Mateusz Iwiński*, Adam Zydroń*, Dariusz Kayzer*, Janusz Dąbrowski**
*Poznań University of Life Sciences, Poland; **Koszalin University of Technology, Poland
corresponding author’s e-mail:
Collisions with forest animals constitute a considerable percentage of all events recorded for railway lines. Due to the development of infrastructure and higher technical parameters of tracks, accidents involving forest animals cause considerable economic losses and lead to significant delays in railway service. Appropriate management of areas neigh-bouring transport networks as well as application of both active and passive methods preventing animals from suddenly entering the road may limit a further increase in the number of such events and reducing negative environmental effects due to animal mortality. The paper presents an analysis whether and if so, which elements of landscape affect the fre-quency of collisions involving forest animals and whether by properly managing and securing these areas there is a pos-sibility to prevent animal intrusion on communication networks. Using spatial analyses in the GIS environment (spatial nodes, topological relationships) as well as statistical analyses (contingency tables) it was indicated which pairs of fac-tors determine an increase in the frequency of events involving roe deer, red deer and wild boars. It was proposed to ap-ply the 0.1 system to determine the incidence of observed collisions in the buffer zone characterised by a given manage-ment function (forested areas, surface waters, arable land, urbanised areas). When analysing the database on collisions of the Polish State Railways providing information on 561 collisions recorded in the years 2007-2016 as well as cartograph-ic materials publicised as open-layers it was shown that an increase in the frequency of collisions involving animals treated jointly, with no division into individual species, is influenced by the presence of developed areas in the buffer zones adjacent to railway tracks. In turn, an analysis of animal species participating in these events shows a certain simi-larity in the case of deer (roe deer and red deer). For these species one identical pair of mutually dependent variables was observed, i.e. surface waters and forested areas. Additionally, for collisions with roe deer another pair of dependent at-tributes was found, i.e. forested areas and agriculturally utilised areas. For wild boars an increase was observed in the frequency of events in the buffer zones, where surface waters were found together with forests and connected with ur-banised areas. In the case of analyses concerning collisions with roe deer an interdependence was observed between two pairs of variables – forested areas and surface waters, and forested areas and agriculturally utilised areas. Red deer ac-counted for the smallest percentage of all collisions and for them one pair of mutually dependent variables involved buffers comprising surface waters and forested areas. Results of these studies confirmed the need to conduct analyses for each species separately. It was observed that the character of sites, where collisions involving individual even-toed ungu-late species were recorded, depends on the specific nature of a given species. Moreover, these studies showed that by appropriate management of areas adjacent to railway infrastructure provided by spatial planning, especially through the use of line protections we may reduce the frequency of collisions. It was observed that the proposed analytical methods (statistical and spatial) facilitate realisation of proposed research aims and can serve as a tool in planning the location of devices against the incursion of forest animals, which would reduce the negative environmental effects caused by animal mortality on railroad tracks.
spatial planning, ecological barriers, landscape fragmentation, railways, infrastructure and the environment