The risk of moose Alces alces collision: A predictive

logistic model for moose-train accidents

Authors: Gundersen, Hege, and Andreassen, Harry P.

Source: Wildlife Biology, 4(2) : 103-110
Published By: Nordic Board for Wildlife Research
URL: https://doi.org/10.2981/wlb.1998.007

BioOne Complete (complete.BioOne.org) is a full-text database of 200 subscribed and open-access titles
in the biological, ecological, and environmental sciences published by nonprofit societies, associations,
museums, institutions, and presses.

Your use of this PDF, the BioOne Complete website, and all posted and associated content indicates your
acceptance of BioOne’s Terms of Use, available at www.bioone.org/terms-of-use.

Usage of BioOne Complete content is strictly limited to personal, educational, and non - commercial use.
Commercial inquiries or rights and permissions requests should be directed to the individual publisher as
copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit
publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to
critical research.

Downloaded From: https://bioone.org/journals/Wildlife-Biology on 02 Sep 2022

Terms of Use: https://bioone.org/terms-of-use
 SHORT S hort com m u n icatio n articles are short scientific entities often dealing
w ith m ethodolo g ical p ro b lem s or w ith bypro d u cts o f larg er research
COMMUNICATION projects. T he style should b e the sam e as in original articles.

The risk of moose Alces alces collision: A predictive logistic model

for moose-train accidents
Hege Gundersen & Harry P. Andreassen

Gundersen, H. & Andreassen, H.P. 1998: The risk o f m oose Alces alces col­
lision: A predictive logistic model for moose-train accidents. - Wildl. Biol.
4: 103-110.

We used logistic m odels to estim ate the risk of m oose-train collisions for the
Rprosbanen railway in Norway. During 1990-1997, a total of 13,506 train
departures w ere registered along R prosbanen during the m onths w hen the
risk o f collision was highest (Decem ber to March). The statistical model
selected to predict the risk o f moose-train collisions included train route,
tim e o f day, lunar phase and average train speed, as w ell as two climatic
covariables, i.e. snow depth and temperature. Trains running at night, in the
m orning or in the evening experienced a higher risk o f collision with moose
A lces alces than day trains. The probability of collision was also higher dur­
ing nights of full m oons than during nights o f h alf or no m oons. As observed
previously with trains in Norw ay m oose-kills increased with increasing
snow depth and decreasing tem peratures. To test the predictability o f the
model, we used a logistic model based on train departures during 1990-1996
to predict the num ber of m oose-train accidents during w inter 1996/97.
Although the model had a satisfactorily high predictability, the best models
would probably be those based on a com bination o f both tem poral and spa­
tial aspects. We discuss how logistic m odels may be applied to introduce
rem edial actions on high-risk routes or during high-risk periods.

Key words: A lces alces, gam e-vehicle accidents, logistic m odel, railway

H ege Gundersen, D epartm ent o f Biology, D ivision o f Zoology, University o f

Oslo, P.O.Box 1050 Blindern, N-0316 Oslo, Norway
H arry P. A ndreassen*, H edm ark College, Evenstad, N-2480 Koppang,
N orw ay

*Present address: D epartm ent o f Biology, D ivision o f Zoology, University

o f Oslo, P.O.Box 1050 Blindern, N -0316 Oslo, Norw ay -
e-mail: harrya@ bio.uio.no

Received 29 A ugust 1997, accepted 26 January 1998

Associate Editor: H eikki H enttonen

© W ILDLIFE BIOLOGY • 4:2 (1998) 103

Downloaded From: https://bioone.org/journals/Wildlife-Biology on 02 Sep 2022

Terms of Use: https://bioone.org/terms-of-use
 In m o st parts o f th e w o rld th e re h as been an increase
in th e n u m b e r o f g am e-v eh ic le accid en ts d uring the
la st d ecad es (G ro o t B ru in d erin k & H az eb ro ek 1996,
R o m in & B isso n ette 1996), resu ltin g in large so c io ­
ec o n o m ic costs (Jaren, A n d ersen , U lleberg, P ed ersen
& W iseth 1991, G ro o t B ru in d e rin k & H azeb ro ek
1996, R o m in & B isso n ette 1996) and p sy ch o lo g ical
p ro b le m s fo r p e o p le in v o lv e d in th e a c c id e n ts
(V atshelle 1995, G u n d ersen , A n d reassen , H aav e &
S to raas 1997). G am e -v eh ic le accid en ts also have
co n seq u e n ces fo r w ild life m an ag em en t, as accidental
d eath o f an im als m ay have sev ere and unpred ictab le
effects on p o p u la tio n d ev e lo p m e n t (P etersso n &
D an ell 1992).
T h e in creasin g n u m b e r o f g am e-v eh ic le accidents
has p ro m o ted an in te n sific atio n o f rem ed ial actions,
an d studies o n the effec tiv e n ess o f d ifferen t m itiga- Figure 1. Location of the R0rosbanen railway study area in
Hedmark county, Norway, with indications of the stations men­
tiv e te ch n iq u es (e.g. S ch o b er & S o m m er 1984, Ja re n tioned in the text.
e t al. 1991, W aring, G riffis & V aughan 1991, R om in
& D alto n 1992, G lea so n & Jen k s 1993, R eev e &
A n d erso n 1993, L u tz 1994, G ro o t B ru in d erin k & acco rd in g to tim e and v ario u s en v iro n m en tal and
H azeb ro ek 1996, R o m in & B isso n ette 1996). T he tech n ical facto rs has, to o u r k n o w led g e, n ev e r been
in crease in th e n u m b e r o f g am e-v eh icle accid en ts has estim ated .
also m o tiv ate d studies addressin g g am e beh av io u ral W e u se d a statistical lo g istic m o d el to estab lish the
an d en v iro n m en tal facto rs w h ich m ay cau se anim als m o st risk y train d ep artu res fo r th e R p ro sb an en ra il­
to b e clo se to traffic arteries. Im p o rtan t g am e b e h a v ­ w ay w h ich has th e h ig h e st risk o f m o o se -tra in co lli­
io u ral aspects in c lu d e m ig ra tio n (A llen & M c C u l­ sions p er km in N o rw ay (G u n d ersen et al. in press).
lo u g h 1976, G o o d w in & W ard 1976, A ndersen, In the m o d el w e in clu d ed train speed, ty p e o f train,
W iseth , P ed erse n & Ja re n 1991, L av su n d & S ande- tim e o f day an d lu n ar ph ase, as w ell as clim atic
g ren 1991, G u n d ersen , A n d reassen & S toraas in c o v ariab les k n o w n to be co rrelated w ith m o o se-train
p ress) an d daily activity rh y th m (C arbaugh, V aughan, co llisio n s.
B eilis & G ra v es 1975, A llen & M cC u llo u g h 1976),
w h ile in flu en tial en v iro n m en ta l factors are snow
dep th (A n d ersen et al. 1991, G u n d ersen et al. in Material and methods
press), te m p eratu re (A n d ersen et al. 1991, G u n d ersen
et al. in press) and lan d scap e featu res (C arbaugh et Study area
al. 1 975, B a sh o re , T z ilk o w s k i & B e ilis 1985, T h e study w as restricted to the R p ro sb a n en railw ay
F eld h am er, G ates, H arm an , L o ra n g er & D ix o n 1986, w h ich ru n s th ro u g h the v alley o f 0 s te rd a le n from
G leaso n & Jen k s 1993, G u n d ersen e t al. in press). E lv e ru m (6 0 °5 3 ’N , H ° 3 4 'E ) to R p ro s (6 2 °3 5 'N ,
H o w ever, it is n o t p o ssib le to m o d ify gam e b eh a v io u r 11°20’E ) (Fig. 1). R p ro sb an en is 240 km lo n g , an d
an d en v iro n m en tal facto rs so that th e n u m b e r o f h as th e h ig h e st freq u en cy o f m o o se -tra in co llisio n s
g am e-v eh icle accid en ts w ill be reduced. (0.36 m o o se k illed /y ear/k m ; G u n d ersen et al. in
S o m e tech n iq u es u sed to p rev e n t g am e-train ac ci­ p ress) in N orw ay. T h e railw ay ru n s th ro u g h th e v al­
d en ts, such as red u c ed speed, ultraso n ic w arning ley, w h ich is su rro u n d ed by h ills o f b o real forest,
so u n d s, as w ell as the n ew ly in tro d u ced p ilo t ca r d o m in ated by N o rw ay sp ru ce P icea a b ies an d S cots
w h ich runs ah ead o f train s frig h ten in g g am e aw ay p in e P in u s sylvestris, in tersp ersed w ith a few b o real
fro m th e railw ay (T. S teph en so n , pers. com m .), m ay d ecid u o u s species such as b irch B etu la p u b esce n s
be lim ited to certain p erio d s o f the year, o r to p artic­ an d w illo w S a lix spp.
u lar routes, w ith a hig h p ro b ab ility o f c o llisio n s w ith
g am e. H ow ever, the p ro b ab ility o f m o o se-co llisio n Data material
alo n g a sp ecific train route, and how the risk varies D ata on train k ills and o n train d ep artu res w ere

104 © W IL D LIFE BIO LO GY • 4 :2 (1998)

Downloaded From: https://bioone.org/journals/Wildlife-Biology on 02 Sep 2022

Terms of Use: https://bioone.org/terms-of-use
 o b ta in e d fro m T h e N o rw e g ia n S tate R a ilw a y s Table 1. Number of train departures analysed, numbers of moose-
train collisions and risks of moose-train collisions along eight
(N S B ), T he N orw eg ian N atio n al R ail A d m in istra­ routes at the Rprosbanen railway. Route indicates the section of
tion, and fro m the local W ildlife C o m m ittees in the the Rprosbanen railway run by the specific train route, but no dis­
m u n icip alities o f S tor-E lvdal and R en d alen fro m 1 tinctions were made between north and south going trains. In
parentheses the distances of the routes are given. See map of the
D ecem b er 1990 till 31 M arch 1997. E ach reco rd o f a Rprosbanen railway line in Figure 1.
m o o se-train co llisio n inclu d ed the train ro u te on
No of train No of
w h ich the co llisio n to o k place. D ata concern in g c li­ Route (distance in km) departures collisions Risk (95% Cl)
m ate, i.e. daily averag e tem p eratu re and snow depth,
Alvdal - Tynset (23) 160 0 0
w ere o btained from the m eteo ro lo g ical station in Tynset - Rena (157) 15 0 0
E v en stad (61°4'N , 11°7'E) (see F ig. 1). Koppang - Tynset (100) 15 0 0
Elverum - Koppang (88) 2317 31 0.013 (0.009, 0.018)
Elverum - Rena (32) 1523 1 0.001 (0.000, 0.003)
Analyses Elverum - Rdros (240) 7201 317 0.045 (0.040. 0.051)
Elverum - Tynset (189) 1405 48 0.035 (0.026, 0.046)
W e h ave prev io u sly show n that 79% o f all m oose- Tynset - Rdros (52) 870 10 0.012 (0.006, 0.020)
train accidents along the R p ro sb an en railw ay occu r
fro m D ecem b er to M arch, p robably due to m igratory
b eh a v io u r o f m o o se m oving from the hills w ith large m o d el w as ap p lied in clu d in g all av ailab le train te ch ­
am o u nts o f snow do w n to the bo tto m o f the v alley o f nical p red ic to r v ariab les, i.e. av erag e train speed,
0 ste rd a le n w here th e railw ay runs (G undersen et al. ty p e o f train (freig h t or p assen g er train) and tim e o f
in press). T hus, w e chose to inco rp o rate only the four day (coded as n ig h t 2 1 :0 0 -3 :0 0 , m o rn in g 3 :0 0-9:00,
m o n th s, D ecem b er - M arch, w ith the h ig h e st risk o f day 9 :0 0 -1 5 :0 0 , ev en in g 15:00-21:00, resp ectiv ely ).
m o o se-train collisio n s in the p resen t analyses. T he We in clu d ed d aily snow d epth an d d aily av erag e te m ­
p red icted risk o f m o o se-train collisio n s fro m A pril to p eratu re as co v ariab les, since th ese tw o clim atic v ari­
N o v em b er (i.e. th e p art o f the y ea r ex clu d ed fro m the ab les p rev io u sly h av e been show n to be g o o d p red ic­
analyses) ranges from 0.001 (July and A ugust) to tors fo r th e y early v ariatio n in the n u m b er o f m o o se
0 .0 0 9 (N ovem ber). T h erefo re, inclusion o f these k illed by trains in N o rw ay (A n d ersen et al. 1991,
m o n th s in th e analyses w o u ld resu lt in low p re ­ G u n d ersen et al. in press). P erso n al o b serv atio n s su g ­
d ictab ility o f the ch o sen statistical m o d el due to the g ested th at lunar p h ase m ig h t affect the n u m b er o f
low explain ab le varian ce o f the bin o m ial response. m o o se-train co llisio n s along the R p ro sb an en railw ay
F o r each train dep artu re on the R pro sb an en railw ay (S. S letten, pers. com m .). W e th erefo re tested this
d u rin g the fo u r m onths fro m D ecem b er 1990 to su g g estio n by in clu d in g lu n ar p h ase (full m oon: ± 3
M arch 1997 w e reg istered w h eth e r the train h it a days, h a lf m oon: ± 3 days, n ew m oon: ± 3 d ay s) in
m o o se (coded as 1) o r n o t (coded as 0). A logistic the m odel. H ow ever, n ot all train ro u tes run all the

Table 2. The logit model chosen to best explain the risk of moose-train collisions. (AIC = 3228.2; AIC without train speed = 3228.7; AIC
for other models > 3230.6).

Predictor Estimate df xJ P

Intercept -7.033 ± 0.880

Train route 4 110.75 <0.001
Route Elverum - Koppang 0.554 + 0.375
Route Elverum - Rena -1.488 ± 1.057
Route Elverum - Rpros 1.705 ±0.331
Route Elverum - Tynset 1.173 ±0 .3 6 5
Route Tynset - Rdros 0
Time of day 3 94.44 <0.001
Morning -0.163 ±0 .1 4 7
Day -1.625 ± 0.229
Evening 0.101 ± 0 .1 4 2
Night 0
Lunar phase 2 9.05 0.011
Full 0.358 ± 0 .1 4 2
Half 0.010 ± 0 .1 2 9
New 0
Snow depth (cm) 0.017 ± 0 .0 0 2 1 107.74 < 0.001
Average daily temperature (°C) -0.039 ± 0.007 1 30.98 < 0.001
Average train speed (km/h) 0.016 + 0.010 1 2.53 0.111

W IL D LIFE BIOLOGY • 4:2 (1998) 105

Downloaded From: https://bioone.org/journals/Wildlife-Biology on 02 Sep 2022

Terms of Use: https://bioone.org/terms-of-use
 w ay fro m E lv e ru m to R 0ros, and as th e risk o f
0.06 '
m o o se -tra in accid en ts varies along th e railw ay line
0.05 '
(G u n d ersen et al. in p ress) w e also in clu d ed th e sp e­ ce
0.04
cific section o f th e railw ay lin e co v ered by the ro u te Z '

o 0.03
in the m o d el (h ereafter te rm ed ro u te factor). T he NH
C/D
0.02 '
m o st p arsim o n io u s m o d el ac co rd in g to A k a ik e ’s 3
In fo rm a tio n C rite rio n (A IC v alu es; B u rn h am & -4 0.01 '
1
A n d erso n (1993)) w as ch o sen to p red ic t the risk o f o 0 '
u N ight M o rn in g D ay E vening
m o o se -tra in collisions. z T IM E O F DAY

B
Results 0.04
w
C/D
O u t o f a to tal o f 13,506 train d ep artu res alo n g the O
O 0.03
R p ro sb a n en railw ay in the perio d s an alysed, there
w ere 4 0 6 m o o se -tra in collisio n s, w hich k illed 466
fa
m o o se. P re lim in ary analy ses sh o w ed th a t th e p ro b a ­ o 0.02
bility o f co llid in g w ith a m o o se w as 0 .030 (95% C l = F u ll H a lf New
H
0.0 2 8 , 0 .0 3 4 ), b u t it v arie d co n sid erab ly b etw een L U N A R PH A S E
ro u tes (u n iv ariate m odel: x \ , i 3,498 = 174.73, P < 0.001) 0.1 0 1
CO
(Table 1). T he ro u te sp ecific risk o f k illin g a m oose
2
s
co u ld m o st prob ab ly b e ascrib ed b oth to v aria tio n s in
0.05 -
th e d istan ce o f the ro u te (u n iv ariate m odel: x \ 13,504 =
153.56, P < 0 .001), alth o u g h the slo p e o f the linear fa
m o d el w as ra th e r flat (d istan ce in km , estim a te =
0 .0 1 0 (S E = 0 .0 0 1 )), an d spatial v ariatio n s in co lli­
sion risk along th e local areas trav e rsed by the d iffe r­ A V E R A G E T R A IN SP E E D
en t ro utes (e.g. due to d ifferen t fo o d av ailab ility and (k m /h )
to p o g rap h y as show n by G u n d ersen et al. (in press)).
To attain co n v erg en ce fo r a global m o d el (w ith no Figure 2. Probability of moose-train collisions in relation to A) the
time of day, B) lunar phase and C) average train speed (see statis­
in teractio n s) w hen including th e ro u te factor, w e had
tics in Table 2).
to ex c lu d e th ree ro u tes w ith o u t any train kills, c o n ­
sistin g o f 190 d ep artu res (see T able 1). T he m odel
sh o w ed n o sig n ific an t asso ciatio n b etw e en m oose- m o o se-k ills an d av erag e train sp eed ( x \ I3,303 = 2.53,
k ills and train ty p e ( x \ ,3,302 = 0.02, P = 0 .857), or P = 0.111). T h e m o st p arsim o n io u s m o d el, w ith only

Table 3. The logit model used to predict the number of train-killed moose for trains running in the morning, day, evening and at night the
whole distance between Elverum and Rpros in winter 1996/97. (AIC = 1752.2; AIC without train speed = 1752.3; AIC for other models
> 1754.0).
Predictor Estimate df x2 P

Intercept -6.817 ± 1.836

Train 3 68.77 <0.001
Morning -0.175 ± 0 .1 9 3
Day -1.745 ± 0 .2 9 5
Evening 0.098 ±0.181
Night 0
Lunar phase 2 3.57 0.059
Full 0.297 ± 0 .1 5 4
H alf / New 0
Snow depth (cm) 0.018 ± 0 .0 0 2 1 68.13 <0.001
Average daily temperature (°C) -0.031 ± 0.009 1 10.60 0.001
Average train speed (km/h) 0.036 ±0.021 1 2.14 0.144

106 © W IL D LIFE BIO LO GY • 4:2 (1998)

Downloaded From: https://bioone.org/journals/Wildlife-Biology on 02 Sep 2022

Terms of Use: https://bioone.org/terms-of-use
 Table 4. Observed and predicted number of train accidents involving moose-train collisions during winter 1996/97 for departures of pas­
senger trains leaving in the morning, during the day, in the evening and at night and running the whole distance between Elverum and
Rpros. G-statistics are presented for the goodness of fit test.

No of accidents
Train departure No o f departures Observed Predicted G P

M orning 201 14 6.44 3.46 0.063

Day 222 2 1.63 0.04 0.842
Evening 242 11 10.03 0.05 0.823
Night 60 2 2.03 0.00 0.992

statistically significant term s, included route, tim e o f
day, lu n ar phase and the covariables snow depth and
tem p erature (Table 2). H ow ever, according to A IC ,
this m odel w as in distinguishable from a m odel w hich
also included average train speed (see Table 2).
T h e risk o f collision w as 5-6.8 tim es higher during
LU N A R PH A SE
H a lf/N e w
the night, m orning o r ev en in g than during the d ay ­
tim e (Fig. 2A ), and 1.3 tim es h ig h er during periods
w ith full m o o n s than during p erio d s w ith new o r h alf
m oons (see Fig. 2B). T h e pro b ab ility o f m oo se-train
co llisio n s in creased w ith in creasin g train speed,
although n ot significantly so (see Fig. 2C). F or in ­
stance, an increase in train speed
from 50 to 100 k m /h o u r d oubled the
risk o f collision. T h e p ro b ab ility o f
m oose-train co llisions also increased
w ith increasing snow depth and d e­
creasin g tem p eratu re (see T able 2 for
estim ates, and G und ersen et al. in
press).

Train specification
T h e above analyses included all train
departures in 0 ste rd a le n . To depict
the risk o f different departures, w e
m ade a m o d el containing only p a s­
senger trains running the w hole d is­
tance betw een E lverum and R pros
(there are tw o trains (one going south
and one g oing north) in the m orning,
day, ev en in g and night). W e included
snow depth, daily average tem p era­
tu re , lu n a r p h ase (fu ll m o o n vs
half/new m oon), train speed and train
(i.e. tim e o f day) in the m odel. To test
the p redictability o f the m odel w e
ex clu d ed the w inter o f 1996/97 and
used the m odel b ased on the data
from 1990 to 1996 to p red ict the
n u m b e r o f train -k illed m o o se fo r
-30j.V ^ each train (m orning, day, even in g and
-20 n ig h t trains) during w in ter 1996/97.
ofS
<0 * F ro m 1990 to 1996 th ere w ere
<c>
5,371 departures o f the train s in clu d ­
ed in our analyses, and 229 m oose
w ere killed. T he best m o d el acco rd ­
Figure 3. Probability of moose-train collisions of passenger trains running the whole dis­
tance between Elverum and Rpros in the morning, day, evening and at night, in relation ing to the A IC criteria in clu d ed all
to lunar phase, snow depth and temperature (see statistics in Table 3). p red icto r variables (Table 3). In w in ­

© W ILDLIFE BIOLOGY • 4:2 (1998) 107

Downloaded From: https://bioone.org/journals/Wildlife-Biology on 02 Sep 2022

Terms of Use: https://bioone.org/terms-of-use
 te r 1 996/97 th e re w ere 2 9 m o o se -tra in co llisio n s
(T able 4). T h e p red ic ta b ility o f th e m o d e l w as g o o d
fo r all train s ex c e p t the m o rn in g train , w h ich e x p e ri­
e n c e d an u n e x p e c te d h ig h n u m b e r o f tra in -k ills d u r­
ing w in te r 1996/97 (see T able 4). H o w ev er, 6 o u t o f
th e 14 c o llisio n s b e tw e e n m o o se an d the m o rn in g
train in w in te r 1996/97 o c c u rre d w ith in a 2 k m long
se c tio n o f th e ra ilw a y c a lle d S to rh o lm en . N o n e o f the
c o llisio n s w ith o th e r train s o c c u rre d w ith in th is area.
In p re v io u s y ears, 1990-1 9 9 6 , o n ly 4 o u t o f 37 c o lli­
sio n s b e tw e e n m o o se an d th e m o rn in g train o c c u rre d
a t S to rh o lm en . S to rh o lm e n w as id e n tified as an
e x c e p tio n a l h ig h -risk a re a in 1996/97 d u e to lo g g in g
activ ity (th e risk o f m o o se -tra in ac cid en ts w as 1 0
tim es h ig h e r in 1996/97 th a n in p rev io u s years,
A n d re a sse n , G u n d e rse n & S to ra as 1997, G u n d ersen
e t al. in press). W h e n e x c lu d in g th e six c o llisio n s
o cc u rrin g at S to rh o lm en , th e o b se rv e d n u m b e r o f
co llisio n s ( 8 ) w as w ell p re d ic te d b y th e m o d e l (G =
0 .1 8 , P = 0 .6 7 ), also fo r th e m o rn in g train.
M o rn in g , day, e v e n in g an d n ig h t p a s se n g e r train s
h ad a v e rag e p ro b a b ilitie s o f 0 .0 4 8 , 0 .0 1 0 , 0 .0 6 2 and
0 .0 5 2 , resp e ctiv ely , o f h ittin g a m o o se , b u t th e p ro b ­
ab ility v arie d ac co rd in g to sn o w d ep th , te m p e ra tu re
an d to so m e e x te n t lu n a r p h a se (F ig. 3).
Discussion
In N o rw ay , th e n u m b e r o f m o o se k ille d in m o o se -
train c o llisio n s h a v e in c re a se d fro m c a 5 0 train -k illed
m o o se p e r y e a r in th e 1950s (C h riste n se n 1956) to a
y ea rly a v e ra g e o f 67 6 d u rin g 1 9 9 0 -1 9 9 6 (G u n d e rsen
e t al. 1997). T h is h a s re su lte d in e c o n o m ic co sts d u e
to m a teria l d am ag e a n d lo ss o f in c o m e g a in e d by
se llin g m e a t a n d h u n tin g lic en ses (Jaren e t al. 1991,
G ro o t B ru in d e rin k & H a z e b ro e k 1996, R o m in &
B isso n e tte 1996). T h e to ta l e c o n o m ic co sts w ere e s ti­
m a te d b y Ja re n e t al. (1 9 9 1 ) to b e c a $ 2 ,9 0 0 p e r train -
k ille d m o o se in N orw ay. O n a v e rag e, 87 m o o se are
k ille d alo n g th e R p ro sb a n e n railw a y ea c h year, y ie ld ­
in g a so c io e c o n o m ic lo ss o f c a $ 2 5 0 ,0 0 0 . B e ca u se o f
th e se v ere so c io e c o n o m ic co sts o f g a m e -v e h ic le
ac cid en ts, p ro b le m s a s so c ia te d w ith g am e m a n a g e ­
m e n t, an d th e w e lfa re o f train p e rso n n e l, n ew k n o w l­
ed g e ab o u t g a m e -v e h ic le ac cid en ts is n e c e ssa ry in
o rd e r to d im in ish the p ro b lem .
B y a p p ly in g a lo g istic m o d e l to e stim a te th e p ro b ­
ab ility th a t a c o llisio n w ill ta k e p la ce alo n g a c e rta in
train ro u te, w e h av e in tro d u c e d a n ew ap p ro a ch to
stu d y g a m e -v e h ic le ac cid en ts. U n lik e p rev io u s stu d ­
108
Downloaded From: https://bioone.org/journals/Wildlife-Biology on 02 Sep 2022
Terms of Use: https://bioone.org/terms-of-use
ies w h ich fo c u se d o n th e fac to rs th a t ca u se g am e to
b e c lo se to traffic arte ries, w e fo c u se d o n th e facto rs
th a t ca u se v e h ic le s to c o llid e w ith g am e, b ec au se w e
b e lie v e th a t th e p ro b le m ca n n o t b e efficien tly ta ck le d
fro m th e fo rm e r p o in t o f view . A p p ro a ch in g th e c o n ­
flic t b etw e en th e g a m e ’s n atu ral b e h a v io u r an d m a n ’s
n ee d fo r tra n sp o rta tio n , b y an a ly sin g m a n ’s d ec isio n
o n tra n sp o rt rem e d ies, m ig h t fac ilita te th e in tro d u c ­
tio n o f n ew m itig a tin g ac tio n s. In d eed , it seem s m o re
fe a sib le to try to c h a n g e h u m a n b eh a v io u r ra th e r th an
th e in h e rited b e h a v io u r o f g am e.
A lth o u g h o u r an a ly sis w as m o tiv a te d b y m a n a g e ­
m e n t p u rp o ses, th e p re se n t stu d y rev e als in te restin g
asp ec ts o f m o o se b io lo g y , es p e c ia lly its tem p o ral
ac tiv ity p attern . C o llisio n p ro b a b ilitie s v arie d b e ­
tw ee n th e d iffe re n t train d ep a rtu re s. T rain s ru n n in g at
n ig h t, in th e m o rn in g an d in th e ev e n in g ex p e rie n c e d
a h ig h e r ris k o f c o llis io n th a n th e d ay tra in s.
T em p o ra l activ ity p attern s o f g am e sp ecies h a v e b een
sh o w n to b e an im p o rta n t fa c to r affec tin g g a m e -v e h i­
cle ac cid en ts (P e ek & B eilis 1969, C a rb a u g h e t al.
1975, A lle n & M c C u llo u g h 1976, G o o d w in & W ard
1976, V in cen t, B id ea u , C ib ien & Q u ere 1988, L av -
su n d & S an d eg re n 1991, G lea so n & Je n k s 1993,
W a h lstro m & L ib e rg 1995). T h u s, the lo w risk o f c o l­
lisio n d u rin g th e d ay tim e m ay b e ca u se d by th e re la ­
tiv e ly lo w activ ity o f m o o se d u rin g th e se h o u rs. It
m a y also b e a s su m e d th a t th e o b se rv e d effec t o f th e
lu n a r p h ase on th e c o llisio n p ro b a b ility m a y b e d u e to
h ig h e r m o o se activ ity d u rin g n ig h ts o f fu ll m o o n s
th a n d u rin g n ig h ts o f h a lf o r n ew m o o n s. T h e re is,
h o w ev e r, a n ee d fo r a m o re d e ta ile d d e sc rip tio n o f
m o o se activ ity d u rin g th e lu n a r p h ase to fu lly u n d e r­
sta n d th e c o rre la tio n b e tw e e n th e lu n a r p h ases and
m o o se -tra in co llisio n s.
F u rth e rm o re , w e sh o w ed h o w m o o se -tra in c o lli­
sio n s co rrela te w ith clim a tic se aso n a l fac to rs (A n d e r­
sen et al. 1991, G u n d e rse n et al. in p ress). T h e h ig h
risk o f m o o se -tra in c o llisio n s d u rin g p e rio d s w ith
larg e am o u n ts o f sn o w h as b ee n su g g e sted to be
ca u se d by m ig ra to ry b eh a v io u r, w h erea s th e lo w risk
at h ig h e r a m b ie n t te m p eratu re s h as b ee n s u g g e ste d to
b e re la te d to lo w e r m o o se activ ity d u rin g th e se p e ri­
od s to av o id o v erh e atin g (see A n d ersen et al. 1991,
G u n d e rse n e t al. in p ress).
F ro m a m a n a g e m e n t p e rsp e c tiv e , th e p re se n t stu d y
su g g e sts th a t rem e d ial ac tio n s sh o u ld b e a p p lie d o n ly
o n p a rtic u la r h ig h -ris k ro u te s. R e m e d ia l ac tio n s
co u ld fo r in stan c e b e ch a n g in g ro u tes, i.e. av o id th at
train s p ass th ro u g h h ig h -risk areas at h ig h -risk tim es,
e.g. at n ig h t. In ad d itio n , a lth o u g h th e e stim a te d p o s ­
© W IL D L IF E B IO LO G Y ■ 4 :2 (1998)
 itive slope o f the association betw een train speed and
m oose collisions w as n o t significant, w e believe that
red u ced speed could serve as a m eans to reduce the
n u m b er o f m oose-train collisions in certain areas du r­
ing h igh-risk periods. A lthough the relation betw een
train speed and m oose collision is considerably u n ­
certain statistically, the relation should nevertheless
b e co nsidered carefully in the future prior to the
in tro d uction o f faster trains. H igh-speed trains m ay
increase the n um ber o f m oose-train collisions co n ­
siderably in the future, and in p articular if careful
atten tion to the tim e schedule w hen high-speed trains
pass high-risk areas is not paid. Finally, techniques
u sed to frighten m oose aw ay before the train passes
th ro u gh p articular areas (e.g. pilot cars driven in front
o f trains) m ay be restricted to high-risk periods, e.g.
n ig h t trains during periods w ith large am ounts o f
snow. L im ited use o f m itigative efforts w ill also
reduce the costs connected w ith such actions, and
thus, it m ay be easier to find the political w ill to
introduce such m ethods.
T h e predictability o f our m odel w as satisfactory,
except for the m orning train. T his m ay have been due
to the exceptionally high n um ber o f m oose-train
accidents at S torholm en. S torholm en is an area close
to the railw ay w hich is used by brow sing m oose dur­
ing w inter (S. H anestad, pers. com m .). D uring sum ­
m er 1996 this area w as clear-cut fo r cultivation.
D uring w inter 1996/97 m oose w ere observed to
m ove from S torholm en to the nearest brow sing area
alo n g th e railw ay, w h ich resu lted in a 1 0 -fold
increase in the n um ber o f m oose-train accidents
w ithin a few kilom etres, com pared to w hat w ould be
ex p ected from experiences in previous years (A ndre­
assen et al. 1997, G undersen et al. in press). We sus­
p ect that the tim e w hen the m orning train passed this
area coincided w ith the tim e w hen m oose w ere
b row sing along the railw ay. O bviously, predictable
m odels should therefore include both tem poral and
spatial aspects.
Acknowledgements - we thank The Norwegian National
Rail Administration for financial support to this study. We
would also like to thank H. Brenden, H. Haave, S.
Hanestad, H. Haug, L. Hesthagen, O. Kristiansen, K.
Nicolaysen, S. Sletten, T. Storaas, L. Saslthun, G. Vestby, T.
0fstaas for data, comments and ideas, and S.-B. Nielsen,
R. A. Ims and two anonymous referees for comments on
the manuscript.
© W ILDLIFE BIOLOGY • 4:2 (1998)
Downloaded From: https://bioone.org/journals/Wildlife-Biology on 02 Sep 2022
Terms of Use: https://bioone.org/terms-of-use
References
Allen, R.E. & McCullough, D.R. 1976: Deer-car accidents
in southern Michigan. - Journal of Wildlife Management
40: 317-325.
Andersen, R., Wiseth, B., Pedersen, P.H. & Jaren, V. 1991:
Moose-train collisions: Effects of environmental condi­
tions. -A lces 27: 79-84.
Andreassen, H.P., Gundersen, H. & Storaas, T. 1997: Vilt-
trafikk i 0sterdalen. Del 1: Tiltak for &.begrense elg naer
jembanelinjen. (In Norwegian with English summary:
Game-vehicles in 0sterdalen. Part 1: Remedial actions
to reduce the presence of moose along the railway). -
Hpgskolen i Hedmark, Rapport nr. 5, 40 pp.
Bashore, T.L., Tzilkowski, W.M. & Beilis, E.D. 1985:
Analysis of deer-vehicle collision sites in Pennsylvania,
USA. - Journal of Wildlife Management 49: 770-774.
Burnham, K.P. & Anderson, D.R. 1993: Databased selec­
tion of an appropriate biological model: the key to mod­
em data analysis. - In: McCullough, D.R. & Barrett,
R.H. (Eds.); Wildlife 2001: populations. Elsevier Ap­
plied Science, London, pp. 16-30.
Carbaugh, B., Vaughan, J.P., Beilis, E.D. & Graves, H.B.
1975: Distribution and activity of white-tailed deer along
an interstate highway. - Journal of Wildlife Management
39:570-581.
Christensen, H. 1956: Innstilling fra komiteen til k utrede
spprsmMet om hva som kan gjpres for k hindre at elg og
rein blir pikjprt og skadet eller drept ved jembanetra-
fikken. - Avskrift, Oslo 1956, 22 pp. (In Norwegian).
Feldhamer, G.A., Gates, J.E., Harman, D.M., Loranger,
A.J. & Dixon, K.R. 1986: Effects of interstate highway
fencing on white-tailed deer activity. - Journal of Wild­
life Management 50: 497-503.
Gleason, J.S. & Jenks, J.A. 1993: Factors influencing deer-
vehicle mortality in east central South Dakota. - Prairie
Naturalist 25: 281-288.
Goodwin, G.A. & Ward, A.L. 1976: Mule deer mortality on
interstate 80 in Wyoming: causes, patterns, and recom­
mendations. - USDA Forest service research note, Rocky
Mountain Forest and Range Experiment Station, 4 pp.
Groot Bruinderink, G.W.T.A. & Hazebroek, E. 1996:
Ungulate traffic collisions in Europe. - Conservation
Biology 10: 1059-1067.
Gundersen, H., Andreassen, H.P., Haave, H.M. & Storaas,
T. 1997: Vilt-trafikk i 0sterdalen. Del 2: Tiltak ved
pikjprsler og nestenpikjprsler av elg. (In Norwegian
with English summary: Game-vehicles in 0sterdalen.
Part 2: Remedial actions to reduce moose train colli­
sions). - Hpgskolen i Hedmark, Rapport nr. 8, 55 pp.
Gundersen, H., Andreassen, H.P. & Storaas, T. in press:
Spatial and temporal correlates to Norwegian moose-
train collisions. - Alces.
Jaren, V., Andersen, R., Ulleberg, M., Pedersen, P.H. &
Wiseth, B. 1991: Moose-train collisions: The effects of
109
 vegetation removal with a cost-benefit analysis. - Alces
27: 93-99.
Lavsund, S. & Sandegren, F. 1991: Moose-vehicle rela­
tions in Sweden: a review. - Alces 27: 118-126.
Lutz, W. 1994: Trial results of the use of a “Duftzaun”
(scent fence) to prevent game losses due to traffic acci­
dents. - Zeitschrift fur Jagdwissenschaft 40: 91-108.
Peek, F.W. & Beilis, E.D. 1969: Deer movements and
behavior along an interstate highway. - Highway Re­
search News 36: 36-42.
Petersson, C.J. & Danell, O. 1992: Simulated production
losses in reindeer herds caused by accidental death of
animals. - Rangifer 12: 143-150.
Reeve, A.F. & Anderson, S.H. 1993: Ineffectiveness of
Swareflex reflectors at reducing deer-vehicle collisions.
- Wildlife Society Bulletin 21: 127-132.
Romin, L.A. & Bissonette, J.A. 1996: Deer-vehicle colli­
sions: status of state monitoring activities and mitigation
efforts. - Wildlife Society Bulletin 24: 276-283.
Romin, L.A. & Dalton, L.B. 1992: Lack of response by
110
Downloaded From: https://bioone.org/journals/Wildlife-Biology on 02 Sep 2022
Terms of Use: https://bioone.org/terms-of-use
mule deer to wildlife warning whistles. - Wildlife Socie­
ty Bulletin 20: 382-384.
Schober, F. & Sommer, F. 1984: Study of acoustic game
warning devices for vehicles. - Zeitschrift fur Jagdwis­
senschaft 30: 164-176.
Vatshelle, A. 1995: Arbeidsmiljp og helse hos norske loko-
motivfprere. Seksjon for arbeidsmiljp. - Universitetet i
Bergen, Rapport 1995, pp. 122. (In Norwegian).
Vincent, J.P., Bideau, E., Cibien, C. & Quere, J.P. 1988:
Traffic deaths in roe deer Capreolus-capreolus example
of woodland area in the Paris France basin. - Zeitschrift
fur Jagdwissenschaft 34: 63-68.
Wahlstrom, L.K. & Liberg, O. 1995: Patterns of dispersal
and seasonal migration in roe deer (Capreolus capreolus)
- Journal of Zoology 235: 455-467.
Waring, G.H., Griffis, J.L. & Vaughn, M.E. 1991: W hite­
tailed deer roadside behavior, wildlife warning reflec­
tors, and highway mortality. - Applied Animal Behaviour
Science 29: 215-223.
© W IL D L IF E B IO LO G Y ■ 4 :2 (1998)