Traffic mortality of four ungulate species in southern Finland

T a ic mo ali y o ou ungula e species in sou he n Finland 13 T a ic mo ali y o ou ungula e species in sou he n Finland Milla Niemi1, Juho Ma ala2, Ma kus Melin3, Visa E onen4, Hannu Jä enpää5 1 Uni e si y o Helsinki, Depa men o Fo es Sciences, P.O. Box 27, FI-00014 Uni e si y o Helsinki, Finland 2 The Na u al Resou ces Ins i u e Finland, P.O. Box 68, FI-80101 Joensuu, Finland 3 Uni e si y o Eas e n Finland, School o Fo es Sciences, P.O. Box 111, FI-80101 Joensuu, Finland 4 Finnish Wildli e Agency, Sompion ie 1, FI-00730 Helsinki, Finland 5 Hy inkää Game Managemen Associa ion, Suopellon ie 258, FI-05720 Hy inkää, Finland Co esponding au ho : Milla Niemi ([email p o ec ed]) Academic edi o : A. Seile |Recei ed31 Decembe 2014|Accep ed 18 June 2015|Published 28 July 2015 h p://zoobank.o g/A6107B77-7D58-4FCB-A0DE-BFB67A34C14D Ci a ion: Niemi M, Ma ala J, Melin M, E onen V, Jä enpää H (2015) T a ic mo ali y o ou ungula e species in sou he n Finland. In: Seile A, Helldin J-O (Eds) P oceedings o IENE 2014 In e na ional Con e ence on Ecology and T anspo a ion, Malmö, Sweden. Na u e Conse a ion 11: 13–28. doi: 10.3897/na u econse a ion.11.4416 Abs ac Ungula e– ehicle collisions a e in ensi ely s udied in many coun ies. Howe e , limi ed knowledge exis s on how many animals s uck ac ually die due o collisions and whe he di e ences in a ic mo ali y occu be ween species li ing in he same a ea. In his s udy, we es ima ed a kill a e ( he p opo ion o indi iduals killed/s uck) and, in ela ion o hei win e popula ion sizes, he collision and a ic mo ali y a es o ou ungula e species (moose Alces alces, whi e- ailed dee Odocoileus i ginianus, oe dee Cap eolus cap eolus, and allow dee Dama dama). We used an uno icial collision egis e collec ed be ween 2001 and 2012 (a o al o 12 yea s) by olun a y hun e s om he Hy inkää Game Managemen A ea (323 km2) loca ed in sou he n Finland. The popula ion es ima es used we e based on annual snow ack censuses. A o al o 497 ungula es we e in ol ed in collisions du ing he s udy pe iod. O hese, 76% we e killed di ec ly o pu down a e wa ds. Roe dee had he highes kill a e; 95% o s uck indi iduals died. Whi e- ailed dee had he highes collision and a ic mo ali y a es (8.0% and 6.5% o he win e popula ion, espec i ely), ollowed by moose (6.5 % and 4.5%), oe dee (3.9% and 3.7%), and allow dee (3.2% and 2.1%). As we ound he collision and a ic mo ali y a es o be unequal be ween species, we ecommend sepa a ely epo ing all ungula e species when compiling collision s a is ics. We addi ionally sugges ha local manage s should be awa e o ungula e collision and a ic mo ali y a es in hei a eas and should use his knowledge when planning annual ha es . Na u e Conse a ion 11: 13–28 (2015) doi: 10.3897/na u econse a ion.11.4416 h p://na u econse a ion.penso .ne Copy igh Milla Niemi e al. This is an open access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (CC BY 4.0), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal au ho and sou ce a e c edi ed. RESEARCH ARTICLE Launched o accele a e biodi e si y conse a ion A pee - e iewed open-access jou nal Milla Niemi e al. / Na u e Conse a ion 11: 13–28 (2015) 14 Keywo ds Dee – ehicle collision, moose– ehicle collision, popula ion size, collision a e, a ic mo ali y a e, game managemen In oduc ion Expanding oad ne wo ks and globally inc easing a ic olumes ha e many nega i e e ec s on he en i onmen and animals (e.g. Fo man and Alexande 1998, Fah ig and Ry winski 2009). Road-induced mo ali y is one o hese impac s. Road kills a e he single mos impo an cause o dea h e.g. o Eu asian badge s (Meles meles) in B i ain (Cla ke e al. 1998) and o Flo ida Key dee (Odocoileus i gianus cla ium) in Flo ida (Lopez e al. 2003). The Eu opean o e (Lu a lu a) is also an example o a species su e ing om high a ic mo ali y (Philcox e al. 1999, Haue e al. 2002). T a ic can also be a signi ican cause o dea h in many common and abundan species, e.g. many ungula es, wi hou di ec ly h ea ening hei popula ion pe sis ence (Seile and Helldin 2006). Ungula e– ehicle collisions (UVCs) a e a no able and inc easing a ic sa e y p ob- lem in Eu ope, No h Ame ica, and Japan, and a e he e o e in ensi ely s udied in many coun ies (G oo B uinde ink and Hazeb oek 1996, Romin and Bissone e 1998, Seile 2004, Huijse e al. 2009, Mo elle e al. 2013). Ne e heless, human inju ies and a ali ies (e.g. Joyce and Mahoney 2001) o he economic consequences caused by UVCs (e.g. Bissone e e al. 2008) a e no he only aspec s esea che s ha e conside ed. Se e al s udies ha e ocused on he empo al and/o spa ial pa e ns o UVCs (Finde e al. 1999, Haikonen and Summala 2001, Danks and Po e 2010, Roland- sen e al. 2011, Niemi e al. 2013a, Rea e al. 2014, S eine e al. 2014) and de eloped models o he pu pose o p edic ing collision si es (Seile 2005, Found and Boyce 2011). Di e en mi iga ion measu es such as o e passes (Olsson e al. 2008), encing (Cle enge e al. 2001), o wa ning signs (K isp and Du o 2007) ha e u he mo e been de eloped wi h he aim o educing he numbe o se e i y o acciden s. Ungula e a ic mo ali y in ela ion o hei popula ion sizes is known in many coun ies. Fo example, he a ic mo ali y o oe dee (Cap eolus cap eolus) has a ied be ween 13% and 16%, depending on he coun y (G oo B uinde ink and Haze- b oek 1996, Seile e al. 2004, Poko ny 2006). Joyce and Mahoney (2001) calcula ed he same a io o moose (Alces alces) in New oundland, Canada, and ound ha he a ic mo ali y was only app oxima ely 3% o he annual allowable ha es quo a o 0.6% o he o al popula ion. In Sweden, hese numbe s we e 10.1% and 4.0%, e- spec i ely (Seile e al. 2004). Al hough he ela i e impo ance o oad kills seems low in many ungula e popu- la ions, egional a ia ion could be no able. Fo Finland, G oo B uinde ink and Hazeb oek (1996) epo ed he annual a ic mo ali y o be 1.2% o he o al moose popula ion, bu poin ed ou ha he po ion could be 10% in he sou he n pa o he coun y wi h dense human popula ions and highe a ic olumes. T a ic mo ali y o ou ungula e species in sou he n Finland 15 Finland has h ee abundan ungula e game species, whose popula ions a e mainly egula ed by hun ing: he moose, which is sp ead ac oss he coun y; he in oduced whi e- ailed dee (Odocoileus i ginianus), which has a dense popula ion concen a ed in sou hwes e n Finland and he oe dee , which li es in he sou he n hal o he coun- y (second la ges dis ibu ion) a low densi ies (see Pulliainen 1980, La sund e al. 2003, Kekkonen e al. 2012, Finnish Wildli e Agency and Finnish Game and Fishe - ies Resea ch Ins i u e 2014). In addi ion, wo small na i e popula ions o wild o es eindee (Rangi e a andus ennicus) li e in cen al and no heas e n Finland and a ew small, in oduced local popula ions o allow dee (Dama dama) in sou he n Finland. The hun ing o all hese species, excep o he oe dee , is based on licenses g an ed by wildli e au ho i ies (Hun ing Ac 615/1993, Hun ing Dec ee 666/1993). Du ing he 21s cen u y, he annual coun y-le el ha es o moose, whi e- ailed dee , and oe dee has a ied be ween 38–76 000, 14–26000, and 1–4 000 indi- iduals, espec i ely. O he species a e hun ed only ma ginally; he annual hun ing bag has a ied be ween 20–130 indi iduals o wild o es eindee and 50–200 indi iduals o allow dee (Finnish Wildli e Agency and Finnish Game and Fishe - ies Resea ch Ins i u e 2014). A he same ime, he o al amoun o annual colli- sions has a ied be ween 1200–3000 o moose and 2600–4300 o he dee species (Finnish T a ic Agency 2014). T a ic is p obably a signi ican cause o mo ali y o Finnish ungula e species, hus playing an impo an ole when planning he an- nual ha es . This is pa icula ly ue o whi e- ailed dee , oe dee , and allow dee , whose dis ibu ions a e loca ed nea he mos densely popula ed human se lemen s in sou he n Finland. S udy aim The aim o ou s udy was o es ima e he a ic mo ali y o ou ungula e species li ing in he same a ea and o disco e possible in e -species di e ences. Fi s , we we e in e - es ed in how many pe cen o animals s uck ac ually die due o collisions (la e e e ed o as kill a e). Second, we wan ed o in es iga e how many collisions ha e occu ed in ela ion o species popula ion sizes (la e e e ed o as collision a e). Finally, we wan ed o explo e how la ge a p opo ion o he popula ions ha e died in he collisions (la e e e ed o as a ic mo ali y a e). We es ed he hypo heses ha 1) he collision a e is equal o all species and 2) he a ic mo ali y a e is equal o all species. Ma e ials and me hods S udy a ea The s udy was conduc ed in he a ea o he Hy inkää Game Managemen Associa ion (la e e e ed o as Hy inkää GMA) (Fig. 1), which co e s he Hy inkää municipali y Milla Niemi e al. / Na u e Conse a ion 11: 13–28 (2015) 16 in sou he n Finland, in he Uusimaa egion. Hy inkää (land a ea 323 km2) is loca ed in a densely popula ed pa o he coun y wi h app oxima ely 46 000 inhabi an s in 2012 (S a is ics Finland 2014). Mos o he landscape is highly domina ed by humans; he ci y o Hy inkää co e s he cen al pa o he a ea and se e al smalle illages exis especially in he sou h. Landscape s uc u es ou side hese popula ion cen e s ange om a mosaic o cul i a ed a eas and se lemen s o mo e o es ed a eas ound mainly in he wes e n pa s o he municipali y. The a ea is di ided by enced Na ional Highway 3 (depending on he oad sec ion, he a ic olume was app oxima ely 20–30 000 ehicles/day in 2010; Finnish T ans- po Agency s a is ics; hea y a ic included), ou e 130 ( unning pa allel o Highway 3; 3200–3500 ehicles/day) and a ailway. These all un sou h o no h, while in he eas -wes di ec ion he a ea is limi ed by Highway numbe 25 (5–10 000 ehicles/day), which uns h ough he sou he n pa o he a ea. Public oad densi y is app oxima ely 0.7 km/km2, wi h an annual a ic low app oxima ely 330 millions o kilome e s in 2010 (Finnish T anspo Agency s a is ics). The e is addi ionally a dense ne wo k o mino oads and o es y acks. The speed limi on he main oads a ies depending on he oad sec ion and season, being 100 o 120 km/hou on Highway 3 and 60 o 80 km/hou on he o he main oads. Ungula e da a A o al o ou ungula e species (moose, whi e- ailed dee , oe dee , and allow dee ) exis in he a ea o Hy inkää GMA. All a e game animals, i.e. hei popula ions a e managed by hun ing. The annual maximum hun ing quo as o moose, whi e- ailed dee , and allow dee a e de ined and con olled by licenses g an ed by he Finnish Wildli e Agency. Roe dee ha es is no egula ed by he au ho i ies, bu hun e s ha e Figu e 1. The map o ou s udy a ea. The land use map is simpli ied om CORINE Land Co e 2006 da a (Finnish En i onmen Ins i u e 2009, CLC2006). Coun y bo de s: Eu os a . T a ic mo ali y o ou ungula e species in sou he n Finland 17 o epo hei bag (Hun ing Ac 615/1993, Hun ing Dec ee 666/1993). The numbe o hun ed indi iduals is hus known o each species. The popula ion es ima e used in ou s udy was based on an annual snow acking census (Finnish Wildli e Agency and Finnish Game and Fishe ies Resea ch Ins i u e 2014) coo dina ed by he Finnish Wildli e Agency and conduc ed by olun a y hun - e s. In he Uusimaa egion, whe e ou s udy a ea is loca ed, each hun ing club ies o assess all he ungula e indi iduals li ing in hei hun ing a ea. To a oid double coun ing, he census is ca ied ou e e ywhe e du ing he same weekend. Whe e ani- mal popula ions a e dense o snow condi ions poo , he esul s o ha census can be complemen ed wi h supplemen al in o ma ion om came a aps o o he sou ces. The ends o an annual snow census om ou s udy a ea a e p o ided in Suppl. ma e- ial1: Annual ends in popula ion size and collisions. UVC da a collec ion Moose– ehicle acciden s a e egis e ed a he species-le el in he Finnish collision s a is ics, bu c ashes wi h o he wild ungula es a e ea ed as dee – ehicle collisions ega dless o he species. Because we we e in e es ed in explo ing he possible in e -spe- cies di e ences, he exis ing na ion-wide collision da abase was no usable. We ins ead used a local da ase collec ed om he Hy inkää GMA a ea by olun a y hun e s who wo k as o icial assis an s o he police. UVCs in Finland ha e o be epo ed o he police, bu he c ash si es a e usually isi ed by he police only in cases whe e pe sonal inju ies o damage o he ehicle has occu ed. The collision si es a e ins ead checked by local hun e s, who wo k as an execu i e assis ance o he police. These olun ee s isi e e y UVC si e, pu he in ol ed animal down i needed, and anspo he ca cass away om he oad a ea. The olun ee s do no ha e any egis e ing du ies, bu will some imes collec uno icial s a is ics o hei own in e es s. Fo ou s udy, we used a speci ic UVC da ase collec ed by olun a y hun e s and main ained by he chie o he Hy inkää GMA. This egis e con ained de ailed in o ma ion conce ning e.g. he species and pos -collision condi ion o an animal. The egis e con ained UVCs om be ween 2001 and 2012 (12 yea s in o al). Da a analyses F om he da a collec ed by he olun a y hun e s, we calcula ed a kill a e, a collision a e, and a a ic mo ali y a e o each species. The kill a e was simply calcula ed om he animals s uck (how many pe cen o animals s uck died in he collision o we e a ally inju ed and pu down a e wa ds). The collision a e was calcula ed by combining he collision da a and he esul s o an annual snow census (i.e. how many collisions occu o each 100 indi iduals assessed in he snow census). The a - Milla Niemi e al. / Na u e Conse a ion 11: 13–28 (2015) 18 ic mo ali y a e was also based on he collision and snow census da a (how many indi iduals died in collisions o each 100 indi iduals assessed in he snow census). We ha e con e ed ou esul s o pe cen ages (e.g. a calcula ed a e o 0.05 = 5%) o simpli y he ex . We used Fishe ’s exac es (e.g. Ran a e al. 1999) o es possible di e ences be ween species. Con ingency ables used o he analyses a e p esen ed in Suppl. ma e ial 2: Con ingency ables. We used Fishe ’s exac es wi h Bon e oni co ec- ions o p- alues o he pos hoc analyses (MacDonald and Ga dne 2000). Analyses we e conduc ed using R so wa e, e sion 3.1.3 (R De elopmen Co e Team 2015). Resul s A o al o 497 ungula es we e in ol ed in 493 collisions du ing he 12-yea s udy pe- iod (Table 1, Suppl. ma e ial 1). One ou o wo collisions (N = 245; 50%) was a c ash in ol ing whi e- ailed dee , ollowed by moose (118; 24%), oe dee (75; 15%), and allow dee (40; 8%). The species was unknown in 15 cases (3%). A o al o 378 indi iduals (76%) we e killed di ec ly in he collisions o pu down a e wa ds (la e e e ed o as oad-killed) (Table 1). Roe dee was he mos ulne able species: 95% o indi iduals in ol ed in c ashes we e killed and only one was ound uninju ed. The lowes kill a e (65%) was eco ded o allow dee , bu concu en ly he numbe o disappea ed indi iduals was high. In compa ison o popula ion es ima es de i ed om he snow ack census da a, whi e- ailed dee had he highes collision a e: 8.0% (8.0 collisions/100 indi iduals), ollowed by moose (6.5%), oe dee (3.9%), and allow dee (3.2%) (Fig. 2). A s a is i- cally signi ican di e ence was obse ed be ween species (DF = 3, p < 0.001). A pai ed pos hoc compa ison showed all species pai s excep whi e- ailed dee & moose and moose & oe dee o di e (0.05 a he α-le el) a e he Bon e oni co ec ion was applied (Table 2). Table 1. Road-killed and s uck bu uninju ed ungula es in he Hy inkää GMA be ween 2001 and 2012 (a o al o 12 yea s). Column “Condi ion unknown” con ains animals ha ha e disappea ed om he col- lision si e and ha e no been ound la e by acking, and animals whose condi ion has no been eco ded in he da abase used. Road-killed indi iduals Uninju ed indi iduals Condi ion unknown To al numbe o indi iduals s uck Whi e- ailed dee 198 (80%) 4 (2%) 46 (19%) 248 (50% o all) Moose 82 (69%) 12 (10%) 24 (20%) 118 (24%) Roe dee 72 (95%) 1 (1%) 3 (4%) 76 (15%) Fallow dee 26 (65%) 1 (3%) 13 (33%) 40 (8%) Unknown 0 (0%) 1 (7%) 14 (93%) 15 (3%) To al 378 (76%) 19 (4%) 100 (20%) 497 (100%) T a ic mo ali y o ou ungula e species in sou he n Finland 19 Table 2. The esul s o he pai wise compa isons be ween species’ collision a es in he Hy inkää GMA be ween 2001 and 2012 (a o al o 12 yea s). Compa isons we e made by using Fishe ’s exac es and p- alues we e adjus ed using he Bon e oni co ec ion. Species 1 Species 2 DF p- alue Whi e- ailed dee Moose 1 0.048 Whi e- ailed dee Roe dee 1 <0.001*** Whi e- ailed dee Fallow dee 1 <0.001*** Moose Roe dee 1 <0.001** Moose Fallow dee 1 <0.001*** Roe dee Fallow dee 1 0.332 *** p < 0.001 a e Bon e oni adjus men ** p < 0.01 a e Bon e oni adjus men * p ≤ 0.05 a e Bon e oni adjus men Figu e 2. Annual a ia ion in collision and a ic mo ali y a es o ou ungula e species in he Hy- inkää GMA be ween 2001 and 2012 (a o al o 12 yea s). Species P e cen age o he win e popula ion 0 5 10 Whi e- ailed dee Moose Roe dee Fallow dee Whi e- ailed dee Moose Roe dee Fallow dee Collision a e T a ic mo ali y a e When analyzing oad-killed indi iduals only (i.e. excluding animals ha had disappea ed a e he collision o we e ound uninju ed), i became appa en ha whi e- ailed dee had he highes a ic mo ali y a e (6.5% o 6.5 oad-killed indi- iduals/100 indi iduals) ollowed by moose (4.5%), oe dee (3.7%), and allow dee (2.1%) (Fig. 2). Again, a s a is ically signi ican di e ence was ound (DF = 3, p < 0.001), and in a pai ed compa ison all species pai s excep moose & oe dee di e ed (0.05 a he α-le el) a e he Bon e oni co ec ion was applied (Table 3). We calcula ed he a io be ween oad-killed indi iduals and he annual ha es o each species. The p opo ion o oad-killed whi e- ailed dee was 10.3% o he annual hun ing bag. The same p opo ions o moose, oe dee , and allow dee we e 6.9%, 30.9%, and 49.1%, espec i ely. Milla Niemi e al. / Na u e Conse a ion 11: 13–28 (2015) 20 Discussion Collision a ali y o ungula es UVCs, especially dee – ehicle collisions, a e ela i ely a ely a al o humans. The op- posi e is ue o animals. In ou da a, he smalles species, oe dee , was he mos ul- ne able: 95% o c ashes lead o he dea h o he animal. This pe cen age is almos he same (94%) as ha ound by Almk is e al. (1980) in Sweden. This numbe was lowe o o he species, bu he numbe o disappea ed indi iduals was concu en ly highe . I is no known how la ge a p opo ion o hese animals has been wounded and would ha e died la e due o he consequences o he collisions. Howe e , when igno ing hese disappea ed indi iduals, he la ges ungula e species, he moose, has he bes pos- sibili y o su i ing a collision: 10% o indi iduals s uck we e ound o be uninju ed. This is simila o indings om Sweden (Almk is e al. 1980; 8%) and New oundland, Canada (Joyce and Mahoney 2001; 11%). We hus no e ha he numbe o collisions wi h wild ungula es is mo e o less he same as he amoun o oad-killed animals. I is hence good o keep in mind ha no all acciden s a e epo ed (e.g. Almk is e al. 1980), and he e o e he eal numbe o collisions and u he , he numbe o oad- killed animals, may be la ge han he numbe o egis e ed acciden s. Al hough he size o he s uck animal seemed o be an impo an ac o a ec ing i s possibili y o su i ing a collision, i is no necessa ily he only one. Vehicle speed is he mos impo an single a iable ha is connec ed o he se e i y o ungula e– e- hicle collisions om he human poin o iew (Ga e and Conway 1999, Joyce and Mahoney 2001), i.e. inc easing speed inc eases he isk o human inju ies o a ali ies due o collisions. I is hus logical o assume ha he p obabili y ha an animal s uck would die in a collision is la ge on highways wi h high speed limi s han on second- a y oads. Un o una ely, he da a we used con ained no exac spa ial in o ma ion o collision si es o hei speed limi s so we we e unable o es he possible e ec o speed on he kill a e o animals. Table 3. The esul s o he pai wise compa isons be ween species’ a ic mo ali y a es in he Hy inkää GMA be ween 2001 and 2012 (a o al o 12 yea s). Compa isons we e made using Fishe ’s exac es and p- alues we e adjus ed using he Bon e oni co ec ion. Species 1 Species 2 DF p- alue Whi e- ailed dee Moose 1 0.003* Whi e- ailed dee Roe dee 1 <0.001*** Whi e- ailed dee Fallow dee 1 <0.001*** Moose Roe dee 1 0.249 Moose Fallow dee 1 <0.001** Roe dee Fallow dee 1 0.008* *** p < 0.001 a e Bon e oni adjus men ** p < 0.01 a e Bon e oni adjus men * p ≤ 0.05 a e Bon e oni adjus men T a ic mo ali y o ou ungula e species in sou he n Finland 21 Species-speci ic collision and a ic mo ali y a es Whi e- ailed dee had he highes collision and a ic mo ali y a es: eigh ou o one hund ed animals (in he win e ing popula ion) we e in ol ed in collisions, and he a ic mo ali y a e was 6.5% o he popula ion. E e e al. (2002) s udied he su - i al a e o whi e- ailed dee in subu ban Chicago and ound a ic-induced mo ali y o be almos wice as high as ou esul s (a a e o 0.10 o does and 0.17 o bucks compa ed o 6.5% o 0.065 in ou da a), while Dusek e al. (1989) epo ed a ic- ela ed mo ali y o only 2% in au umn popula ions along he Lowe Yellows one Ri e . We ound he collision and a ic mo ali y a e o moose o be 6.5% and 4.5% o he popula ion, espec i ely. This was simila o wha Seile e al. (2004) es ima ed in Sweden, bu lowe collision a es ha e been ound elsewhe e (G oo B uinde ink and Hazeb oek 1996, Joyce and Mahoney 2001). The oe dee a ic mo ali y a e ound by us was in concu ence wi h wha o he Eu opean coun ies epo ed in he ea ly 1990s, while he allow dee mo ali y a e was somewha highe han epo ed elsewhe e (G oo B uinde ink and Hazeb oek 1996). Howe e , a s aigh compa ison be ween collision o a ic mo ali y a es om di e en a eas wi hou knowledge o o he explana o y ac o s does no necessa ily il- lus a e he whole pic u e. The ac ual amoun o collisions, and hence he amoun o oad-killed animals, is a ec ed by se e al ac o s. Popula ion size is one o he impo - an a iables explaining he numbe o UVCs (La sund e al. 2003, Seile 2004, Ro- landsen e al. 2011). In addi ion, hough no always simple and linea , a ic olume is p obably one o he key ac o s a ec ing he numbe o UVCs (e.g. Seile 2004, Seile 2005, Balčiauskas 2009). O he ac o s such as wildli e ences, unde - and o e passes, and he dis ibu ion o eeding si es can also ha e an e ec on he amoun o collisions especially a local scales, and u he , he numbe o oad-killed animals. Thus, i is e y likely ha collision and a ic mo ali y a es a y be ween di e en a eas and/o o e ime, e en a he same popula ion densi y. In his s udy, we we e in e es ed in he di e ences conce ning collision and a ic mo ali y a es be ween species concu en ly li ing in he same a ea. The a ic low and en i onmen al a iables we e hus same o all species, gi ing us he possibili y o discussing and compa ing species beha io - ela ed ac o s. We ound whi e- ailed dee o ha e he highes collision and a ic mo ali y a es, ollowed by he moose. Howe e , a e calcula ing Bon e oni co ec ions o p- alues, he s a is ical di e ence be ween species collision a es disappea ed while di e ence be ween species a ic mo ali y a es emained. This is likely o be due o he smalle body size o whi e- ailed dee , and u he o he la ge oad kill a e ound in ou s udy. Compa ing his species pai in a mo e de ailed ashion would be in e es ing in he u u e, o in es iga e whe he hei collision a es eally di e . Moose a e known o ha e la ge home anges and some o he animals implemen seasonal mig a o y be- ha io (Heikkinen 2000, Singh e al. 2012). This esul s in moose being mo e likely o c oss se e al oads du ing hei ou ine mo emen s. Lau ian e al. (2008) on he o he hand ound ha moose end o a oid oad c ossings al hough hey occasionally isi Milla Niemi e al. / Na u e Conse a ion 11: 13–28 (2015) 28 Supplemen a y ma e ial 1 Annual ends in popula ion size and collisions Au ho s: Milla Niemi, Juho Ma ala, Ma kus Melin, Visa E onen, Hannu Jä enpää Da a ype: species da a Explana ion no e: Annual ends in popula ion size, ha es , and collisions. Copy igh no ice: This da ase is made a ailable unde he Open Da abase License (h p://openda acommons.o g/licenses/odbl/1.0/). The Open Da abase License (ODbL) is a license ag eemen in ended o allow use s o eely sha e, modi y, and use his Da ase while main aining his same eedom o o he s, p o ided ha he o iginal sou ce and au ho (s) a e c edi ed. Supplemen a y ma e ial 2 Con ingency ables used in he analysis o collision and a ic mo ali y a es Au ho s: Milla Niemi, Juho Ma ala, Ma kus Melin, Visa E onen, Hannu Jä enpää Da a ype: species da a Explana ion no e: Con ingency ables. Copy igh no ice: This da ase is made a ailable unde he Open Da abase License (h p://openda acommons.o g/licenses/odbl/1.0/). The Open Da abase License (ODbL) is a license ag eemen in ended o allow use s o eely sha e, modi y, and use his Da ase while main aining his same eedom o o he s, p o ided ha he o iginal sou ce and au ho (s) a e c edi ed.