In northern Sweden, railways are often fenced to prevent reindeer (semi-domestic but free-ranging) from getting train-killed. Accordingly, the new Bothnia Line railway along the north Swedish coast was fenced in connection with its completion in 2010. This 190 km long railway runs parallel to the already fenced motorway E4 along the coast, partially within meters and partially separated by up to ca 3 km. The proximity between these two main traffic arteries has produced a complex fencing setup, with single-sided fencing of railway sections close to the motorway (within 1 km), and double-sided fencing on more distant sections. We studied the impact of the railway fencing on ungulate movements and collision rate, particularly in relation to the single-sided fencing, the fence openings, and the proximity to the fenced motorway. We monitored moose and deer movements by snowtracking a 47.5 km section of the railway in the winters 2013-2014 and 2014-2015, and we compiled data on wildlife collissions on the same section from the years 2010-2015.The studied railway has 24 bridges for waters and local roads, and two longer tunnels constructed for topographic reasons. In addition, 17 fence openings of various length (50-450 m) are left to allow wildlife movements over the railway. The railway stopped on average ca 76 % of the animals that, according to our interpretation, were intending to pass it. This barrier effect however varied between sections, from 100 % along fenced sections (single- or double-sided), 33 % in fence openings, and 0 % in bridges and tunnels. Only few of the fence openings were used by wildlife, suggesting that also the unfenced railway has a barrier effect, or alternatively that most openings were misplaced for wildlife. Along single-sided fencing several animals were stopped on the unfenced side, i.e. on or very near the track, potentially increasing the collision risk. Observed collisions (n=27) derived primarily from two shorter sections, single-sided fenced, and with two and three fence openings, respectively. We conclude that the fence and the Bothnia Line, jointly and individually, contributes to a demographic, and with time potentially also a genetic, isolation of moose and deer. However, it is difficult to separate the barrier effect of the railway from that caused by the motorway, and likely these work in concert. The results indicate that the single-sided fencing along the already fenced motorway is not effective in preventing wildlife collision. We therefore suggest installation of jump-out ramps, and that some particularly problematic single-sided sections are completed with double-sided fencing. We further suggest that the fence openings are adjusted, for example with active warning systems and prevention from entering into the fenced track. We suggest that the match between fence openings in the railway and potential wildlife passages along the motorway is assessed and coordinated, to ensure that no animals get caught in the habitat strip between the railway and the motorway.
Barrier effects, Collision risk, Moose, Railway fencing, Reindeer, Ungulates