Corresponding author: Douglas William Cirino (
Academic editor: C-R Papp
While road network expansion connects human settlements between themselves, it also leads to deforestation and land use changes, reducing the connectivity between natural habitat patches, and increasing roadkill risk. More than 30% of registered mammal roadkills in Brazil are concentrated in four species:
Cirino DW, Lupinetti-Cunha A, Freitas CH, de Freitas SR (2022) Do the roadkills of different mammal species respond the same way to habitat and matrix? In: Santos S, Grilo C, Shilling F, Bhardwaj M, Papp CR (Eds) Linear Infrastructure Networks with Ecological Solutions. Nature Conservation 47: 65–85.
Road ecology is a research area that aims to understand the impact of highways and railways on natural ecosystems, economics, and society. Many studies on this subject focus on one of the most conspicuous effect of roads: wildlife mortality by vehicle collisions (
Changes in landscape composition and its structure are some of the main factors leading to biodiversity loss (
The majority of studies on road mortality focus on a small region, studying a road or a portion of it. Those studies are important to understand the local impacts of roads, and to search for patterns on specific landscape configuration. When we search for similar studies in Brazil, it is notable that some species are constantly found on roadkill registers (
The crab-eating-fox (
The six-banded-armadillo (
Both anteater species (
Most road impacts mitigation measures focus on general recommendations, such as implementation of underpasses or fencing in roadkill hotspots, which usually comes in association with native or riparian vegetation, assuming that most animals would use those areas to move and cross the road. However, we cannot assume that all species have the same habitat requirements and patterns of space usage, since it is known that the rate of underpasses usage differs among species (
Understanding the landscape patterns linked to road mortality of those species can provide guidance for protection and conservation efforts aiming to mitigate the road impacts on wildlife. Together, these four species presented here represent between 34,7% and 38,8% of the total roadkills of medium-large sized mammals in Brazil (
We collected a sample of georeferenced roadkill data from two main sources: (1) monitoring studies across the country; and (2) the “Banco de Dados Brasileiro de Atropelamento de Fauna Selvagem” (
For land cover and land use, we utilized the serial time data from MapBiomas (
For each species, we considered a different influence buffer radius starting from the place of the roadkill, since each one has different home ranges, body sizes, and habits requirements. We estimated the mean home range for
If the area of a circle is given by:
where r is the radius of the circle, so the double of a radius of a circle of a given area is:
We used this radius size because the roadkill point may have occurred on the center of the home range, or on its border (Fig.
Scheme exemplifying the radius of roadkill influence chosen. For a given roadkill point using the simple radius of home range (r), we might exclude some of the landscape characteristics if the roadkill occurred in the border of the home range. Including the possible home ranges (approximated to a circular shape), and doubling the radius (φ), we ensure that all landscape composition associated with the roadkill occurrence is incorporated within the analysis.
For each presence or absence of roadkills we calculated the proportion of land use and land cover inside the buffer based on MapBiomas land cover map for the corresponding year of the roadkill. The classes of land use and land cover considered in the analysis were: (1) forest; (2) savanna; (3) natural open areas; (4) forestry; (5) agriculture; (6) pasture; (7) farming; and (8) water. Farming represents the sum of agriculture and pasture in addition to mosaics or rotation of both classes in the same area. We conduct all landscape analysis and data extraction on ArcGIS v10.3 environment.
To estimate the relative chance of roadkill of each species we constructed binomial generalized linear models (
All models were ranked by Akaike Information Criteria (
We collected a total of 2698 georeferenced roadkill records across the country (
Roadkill samples distribution for the species studied.
For
Best model selected by
Model selected by species according to Akaike criteria. dAICc represents the
Species | Model |
|
dAICc | df | Weight | Evidence |
---|---|---|---|---|---|---|
|
Pasture + Agriculture | 3545.9 | 0.0 | 3 | 0.3080 | 1.00 |
|
Farming + Forestry | 1612.4 | 0.0 | 3 | 0.3772 | 1.00 |
Pasture + Agriculture | 1612.8 | 0.3 | 3 | 0.3173 | 1.19 | |
|
Forest + Pasture + Forest:Pasture | 1170.6 | 0.0 | 4 | 0.2057 | 1.00 |
Forest + Savanna + Forest:Savanna | 1171.5 | 0.9 | 4 | 0.1309 | 1.57 | |
|
Savanna + Agriculture + Savanna:Agriculture | 1111.4 | 0.0 | 4 | 0.8495 | 1.00 |
Besides giving information on the studied animal mortality, roadkill records are also useful for assessing a species occurrence. We found registers of
Roadkill records of
As a generalist species,
Best model selected by
This species inhabits a vast number of natural formations, but also human-modified landscapes, such as sugar cane plantations (
Best model selected by
Depending on the amount of matrix in the landscape, the direction of the effect of habitat on roadkill risk changes. In other words, when there are small quantities of forest, the effect of pasture is positive to predict the roadkill, while when there is an increased proportion of forest in the region the effect changes, and the roadkill risk decreases with the increase of pasture areas. This could be related to the species’ habits: when the landscape is mostly composed of pastureland, the animals need to move more in search of shaded shelter. This movement decreases in frequency when there are some forested areas in the landscape, allowing the individuals to rest, and therefore, decreasing the chance of encountering a road and consequently being roadkilled. This shows the importance of maintaining habitat patches in the landscape, such as riparian forests or even native vegetation fragments inside private rural property, as established by the Brazilian Forest Code (
Best model selected by
This habitat dependence reflects on the best model selected to predict the collared-anteater roadkill risk: the presence of savanna formations modulates the effect of agriculture. When a landscape has no natural formations cover, the effect of agriculture is negative, since the species probably do not occur in the area; and with the increment of habitat areas, the roadkill risk increases rapidly, reaching our model’s peak when we have at least 40% of savanna and 50% of agriculture.
On the other hand, the roadkill risk when the landscape is entirely comprised of savanna, without agriculture, is very low, and it increases very fast when there is an increase of agricultural coverage. As it is a forest dependent animal, it was unexpected that its roadkill response was better suited to savanna than to dense forests, but that can be explained by this animal’s movement pattern. In areas with continuous dense forests the locomotion of individuals occurs mainly through canopies, but in areas with low density of trees, as open areas, savannas and monocultures, it moves by ground. It can also move more often in search for sheltering trees, therefore increasing the chance of being roadkilled.
It is already known that many factors affect the roadkill risk of a species, such as species density and movement patterns (
For habitat dependent and more sensitive species like anteaters, the effect of the matrix on the roadkill risk depends on habitat availability in the landscape. It changes the strength and direction of the effect according to the proportion of natural areas in the region. As for generalist species, the quantity of human-modified land uses increases the roadkill risk regardless of the habitat availability or natural formations. It also indicates the occurrence of these species in those anthropic areas.
Therefore, the habitat and matrix composition impacts the studied species differently, depending on their demand and habitat dependence. Each species showed different prediction factors regarding their roadkill risk. Overall, all four target species had some dependency on the habitat, but two of them (
The habitat dependent species have more complex models predicting their roadkill risk, including an interaction component between habitat and matrix. It shows the importance of maintaining the natural coverage of rural properties that, as indicated by Brazilian Forest Code, can potentially decrease the risk of roadkill, connect habitat areas, and increase habitat quality. Given that, areas with vast cover of monoculture and pasture can both decrease the natural populations’ size and increase the movement of individuals that can be roadkilled while they are searching for best habitats on the landscape. Since we have shown that not only riparian corridors or continuous habitats are associated with roadkill, but also areas out of protected areas we suggest that more studies investigating the effect of movement in roadkill should be performed. We also highlight the need to consider the landscape as a whole while assessing species protection.
We thank FAPESP for the financial support (Process n° 16/12785-0) through DWC research grant, NERF/UFRGS (“Núcleo de Ecologia de Ferrovias e Rodovias” – Universidade Federal do Rio Grande do Sul) and others involved in field data collection, such as Andreas Kindel, Arnaud Desbiez, Janaina Casella and Sidnei Dornelles. We also thank Fernanda Delborgo Abra and ViaFauna for providing the illustrations in this article and contributions to data interpretation; and Guillermo Flórez for statistical contributions. We also thank the MapBiomas initiative, for the land use and land cover mapping and a reward attributed to a preliminary version of this study (CIRINO, 2018).
Correlation plot and R script for building and selecting best models
R script (text file)
Plot of correlations between predictor variables. The script used for reading variables, building statistical models and selecting the best model by Akaike Information Criteria.