Relative effects of road risk, habitat suitability, and connectivity on wildlife roadkills: the case of tawny owls (Strix aluco)

Assessing Contamination Profiles in Livers from Road-Killed Owls

Raptors are recognized as valuable sentinel species for monitoring environmental contaminants owing to their foraging behavior across terrestrial and aquatic food webs and their high trophic position. The present study monitored environmental contaminants in livers from road-killed owls to evaluate differences in the exposure patterns due to factors such as species, age, and sex of individuals. Carcasses of road-killed individuals of eagle owl (Bubo bubo), long-eared owl (Asio otus), little owl (Athene noctua), tawny owl (Strix aluco), and barn owl (Tyto alba) were collected in Alentejo (Portugal). Eighty-one organic contaminants were analyzed, including organochlorine pesticides (OCPs), per- and polyfluoroalkyl substances (PFAS), polychlorinated biphenyls (PCBs), polycyclic aromatic hydrocarbons (PAHs), pharmaceuticals, in-use pesticides, and organophosphate esters (OPEs). Overall, 21 contaminants were detected. In all species ∑OCPs were prevalent at concentrations from 3.24 to 4480 ng/g wet weight, followed by perfluorooctane sulfonic acid (PFOS), the only PFASs detected (from 2.88 to 848 ng/g wet wt) and ∑PCBs (1.98-2010 ng/g wet wt); ∑PAHs were ubiquitous but detected at the lowest concentrations (7.35-123 ng/g wet wt). Differences among species were observed according to principal component analysis. Eagle owl and long-eared owl presented the highest levels of ∑OCPs, ∑PCBs, and PFOS, consistent with its higher trophic position, while ∑PAHs prevailed in tawny owl, barn owl, and little owl, related to their frequent use of urban areas for nesting and roadsides for hunting. Adults presented higher concentrations of ∑OCPs and ∑PCBs than juveniles, while no differences were observed for PFOS and ∑PAHs. Pharmaceuticals, in-use pesticides, and OPEs were not detected. Overall, the present study shows specific contamination patterns in five species with similar diet but with differences in habitat preferences. Environ Toxicol Chem 2024;00:1-12. © 2023 The Authors. Environmental Toxicology and Chemistry published by Wiley Periodicals LLC on behalf of SETAC.

Aliens on the Road: Surveying Wildlife Roadkill to Assess the Risk of Biological Invasion

Monitoring the presence and distribution of alien species is pivotal to assessing the risk of biological invasion. In our study, we carried out a worldwide review of roadkill data to investigate geographical patterns of biological invasions. We hypothesise that roadkill data from published literature can turn out to be a valuable resource for researchers and wildlife managers, especially when more focused surveys cannot be performed. We retrieved a total of 2314 works published until January 2022. Among those, only 41 (including our original data) fitted our requirements (i.e., including a total list of roadkilled terrestrial vertebrates, with a number of affected individuals for each species) and were included in our analysis. All roadkilled species from retrieved studies were classified as native or introduced (domestic, paleo-introduced, or recently released). We found that a higher number of introduced species would be recorded among roadkill in Mediterranean and Temperate areas with respect to Tropical and Desert biomes. This is definitely in line with the current knowledge on alien species distribution at the global scale, thus confirming that roadkill datasets can be used beyond the study of road impacts, such as for an assessment of different levels of biological invasions among different countries.

The importance of population contextual data for large-scale biomonitoring using an apex predator: The Tawny Owl (Strix aluco)

Top predators are often used as sentinel species in contaminant monitoring due to their exposure and vulnerability to persistent, bioaccumulative and, in some cases, biomagnificable contaminants. Some of their ecological traits can vary in space and time, and are known to influence the contamination levels and therefore information on ecological traits should be used as contextual data for correct interpretation of large-scale contaminant spatial patterns. These traits can explain spatiotemporal variation in contaminant exposure (traits such as diet and dispersal distances) or contaminant impacts (traits such as population trend and clutch size). The aim of our research was to review the spatial variation in selected contextual parameters in the Tawny Owl (Strix aluco), a species identified by the COST Action European Raptor Biomonitoring Facility as one of the most suitable candidates for pan-European biomonitoring. A considerable variation in availability of published and unpublished contextual data across Europe was found, with diet being the most extensively studied trait. We demonstrate that the Tawny Owl is a suitable biomonitor at local scale but also that taking spatial variation of other contextual data (e.g. diet) into account is necessary. We found spatial gaps in knowledge about the species ecology and biology in Southern Europe, along with gaps in certain population parameters (e.g. population trends) in several countries. Based on our findings, we proposed a minimal recommended scheme for monitoring of population contextual data as one of the first steps towards a pan-European monitoring scheme using the Tawny Owl.

A spatiotemporal risk prediction of wildlife-vehicle collisions using machine learning for dynamic warnings

INTRODUCTION

The technology in the automotive industry is becoming increasingly safer in the age of automated driving, but the number of accidents is still high, especially in wildlife-vehicle collisions (WVCs). To better avoid these accidents, patterns involved in these accidents must be detected.

METHOD

This paper presents a spatiotemporal risk prediction of WVCs, including various road and environmental characteristics. A process of data preparation using GIS automated by Python scripts was developed to enable a spatiotemporal link of diverse features for the subsequent predictive data analysis. Different machine learning (ML) approaches were applied- random forest (RF), feedforward neural networks (FNN), and support vector machine classifier (SVM) - including automated ML to predict the risk of WVCs. Therefore, a dataset of approximately 731,000 accidents reported to the police in Bavaria over a period of 10 years (2010-2019) was used. In addition, non-accidents were randomly generated in Python over time and space for the supervised ML processes. As the actual risk probability for WVCs and non-WVCs is not entirely known, the impact of different training ratios between accidents and non-accidents was tested on the risk prediction quality (RPQ) (25%, 50%, 75%, 90% WVCs) of the double-weighted sensitivity and single-weighted specificity rate.

RESULTS

The test yielded high mean values of RPQ as an indicator for a suitable WVC prediction. Both RF (86.6%) and FNN (86.7%) were identified as suitable choices for WVC risk prediction in terms of RPQ. The SVM yielded a lower prediction quality, even though acceptable results could be achieved within a shorter runtime.

PRACTICAL APPLICATIONS

Spatial transferability was verified since the algorithm was trained on the dataset of Bavaria (excluding Upper Bavaria) and successfully tested in Upper Bavaria. WVC forecasts were also proven through training with datasets from 2010-2017 and in prediction for 2018 and 2019.

The Influence of Landscape Structure on Wildlife–Vehicle Collisions: Geostatistical Analysis on Hot Spot and Habitat Proximity Relations

Vehicle collisions with animals pose serious issues in countries with well-developed highway networks. Both expanding wildlife populations and the development of urbanised areas reduce the potential contact distance between wildlife species and vehicles. Many recent studies have been conducted to better understand the factors that influence wildlife–vehicle collisions (WVCs) and provide mitigation methods. Most of these studies examined road density, traffic volume, seasonal fluctuations, etc. However, in analysing the distribution of WVC, few studies have considered a spatial and significant distance geostatistical analysis approach that includes how different land-use categories are associated with the distance to WVCs. Our study investigated the spatial distribution of agricultural land, meadows and pastures, forests, built-up areas, rivers, lakes, and ponds, to highlight the most dangerous sections of roadways where WVCs occur. We examined six potential ‘hot spot’ distances (5–10–25–50–100–200 m) to evaluate the role different landscape elements play in the occurrence of WVC. The near analysis tool showed that a distance of 10–25 m to different landscape elements provided the most sensitive results. Hot spots associated with agricultural land, forests, as well as meadows and pastures, peaked on roadways in close proximity (10 m), while hot spots associated with built-up areas, rivers, lakes, and ponds peaked on roadways farther (200 m) from these land-use types. We found that the order of habitat importance in WVC hot spots was agricultural land < forests < meadows and pastures < built-up areas < rivers < lakes and ponds. This methodological approach includes general hot-spot analysis as well as differentiated distance analysis which helps to better reveal the influence of landscape structure on WVCs.

Implications for conservation and game management of the roadkill levels of the endemic Iberian hare (Lepus granatensis)

The Iberian hare (Lepus granatensis) is an important small game species endemic to the Iberian Peninsula for which the incidence of roadkill is unknown. We surveyed Iberian hare-vehicle accidents on road networks in southern Spain, focusing on roads that mainly run through favorable habitats for this species: Mediterranean landscapes with plots of arable crops, olive groves, and vineyards. We recorded roadkills over a 5-month period, estimated hare accident densities on roads, and compared these numbers to hare hunting yields in adjoining hunting estates. We also analyzed the spatial patterns of and potential factors influencing hare roadkills. We detected the existence of black spots for hare roadkills in areas with high landscape heterogeneity that also included embankments and nearby crossroads and had high traffic intensity. Hare roadkill levels ranged from 5 to 25% of the annual harvest of hares killed on neighboring hunting estates. We suggest that road collisions should be considered in Iberian hare conservation in addition to hunting, since they may represent an additive source of mortality. Game managers should address the issue of hare roadkill in harvest planning to compensate for hare accidents, adjusting hunting quotas to account for this unnatural source of mortality. Our results suggest future directions for applied research in road ecology, including further work on demographic compensation and roadkill mitigation.

Colonization of Urban Habitats: Tawny Owl Abundance Is Conditioned by Urbanization Structure

Natural habitats are being altered and destroyed worldwide due to urbanization, leading to a decrease in species abundance and richness. Nevertheless, some species, including tawny owls, have successfully colonized this novel habitat. Consequences at the population level have not been described; thus, our main objective was to describe the effects that urban structure have on the tawny owl population at local and landscape levels. Data were obtained from 527 survey points over 7 months in a large-scale owl survey in the Basque Country (northern Spain) in 2018. At the local scale, the interaction between forest and urban cover affected tawny owl abundance, the optimum being in medium forested areas. The interaction between urban cover and clumpiness index (urban patch distribution) showed a generally negative effect. At the landscape scale, its abundance decreased in complex-shaped urban patches and when distance between them was greater. In conclusion, at the local scale, when a minimal forest structure is present in urbanized areas, the species can exploit it. At the landscape scale, it prefers smaller urban towns to cities. Thinking ahead, the current tendency toward "green capitals" should benefit tawny owl populations.

Land cover, individual's age and spatial sorting shape landscape resistance in the invasive frog Xenopus laevis

The description of functional connectivity is based on the quantification of landscape resistance, which represents species-specific movement costs across landscape features. Connectivity models use these costs to identify movement corridors at both individual and population levels and provide management recommendations for populations of conservation interest. Typically, resistance costs assigned to specific land cover types are assumed to be valid for all individuals of the population. Little attention has been paid to intraspecific variation in resistance costs due to age or dispersal syndrome, which may significantly affect model predictions. We quantified resistance costs in an expanding invasive population of the African clawed frog Xenopus laevis in Western France. In this principally aquatic amphibian, juveniles, sub-adults and adults disperse overland. The enhancement of dispersal traits via spatial sorting has been also observed at the range periphery of the population. Resistance costs, and thus connectivity, might vary as a function of life stage and position within the invaded range. We assessed multiple dimensions of functional connectivity. On various land cover types, we measured locomotion, as crossing speed, in different post-metamorphic age classes, and dehydration, sensitivity of locomotion to dehydration and substrate preference in juveniles. We also tested the effect of the position in the invaded range (core vs. periphery) on individual performances. In juveniles, general trends towards higher resistance costs on grass and lower resistance costs on bare soil and asphalt were observed, although not all experiments provided the same cost configurations. Resistance to locomotion varied between age classes, with adults and sub-adults facing lower costs than juveniles, particularly when crossing structurally complex land cover types such as grass and leaf litter. The position in the range had a minor effect on landscape resistance, and only in the dehydration experiment, where water loss in juveniles was lower at the range periphery. Depicting functional connectivity requires (a) assessing multiple dimensions of behavioural and physiological challenges faced by animals during movement; (b) considering factors, such as age and dispersal syndrome, that may affect movement at both individual and population levels. Ignoring this complexity might generate unreliable connectivity models and provide unsupported management recommendations for conservation.

Potential Movement Corridors and High Road-Kill Likelihood do not Spatially Coincide for Felids in Brazil: Implications for Road Mitigation

The negative effects of roads on wildlife populations are a growing concern. Movement corridors and road-kill data are typically used to prioritize road segments for mitigation measures. Some research suggests that locations where animals move across roads following corridors coincide with locations where they are often killed by vehicles. Other research indicates that corridors and road-kill rarely occur in the same locations. We compared movement corridor and road mortality models as means of prioritizing road segments for mitigation for five species of felids in Brazil: tiger cats (Leopardus tigrinus and Leopardus guttulus were analyzed together), ocelot (Leopardus pardalis), jaguarundi (Herpailurus yagouaroundi), and puma (Puma concolor). We used occurrence data for each species and applied circuit theory to identify potential movement corridors crossed by roads. We used road-kill records for each species and applied maximum entropy to determine where mortality was most likely to occur on roads. Our findings suggest that movement corridors and high road mortality are not spatially associated. We suggest that differences in the behavioral state of the individuals in the species occurrence and road-kill data may explain these results. We recommend that the road segments for which the results from the two methods agree (~5300 km for all studied species combined at 95th percentile) should be high-priority candidates for mitigation together with road segments identified by at least one method in areas where felids occur in low population densities or are threatened by isolation effects.

Predicting hedgehog mortality risks on British roads using habitat suitability modelling

Road vehicle collisions are likely to be an important contributory factor in the decline of the European hedgehog (Erinaceus europaeus) in Britain. Here, a collaborative roadkill dataset collected from multiple projects across Britain was used to assess when, where and why hedgehog roadkill are more likely to occur. Seasonal trends were assessed using a Generalized Additive Model. There were few casualties in winter-the hibernation season for hedgehogs-with a gradual increase from February that reached a peak in July before declining thereafter. A sequential multi-level Habitat Suitability Modelling (HSM) framework was then used to identify areas showing a high probability of hedgehog roadkill occurrence throughout the entire British road network (∼400,000 km) based on multi-scale environmental determinants. The HSM predicted that grassland and urban habitat coverage were important in predicting the probability of roadkill at a national scale. Probabilities peaked at approximately 50% urban cover at a one km scale and increased linearly with grassland cover (improved and rough grassland). Areas predicted to experience high probabilities of hedgehog roadkill occurrence were therefore in urban and suburban environments, that is, where a mix of urban and grassland habitats occur. These areas covered 9% of the total British road network. In combination with information on the frequency with which particular locations have hedgehog road casualties, the framework can help to identify priority areas for mitigation measures.

Simulating the Consequences of Roads for Wildlife Population Dynamics

Rapidly expanding road networks have been a key driver of the fragmentation and isolation of many wildlife species, and are a source of significant mortality due to collisions with vehicles. But not all animals are affected equally by transportation infrastructure, and in most cases little is known about the population-scale consequences of roads for wildlife. Even less information is available to characterize species' behavioral responses to roads. Although research shows that maned wolves (Chrysocyon brachyurus) in Brazil are experiencing considerable fragmentation and road mortality, the degree to which these disturbances are impacting the species' viability is unknown. The goal of this study was to analyze the potential effects of roads on maned wolf population size and structure. We used a simulation model to evaluate the population-scale consequences of individual maned wolf interactions with roads, which can result in road crossing, avoidance, or mortality due to a collision with a vehicle. We also forecasted where in Brazil these impacts might be most significant. Our model incorporated species demographic and movement parameters, plus habitat quality and a map of the road network. We found that even moderate rates of road mortality led to severe declines in population size, and that four specific locations accounted for a disproportionate fraction of roadkill events. Our approach will be generally useful for evaluating the relative importance of road effects on species conservation in many ecological systems, for prioritizing data collection efforts, and for informing conservation policies and mitigation strategies.

Predicting spatiotemporal patterns of road mortality for medium-large mammals

We modelled the spatiotemporal patterns of road mortality for seven medium-large mammals, using a roadkill dataset from Mato Grosso do Sul, Brazil (800 km of roads surveyed every two weeks, for two years). We related roadkill presence-absence along the road sections (1000 m) and across the survey dates with a collection of environmental variables, including land cover, forest cover, distance to rivers, temperature, precipitation and vegetation productivity. We further included two variables aiming to reflect the intrinsic spatial and temporal roadkill risk. Environmental variables were obtained through remote sensing and weather stations, allowing the estimate of the roadkill risk for the entire surveyed roads and survey periods. Overall, the models could explain a small fraction of the spatiotemporal patterns of roadkills (<0.23), probably due to species being habitat generalists, but still had reasonable discrimination power (AUC averaging 0.70 ± 0.07). The intrinsic spatial and temporal roadkill risk were the most important variables, followed by land cover, climate and NDVI. We show that identifying spatiotemporal roadkill patterns may provide valuable information to define specific management actions focused on road sections and time periods, in complement to permanent road mitigation measures. Our approach thus offers a new insight into the understanding of road effects and how to plan and strategize monitoring and mitigation.

Factors influencing predator roadkills: The availability of prey in road verges

Road mortality is the most noticeable effect of roads on wildlife. Road verges may provide important refuges for small mammals and rabbits, particularly when roads cross intensive agricultural or grazed areas. In these circumstances, the increasing use of verges by prey species may attract predators to road surroundings increasing the risk of roadkill. The aim of this study was to quantify the role of prey availability (small mammals and rabbits) on predator road casualties, taking into account road and surrounding landscape characteristics. We analyzed this effect on different predator species, such as, snakes, owls and mammal carnivores. The study took place in a 10 km stretch of a National Road (EN4) in southern Portugal. Relationships among predator mortality and explanatory variables (prey abundance, landscape characteristics, and road verge features) were evaluated using Multivariate Redundancy Analysis (RDA) followed by a variation partitioning. Our results show that, although landscape features explained most of the mortality variation, the prey availability was also very important. Roadkills of Montpellier snake and Egyptian mongoose are strongly associated with wild rabbit abundance on verges, while mortality of stone marten, barn owl and tawny owl is related with wood mouse abundance, Mediterranean forest (montado) density, and verge shrub density. Implications for verge management and implementation of mitigation actions are discussed. We suggest vegetation removal in verges to decrease shelter and food availability for prey, and/or the promotion of habitat for prey in areas distant from roads.

Accounting for Connectivity Uncertainties in Predicting Roadkills: a Comparative Approach between Path Selection Functions and Habitat Suitability Models

Functional connectivity modeling is increasingly used to predict the best spatial location for over- or underpasses, to mitigate road barrier effects and wildlife roadkills. This tool requires estimation of resistance surfaces, ideally modeled with movement data, which are costly to obtain. An alternative is to use occurrence data within species distribution models to infer movement resistance, although this remains a controversial issue. This study aimed both to compare the performance of resistance surfaces derived from path versus occurrence data in identifying road-crossing locations of a forest carnivore and assess the influence of movement type (daily vs. dispersal) on this performance. Resistance surfaces were built for genet (Genetta genetta) in southern Portugal using path selection functions with telemetry data, and species distribution models with occurrence data. An independent roadkill dataset was used to evaluate the performance of each connectivity model in predicting roadkill locations. The results show that resistance surfaces derived from occurrence data are as suitable in predicting roadkills as path data for daily movements. When dispersal was simulated, the performance of both resistance surfaces was equally good at predicting roadkills. Moreover, contrary to our expectations, we found no significant differences in locations of roadkill predictions between models based on daily movements and models based on dispersal. Our results suggest that species distribution models are a cost-effective tool to build functional connectivity models for road mitigation plans when movement data are not available.

The impact of human activities on Australian wildlife

Increasing human population size and the concomitant expansion of urbanisation significantly impact natural ecosystems and native fauna globally. Successful conservation management relies on precise information on the factors associated with wildlife population decline, which are challenging to acquire from natural populations. Wildlife Rehabilitation Centres (WRC) provide a rich source of this information. However, few researchers have conducted large-scale longitudinal studies, with most focussing on narrow taxonomic ranges, suggesting that WRC-associated data remains an underutilised resource, and may provide a fuller understanding of the anthropogenic threats facing native fauna. We analysed admissions and outcomes data from a WRC in Queensland, Australia Zoo Wildlife Hospital, to determine the major factors driving admissions and morbidity of native animals in a region experiencing rapid and prolonged urban expansion. We studied 31,626 admissions of 83 different species of native birds, reptiles, amphibians, marsupials and eutherian mammals from 2006 to 2017. While marsupial admissions were highest (41.3%), admissions increased over time for all species and exhibited seasonal variation (highest in Spring to Summer), consistent with known breeding seasons. Causes for admission typically associated with human influenced activities were dominant and exhibited the highest mortality rates. Car strikes were the most common reason for admission (34.7%), with dog attacks (9.2%), entanglements (7.2%), and cat attacks (5.3%) also high. Admissions of orphaned young and overt signs of disease were significant at 24.6% and 9.7%, respectively. Mortality rates were highest following dog attacks (72.7%) and car strikes (69.1%) and lowest in orphaned animals (22.1%). Our results show that WRC databases offer rich opportunities for wildlife monitoring and provide quantification of the negative impacts of human activities on ecosystem stability and wildlife health. The imminent need for urgent, proactive conservation management to ameliorate the negative impacts of human activities on wildlife is clearly evident from our results.

Influence of Landscape Factors on Amphibian Roadkills at the National Level

Roads exert multiple effects on wildlife, from animal mortality, habitat and population fragmentation, to modification of animal reproductive behaviour. Amphibians are the most frequently road-killed animal group. Many studies have attempted to analyse the factors driving amphibian casualties on roads, but these factors are limited to the roads themselves (e.g., traffic, type of roads, roads crossing water bodies) or to structures along them (e.g., ditches, walls). Sometimes, roadkills are related to land use along the roads. We analysed the influence of landscape factors on roadkill hotspots at the national level (Slovenia). Specifically, we aimed at: (1) identifying hotspots of roadkills, (2) analysing whether records of amphibian presence on roads are related to the distribution of water bodies and (3) analysing which factors (proximity to water bodies or human factors) explain the distribution of hotspots. Hotspots were identified by Malo’s method. Roadkills were modelled with Maxent for the first time in Slovenia. The relationships between amphibian presence and hotspots with factors were analysed with GLM. A total of 237 road sections were identified as hotspots, corresponding to 8% of road sections and containing 90% of road-killed individuals. Proximity to forests, meadows and agricultural land were the most important variables in Maxent models. The number of roadkills depended on the proximity to agricultural land, forests, water bodies and wetland areas, while the number of hotspots additionally depended on the proximity to urban settlements.

Infrastructure features outperform environmental variables explaining rabbit abundance around motorways

Human disturbance is widespread across landscapes in the form of roads that alter wildlife populations. Knowing which road features are responsible for the species response and their relevance in comparison with environmental variables will provide useful information for effective conservation measures. We sampled relative abundance of European rabbits, a very widespread species, in motorway verges at regional scale, in an area with large variability in environmental and infrastructure conditions. Environmental variables included vegetation structure, plant productivity, distance to water sources, and altitude. Infrastructure characteristics were the type of vegetation in verges, verge width, traffic volume, and the presence of embankments. We performed a variance partitioning analysis to determine the relative importance of two sets of variables on rabbit abundance. Additionally, we identified the most important variables and their effects model averaging after model selection by AICc on hypothesis‐based models. As a group, infrastructure features explained four times more variability in rabbit abundance than environmental variables, being the effects of the former critical in motorway stretches located in altered landscapes with no available habitat for rabbits, such as agricultural fields. Model selection and Akaike weights showed that verge width and traffic volume are the most important variables explaining rabbit abundance index, with positive and negative effects, respectively. In the light of these results, the response of species to the infrastructure can be modulated through the modification of motorway features, being some of them manageable in the design phase. The identification of such features leads to suggestions for improvement through low‐cost corrective measures and conservation plans. As a general indication, keeping motorway verges less than 10 m wide will prevent high densities of rabbits and avoid the unwanted effects that rabbit populations can generate in some areas.

Spatial patterns of road mortality of medium–large mammals in Mato Grosso do Sul, Brazil

Context Brazil has one of the richest biodiversity and one of the most extensive road networks in the world. Several negative impacts emerge from this interaction, including wildlife–vehicle collisions (WVC), which may represent a significant source of non-natural mortality in several species. The understanding of the main drivers of WVC is, therefore, crucial to improve the safe coexistence between human needs (transportation of goods and people) and animal populations. Aims We aimed to (1) evaluate the relative influence of land-cover patterns on the distribution of WVC, (2) assess whether WVCs are clustered forming hotspots of mortality, and, if so, (3) evaluate the benefits of mitigating only hotspot sections. Methods We collected WVC data involving medium–large mammals (4–260kg) along three road transects (920km), fortnightly over 1 year (n=1006 records). We used boosted regression trees to relate the WVC locations with a set of environmental variables including a roadkill index, reflecting overall habitat suitability and landscape connectivity, while accounting for spatial autocorrelation effects. We identified hotspots of mortality using Ripley’s K statistic and testing whether data follow a random Poisson distribution correcting for Type I error. Key results We found a strong association between WVC probability and roadkill index for all focal species. Distance to riparian areas, tree cover, terrain ruggedness and distance to urban areas were also important predictors, although to a lesser extent. We detected 21 hotspots of mortality, yet with little spatial overlapping as only four road sections (2%) were classified as hotspot for more than one species. Conclusions Our results supported that WVC mainly occur in road sections traversing areas with more abundant and diverse mammal communities. Hotspots of mortality may provide important information to prioritise road sections for mitigation, but this should be used in complement with roadkill indexes accounting for overall mortality. Implications The results support focusing on hotspots and habitat quality and landscape connectivity for a better assessment of road mortality. At the local scale, a larger number and improved road passages with exclusionary fencing of appropriate mesh size in riparian areas may provide safe crossings for many species and constitute a promising mitigation measure.

A simple framework for a complex problem? Predicting wildlife-vehicle collisions

Collisions of vehicles with wildlife kill and injure animals and are also a risk to vehicle occupants, but preventing these collisions is challenging. Surveys to identify problem areas are expensive and logistically difficult. Computer modeling has identified correlates of collisions, yet these can be difficult for managers to interpret in a way that will help them reduce collision risk. We introduce a novel method to predict collision risk by modeling hazard (presence and movement of vehicles) and exposure (animal presence) across geographic space. To estimate the hazard, we predict relative traffic volume and speed along road segments across southeastern Australia using regression models based on human demographic variables. We model exposure by predicting suitable habitat for our case study species (Eastern Grey Kangaroo Macropus giganteus) based on existing fauna survey records and geographic and climatic variables. Records of reported kangaroo-vehicle collisions are used to investigate how these factors collectively contribute to collision risk. The species occurrence (exposure) model generated plausible predictions across the study area, reducing the null deviance by 30.4%. The vehicle (hazard) models explained 54.7% variance in the traffic volume data and 58.7% in the traffic speed data. Using these as predictors of collision risk explained 23.7% of the deviance in incidence of collisions. Discrimination ability of the model was good when predicting to an independent dataset. The research demonstrates that collision risks can be modeled across geographic space with a conceptual analytical framework using existing sources of data, reducing the need for expensive or time-consuming field data collection. The framework is novel because it disentangles natural and anthropogenic effects on the likelihood of wildlife-vehicle collisions by representing hazard and exposure with separate, tunable submodels.

Sampling effects on the identification of roadkill hotspots: Implications for survey design

Although locating wildlife roadkill hotspots is essential to mitigate road impacts, the influence of study design on hotspot identification remains uncertain. We evaluated how sampling frequency affects the accuracy of hotspot identification, using a dataset of vertebrate roadkills (n = 4427) recorded over a year of daily surveys along 37 km of roads. "True" hotspots were identified using this baseline dataset, as the 500-m segments where the number of road-killed vertebrates exceeded the upper 95% confidence limit of the mean, assuming a Poisson distribution of road-kills per segment. "Estimated" hotspots were identified likewise, using datasets representing progressively lower sampling frequencies, which were produced by extracting data from the baseline dataset at appropriate time intervals (1-30 days). Overall, 24.3% of segments were "true" hotspots, concentrating 40.4% of roadkills. For different groups, "true" hotspots accounted from 6.8% (bats) to 29.7% (small birds) of road segments, concentrating from <40% (frogs and toads, snakes) to >60% (lizards, lagomorphs, carnivores) of roadkills. Spatial congruence between "true" and "estimated" hotspots declined rapidly with increasing time interval between surveys, due primarily to increasing false negatives (i.e., missing "true" hotspots). There were also false positives (i.e., wrong "estimated" hotspots), particularly at low sampling frequencies. Spatial accuracy decay with increasing time interval between surveys was higher for smaller-bodied (amphibians, reptiles, small birds, small mammals) than for larger-bodied species (birds of prey, hedgehogs, lagomorphs, carnivores). Results suggest that widely used surveys at weekly or longer intervals may produce poor estimates of roadkill hotspots, particularly for small-bodied species. Surveying daily or at two-day intervals may be required to achieve high accuracy in hotspot identification for multiple species.