Mortality and survival of semi-domesticated reindeer (Rangifer tarandus tarandus L.) calves in northern Finland

Rangi e , 31 (1), 2011 71 Rangi e , 31 (1): 71 - 84 Mo ali y and su i al o semi-domes ica ed eindee (Rangi e a andus a andus L.) cal es in no he n Finland Mau i Nieminen1, Ha i No be g2,1, & Veikko Maijala3,1 1 Finnish Game and Fishe ies Resea ch Ins i u e, Toi oniemen ie 246, FIN-99910 Kaamanen, Finland (co esponding au ho : [email p o ec ed]). 2 Finnish Wildli e Agency, Vanamoka u 3 D, FIN-96500 Ro aniemi, Finland. 3 Uni e si y o Applied Sciences, Joki äylä 11 C, FIN-96300 Ro aniemi, Finland. Abs ac : Du ing he pe iod 1999 o 2004 he eindee cal su i al and mo ali y we e s udied in wo eindee -he ding coope a i es and in i e he ding-g oups in no he n Finland, whe e in o al 1725 cal es we e i ed wi h mo ali y indica ing adio- ansmi e s ixed on expandable neck colla s. The cal es we e weighed and ma ked a he age o 2-5 days in cal ing co als and also du ing ea ma king in June/July, when he age o cal es was 2-8 weeks. The a e, iming and causes o mo ali y o cal es we e in es iga ed. In 1999-2001 in I alo eindee -he ding coope a i e 4.6% o adio- colla ed cal es and in 2002-04 in Käsi a si eindee coope a i e 5.2% was ound dead. The a e age mo ali y o he cal es adio-colla ed du ing cal ing ime in May, and moni o ed o he end o Oc obe , was 6.7% in I alo and 9.0% in Käsi a si. F om July on, he a e age mo ali y a es a ied be ween 1.8-5.7% among eindee he ding-g oups. On a e - age 54 and 42% o all adio-colla ed cal es ound dead in I alo and Käsi a si coope a i es we e a ibu ed o p eda ion, and golden eagle was he mos signi ican cause o dea h in bo h coope a i es killing 0-3.5% o adio-colla ed cal es in di e en s udy a eas and yea s. Golden eagle p eda ion accoun ed o 33-43% o all adio-colla ed cal es ound dead, 55-59% o he cases wi h iden i ied cause o dea h and 80% o all iden i ied p eda ion. Mos o he cal es killed by golden eagle we e ound du ing July and Augus mainly in he open a eas, as in highlands, bogs and clea -cu o es a eas. The mean body weigh o he cal es adio-colla ed in May (weigh s adjus ed on June 1s ) and ound dead du ing he summe was signi ican ly (P<0.01) ligh e han he mean weigh o su i o s bo h in I alo and Käsi a si. Fu he - mo e, he midsumme body weigh s o he cal es (weigh s adjus ed on July 1s ) killed by all p eda o s and by golden eagles we e signi ican ly (P<0.001) lowe han he mean weigh o su i ing cal es in bo h coope a i es. Howe e , he weigh s did no di e be ween dep eda ed cal es and hose cal es ha succumbed due o o he causes han p eda ion. The esul s o his s udy emphasize he ela i e impo ance o golden eagle as a mo ali y ac o o eindee cal es in he no he n pa o he Finnish eindee husband y a ea. Key wo ds: Semi-domes ica ed eindee ; Rangi e ; cal mo ali y; p eda ion; golden eagle; su i al; adio eleme y. In oduc ion Knowledge o ju enile su i al and he in lu- ences o many in insic and en i onmen al ac o s on mo ali y and su i al a es a e es- sen ial o imp o ed unde s anding and model- ling o dynamics o ee- anging popula ions. Many s udies wi h no he n dee species ha e demons a ed ha low bi h-mass, dep essed condi ion, inabili y o nu se, and ma e nal e- jec ion may p edispose o ea ly mo ali y (see Langenau & Le g, 1976; Be ge ud, 1980; Sams e al., 1996; T e aa e al., 2003; Ca s ensen Powell & DelGiudice, 2005). Du ing he li e cycle he isk o dea h is usually g ea es du ing he i s yea o li e. I has been also sugges ed ha he p obabili y ha a p eda o kills an in- di idual is dependen on he p ey´s nu i ional s a us (DelGiudice e al., 2006). Rangi e , 31 (1), 201172 Many s udies ha e shown he majo ole o la ge e es ial ca ni o es on he neona e mo - ali y o la ge and medium-sized ungula es (see Linnell e al., 1995). Cal mo ali y is also an impo an ac o in he popula ion dynamics o many eindee and ca ibou he ds, and p eda- ion is suspec ed as he p ima y cause o mo - ali y (Mille & B ough on, 1974; Page, 1985; Mahoney e al.; 1990, Skogland, 1991; Whi - en e al., 1992; Adams e al., 1995; Valken- bu g e al., 2004; No be g e al., 2005, 2006; Nieminen, 2010). Ca ibou he ds exposed o p eda ion may lose e en 50% o he annual cal c op p io o six mon hs o age (Be ge ud, 1980). Boe je (1984) sugges ed also ha ea ly p eda ion o cal es, independen o nu i ional condi ion, was he key ac o limi ing he pop- ula ion in he Denali ca ibou he d in Alaska. In he no he n pa s o Fennoscandia semi- domes ica ed eindee Rangi e a andus a an- dus L. comp ises an impo an sou ce o p ey o many p eda o s. Apa o many he ding ac i i ies and inc easing supplemen a y eed- ing du ing win e mon hs, eindee a e ee- anging mos o he yea also in Finland, es- pecially du ing he summe , and sha e pa s o hei ange wi h di e en p eda o s including wol e ine Gulo gulo, Eu asian lynx Lynx lynx, b own bea U sus a c os, wol Canis lupus and golden eagle Aquila ch ysae os. I is also well accep ed ha la ge ca ni o es may cause sub- s an ial losses in semi-domes ica ed eindee by p eying bo h adul s and ju eniles (Bjä all e al., 1990; Nybakk e al., 2002; Danell e al., 2006; Nieminen, 2009, 2010). Fu he mo e, also golden eagle and ed ox Vulpes ulpes Fig. 1. S udy a eas wi hin he Finnish eindee husband y a ea. The eindee -he ding coope a i e o I alo (B) is di ided in o wo he ding-g oups (Nellim and sou he n a ea o I alo) and Käsi a si (A) in o h ee he ding-g oups (Palojä i, Ko a-Labba and Rai ijä i). Reindee cal es we e i ed wi h adio-colla s in di e en co als () in 1999-2004. The illages o I alo and Kaa esu an o a e ma ked on he map (). Rangi e , 31 (1), 2011 73 ha e been shown o p ey upon eindee cal es (Nybakk e al., 1999; T e aa e al., 2003; No - be g e al., 2006). The ole o golden eagle as p eda o o eindee cal es is, howe e , no so well known as ha o la ge ca ni o es. Only ew s udies (see Whi en e al., 1992; K am e al.,1998; Nybakk e al., 1999; Valkenbu g e al., 2004; No be g e al., 2006) ha e shown p eda ion by golden eagle o comp ise a sub- s an ial pa o he o al i s summe mo ali y in s udied popula ions. The aim o his s udy was o in es iga e he su i al and he ex en , iming and causes o eindee cal mo ali y in wo eindee -he ding coope a i es, I alo and Käsi a si, bo h si ua - ing in no he n Finnish Lapland. We hypo h- esized ha 1) mo ali y peaks a he i s sum- me mon hs, 2) golden eagle p eda ion on semi-domes ic eindee cal es is ema kable in hese coope a i es, 3) i occu s usually in mo e open habi a s and highe ele a ions compa ed o o he mo ali y (see No be g & Nieminen, 2008), and 4) smalle han a e age cal es a e in g ea e isk o succumb and ge p eyed by golden eagles. This s udy is based pa ly on ou ea lie epo s in Finnish (see Maijala e al., 2002; No be g & Nieminen, 2004; No be g e al., 2005). S udy a eas The eindee -he ding coope a i e o I alo (ca. 130 eindee owne s), si ua ed in he munici- pali y o Ina i, no he n Finland (I alo illage, 68o39´N, 27o33´E) co e s a o al land a ea o 2626 km2. In I alo coope a i e, he e a e wo pa ly by a ence sepa a ed he ding-g oups: 1) Nellim and 2) Sou he n a ea (see Fig. 1). The eindee -he ding coope a i e o Käsi a si (ca. 160 eindee owne s), si ua ed in he mu- nicipali y o Enon ekiö (Kaa esu an o illage, 68o27´N, 22o29´E), co e s a o al land a ea 4658 km2, and i is he second la ges among he 56 coope a i es in Finland. The coope a- i e o Käsi a si is di ided by ences in o h ee sepa a e he ding-g oups: 1) Palojä i, 2) Ko a- Labba and 3) Rai ijä i (Fig 1). Bo h o he s udy coope a i es a e si ua ed in he no h bo eal ege a ion zone (Ah i e al., 1964). In I alo coope a i e e ain is domina ed by oll- ing hills wi h di e en aged o es s ands o mainly Sco s pine Pinus syl es is. Moun ain bi ch Be ula pubescens cze epanowii g ows in he slopes o he highes hills and only he ops o he highes ells a e ba en. In Käsi a si, he a ea o coni e ous o es s is ela i e small and loca es in he sou he n pa o he coope a i e. The moun ains domina e he landscape, and he ele a ion anges be ween 275 and 1320 m abo e sea le el. Du ing his s udy he mean snow dep h in Ap il was 56-99 cm in I alo and 59-68 cm in Käsi a si. The mean empe a u e in July was 13.6-14.3 oC in I alo and 13.5- 15.7 oC in Käsi a si coope a i e, espec i ely (Finnish Me eo ological Ins i u e). The e we e on a e age 5660 eindee (adul s and cal es a e slaugh e ing) in he eindee - he ding coope a i e o I alo du ing he s udy in 1999-2001, and 10 690 eindee in Käsi- a si du ing 2002-04. The maximal eindee numbe can be in I alo 6000 and in Käsi a si 10 000. Reindee densi ies (win e s ock/km2 land a ea) we e almos equal, 2.1 in I alo and 2.3 in Käsi a si. Du ing 2000-03 in he whole no he n eindee husband y a ea (including in o al 13 coope a i es) he minimum numbe s o la ge ca ni o es we e: 45-50 b own bea s, 40-45 wol e ines, 5-10 lynx and 3 wol es (Kojola & Mää ä, 2004). The mos abundan among la ge ca ni o es was b own bea in I alo and wol e ine in Käsi a si coope a i es. How- e e , he exac numbe o hese elusi e species wi hin he s udy a eas is no known. 90% o he known 400 golden eagle e i o ies in Finland we e si ua ed in he eindee husband y a ea and he es ima e o indi idual eagles be o e all mig a ion was a ound 1000 (Ollila, 2003). Mos e i o ies in I alo and Käsi a si we e si ua ed in he alpine highlands, and abou 16 Rangi e , 31 (1), 201174 golden eagles we e li ing du ing his s udy in I alo and 21 in Käsi a si eindee -he ding co- ope a i es (Ollila, pe s. comm.). Ma e ial and me hods In o al 1725 eindee cal es (806 in I alo and 919 in Käsi a si coope a i es, annually ca. 3% o all cal es in he s udy a eas) we e i ed wi h mo ali y indica ing adio- ansmi e s (Tele il Inc., Lindesbe g, Sweden; equency 230 MHz) ixed on expandable neck colla s in di e en s udy a eas du ing 1999-2004. Radio-colla s weighed abou 100 g ams and co esponded 0.3-2.5% o he body weigh o he cal es a ma king. Cal su i al and cause- speci ic mo ali y we e s udied wo yea s in he he ding-g oups o Nellim (1999-2000) and Rai ijä i (2002-03), and h ee yea s in o he he ding-g oups (Sou he n a ea o I alo 1999- 2001, Palojä i and Ko a-Labba 2002-04). 11 cal es om Nellim he d we e adio-colla ed a midsumme ea ma king in 2001, and su i al o hese cal es was included in he su i al es- ima es o Nellim he d and I alo coope a i e. Cal es we e weighed and ma ked a he age o 2-5 days in cal ing co als in May/June. In he co als he emales we e ed o 1.5 mon h wi h silage and concen a es du ing sp ing and cal ing pe iod. Midsumme ea ma king ook place in he las weeks o June and he i s weeks o July, when he age o cal es was 2-8 weeks. Cal es we e ma ked in i e di e en co als in I alo and six in Käsi a si eindee - he ding coope a i es (see Fig. 1). Cal es we e sexed (950 emales, 831 males and 17 sex no known) and weighed, and pel colou was e- co ded. A e secu ing ha he adio-colla ed cal es we e euni ed wi h hei mo he s, ein- dee we e eleased o he summe pas u es. A e ma king, signals o mo ali y ans- mi e s, which ac i a e a e being 2.5 hou s mo ionless, we e adio- acked bo h om he ai ( ixed-wing ai c a and helicop e ) and by g ound iangula ion o loca e dead cal es. T acking was pe o med in 2-3 day in e als du ing summe un il he end o Augus , and once pe week in Sep embe and Oc obe . On he ield, ac i a ed adio-colla s we e loca ed by using hand ecei e s (Tele il RX-8910®, Tele il Inc., Lindesbe g, Sweden and T ack- e Maxima® and hound ada s, T acke Inc., Oulunsalo, Finland). Field obse a ions o he si e and ca cass we e eco ded and pho o- g aphed. Cause o dea h was i s in es iga ed in he ield (e.g. e idence suppo ing p esence o p eda o /sca enge species, such as acks, sca s and ea he s/downs) and hen suppo ed by nec opsies conduc ed by biologis s in he labo a o y o he Reindee Resea ch S a ion a Kaamanen. The p esence o haemo hages and pe o a ions, bo h in he skin and so issues o he dead cal , we e c i ical o de e mining he cause o dea h when dep eda ion was suspec - ed. I he combined e idence om he ield si e and he nec opsy was inadequa e, usually due o la e disco e y o ca cass, he cause o dea h was classi ied as unknown (see also Bjä all e al., 1990; No be g e al., 2005, 2006). S a is ical analysis Due o he di e ence in he ma king ime and age o cal es he weigh o he cal es a ma k- ing anged be ween 4.2-32 kg, and he e o e o s a is ical analysis he weigh s we e adjus ed o 1s o June and o 1s o July by using a daily g ow h a e o 270 g ams (see Timisjä i e al., 1982) o all cal es weighed in he cal ing co als in May and a daily g ow h a e o 302 g ams o emale and 315 g ams o male cal es weighed la e du ing ea ma king (No be g e al., 2005, 2006). The daily su i al es ima es and ´ eindee days´ (one ´ eindee day´= one adio colla ed eindee ou o one day) o he adio-colla ed cal es we e calcula ed using he Kaplan-Meie p oduc /limi me hod (Kaplan & Meie , 1958) and using he compu e p og am ´Kaplan- Meie su i o ship analysis e sion 1.0´ (Pol- Rangi e , 31 (1), 2011 75 lock e al., 1989) o ob ain dai- ly and o al su i al es ima es o he s udy pe iods. Daily su i al es ima es we e used o p esen su i o ship cu es be- ween May/June and Oc obe . Fo calcula ing mon hly su - i al es ima es, cause-speci ic mo ali y a es and 95% con i- dence limi s, he p og am ´Mi- c omo e sion 1.3´ (Heisey & Fulle , 1985) was used. The su i al es ima es we e calcula ed based on documen - ed cases, i.e. only he cal es ha 1) we e ound dead, 2) had d opped hei adio-colla s du ing he s udy o 3) we e e- co e ed in he au umn/win e ound-ups (su i o s) when a- dio-colla s we e aken o we e included in he su i al analy- sis. The s a is ical di e ences in cal weigh s in di e en g oups we e es ed using - es and s epwise logis ic eg ession. In addi ion o weigh , also he e - ec o sex, pel colou , s udy yea and possible in e ac ions on su i al p obabili y we e in es- iga ed using logis ic eg ession. S a is ical es s we e ca ied ou by use o SPSS e . 7.0 o Windows. The da a we e examined o s a is i- cal signi icance a P<0.05. Resul s O 806 eindee cal es adio-colla ed in I alo eindee -he ding coope a i e du ing 1999- 2001 in o al 4.6% (37 cal es) was ound dead, 90.4% su i ed and 5.0% was no ecap u ed un il he end o he s udy (annual moni o - ing om ma king un il he end o Oc obe ). O 919 adio-colla ed cal es in Käsi a si co- ope a i e du ing 2002-04 in o al 5.2% (48 cal es) was ound dead, 87.4% su i ed and 7.4% was no ecap u ed, espec i ely. Highes a ea-speci ic annual mo ali y occu ed in I alo in 2000, when in o al 19 (51.4% o all dead cal es) we e ound dead, and in Käsi a si in 2004, when 20 adio-colla ed cal es (41.7%) we e ound dead. Pooled su i al es ima es in he coope a i e o I alo (including all adio-colla ed cal es du ing yea s 1999-2001) and in he coope a- i e o Käsi a si (yea s 2002-04) we e 0.937 (SE=0.011) and 0.885 (SE=0.011), espec i e- ly. Su i al unc ion cu es (pooled da a o he whole s udy) o he coope a i es o I alo and Käsi a si as well as o he i e di e en he ding g oups s udied a e p esen ed in Figs. 2 A and B. Fig. 2. Su i o ship cu es o adio-colla ed eindee cal es (A) in I alo coope a i e and in i s he ding-g oups du ing 1999-2001, and (B) in Käsi a si and i s he ding-g oups du ing 2002-2004. Day 1 is 1s o May. (Mo ali y (M) = 1 – Su i al (S)). Rangi e , 31 (1), 201176 The mo ali y es ima e (M=1-Su i al) o he cal es adio-colla ed in Nellim a ea (coope a- i e o I alo) du ing cal ing ime in May/June and moni o ed o end o Oc obe was 0.051 (SE=0.026) in 1999 and 0.086 (SE=0.038) in 2000. Pooled mo ali y o bo h s udy yea s in Nellim was 0.067 (SE=0.022). The annual mo ali y es ima es o cal es adio-colla ed in cal ing co als in Ko a-Labba he ding-g oup (coope a i e o Käsi a si) in 2002-04 we e 0.051 (SE=0.021), 0.093 (SE=0.027) and 0.118 (SE=0.042), espec i ely. Including also addi ional cal es adio-colla ed in midsumme ea ma kings in Ko a-Labba, he annual mo - ali y es ima es we e 0.049 (SE=0.018), 0.115 (SE=0.029) and 0.146 (SE=0.036) in 2002- 04, espec i ely. The pooled h ee-yea mo al- i y in Ko a-Labba was 0.090 (SE=0.017). In o al 52% o mo ali y among cal es adio- colla ed du ing cal ing in Nellim and 87% o mo ali y in Ko a-Labba ook place be o e he end o June. The pooled mo ali y es ima es o cal es adio-colla ed du ing ea ma king in June/ July we e 0.048 (SE=0.012) in Sou he n a ea o I alo (yea s 1999-2001), 0.018 (SE=0.013) in Rai ijä i (yea s 2002-03) and 0.057 (SE=0.013) in Palojä i (yea s 2002-04), e- spec i ely (Figs. 2A and B). Pooled mo ali y es ima es o co esponding ime pe iods ( om 15 June un il he end o Oc obe ) we e 0.047 (SE=0.018) in Nellim and 0.030 (SE=0.009) in Ko a-Labba a eas. The annual mo ali y es ima es analysed om midsumme ea ma kings un il he end o Oc obe in Sou he n a ea o I alo coope a i e we e 0.080 (SE=0.037), 0.085 (SE=0.027) and 0.016 (SE=0.010) o yea s 1999-2001, espec- i ely. In Palojä i, he sou he nmos he ding- g oup o Käsi a si, annual mo ali y es ima es we e 0.051 (SE=0.022), 0.056 (SE=0.024) and 0.074 (SE=0.024) o yea s 2002-04, e- spec i ely. Ma king ime o cal es was la es in Rai ijä i, he no he nmos he ding-g oup o Käsi a si, and ook place mainly du ing he second week o July bo h in 2002 and 2003. In 2002 no any o he adio-colla ed cal es we e ound dead in Rai ijä i a ea, and cal es su i ed well also in 2003 mo ali y es ima e being 0.034 (SE=0.24). Fo compa ison wi h o he he ding-g oups o Käsi a si ( ime pe iod July-Oc obe ), he annual mo ali y es ima es Dis ic Yea s Radio days Su i al es ima e 95% CI Eagle (n) O he p e- da ion (n) O he causes (n) Unknown (n) I alo 1999-2001 96 106 93.2 90.7-95.7 2.1 (16) 0.8 (4) 1.6 (7) 2,3 (10) Nellim11999-2001 26 918 93.0 89.4-96.7 3.0 (6) 1.1 (2) 0.5 (1) 2,3 (4) I alo sou h 1999-2001 69 188 92.6 87.9-97.6 1.7 (10) 0.3 (2) 4.3 (6) 1,1 (6) Käsi a si 2002-2004 107 885 88.0 83.5-92.7 3.0 (16) 0.6 (4) 4.0 (9) 4.3 (19) Ko a-Labba 2002-2004 58 492 89.0 84.3-93.8 3.1 (11) 0.5 (2) 3.7 (5) 3,6 (11) Palojä i 2002-2004 37 065 94.3 91.7-97.0 1.7 (5) 0.6 (2) 1.4 (4) 2,0 (6) Rai ijä i 2002-2003 12 328 98.2 95.7-1.0 0.0 (0) 0.0 (0) 0.0 (0) 1.8 (2) 1 in 2001 only 1400 adio days o Nellim he d (no dead cal es we e ound) Table 1. Su i al and cause-speci ic mo ali y es ima es o adio-colla ed eindee cal es in he s udied eindee he ding-coope a i es and he ding-g oups du ing s udy pe iods in 1999-2004. The su i al and mo ali y es ima es wi h 95% CI a e calcula ed acco ding o 2-week in e als using so wa e Mic omo 1.3. Numbe o dea h cases is in pa en hesis. Rangi e , 31 (1), 2011 77 among all cal es ma ked in Ko a-Labba (bo h in he cal ing co als and in midsumme ea - ma kings), we e 0.007 (SE=0.007), 0.042 (SE=0.019) and 0.043 (SE=0.022) in 2002- 04, espec i ely. Golden eagle was he mos signi ican cause o dea h in I alo and Käsi a si coope a i es killing 0-3.5% o adio-colla ed cal es in di - e en s udy a eas and yea s. Mos cal es killed by golden eagle ( o ally 11) we e ound in Ko a-Labba he ding-g oup in Käsi a si du ing 2002-04. Fi e eagle-killed cal es we e ound in Palojä i he ding g oup, while in Rai ijä i no cal es killed by golden eagle we e ound. In I alo coope a i e golden eagle killed du ing 1999-2001 in o al 16 cal es, o which 6 we e cal es om Nellim he d. In pooled da a including all adio- colla ed cal es du ing he whole s udy, p eda ion by golden eagle comp ised on a e age 2% mo ali y a e in I alo and 3% mo ali y a e in Käsi a si (Table 1). Golden eagle p eda ion was ob- se ed al eady on cal ing g ounds, bu peaked in July-Augus , when eindee p e e ed highe ele a ions o g azing and escaping insec ha assmen . Mos cal es (24 om 32) killed by golden eagle we e ound in he open a eas, as in highlands, bogs and clea -cu o - es a eas. In I alo coope a i e, 72 % o he golden eagle p eda ion occu ed in he o es ed habi a s, including 22% o open clea -cu o seedling s ands a eas. In Käsi a si coope a i e, due o highe p opo ion o open moun ainous land- scape, 85% o golden eagle p eda ion and 79% o o he mo ali y occu ed in ell habi a s. The es o eagle p eda ion (15%) occu ed on mi es. On a e age 54% o he adio-colla ed cal es ound dead in I alo and 42% in Käsi a si coope a i es we e a ibu ed o p eda ion, espec i ely. Golden eagle p eda ion accoun ed o 33-43% o he cases and 80% o all iden i ied p eda ion (Figs. 3A and B). In I alo some cal es we e killed also by b own bea (8% o all cal es ound dead) and ed ox (3%), and in Käsi a si by wol e ine (6%) and uniden i ied p eda o (2%). Mo al- i y a es caused by o he p eda o s han golden eagle we e on a e age less han one pe cen in bo h s udy coope a i es. Ye , bea -caused mo - ali y was highe in Nellim han in he sou he n a ea o I alo (Table 1). The sha e o o he iden i ied causes o dea h (acciden s, a ic, o he s) was in bo h coop- e a i es 19%, while 27-40% o he succumbed Fig. 3. The dis ibu ion o all dea h causes o adio-colla ed cal es ound dead (A) in I alo and (B) in Käsi a si eindee - he ding coope a i e du ing he s udy. Rangi e , 31 (1), 201178 cal es we e associa ed wi h unknown causes o dea h. In I alo du ing 1999-2001 a o al o 8% o all cal es ound dead, succumbed in acciden s and in a ic, while 6% succumbed in acciden s in Käsi a si, espec i ely. O he causes (11-13%) included in bo h coope a i es disease, s ess and poo condi ion o cal es. When cal es wi h uniden i ied cause o dea h we e excluded om dis ibu ion, p eda ion comp ised on a e age 74% o he obse ed mo ali y (i.e. o cal es wi h iden i ied cause o dea h) in I alo and 69% in Käsi a si. The a e age adjus ed (on June 1s , weighed in he cal ing co als in May) weigh o hose a- dio-colla ed cal es ha su i ed in I alo coop- e a i e was signi ican ly highe (mean 9.8 kg, SD=2.0 kg, n=169) han weigh o succumbed cal es (mean 7.8 kg, SD=2.2 kg, n=12; =2.79, d =179, P=0.006). Acco ding o he s epwise logis ic eg ession, male cal es had highe isk o die han emale cal es (Wald=9.88, d =1, P=0.002), and weigh had signi ican e ec o su i al o male cal es (Wald=8.30, d =1, P=0.004), bu no on su i al o emale cal es. Cal es killed by golden eagles we e signi i- can ly and 1 kg ligh e (mean 7.2 kg, SD=2.2 kg, n=5; =2.47, d =172) han cal es ha su - i ed, and also ligh e han hose cal es ha succumbed due o o he causes (mean 8.2 kg, SD=2.4 kg, n=7). Also in Ko a-Labba he ding g oup in Käsi a si coope a i e cal es killed by golden eagles we e 1.7 kg ligh e (mean 10.7 kg, SD=1.4 kg, n=5) han hose cal es ha su i ed (mean 12.4 kg, SD=1.9 kg, n=259; =1.9, d =262, P=0.053). The a e age adjus ed (on July 1s , weighed in ea ma king in June/July) weigh o he adio- colla ed cal es ha we e killed by golden eagles in he I alo coope a i e was signi ican ly lowe (mean 12.8 kg, SD=1.7 kg, n=11) han he mean weigh o su i o s (16.6 kg, SD=2.5 kg, n=560; =760, d =10.84, P<0.001). Like- wise, he mean weigh o cal es killed by all p eda o s was signi ican ly lowe (13.1 kg, SD=1.7 kg, n=13) han he mean weigh o he su i ing cal es ( =7.05, d =13.17, P<0.001). Simila ly, he weigh o he cal es killed by golden eagles in he Palojä i he ding-g oup in Käsi a si was signi ican ly lowe (mean 13.0 kg, SD=1.7 kg, n=5) han he mean weigh o su i o s (19.3 kg, SD=3.4 kg, n=280; =409, d =283, P<0.001). Fu he mo e, he mean weigh o he cal es killed by all p eda o s was signi ican ly lowe (12.2 kg, SD=2.2 kg, n=7) han he mean weigh o he su i ing cal es ( =5.46, d =285, P<0.001). The weigh o cal es killed by golden eagles in Ko a-Labba he ding g oup was e y low and signi ican ly lowe (mean 7.2 kg, SD=3.2 kg, n=6) han he mean weigh o he su i ing cal es (13.4 kg, SD=4.3 kg, n=28; =3.03, d =31, P=0.005). The weigh s o cal es killed by golden eagles o o he p eda o s did no di e om he weigh s o cal es ha succumbed due o o he causes in I alo no in Käsi a si. Mo eo e , no signi ican e ec o pel colou on cal su i al was ound in he p esen s udy. Discussion In mos mammal species he mo ali y cu e is ypically U-shaped and he neona es and ju e- niles su e he highes mo ali y, which hen g adually dec eases owa ds he p ime-aged indi iduals (Caughley, 1966). Also in S alba d wild eindee (Rangi e a andus pla y hynchus V olik), exis ing in an en i onmen almos ee o p eda o s, mo ali y a es ollow his U-shaped pa e n, wi h highe mo ali y a e among cal es and old animals han in middle aged indi iduals. Mo ali y a es in cal es o S alba d eindee a e calcula ed o 1% and 19% in he age in e als 0-6 and 6-12 mon hs (Reime s, 1983). The mo ali y a es among 0-6 mon hs-old eindee cal es is expec ed o be wi hin he ange o 6-21% ound in he p eda o - ee Sou h-Geo gia (Leade -Wil- liams, 1980), and lowe han 45-60% ound in many Rangi e he ds subjec ed o p eda ion Rangi e , 31 (1), 2011 79 (Rehbinde , 1975; Be ge ud, 1980). Wol es, bea s and golden eagles ha e been he mos impo an p eda o s o adio- agged eindee / ca ibou cal es in bo h No h Ame ica and Rus- sia, and in a s udy by Be ge ud (1988) e en 80-89% o dea hs o ca ibou we e caused by p eda ion. Many s udies on eindee /ca ibou ha e dem- ons a ed ha cal mo ali y is usually highes du ing he i s days and weeks a e cal ing and hen dec eases conside ably du ing sum- me and au umn (see Elo an a & Nieminen, 1986; Whi en e al., 1992; Adams e al., 1995; No be g e al., 2005, 2006). I is impo an o no e he e ha , he shape o he su i o ship unc ion cu e is associa ed wi h he s a ing da e o he s udy and composi ion o local p eda o guild, i.e. which species and in which numbe s a e ound in he a ea o in e es . Fo example, he s udy in he sou he n he d o I alo was s a ed om he second week o June, while bea p eda ion commonly peaks al eady a he u n o May and June (see Ad- ams e al., 1995; Young & McCabe, 1997). Thus, by being unable o ma k cal es al eady a cal ing may lead o unde es ima es o o al cal mo ali y and lea e some mo ali y-agen s, such as bea p eda ion, incomple ely obse ed. Indeed, mos bea p eda ion in ou s udy was obse ed in Nellim a ea, whe e cal es we e a- dio-colla ed in cal ing co als. Mo ali y es ima es ound in his s udy we e signi ican ly lowe han he o al mo ali y om Augus 1995 o Ap il 1996 in No h- T øndelag in No way (31%) (Nybakk e al., 2002). The es ima es we e also lowe han in he s udy conduc ed in he 1980s in Umbyn, Sweden, whe e mo ali y om July o Ap il was 14%, and was mainly caused by lynx and wol- e ine (Bjä all e al., 1990). Howe e , in he p esen s udy cal es we e adio-colla ed pa ly a he cal ing a eas in May and adio- acking was conduc ed un il he ea ly win e ound- ups, when adio-colla s we e emo ed om he su i o s. Win e mo ali y, which may com- p ise subs an ial losses in Rangi e he ds (e.g. Reime s, 1983), due o se e i y o he win e o p esence o la ge ca ni o es, was no in es- iga ed in he p esen s udy. By compa ing he esul s o Bjä all e al. (1990) wi h he p esen s udy using co esponding pe iod o ollow- up, one no ices, ha he mo ali y es ima e be ween July and Oc obe is simila o e en lowe in he Swedish s udy being 4-5%. The mo ali y es ima es om he Swedish s udy i wi hin he annual a ia ion in su i al bo h in I alo and Käsi a si. The closes e e ence o I alo and Käsi a si, howe e , is om he ea lie Finnish s udy by No be g e al. (2006). The mo ali y es ima e om July un il he end o subsequen Janua y was ca. 8.5% in bo h yea s, and 10.7% in pooled da a comp ising ime pe- iod om cal ea ma king in la e June un il he end o Oc obe . The majo i y o semi-domes ica ed eindee cal es a e bo n in no he n Finland in May (peak cal ing aking place on 18-23 May: Elo an a & Nieminen, 1986; Weladji e al., 2006), and in p esen s udy, o ally 52% o cal mo ali y in Nellim and 87% in Ko a- Labba eindee -he ding g oups ook place be- o e he end o June. The annual mo ali y o he cal es adio-colla ed du ing cal ing ime in May and moni o ed o he end o Oc obe was maximally 11.8% (Ko a-Labba in 2004) and o he cal es adio-colla ed du ing ea ma king in June/July 8.5% (sou he n a ea o I alo in 2000). When including cal es bo n la e and ma ked a he midsumme ea ma king, he mo ali y in Ko a-Labba inc eased o 14.6% due o he ac ha many o he adio-colla ed cal es a ha speci ic momen and place we e cal es o young emales and hus bo n la e and ligh e han a e age. These cal es had a highe han a e age p obabili y o succumbing due o p eda ion and o he causes. As hypo hesized, he mo ali y in he p esen s udy was highes du ing he i s mon h(s) a e ma king.