A forecasting map of avian roadkill-risk in Europe: A tool to identify potential hotspots
Introduction
The urbanization process is characterized by the concentration and expansion of human settlements. Consequently, the development of interconnectivity tissue (the road network) guarantees the connectivity among urban centers (Marzluff et al., 2008). There is currently more than 36,000,000 km of roads on the global scale, following the estimates of the World Development Index of World Bank, World Road Statistics of the International Road Federation and World Factbook of the Central Intelligence Agency on the World. Road networks are a prominent feature of the capacity of humans for penetrating the last remaining natural areas and landscapes, causing severe habitat loss, fragmentation, and many other well-known disturbances (Bennett, 1991; Forman et al., 2003). Specifically, roads are commonly associated with a significant decline of biodiversity (Benítez-López et al., 2010; Kociolek and Clevenger, 2009; Loss et al., 2014a; van der Ree et al., 2011). The main adverse ecological effects of roads on vertebrate wildlife are related to habitat loss associated to the construction and spread of road networks (Davenport and Davenport, 2006; Kociolek et al., 2011); to traffic disturbance, as noise pollution (Kaseloo and Tyson, 2005; McClure et al., 2013; Zhou et al., 2020), light pollution (De Molenaar et al., 2006), and even fostering changes in species ranges, habitat use, local abundance and distribution (Cooke et al., 2020; Delgado et al., 2008). Roads can also represent ecological barriers for non-flying terrestrial animals, reducing the species' range or increasing the isolation of some populations (van der Ree et al., 2011). However, perhaps the most immediate and critical effect of roads is the direct mortality due to collisions with the traffic (Collinson et al., 2014; Reijnen et al., 2008), and at the same time the road barrier effect (Jaeger et al., 2005). In addition, roads and traffic may be interacting with other drivers, such as climate change, leading to the decline of bird diversity in Europe (Huntley et al., 2008; Morelli et al., 2020).
In Europe, several bird species use roads and roadsides for different types of activities, from hunting (e.g., carnivorous species as shrikes or raptors) to breeding or nesting activities (Helldin and Seiler, 2003; Morelli et al., 2015, Morelli et al., 2014). Scavenger species such as vultures and kites, or even strigids (nocturnal raptors), are likely to become road casualties when they are attracted to corpses on the road (Vidal-Vallés, 2018). For decades, birds were considered less affected by roadkill than mammals because of their ability to fly. However, there is increasing evidence about birds as road casualties indicating that vehicle collision is a significant cause of death (Kociolek et al., 2015). Some studies suggest that the number of birds killed by vehicle collisions in Europe is vast (up to 5–10% of bird mortality caused by roads and traffic for W Palearctic; Møller et al., 2011). In Germany, there is an estimated average of more than 9 million birds killed on the roads every year (Fuellhaas et al., 1989). Variable estimations of annual averages come from Spain (12 million) (PMVC, 2003), Bulgaria (7 million) (Nankinov and Todorov, 1983), the Netherlands (2 million) (Schrijver, 1993), England (27 million), Denmark (1.1 million) and Sweden (8.5 million) (Erritzøe et al., 2003). We do not know if there is an increasing or decreasing trend in road–caused mortality for European birds, since we lack long-term monitoring. However, these numbers are likely to increase if the road network continues to develop across the continent.
Numerous factors would explain bird susceptibility to become roadkill. Aside from road and traffic design variables (Clevenger et al., 2002; Husby, 2016), research has focused on bird responses, and functional traits have been analyzed as reasons for susceptibility to road mortality (Santos et al., 2015). Firstly, road mortality depends on bird escape and vehicle avoidance behavior (Husby and Husby, 2014). Though, some species seem to be more susceptible than others to the roadkill, showing a frequency as road casualties higher than expected (Erritzøe et al., 2003; Santos et al., 2015). It is assumed that the relative frequency of traffic casualties is associated with the eco-functional traits of the species. For example, a higher incidence of car-collisions is expected for birds that are abundant near to roads (Santos et al., 2016), for frequent roadside users (Morelli et al., 2014), for birds more prone to cross the roads (Johnson et al., 2017), but also merely for species that occupy landscapes where the road networks are more spread (e.g., lowland bird populations, agricultural, urban and periurban areas, etc.). Interspecific differences could also be related to different sensitivity to collision risk (Husby and Husby, 2014; Møller et al., 2011).
With this in mind, we can hypothesize that avian communities overall more prone to roadkill than others could characterize some geographical areas in Europe. For example, a bird community composed of five species often recorded in road casualties should have a cumulative roadkill risk higher than that of another community consisting of five species never recorded as road casualties. Then, combining areas with bird communities composed by species with a high frequency of roadkills and simultaneously characterized by a dense road network, especially with high-speed roads or highways, we can expect an increased overall risk of bird-car collisions. To predict zones of potentially elevated collision risk would help to identify areas posing a consistent threat for the conservation of bird diversity (Gunson et al., 2011).
The accurate detection of the potential hotspot of wildlife-vehicle collisions should be the first step to elaborate regional and local strategies aiming to reduce the problem and to identify areas (or in some cases, species) with higher mitigation needs. Different mitigation strategies and measures can be developed and applied, depending on the area and road characteristics, but to mitigate the ecological impacts of roads effectively, continued actions and monitoring are necessary (Karlson et al., 2014; Trombulak and Frissell, 2000). The use of forecasting models, which can combine different layers of information to predict an event, primarily used in ecology (Guisan and Thuiller, 2005), constitutes a valid tool also for transportation mitigation projects as well as urban and road planning (Gunson et al., 2011; Santos et al., 2013; Soldatini et al., 2011). A good example is constituted by the use of ecological modelling to create risk maps (Yañez-Arenas et al., 2014).
In this study, we combined data on the spatial distribution of bird species characterized by high incidence in roadkill events in Europe, with data on road density across the continent, by using a fixed spatial unit. We compared the spatial congruence/mismatch between the road network, the bird species richness, by considering the different types of environments. Then, we elaborated, on a large spatial scale, a forecasting map of the potential risk of collisions between cars and birds, across Europe. We tried to identify potential hotspots where the risk of avian roadkill is higher, to be considered as areas where it should be necessary to focus on mitigation strategies.
Section snippets
The cumulative risk of roadkill of bird species assemblages in Europe
We estimated the relative frequency of roadkill for each European breeding bird species from bibliographic bird mortality records along European roads at different latitudes and including data from 9 countries (Table S1). We searched for studies presenting net quantitative data, i.e., the number of bird casualties per species. We excluded from analysis documents, including birds as a general category and those identifying only one bird species. We also excluded sources on bird collisions with
Results
We calculated the standardized roadkill frequency for 209 avian species documented in 37 publications and reports dealing with bird mortality records along roads of up to 9 European countries (Table S2). The top-ten of bird species more often involved in roadkill incidents was: Turdus merula, Passer domesticus, Erithacus rubecula, Tyto alba, Pica pica, Athene noctua, Fringilla coelebs, Hirundo rustica, Strix aluco (Table S2). The bird species characterized by the higher roadkill occurrence are
Discussion
Our findings suggest that the road birds' mortality related to collisions is well widespread across the phylogenetic “landscape” of European breeding birds. Fig. 1 highlights some of the species more often recorded as roadkill in Europe within the phylogenetic context of the continental bird assemblage recorded as road casualties. The figure shows that species intensively impacted by traffic mortality are not clustered in concrete phylogenetic lineages, but rather in a different and distant
CRediT authorship contribution statement
FM: Conceptualization, Methodology, Data curation, Visualization, Writing - Original draft preparation. YB: Visualization, Writing - Original draft preparation. JDD: Validation, Writing - Reviewing and Editing.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
F.M. and Y.B. were financially supported by the Czech Science Foundation GAČR (Project Number 18-16738S).
References (89)
- et al.
Bird collisions at wind turbines in a mountainous area related to bird movement intensities measured by radar
Biol. Conserv.
(2018) - et al.
The impacts of roads and other infrastructure on mammal and bird populations: a meta-analysis
Biol. Conserv.
(2010) - et al.
Vertebrate road-kill patterns in Mediterranean habitats: who, when and where
Biol. Conserv.
(2015) - et al.
The spatial distribution of animal casualties within a road corridor: implications for roadkill monitoring in the southern Iberian rangelands
Transp. Res. Part D Transp. Environ.
(2019) - et al.
Predicting when animal populations are at risk from roads: an interactive model of road avoidance behavior
Ecol. Model.
(2005) - et al.
Road ecology in environmental impact assessment
Environ. Impact Assess. Rev.
(2014) - et al.
An adaptive inverse-distance weighting spatial interpolation technique
Comput. Geosci.
(2008) - et al.
Landscape heterogeneity metrics as indicators of bird diversity: determining the optimal spatial scales in different landscapes
Ecol. Indic.
(2013) - et al.
Can roads, railways and related structures have positive effects on birds? A review
Transp. Res. Part D Transp. Environ.
(2014) - et al.
Testing bird response to roads on a rural environment: a case study from central Italy
Acta Oecol.
(2015)