipages.

• The effort of sheep CDEs is important to Nor- wegiansheepfarmersapplyingforcompensation becauseof theincrease inpercentage ofproven lossescausedbyprotectedcarnivores.

• In the future carcass detection dogs in Norway couldbeusedforwildlifeconservationandman- agementtasks.

Keywords: efficiency,cadaver,predation,rangeland,sheeplosses,sniffingdogs.

Rangelands43(5):194–199 doi10.1016/j.rala.2021.03.004

Introduction

Closeto2millionsheepgrazeonopenrangelandpastures inNorwayeachsummerandaround100,000ofthesearelost asa resultofcarnivores,diseases,andaccidentsevery year.¹ In2019,compensationswerepaidfor17,567ewesandlambs thatweredocumentedorassumedkilledbyprotectedcarni- vores (wolverines[Gulogulo],lynx[Lynxlynx],brownbears

[Ursusarctos],gray wolves [Canis lupus],and goldeneagles [Aquilachrysaetos]).²AlthoughinspectionsofsheepinNor- wegianmountainandforestrangesarefrequent,³ itischal- lengingtoﬁndsheepcarcasseswithinlargegrazingareasand carcassescanbehiddeninvegetationoraninaccessibleloca- tion.Furthermore,carnivores,scavengers,and maggots(i.e., ﬂylarvaefoundindecayingmatter)consumepartsofthecar- cassleavingonlydrybonestobefound.Infact,lambcarcasses canbegonewithin24hoursafterdying.⁴Toreceivecompen- sationsforsheepkilledbylargecarnivores,farmersneedtolo- catecarcassesandtheNorwegianNatureInspectorate(NNI) hastoprovethatprotectedcarnivoresmostlikelyarethecause ofdeath,orthereareothercircumstancesshowingpredom- inanceofprobabilitiesthatthesheeparekilledbylargecar- nivores.⁵However,<10%oflivestocklossescompensatedare provenbytheNNIasdepredated.⁶

Dogs’sense of smell is far better than that of humans.

Humans have about5 million olfactory receptors,whereas dogs(bloodhounds)have>100million.⁷ Dogscansmellan item with particle concentration levels at one to two parts per trillion.⁸ Thus,dogs have been used for olfactory dis- crimination (nonbiologicalscents and biologicalscents)for decades,suchasdetectionofbombs,drugs,cancer,andhuman remains.^9-12 Owing to their extensive olfactorycapabilities, dogsareincreasinglybeingusedinwildlifemanagementand conservationinarangeoftasksincluding1)locatingwildlife andassessingpopulationstatus,2)facilitatingspecimenand carcasscollection,3)detecting scat,4) capturingandmark- ing wildlife,5) studyingwildlifebehavior,and6) managing wildlifedamage.¹³

Dogs searchingfor birdand batcarcassesatwindfarms had a higherdetection ratethan humans (80% vs.20%).¹⁴ In anotherstudy,songbirdcarcasseswererecoveredapprox- imately12:1tothosemissedfordogsand1:1forhumans.¹⁵ However,onlyrecentlyhavedogsbeentrainedtolocatethe remainsoflivestockcarcasses.Whichodorsareusedbydogs intheirsearchforcarcassesarenotknown,butsometypical volatileodorsfromcadaversarecadaverineandputrescin.¹⁶

NorwegianCarcassDetectionDogs¹⁷(NCDD)isanor- ganization aimed at improving the competence of carcass detectionequipages (CDEs,i.e.speciallytrained dogs with guides).Withexperiencefrompreparationofrescueandmil- itarydogs,theassociationhasdevelopedacourseandanap- proval testfor CDEs.The basic courseconsists of3 hours oftheoryand 4hours ofindividualtraining.Tobecomean approvedequipage,onemustcompletethebasiccourse,ﬁnd

(2)

3outof4carcassesorobjectsplacedinalimitedforestarea (approvaltest),anddocument≥30hoursofactivesearching for sheepcarcassesinopen rangelands.Asof August2019, 207 equipages havebeen approved.Thetraining ofdogs is basic with alow threshold,meaningthat classic conditioning/positive reinforcementisusedto encouragethe dogsto ﬁnd carcasses.^18-20 Positivereinforcement strengthensa be- havior byrewardingthedog.Indeed,foracanine,ﬁndinga carcassisapositivereinforcementinitself.Furthermore,itis possible for“ordinary” dogowners tojointhecourse.Thus, thelevelisadjustedtosheepfarmers(andothers)whowould liketoincreasethesuccessofsheepcarcassdiscoveriesintheir own(andothers)grazingareas.

OurstudydocumentedtheefficiencyofCDEsinfinding deadsheep.Ourresearchquestionswere1)howmuchbetter areCDEscomparedwithhumanswithoutdogsindetecting sheepcarcasses,and2)areCDEsasgoodatfindingfreshand hiddencarcassesasoldandopensitedones?

Materialandmethods

Dogsandfieldstaff

Atotalof16trainedCDEsthathadpassedtheNCDD approvaltestand16trainedpersons(sheepfarmersorshep- herds)withoutdogsparticipated.NCDDhelpedusinselect- ing CDEswithinthe geographicalregionsofeachtestsite, andtheregionalagriculturaloﬃceassistedusinﬁndingexpe- riencedsheephandlers.Windspeed(calm,moderate,windy) andprecipitation(none,moderate,heavy)werenoted.Allthe CDEscarriedouttheirsearchesinheadwinds,asfaraspossi- ble,andalldogswereona20-mlongleashduetotheprotec- tionofwildanimalsandgrazing livestockduring1Aprilto 20August.²¹TracksandcarcasspositionswereGPS-logged.

Experimentaltrials

Intensivesearcheswereconductedin10fieldsof0.5km² (123.6 acres) each,at first with a dog (n=10 CDEs) and thenbyapersonwithoutdog(n=10).Withineachfield,8 lambcarcassesweighing2to4kgwereputoutinan2×2 factorialdesign; 2 carcasseswere “fresh” (frozenwhenday- fresh,thawed,andplacedintestfieldthesameday)and“hidden” (undervegetation,stones,etc.);2werefreshandlocated

“open” (laidrightattheground);2were“old” (frozenwhen day-fresh,thawedandlaidoutintestﬁeldafter3-4days)and hidden;and2wereoldandplacedintheopen.Atotalof4 hours searching timewas given(8 hours per km² [8hours per247.1acres]).Allexperimentalﬁeldswerelocatedwithin scatteredmountainforestareas.

Fieldtrials

Partsof6outﬁeldpasturesforsheep,sized4km²(988.4 acres) each, were investigated for dead sheep by CDEs (n=6)andpersonswithoutadog(n=6)simultaneously.The

searchingtimewassetto32hours(intensivesearch;8hours perkm²),butinorderforthepairofpersonnelwithandwith- outadogworkinginthesamefieldnottoseeeachother,9 days separatedthem.Theaverageratesoflambslost inthe grazingareasselectedhadtobe≥10%overthelast3years.All typesofcarcassdiscoverieswererecorded,rangingfromwhole carcassestosmallboneremnantsandpiecesofwool/fur,re- gardlessofanimalspecies(Fig.1).Anycarcassremainswithin a20-meterradiuswasconsideredasonecarcassfinding.Staff fromtheNNItrainedfordocumentingwhetherlivestockare killedbyprotectedcarnivoresornotwereusedtodiscriminate predationfromothercausesofmortality.

Statistics

Allmeasureswerestandardizedto8hourssearchperkm² (i.e.,amount of effort).In orderto measure the search efficiency,themeannumberof kilometersand minutesspent per carcass detectionwere calculated. Number of carcasses foundwithandwithoutdogs,bothintheexperimentaltrials andinthefieldtrialsdidnotfollowthenormaldistribution (Kolmogorov-SmirnovTest).Thus,pairwisedifferenceswith andwithoutdogsregardingthetotalnumberofcarcassde- tectionsandnumberoffindsbycarcasstype(freshandopen;

freshandhidden;oldandopen;oldandhidden)weretested using Wilcoxon SignedRank Test.Chi-squared goodness- of-ﬁt testswereusedto testifthe typesof carcassesfound bydogsintheexperimentaltrialswererandomlydistributed.

Differences indistance(kilometer)searchedwithandwith- outdogsinbothtypeoftrials,andkilometersandminutes per carcassfindinthefield trialswerenormally distributed andtestedwithpairedt-tests.Thesignificancelevelwassetto 0.05.Datawasprocessedwiththestatisticssoftwarepackage Minitab17.²²

Results

Experimentaltrials

TheCDEsfound 19outof80lambcarcasses(detection rate23.8%),whereaspersonswithoutdogsonlyfound2(detection rate2.5%;conﬁdence intervalsdo notoverlap,P<

0.05,Fig.2).Fivecarcassesdetectedbythedogswerefresh andlocatedintheopen,fourwerefreshandhidden,whereas fivewereoldandintheopenandfivewereoldandhidden (χ²=0.158,P >0.5).Thus,the dogsfound carcassesthat wereassumedthemostdifficulttodetect(freshandhidden) just asoftenastheonesthatwereassumedeasier.Thetwo carcassesfoundbypeoplewereoldandlocatedinopensite.

Themeandistancessearchedperfieldwithandwithoutdogs wereindifferent(9.6vs.9.7km,T=–0.1,P>0.5).Thedogs showed individualvariationindetectionrate,rangingfrom twodogsfindingnocarcasses(trialsFandG,Fig.2)totwo dogsthatfound50%ofthecarcasses(trialsAandH,Fig.2).

In general,the track logsshowed extensive andsystem- aticsearches,bothwithandwithoutdogs.However,searchers

(3)

Figure1. Acarcassdetectiondogduringaﬁeldtrialthathasfoundanoldbonefromasheep.TheGPSposition(universaltransversemercator projection(UTM)coordinates)ofthecarcassislogged.

0 2 4 6 8 10 12 14

0 1 2 3 4 5 6 7 8

Trial A Trial B Trial C Trial D Trial E Trial F Trial G Trial H Trial I Trial J

Distance (kilometer) searched

dnuof sessacrac fo rebmuN

Carcasses with dog Carcasses without dog Distance with dog Distance without dog

Figure2. Numberofcarcassesfound,anddistance(kilometers)searchedperﬁeldwithandwithoutdogsinthe10experimentaltrials.

withoutadog,butalsosomeoftheCDEs,passedseveralof thecarcassesclose(<10meters)withoutﬁndingthem.

Fieldtrials

In the six rangeland pastures,CDEsdetected a total of 132carcassremains,whereassearcherswithoutdogsfound76

(Table1).Twoofthedogﬁndswerehidden(buried)and25 werefromthepresentgrazingseason.Allcarcassesfoundby peoplelaidonopengroundand13werefromthepresentyear.

Onlythreecarcasseswerefoundinaconditioninwhichthe causeofdeathcouldbedocumented(i.e.,freshandenough leftofthebody).Twoofthesewerelocatedbypeoplewithout dogs(onedocumentedbyNNIbeingkilledbywolverineand

(4)

Table1

Numberofcarcassesfound,distancesearchedperﬁeldtrial(kilometer),andkilometersandminutesspentpercarcassdetected(mean±SEM),andrange (minimum-maximum),withandwithoutdogsinthe6ﬁeldtrials

Withandwithoutdogs Mean±SEM Range(min-max) Diﬀerencewithand

withoutdogs

Numberofcarcassesfound Withadog 21.7±7.6 4.0-58.0 CIsoverlap

P>0.5

Withoutadog 12.7±5.3 0.0-30.0

Kilometerssearchedperﬁeld Withadog 63.4±6.8 41.0-83.7 T=–1.34

P>0.05

Withoutadog 75.9±4.1 63.3-88.8

Kilometerstraveledper carcassdetected

Withadog 5.8±2.7 0.8-18.6 T=–1.31

P>0.05

Withoutadog 20.4±14.4 2.8-77.5

Minutespercarcassdetected Withadog 161±65.6 33.1-480.0 T=–1.27

P>0.05

Withoutadog 516±356 64.0-1920.0

Conﬁdenceinterval(CI)indicates;SEM,standarderrorofmeans.

onediedinanaccident),andthethird(documentedbyNNI aspossiblykilledbywolverine)wasfoundbya CDE.More than 95% of the ﬁnds conductedboth bydogs and people consistedofsmallbonefragmentsand/orwool.

Bycombiningthedetectionrates(d.r.)intheexperimen- taltrialsforCDEs(23.8%)andpeople(2.5%),weestimated thetotalnumberofcarcassesintheﬁeldstobebetween555 (based onthe CDEs’d.r.)and3,040(basedon thepeoples’

d.r.).The“best”CDEintheﬁeldtrialsdetected13timesmore carcassesthanthepersonwithoutadog(13vs.1),whereasthe

“least” eﬀective CDEdetected 1.4times lesscarcassesthan thesearcheralone(19vs.26).TheotherCDEsfound1.2to 4.2timesmorecarcassesthanpersonswithoutdogs.

Duetothevariationinnumberofcarcassesfoundinthe fieldtrialsandalimitednumberoffields(Table1),nosignif- icantdifferenceswithorwithoutdogswerefoundregarding numberofcarcassesdetected,distancesearchedorkilometers, andminutesspentpercarcassfound.Nevertheless,themean numeric valuesofkilometersandminutesspent percarcass findwere>3xofthatforpeoplewithoutdogs,indicatingthat CDEsweremoreefficientthanpeoplealone(Table1).

Discussion

TheexperimentaltrialsshowedthatCDEswere9.5times better thanpeople aloneatfinding sheepcarcasses(success rate:23.8%vs.2.5%),andthefieldtrialsindicatedthatCDEs were>3times aseffectiveintheirsearch,measuredinkilo- meters,andminutespercarcassdetection.Furthermore,the dogsfoundalltypesofcarcassesjustaseasily.Thedogswere superior to people in finding hidden carcasses.Our results areinaccordancewithotherstudiescomparingdogsandhu- mansforcarcassdetection,suchasfindingpasserinecarcasses in densevegetation¹⁵ andbird and batcarcass detectionat windfarms.¹⁴delValleetal.¹⁴foundthathumansperformed poorlyatdetectingsmallcarcasses(∼20%detectionrate)and moresoindensevegetation.Wedidnotmeasurethesizesof thecarcasses,butpredominantlythecarcassesdetectedinthe fieldtrialsconsistedofsmallboneremainsand/orwool.

Resultsfromtheexperimentaltrials,however,documented thatthe CDEsdidnotfindmorethaneveryfourthcarcass outinthefields.AsdiscussedinthereviewarticlebyTroisi etal.,²³andbasedonexperiencesfromourstudy,webelieve thatwinddirection,overlywidesearchloops,distractionsby otherscents,ormisinterpretedorabsenceofcommunication betweendogandhandlermayexplainsomeofthefailureof CDEstodetectcarcasses.Furthermore,weexperiencedthat some dogs being “on duty” several days in advance of the testfailed.NCDDbelievesaday-longintensivesearchistoo muchfor adog tomaintain fullmotivationand concentration.Thedoghandlershouldknowhowtomotivatethedog andapplythistotraining andworkingsituations.Therela- tionship betweenthehandlerand thedog issignificant for workingdogs.²³^,²⁴ Likewise,hottemperatureswillaffectthe activity levelofdogs andtheir sniffingefficiencynegatively becausedogscannotpantandsniff atthesametime.Inone ofthefieldtrialsthetemperatureincreasedabove30°C(85°F) on8August,andthedoghandlernotedthedogwas“lazy” all day.

Olfactory abilities vary both within and between dog breeds,accordingtothesizeofthecanineolfactorybulband thegenesthatcontroltheolfactoryreception.¹⁰^,¹²^,²⁵Indeed, intheexperimentaltrialswedocumentedvariationindetec- tionratebetweenthetwodogsthatdidnotfindanycarcasses (trials F and G,Fig.2)to the two dogsthat found halfof the carcasses(trials AandH,Fig.2).In thefield trials,the numberofcarcassesdetectedwithdogscomparedwithfinds madebypersonsalone(withineachfield)variedevenmore, from13to0.7.However,thesetestfieldsdifferedinthenum- berofcarcassespresentandthetotalnumberofcarcasseswas notknown,thusitwasnotpossibletocomparethedetection ratesbetweendogs.

ItisimportantthatCDEsmaintainahighdetectionrate, especiallyiftheworkispaidfor.Byusingdogsthatarefos- tered and trained by their owners it is possible to educate CDEs at a large scale with limited resources.²⁶ How high the competencelevel shouldbeis atrade-oﬀ betweenhav- ingmany"mediumquality"CDEsavailableandafewexcel- lentones.NCDDwasnotsatisﬁedwiththeresultsfromthe

(5)

experimental trials. To increase the CDEs’ success rate, NCDDhasnowimprovedtheirtrainingprogrambyoﬀering acomprehensivecourseandbyprovidingapprovedequipages withadditionaltrainingsessions.

Thefactthatfew freshcarcasses,inwhichwewereable to document the cause of death were found,confirms that itisdifficult tobeintherightplace attherighttime.This hasalsobeenconfirmedinothertrials.²⁷Inanacutesituation withcarnivoreattacks(i.e.,massivedepredationoflivestock, limitedintime andspace),theconditionsaredifferent;the CDEswill besearchingfor acluster offresh killswithina smallpartofthegrazingarea,helpingtoidentifytheextent ofthe damage.Whereassearching forcarcassesthroughout an entire rangeland pasture(most of them sized >50km² [2,355 acres]),is likesearching for a needle ina haystack.

Thus,thisprojecthasstrengthenedouropinionthatbestre- sults areachieved byCDEs whenused in acute situations.

Despitethis,dogsareabettertool thanpeople alonehelp- ing farmers toﬁnd dead animals,otherwise every livestock animalwillneedtohaveaGPS-transmitterwithamortality function.

Conclusionsandimplications

CDEsweremoreeffectiveatfindingsheepcarcassesthan searcherswithoutdogs,measuredintotalnumberofcarcasses discoveredaswellaskilometersand minutesspentpercar- cassfind.Thedogsfoundalltypesofcarcassesjustaseasily andwereparticularlybetterthanpeoplealoneinfindinghid- dencarcasses.However,theexperimentaltrialsshowedthat CDEsdetectedonlyoneinfourofthecarcasseslaidout.Our results have led to revisions of NCDD’s training program.

NCDDisnowoﬀeringacomprehensivecourseandtheyare following-uponapprovedCDEsmoreclosely.Itisimportant forCDEstotrainoncarcassesthatarelaidoutintheﬁeldall yearround.Basedonexperiencesfromthisprojectaguidefor CDEswasoutlined.²⁸

TheeﬀortofcarcassdetectiondogsisimportanttoNor- wegiansheepfarmersapplyingforcompensationstoincrease thepercentageofprovenlosses.In thefuturecarcassdetec- tiondogsinNorwaymight alsobeusedtodocument dead and injuredbirdsatwind farms(specially protected species likegoldeneagle,seaeagle[Haliaeetusalbicilla]andEurasian eagle-owl[Bubobubo]),orforotherwildlifeconservationand managementtasks.¹³

DeclarationofCompetingInterest

Theauthorsdeclarethattheyhavenoknowncompeting ﬁnancial interestsor personalrelationships that couldhave appearedtoinﬂuencetheworkreportedinthispaper.

Acknowledgments

WewouldliketothankNilsSteinvikandJohnnyMath- ieseninNorwegianCarcassDetectionDogsfor facilitating thefieldexperiments,aswellasthefieldstaff withandwith- outdogsparticipatinginthetrials.ThankstoGreteMeisfjord JørgensenandJoJoremAarsethforperusalofthetext.

This work was supported by mitigating- and conﬂict- reducing grantsfromTheCountyGovernors ofFinnmark, Troms,Nordland,Sør-Trøndelag,MøreogRomsdalandOp- pland.

References

1. Statistics on losses of livestock during the grazing season.

Accessed 16 December 1, 2020. https://www.nibio.no/tema/

landskap/kart-over-beitebruk-og-seterdrift/beitestatistikk?

locationﬁlter=true

2. Recordsonnumberofsheepthatarecompensated(documented orassumedkilledbyprotectedcarnivores.AccessedDecember 16,2020.https://www.rovbase.no/erstatning/sau.

3. Regulations on welfare of small ruminants.Accessed Febru- ary 22, 2021. https://lovdata.no/dokument/SF/forskrift/

2005-02-18-160

4.Hansen I. Causes of mortality in lambs at rangeland pasture inØrpen-Redalen,2007and2008.BioforskReport.2009;

4(19):1–21.

5. Regulationsoncompensationsoflivestockkilledbyprotected carnivores. Accessed February 22, 2021. https://lovdata.no/

dokument/SF/forskrift/2014-05-30-677

6. InformationaboutlargecarnivorespublishedbytheNorwegian EnvironmentAgency.AccessedFebruary22,2021.http://www.

rovviltportalen.com/

7.Rebmann A, David E, Sorg MH. CadaverDog Handbook:

ForensicTrainingandTacticsfortheRecoveryofHumanRemains. CRCPress;2000.

8.CastaldoNF.SniﬀerDogs:HowDogs(andTheirNoses)Savethe World.1sted.HoughtonMiﬄinHarcourt;2014.

9.BrowneC, StaffordK, FordhamR.Theuseofscent-detec- tiondogs.IrishVeterinaryJournal.2006;59(2):97–104. 10.StejskalSM.Death,Decomposition,andDetectorDogs:FromSci-

encetoScene.CRCPress;2013.

11.NawnKM.SniﬃngOutDecomposition:InvestigatingtheRelia- bilityofHumanRemainsDetectingDogsMasterThesis.Humbolt StateUniversity;2018.

12.MazzolaSM, PironneF, SeddaG, GasparriR, RomanoR, SpaggiariL, MariangelaA.Two-stepinvestigationoflung cancer detection by sniﬃng dogs. JBreath Res. 2020; 14(2) https//doi.org/10.1088/1752-7163/ab716e.

13.Dahlgren DK, Elmore RD, Smith DA, Hurt A, Ar- nettEB, ConnellyJW.Useofdogsinwildliferesearchand management.In:SilvyN,ed.WildlifeTechniquesManual.7thed.

TheWildlifeSocietyInc;2012:140–153.

14. delValleJD,PeraltaFC,ArjonaMIJ.Factorsaﬀectingcar- cassdetectionatwindfarmsusingdogsandhumansearchers.J ApplEcol.2020;57:1926–1935.doi:10.1111/1365-2664.13714. 15.Homan JH, Linz G, Peer BD. Dogs increase recovery of passerine carcasses in dense vegetation.Wildl Soc Bull. 2001;

29(1):292–296.

(6)

16. GossK-U.Thephysicalchemistryofodors-consequencefor theworkwithdetectiondogs.ForensicSciInt.2019;296:110–

114.doi:10.1016/j.forsciint.2019.01.023.

17. Norwegian CarcassDetectionDogs.www.norskekadaverhund er.noInNorwegian.

18.SkinnerBF.TheBehaviorofOrganisms:AnExperimentalAnal- ysis.Appleton-Century;1938.

19. McLeod S.A. Skinner - Operant Conditioning. Accessed December 16, 2020. https://www.simplypsychology.org/

operant-conditioning.html

20.Chance P. Learning and Behaviour. Wadsworth Cengage Learning;2009.

21. Law on dog keeping.Accessed December 16,2020.https://

lovdata.no/dokument/NL/lov/2003-07-04-74/

22. Minitab. Minitab ®17. Accessed April 17, 2020.

https://www.minitab.com/uploadedFiles/Documents/

getting-started/Minitab17_GettingStarted-en.pdf2016 23. Troisi CA, Mills DS,Wilkinson A,Zulch HE. Behav-

ioralandcognitivefactorsthataﬀectthesuccessofscentdetec- tiondogs.CompCognBehavRev.2019;14:51–76.doi:10.3819/

CCBR.2019.140007.

24. DiverioS, MenchettiL, GlacomoR. Dog’scopingstylesand dog-handlerrelationships inﬂuenceavalancesearchteamper- formance.ApplAnimBehavSci.2017;191:67–77.doi:10.1016/j.

applanim.2017.02.005.

25.TacherS, QuignonP, RimbaultM, DreanoS, AndreC, GalbertF.Olfactoryreceptorsequencepolymorphismwithin andbetweenbreedsofdogs.JHered.2005;96(7):812–816. 26. ByosiereS-E,FengLC,RutterNJ.Factorsthatmayaﬀect

thesuccessofscentdetectiondogs:Exploringnonconventional modelsofpreparationanddeployment.CompCognBehavRev. 2019;14:81–86.doi:10.3819/CCBR.2019.140009.

27.NerhoelI, SmestadK.Searchingforsheepcarcassesusingde- tectiondogsinSkaretgrazingarea– apilotproject.FinalReport;

2013:1–17.

28.SteinvikN, MathisenJ, HansenI, WinjeE.Aguideinprac- ticalsearchingfordeadlivestockwithdogs– preparations,prac- ticalworkandreadiness.NIBIOPOP.2017;3(13):1–4. Authorsarefrom:NorwegianInstituteofBioeconomyResearch,˚As, Norway(IngerHansen,ErlendWinje);