Introduction

Wildlife roadkill refers to the inadvertent mortality of animals resulting from collisions with vehicles, a phenomenon that has escalated with the global expansion of road networks and has emerged as a critical concern for ecological, economic, and social systems (Grilo et al. 2021; Medrano-Vizcaíno et al. 2022). This mortality disrupts population dynamics and species distributions by fragmenting habitats and reducing genetic diversity (Jackson and Fahrig 2011), and it imposes significant economic burdens through vehicle damage, traffic delays, and increased public spending on road maintenance (Oddone Aquino and Nkomo 2021; Santos et al. 2011) and wildlife management (Grilo et al. 2021). Ecologically, roadkill serves as an index of broader environmental health, revealing trends in species decline and habitat degradation that may otherwise remain undetected (Schwartz et al. 2020). The utilization of roadkill data has evolved into an instrumental component of road ecology, a discipline that originated as expanding road systems and the concomitant impacts on wildlife prompted early investigations into animal mortality on roadways (Périquet et al. 2018; Schwartz et al. 2020). Foundational studies in this realm have highlighted the utility of standardized roadkill monitoring and citizen science initiatives to delineate spatial and temporal patterns of wildlife mortality, thereby enhancing our understanding of species vulnerability and informing targeted mitigation measures (Périquet et al. 2018; Schwartz et al. 2020). Historically, road ecology emerged in response to the mid-twentieth century’s exponential growth in transportation infrastructure (Bonan and Occhi 2023), which catalyzed landmark investigations demonstrating the relationship between road density and wildlife mortality (Moore et al. 2023). This interdisciplinary studies frame wildlife roadkill as not merely a byproduct of human transportation networks but as an indicator of socio-ecological imbalances that necessitate integrated conservation strategies. Consequently, the challenge of wildlife roadkill has evolved from a peripheral concern to a central topic in conservation biology, underscoring the need for comprehensive, data-driven policies that reconcile infrastructural development with biodiversity preservation (Périquet et al. 2018; Schwartz et al. 2020; Grilo et al. 2021; Medrano‐Vizcaíno et al. 2022).

The escalating global concern regarding road ecology and the increasing volume of studies on roadkill and mitigation measures underscore the necessity of synthesizing the extant literature using a bibliometric approach. Roadkill incidents are widely recognized as a crucial indicator of ecosystem disruption, and the growing body of research reflects heightened environmental awareness and urgency in addressing biodiversity loss (Collinson et al. 2015; Zhao et al. 2023). Although numerous ecological studies have explored the impacts of road networks—such as habitat fragmentation and species mortality—a systematic and quantitative evaluation of this expanding literature remains scarce. To date, bibliometric efforts in this field have been limited, with only one dedicated study identified (Hong et al. 2022), which reviewed roadkill research in South Korea. Their study used VOSviewer to compare bibliographic networks between domestic and international research, revealing a narrower thematic scope and less interconnected scholarly network within the Korean literature. This underscores the global need for broader bibliometric analyses to uncover research trends and gaps across the wider road ecology landscape. Bibliometric analysis, a method well-established in various disciplines, offers a robust framework to quantify publication trends, reveal co-authorship networks, and identify emerging research hotspots (Ninkov et al. 2021). By applying techniques analogous to those used in other fields—for instance, the structured quantitative analyses in ecological studies (Guan et al. 2018)—this review intends to unearth fundamental thematic shifts and interconnections within the vast body of roadkill research. Such an investigation is critical not only for mapping the intellectual landscape but also for identifying influential authors and institutions, thereby providing a comprehensive overview of the field’s evolution. While local and regional case studies have enriched our understanding of road ecology, their fragmented nature necessitates a holistic bibliometric synthesis to delineate broader trends and research gaps (Collinson et al. 2015; Zhao et al. 2023).

This review conducts a comprehensive bibliometric analysis of wildlife road mortality studies to evaluate research trends, identify gaps, and explore future directions. To guide the structure and narrative of this review, we developed a conceptual framework (Fig. 1) that links the primary drivers of wildlife roadkill—such as road expansion, habitat fragmentation, and traffic intensity—to three core research themes: (1) ecological and taxonomic impacts, focusing on mortality patterns, species vulnerability, and biodiversity loss; (2) human–wildlife interactions, including behavioral responses, public awareness, and policy dimensions; and (3) mitigation strategies, emphasizing technological and methodological advancements for monitoring and reducing roadkill. This framework informs our synthesis of historical trends, thematic developments, and future directions, while also providing a structured lens through which bibliometric and content analyses are interpreted. It serves as a scaffold for aligning analytical tools (e.g., keyword co-occurrence, MCA, and content synthesis) with broader conservation outcomes.

Fig. 1
figure 1

Our proposed conceptual framework

Full size image

Methodology

Rationale for bibliometric approach

This study employs a systematic bibliometric analysis to synthesize the extensive and multidisciplinary literature quantitatively and visually on wildlife road mortality research. This approach is particularly valuable for understanding the breadth and evolution of the research landscape, identifying major trends, and revealing knowledge structures within this rapidly evolving field. Unlike systematic reviews or meta-analyses, which focus on synthesizing empirical findings to address narrowly defined questions (Crocetti 2016), bibliometric analysis offers a broader perspective by mapping scholarly trends, relationships, and the intellectual structure of an entire research domain (Huertas-Valdivia et al. 2022). Moreover, bibliometric techniques facilitate the identification of emerging topics, influential contributions, and novel intersections or gaps in the literature, providing a comprehensive understanding of the research landscape and guiding strategic directions for future investigations (Farooq 2022).

Data collection and inclusion criteria

Data for this study were retrieved from the Scopus database, with searches conducted up to April 29, 2025, using a combination of title, abstract, and keyword fields, as outlined in Fig. 2. The full dataset used in the analysis is available in Supplementary I. Search terms included “roadkill” AND wildlife, “wildlife-vehicle collisions,” “animal-vehicle collisions,” “road mortality,” and “road-associated wildlife mortality,” yielding an initial total of 1,506 records. Following a screening process that excluded 15 non-research items—such as notes (6), letters (2), errata (2), editorials (2), a short survey, a report, and a data paper—the dataset was reduced to 1,491 eligible publications. These included 1,318 articles, 66 reviews, 69 conference papers, 37 book chapters, and one conference review. After a manual review of all eligible publications, 38 studies were excluded due to irrelevance (e.g., those focusing on human or pet road mortality). As a result, 1,453 studies with available English-language abstracts were included in the final bibliometric and content analyses. The language distribution of these studies was predominantly English (1,411), followed by Spanish (18), Chinese (7), Portuguese (5), French (4), German (4), and one publication each in Catalan, Croatian, Italian, and Korean.

The selection of the Scopus database for conducting bibliometric analyses in interdisciplinary fields such as environmental science, ecology, and transportation studies is supported by several compelling attributes aligned with rigorous academic standards. Scopus is recognized as the largest abstract and citation database of peer-reviewed literature, encompassing over 20,500 titles from 5,000 publishers. This broad and high-quality coverage ensures comprehensive representation of relevant research and makes Scopus an ideal choice for investigating emerging and interdisciplinary fields such as wildlife road mortality (Alves et al. 2021).

Fig. 2
figure 2

PRISMA flow diagram for systematic literature review

Full size image

Mapping the intellectual, thematic, and geographic landscape of wildlife road mortality research

To capture the intellectual, thematic, and geographic evolution of wildlife road mortality research, a comprehensive bibliometric and performance analysis was conducted on 1,453 relevant publications. The analysis was performed using Biblioshiny 4.1, a web-based application built on the Bibliometrix R package, which offers a robust suite of tools for descriptive statistics, co-authorship networks, conceptual structure mapping, and temporal trend analysis (Gutiérrez-Salcedo et al. 2018). This platform enabled the exploration of vital performance indicators, including annual publication growth, citation dynamics (total and average citations per document), document types, and the influence of high-impact journals and prolific contributors. Techniques such as fractionalized authorship were employed to accurately distinguish between individual and collaborative contributions.

Conceptual analysis focused on high-frequency keyword co-occurrence to identify dominant research areas and emerging themes. Multiple Correspondence Analysis (MCA) of 60 curated keywords revealed distinct thematic clusters that reflect the interdisciplinary and evolving nature of the field. Temporal keyword analysis traced shifts in research focus over time, while geographic mapping visualized regional patterns in research output and international collaboration. Co-authorship network analysis applied metrics such as Betweenness Centrality, which measures how often an author or institution lies on the shortest path between others—indicating its role as a connector or bridge within the network—and PageRank, which evaluates the influence of a node based on both the number and the importance of its connections.

To enhance the interpretability of the thematic clusters, five publications were randomly selected from the 973 articles that met the criteria in the ‘Article by Cluster’ feature of Biblioshiny. These publications were summarized in detail as part of the research content analysis, which is presented in the final section on main findings. Additionally, information on roadkill wildlife taxa and mitigation strategies was manually extracted from publications and organized in Excel for Mac (version 16.78; see Supplementary II). This process enabled a summary of the frequency of affected taxa and proposed mitigation measures. Together, these bibliometric and content-based approaches provide an integrated understanding of the development, structure, and future directions of wildlife road mortality research.

Results

Growth, collaboration, and key insights in wildlife road mortality research

As summarized in Table 1 and illustrated in Fig. 3, wildlife road mortality research has shown steady and substantial growth over the past six decades. Between 1960 and 2025, a total of 1,453 documents were published across 457 sources, with an average annual growth rate of 5.53%. Publication output remained low until the late 1990s, followed by a noticeable increase after 2000 and a more pronounced surge around 2008. The number of publications peaked in 2022 with 130 documents, then slightly declined in 2023 and 2024. As of April 2025, 33 documents had been published, suggesting a potentially lower total for the year—possibly due to a plateau in output or indexing delays. Peer-reviewed journal articles account for many publications (88%, or 1,284 documents), complemented by reviews (4%, or 65 documents), conference papers (5%, or 67 documents), book Chaps. (3%, or 36 documents), and a single conference review.

The field’s scholarly engagement is reflected in an average of 25.48 citations per document and a cumulative total of 61,353 references. A total of 4,211 authors have contributed to this body of work, with an average of 4.1 co-authors per document and 23.68% of publications involving international collaboration. Despite the presence of 115 single-authored papers, the overall trend demonstrates a strong preference for collaborative and interdisciplinary research. Additionally, the dataset includes 4,593 Keywords Plus (ID) and 3,503 Author Keywords (DE), which represent indexed and author-defined terms used to classify and retrieve research topics within the field.

Table 1 Summary information about wildlife road mortality publication sample pool
Full size table
Fig. 3
figure 3

Trend in publication types of wildlife road mortality research

Full size image

Hudson’s early contributions to road ecology, through studies in 1960 and 1966, provided foundational insights into wildlife road mortality in rural Northamptonshire, UK. The 1960 study documented 583 vertebrate casualties—including 288 birds and 295 other animals—on a two-mile stretch of the A6003, with Common Frogs (Rana temporaria) and House Sparrows (Passer domesticus) most affected; mortality peaked in summer and during holidays and decreased with adverse weather and lower vehicle speeds (Hodson 1960). The 1966 follow-up focused on small mammals and herpetofauna, identifying high fatalities among hedgehogs (Erinaceus europaeus), Common Shrews (Sorex araneus), Pigmy Shrews (S. minutus), rabbits (Oryctolagus cuniculus), Brown Hares (Lepus europaeus), and amphibians such as the Common Frog (R. temporaria) during breeding migrations, with risk influenced by habitat features such as hedgerows and water bodies (Hodson 1966). Both studies linked increased roadkill to higher traffic volumes, especially on weekends and holidays, emphasizing the urgent need for mitigation measures such as wildlife crossings.

Global contributions and publication trends in wildlife road mortality research

The ranking of leading sources in Scopus reveals the dominant platforms contributing to global research on wildlife road mortality (Table 2). Biological Conservation stands out with the highest number of publications (62) and holds a Q1 ranking, underscoring its influence and high impact. Following closely is the Journal of Wildlife Management, also a Q1 journal, with 46 publications. Other major contributors include Wildlife Research (Q2) with 39 publications and the European Journal of Wildlife Research (Q2) with 35. Both Journal of Environmental Management and PLoS ONE, each ranked Q1, contribute 31 and 30 articles respectively. Additional sources include the Wildlife Society Bulletin (Q2) with 29 publications, Animals (Q1) with 24, and both the Handbook of Road Ecology (a book chapter) and Herpetological Conservation and Biology (Q3), each also with 24. This spread reflects a growing diversity in publication venues and suggests the interdisciplinary expansion of road ecology research.

Temporal trends shown in Fig. 4 illustrate the evolving landscape of wildlife road mortality research across six major journals. Biological Conservation has seen a marked rise in output since 2006, overtaking the Journal of Wildlife Management around 2020, indicating its increasing centrality in the field. Although the Journal of Wildlife Management showed steady growth after 2006, its pace has been outpaced by Biological Conservation in recent years. Wildlife Research displays consistent contributions since the early 2000s, while the European Journal of Wildlife Research and Journal of Environmental Management show increasing activity from around 2010 and 2012, respectively. PLoS ONE also emerged as a significant contributor starting in the early 2010s.

The past two decades have seen a sharp rise in wildlife road mortality publications, expanding from a few specialized journals to a wide array of conservation-focused outlets (Fig. 4), reflecting the field’s growing ecological and policy relevance.

Table 2 Top 10 journals/ book contributing to global wildlife road mortality studies, ranked by publication output
Full size table
Fig. 4
figure 4

Growth in wildlife road mortality publications across the top six highest producing journals

Full size image

Analysis of leading authors and their contributions to wildlife road mortality research

The scholarly contributions of leading authors in wildlife road mortality research were evaluated using both total publication counts and fractionalized authorship metrics (Table 3). This dual approach provides a nuanced understanding of research productivity, distinguishing between authors who frequently collaborate on multi-author papers and those making more substantial individual contributions per publication. As shown in Table 3, the top 10 authors demonstrate varying patterns of engagement– from Grilo C’s high publication count (25) with moderate fractionalization (5.8), to Fahrig L’s fewer total publications (23) but greater individual contribution per paper (7.8). These metrics broadly reveal the field’s collaborative nature while highlighting the importance of assessing both quantity and quality of scholarly output.

Table 3 Top 10 leading authors contributing to wildlife road mortality research, ranked by total publication output and their fractionalized Article analysis
Full size table

As illustrated in Fig. 5, publication activity and citation influence vary widely among the top 10 authors. For instance, Van der Ree R shows a productivity peak around 2015, marked by a large, dark blue circle—indicating both a high number of publications and strong citation impact. In contrast, authors such as Fahrig L demonstrate steady output over time, while others such as Grilo C exhibit sustained influence through consistent citation density.

Fig. 5
figure 5

Publication output and citation impact of top 10 most productive authors over time. Total number of publications is represented by the size of the circles, with larger circles indicating higher publication volumes. Color intensity reflects citations per year, with deeper blue signifying higher citation concentration

Full size image

Geographic and institutional contributions to research output and impact

Figure 6 highlights the top 10 research institutions ranked by publication output in wildlife road mortality research, with the University of Georgia (USA) leading at 58 publications, followed by Montana State University (USA) with 47. Carleton University (Canada) and the China Academy of Transportation Sciences (China) share third place with 44 publications each, while the University of Oradea (Romania) follows closely with 42. Notably, 40 publications lack reported affiliations, indicating gaps in metadata that could hinder institutional analysis. Other main contributors include Laurentian University (Canada) with 37, the University of California (USA) with 37 publications, the Czech University of Life Sciences Prague (Czech Republic) with 36, and the University of Évora (Portugal) with 34. This distribution reflects strong contributions from North American institutions and growing international engagement, particularly from Europe and Asia, in advancing research on wildlife road mortality.

Fig. 6
figure 6

Top 10 research institutions contributing to wildlife road mortality studies based on publication output

Full size image

The publication output in wildlife road mortality studies, as depicted in Fig. 7, reveals a significant disparity in research activity across continents and countries. North America leads the field, primarily driven by the substantial contributions from the USA and Canada, indicating a strong historical and ongoing focus on this issue in the region. Europe follows with a considerable output, with Spain, the United Kingdom, Portugal, and Germany as leading contributors, highlighting a well-established research presence. Asia also demonstrates a notable publication volume, largely propelled by the increasing research activity in China and India. In contrast, South America and Oceania exhibit moderate levels of publication, with Brazil and Australia respectively leading their continents. Africa shows the lowest overall publication output among the regions visualized, suggesting a relatively less explored area of research in this context. This global distribution of research effort reflects varying levels of awareness, ecological pressures, research funding, and infrastructure across different parts of the world.

Figure 8 reveals striking global disparities in research impact. The USA leads with the highest citation volume (8,858), followed by Canada and Australia, while Portugal and the UK, despite lower output, achieve the highest average citations per article, highlighting the disproportionate influence of certain nations relative to their research volume

Fig. 7
figure 7

Global publication output in wildlife road mortality research by continent and country

Full size image
Fig. 8
figure 8

Global citation impact of wildlife road mortality research by country. The top 10 countries by total citations are listed in the accompanying table

Full size image

Collaboration networks of authors, institutions, and countries in wildlife road mortality research

The collaboration network of authors contributing to wildlife road mortality research, as visualized in Fig. 9A, reveals a well-defined community structure with varying degrees of connectivity and influence among researchers. Distinct clusters are represented by different colors, with each cluster indicating a group of authors who frequently collaborate within their network. A particularly prominent cluster, marked in orange, features influential authors such as Grilo C, Clevenger AP, Van der Ree R, and Fahrig L. These individuals, represented by larger nodes due to their high degree and PageRank values, serve as central figures within the field and act as key drivers of collaborative research. Authors who act as intermediaries between clusters, such as Huijser MP and Perkins SE—linking the orange/brown cluster to the light green cluster containing Bíl M—and Balčiauskas L—connecting to the pink cluster—demonstrate high Betweenness Centrality, playing a pivotal role in facilitating cross-group knowledge exchange. In contrast, peripheral groups, such as the red cluster (e.g., Litzgus JD, Lesbarrères D) and the isolated purple node (Kučas A), appear more disconnected from the core community, reflecting limited engagement in broader collaborative efforts. The network structure emphasizes the importance of both central and bridging authors in fostering an integrated research landscape and highlights potential areas where collaboration could be expanded.

The institutional collaboration network, depicted in Fig. 9B, exhibits a structured configuration characterized by several interconnected clusters of academic and research organizations. Montana State University (USA) is positioned as a central hub within this network, serving as a key bridge among various institutional clusters and underscoring its influential role in facilitating interdisciplinary and cross-institutional collaboration. Another major collaboration cluster is represented by the red grouping, which includes institutions such as the University of Georgia (USA) and Purdue University (USA), indicating strong regional or thematic ties. Similarly, the large blue cluster—comprising institutions such as Carleton University (Canada) and the University of Reading (UK)—demonstrates robust collaborative activity across international lines. Bridging institutions, including Utah State University (USA) and the University of Alberta (Canada), connect multiple clusters and contribute significantly to the cohesion of the broader network. Conversely, institutions located on the periphery or in smaller, less-connected clusters reflect areas where collaboration is more limited, suggesting opportunities for strategic partnership development. This network analysis highlights the centrality of specific institutions in shaping the field’s collaborative dynamics and underscores the value of expanding cooperative relationships to support innovation and knowledge dissemination.

At the international level, the country-level collaboration network illustrated in Fig. 9C reveals a globally interconnected system in which all analyzed countries maintain some level of participation. The United States occupies a dominant position within this network, as evidenced by its highest Betweenness Centrality (223.9) and PageRank (0.1) values (not shown in Fig. 9C), signifying its role as a primary hub for international collaboration. Other countries such as South Africa, Australia, Canada, and Brazil also serve important intermediary functions, contributing to the cohesion and integration of the global research landscape. In contrast, countries such as China, Korea, Argentina, and Colombia play more peripheral roles, while Iran is notably marginal in terms of collaborative influence. These disparities suggest differences in research capacity, international engagement, or thematic focus. Despite these variations, the existence of a single unified collaboration network underscores the global relevance of wildlife road mortality research and reflects a shared commitment across nations to address the ecological consequences of road infrastructure.

Fig. 9
figure 9

Visualized collaboration networks in wildlife road mortality research: (A) author collaboration network, (B) institutional collaboration network, and (C) country collaboration network

Full size image

The analysis of international collaborations in wildlife road mortality research reveals a globally interconnected network, with the USA serving as the central hub (Table 4). American researchers maintain particularly strong ties with major partners across North America, Europe, and Oceania, while European nations demonstrate both regional cooperation and extensive intercontinental linkages. Brazil emerges as a significant node, bridging research efforts between the Americas and Europe, and Australia maintains notable trans-Pacific connections. These collaboration patterns underscore how international partnerships facilitate knowledge-sharing to address wildlife conservation challenges across diverse geographic contexts.

Table 4 Frequency of mutual publications between the top 15 countries contributing to wildlife road mortality research
Full size table

Multifaceted impacts of roads on wildlife: key findings from influential research

Traffic-related mortality poses a serious threat to wildlife, particularly amphibians. Studies by Fahrig et al. (1995) and Hels and Buchwald (2001) demonstrate that increasing traffic intensity has a significant negative effect on local anuran (frog and toad) populations. Both studies found that as traffic volume rose, the number of live and dead individuals observed per kilometer declined, while the proportion of road-killed individuals increased. Additionally, chorus intensity—a proxy for population density—decreased with higher traffic levels, indicating reduced local abundance. These findings suggest that road mortality is a major factor contributing to amphibian population declines, particularly in regions experiencing rapid traffic growth.

Clevenger et al. (2002) and Malo et al. (2004) provide further insights into how road characteristics and surrounding landscapes influence wildlife-vehicle collisions. In Canada’s Bow River Valley, Clevenger et al. (2002) found that small mammals and birds experienced higher roadkill rates on low-traffic parkways than on high-speed highways, with birds especially vulnerable on the latter. Collisions were more likely to occur near vegetative cover and away from wildlife passages or culverts, and less frequently on elevated road sections—highlighting how both road design and adjacent habitats shape mortality patterns. Complementing these findings, Malo et al. (2004) analyzed over 2,000 collision records in Spain to build predictive models at both landscape and local scales. They identified high-risk zones associated with forested areas, low building density, and a lack of roadside barriers. Their models achieved high predictive accuracy—over 70% for broader road sections and 85% for specific hotspots—showing their value in guiding effective mitigation strategies. Together, these studies emphasize the importance of integrating spatial and ecological data into road planning to minimize wildlife mortality and improve road safety.

Mitigation strategies also play a central role in addressing wildlife-vehicle collisions (WVCs), as highlighted by Glista et al. (2009) and Clevenger et al. (2001). Glista et al. (2009) reviewed a range of mitigation measures, noting that roads not only cause direct mortality but also fragment habitats and reduce landscape connectivity. Although wildlife-crossing structures are commonly recommended, a lack of rigorous “before-and-after” studies makes it difficult to assess their effectiveness. Nevertheless, the authors suggest that even simple, low-cost interventions can meaningfully reduce WVCs. Clevenger et al. (2001) evaluated highway fencing along the Trans-Canada Highway in Banff National Park and observed an 80% reduction in ungulate-vehicle collisions post-installation. Most remaining collisions occurred near fence ends, indicating the importance of strategic fence design. The study also found no link between collisions and fence access points but noted that natural features such as major drainages may influence collision hotspots. Both studies call for integrated mitigation planning that combines structural design with behavioral interventions to enhance road safety for wildlife.

Beyond direct mortality, roads act as significant behavioral and physical barriers to wildlife movement. Shepard et al. (2008) and Soulsbury and White (2015) explore the broader impacts of roads and urbanization on animal behavior and population dynamics. Shepard et al. (2008) used radiotelemetry to study three terrestrial vertebrates—the eastern massasauga (Sistrurus catenatus), eastern box turtle (Terrapene carolina), and ornate box turtle (T. ornata)—and found that all species crossed roads significantly less frequently than expected by chance, indicating strong behavioral avoidance likely driven by mortality risk. While such avoidance may reduce short-term fatalities, it can result in long-term genetic isolation and increased extinction risk, particularly for long-lived species. The authors argue that effective conservation requires integrating data on animal behavior, mortality, and genetics to fully address the barrier effects of roads. Expanding on this, Soulsbury and White (2015) highlight the rising frequency of human–wildlife interactions in urban environments. While much of the current research focuses on conflict—due to its relative ease of measurement—they also emphasize the underappreciated benefits of urban wildlife, including ecosystem services and contributions to human well-being. The authors advocate for interdisciplinary collaboration to better manage risks and promote coexistence, underscoring the importance of adaptive, inclusive strategies in an increasingly urbanized world.

Species respond differently to road impacts depending on their ecological traits. Gibbs and Shriver (2002) addressed this variation in the context of the United States’ rich turtle diversity. Using a modeling approach that combined road maps, traffic data, and simulated turtle movements, they evaluated the vulnerability of three turtle groups: (1) small-bodied pond turtles, (2) large-bodied pond turtles, and (3) terrestrial and semi-terrestrial (“land”) turtles. The results showed that road networks in the northeastern, southeastern, and central U.S. pose serious risks to land turtles and, to a lesser extent, large-bodied pond turtles. In contrast, small-bodied pond turtles were not found to be regionally threatened by road mortality. The study concluded that the combination of specific life history traits and limited mobility makes some turtle populations particularly susceptible to decline in heavily roaded landscapes.

In summary, this body of research highlights the complex and far-reaching ecological consequences of road infrastructure. Effective wildlife conservation in the face of expanding transportation networks will require a multifaceted approach—one that integrates traffic regulation, ecologically sensitive road design, species-specific mitigation, and cross-sector collaboration. Table 5 synthesizes main findings from these studies, demonstrating the need for integrated, data-informed strategies to address the diverse impacts of roads on wildlife.

Table 5 Summary of leading authors and their high-citation articles on wildlife road mortality
Full size table

Thematic landscape of wildlife road mortality research: a keyword analysis

The analysis of the tree map illustrating the frequency of the top 50 “Keyword Plus” terms in wildlife road mortality research reveals that the most prominent keywords are: mortality (397), roadkill (388), animals (328), wildlife management (208), article (177), United States (165), collision (162), roads and streets (140), nonhuman (140), traffic accident (133), deer (132), Mammalia (144), road (123), human (119), and biodiversity (114). This distribution highlights a strong emphasis on the direct consequences of roadkill—particularly animal mortality—with significant attention given to mammals and deer. The frequent appearance of “United States” suggests a substantial volume of research originating from or focused on this region. The prominence of terms such as wildlife management and article reflect the applied nature of the field and its increasing publication output.

Other frequently occurring keywords include habitat fragmentation (107), mitigation (92), wildlife (92), Amphibia (91), anthropogenic effect (88), vehicles (87), Aves (86), female (86), male (87), Brazil (82), accidents traffic (84), ecosystem (78), wild animal (77), risk assessment (76), nature-society relations (75), wild population (75), Australia (72), endangered species (71), Reptilia (70), Vertebrata (70), vehicle collision (69), conservation management (63), Canada (65), population density (64), movement (63), spatiotemporal analysis (61), turtle (60), species conservation (57), roadside environment (54), adult (52), environment conservation (50), controlled study (49), and connectivity (47). These terms reflect a growing interest in the broader ecological implications of road mortality, including habitat fragmentation, threats to endangered species, and efforts in mitigation, conservation management, risk assessment, and enhancing landscape connectivity. The presence of geographic terms such as Brazil, Canada, and Australia underscore the international scope of the research. Additionally, the inclusion of specific taxonomic groups such as Amphibia, Aves, Reptilia, and turtle highlights the diverse range of species addressed in road mortality studies.

Temporal evolution of fundamental research themes in wildlife road mortality studies

Applying the specified parameters of a minimum word frequency of 50, the top five words per year (as shown in Fig. 10), and merged synonyms, the analysis of trending topics in wildlife road mortality research reveals a clear temporal evolution of fundamental themes. Early research, particularly between 2011 and 2013, focused primarily on specific taxonomic groups such as amphibians and reptiles while also introducing the emerging concept of habitat fragmentation. Geographic regions such as Australia and Canada appeared during this period, though their prominence was less pronounced compared to thematic keywords due to the frequency threshold.

During the mid-period of research, roughly between 2014 and 2017, core concepts related to road mortality became more dominant. Keywords such as “mortality” and “roadkill” gained central importance alongside a strong geographic focus on the “United States” and frequently studied species such as “deer.” Other notable topics included the “anthropogenic effect” and “collision,” reflecting deeper exploration of the causes of road mortality. Additionally, “wildlife management” emerged as a key area of interest, demonstrating a growing emphasis on mitigation strategies to address these impacts.

More recent trends, evident from 2018 to 2023, show a broadening of research scope. Terms such as “biodiversity” gained prominence, signaling increased recognition of the wider ecological ramifications of roadkill. Methodological advancements were also observed with a rise in the use of “spatiotemporal analysis,” reflecting more sophisticated approaches to studying wildlife road mortality. Geographically, research expanded beyond traditional regions with a notable increase in studies from areas such as “Brazil.” Throughout all periods, general terms such as “animal” and “wildlife” remained consistently relevant, underscoring their foundational role in the research.

Figure 10 tracks the evolution of thematic focus in the field over time. Early emphasis on amphibians and habitat fragmentation (2011–2013) gradually shifted to mortality, roadkill, and wildlife management (2014–2017). From 2018 onward, newer themes such as biodiversity and spatiotemporal analysis gain prominence, showing a methodological and conceptual broadening of the field.

Fig. 10
figure 10

Trends of Keyword Plus terms in wildlife road mortality research

Full size image

Conceptual structure and thematic clusters in wildlife road mortality research

The field of wildlife road mortality research exhibits a clear thematic structure, as revealed through Multiple Correspondence Analysis (MCA) of 60 curated keywords. This analysis identified three major clusters (Fig. 11), each reflecting distinct conceptual and methodological orientations within the literature.

Cluster 1 (red), identified as Core Road Ecology and Applied Conservation, is the largest and most central thematic group. It includes keywords such as “roadkill,” “mortality,” “collision,” “vehicles,” “conservation management,” and “habitat fragmentation.” These terms highlight a strong emphasis on direct ecological impacts of roads on wildlife, including vehicle collisions and population declines. This cluster underscores the practical and applied focus of much road ecology research, which seeks to mitigate harm through conservation strategies, risk assessments, and infrastructure interventions. The presence of geographic references such as “United States” and “Australia” also indicates the broad regional coverage of these applied studies.

Cluster 2 (blue), labeled Human–Wildlife Interactions and Theoretical Perspectives, encompasses terms such as “human,” “nonhuman,” “nature–society relations,” “controlled study,” and “environmental protection.” Situated on the opposite end of the MCA spectrum from Cluster 1, this grouping reflects a more philosophical or interdisciplinary approach. It draws from fields such as environmental ethics, social sciences, and human-animal studies. The cluster explores how societies perceive and engage with wildlife, particularly in contexts where roads intersect with natural habitats. Its distinct location in the MCA plot confirms a conceptual departure from applied ecological studies, emphasizing theoretical inquiry and socio-ethical dimensions.

Cluster 3 (green), or Taxon-Specific and Biodiversity-Oriented Studies, is focused on species-level research and biodiversity assessments. It includes keywords such as “mammal,” “bird,” “Anura,” “Reptilia,” and “Brazil.” These terms reflect studies that are taxonomically focused and often grounded in ecological monitoring within biodiversity hotspots. The cluster’s position along the second MCA dimension suggests that these works are differentiated more by biological focus than by methodological or policy orientation. It highlights the importance of species-specific assessments in understanding the varied impacts of road systems on vertebrate fauna.

Fig. 11
figure 11

Factorial map of the conceptual structure in wildlife road mortality research, generated using Multiple Correspondence Analysis (MCA) of 60 Keyword Plus terms

Full size image

In summary, the MCA results shown in Fig. 11 reveal a multidimensional research landscape organized around three core orientations: applied conservation science, human–wildlife relational theory, and taxonomic biodiversity monitoring. These thematic clusters illustrate how wildlife road mortality research integrates ecological, social, and biological perspectives, contributing to a complex and evolving body of knowledge that encompasses both practical mitigation efforts and broader theoretical inquiry. To further examine the content of each cluster, five articles were randomly selected from each group using the “Article by Cluster” function in Biblioshiny. A complete list of the selected articles is provided in Supplementary II, and detailed descriptions of each cluster are presented below.

Core road ecology and applied conservation research—comprising 95% (923 of 973) of the reviewed articles—has firmly established wildlife road mortality as a critical conservation concern globally. Studies show that expanding transportation infrastructure and increasing vehicle traffic are major drivers of vertebrate mortality and demographic shifts in wildlife populations. For instance, Teixeira et al. (2013a) demonstrated that conventional roadkill surveys often drastically underestimate actual mortality due to unaccounted carcass removal and detection biases. Their differential equation-based models revealed that bird mortality could be underestimated by up to 39 times, and reptile mortality by roughly twice that figure. These findings underscore the necessity of refined mortality estimation techniques for accurate impact assessments. Aresco (2005a) further highlighted the demographic consequences of road mortality through a nine-year study of freshwater turtles in Florida, USA, including Pseudemys floridana, Trachemys scripta, and Sternotherus odoratus. The study found male-biased sex ratios of 65–80% near high-traffic roads, attributed to disproportionately high mortality among nesting females—up to 29% annually—crossing roads to lay eggs. Such skewed sex ratios can threaten population viability over time. Meanwhile, Kociolek et al. (2011) synthesized findings from over 100 studies, noting that birds are impacted not only by direct mortality but also by indirect effects such as noise, habitat fragmentation, and artificial lighting, with species such as the Northern Cardinal (Cardinalis cardinalis) and Eastern Kingbird (Tyrannus tyrannus) particularly vulnerable.

Further evidence of roadkill mitigation and adaptive strategies has emerged from studies leveraging both environmental change and technological innovation. Pokorny et al. (2022) used time-series data from Slovenia during COVID-19 lockdowns to show that reduced human mobility significantly decreased roadkill events in several mammal species, including roe dear (Capreolus capreolus) and wild boar (Sus scrofa), with declines of 156–321 roe deer collisions in spring alone. However, mortality patterns were species-specific, with some, such as the Eurasian badger (Meles meles), even showing increased collisions during certain periods. In response to the need for proactive mitigation, Saxena et al. (2021) proposed an AI-based animal detection and collision avoidance system utilizing SSD and Faster R-CNN neural networks. Using a custom dataset with 31,774 images across 25 animal classes, their system achieved mean average precisions (mAP) of 80.5% at 100 fps (SSD) and 82.11% at 10 fps (Faster R-CNN), enabling real-time animal detection and potential AVC (animal–vehicle collision) mitigation. Overall, these studies emphasize the urgent need for integrative road ecology strategies that combine precise mortality estimation, demographic impact assessment, species-specific behavioral understanding, and advanced technological interventions to reduce wildlife road mortality and ensure sustainable coexistence between road networks and biodiversity.

Human–Wildlife Interactions and Theoretical Perspectives—comprising 1% (13 of 973) of the reviewed articles—address the complex dynamics between wildlife and human activity on roads, which pose significant risks to both biodiversity and public safety. In São Paulo State, Brazil, 18.5% of 2,611 annual animal–vehicle collisions (AVCs) resulted in human injury or death, with associated costs estimated at approximately US$25.1 million per year (Abra et al. 2019). Notably, Brazilian law assigns liability to road administrators in 91.7% of AVCs, whether they involve wildlife or domestic animals such as horses, cattle, and dogs. This legal framework reflects societal expectations for animal-proof roads and underscores the need for integrated solutions, including enhanced infrastructure (e.g., fencing and wildlife crossings), improved vehicle licensing systems, and greater public awareness. In Canada, Vanlaar et al. (2019) identified a significant gap between driver knowledge and behavior during wildlife collisions. Although most drivers understood that slowing down and steering straight is the safest response, many swerved instead, often increasing the severity of collisions. These findings underscore the importance of behavioral interventions and driver education in reducing wildlife–vehicle collisions (WVCs), particularly with large mammals.

Understanding species-specific behaviors and temporal patterns is essential for implementing targeted mitigation measures. In Lithuania, Kučas and Balčiauskas (2020) analyzed 14,989 ungulate–vehicle collisions involving roe deer (Capreolus capreolus), red deer (Cervus elaphus), moose (Alces alces), and wild boar (Sus scrofa), revealing peak incidents at dawn and dusk, with increased frequency on Fridays and notable seasonal variation. In the United Kingdom, Raymond et al. (2021) used over 54,000 citizen-reported records to demonstrate that species such as European badgers (Meles meles) and grey squirrels (Sciurus carolinensis) exhibit seasonal roadkill patterns, while others, including red foxes (Vulpes vulpes) and Reeves’ muntjac deer (Muntiacus reevesi), showed no distinct seasonality. Environmental conditions—particularly rainfall—further influenced the detectability and mortality of species such as Eurasian magpies (Pica pica) and European rabbits (Oryctolagus cuniculus). In Massachusetts, USA, Zeller et al. (2018) found that moose–vehicle collisions (MVCs) involving moose (Alces americanus) accounted for an estimated annual loss of 3% of the state’s moose population. By integrating GPS tracking with collision data, the study identified high-risk segments of roadways and emphasized the importance of fencing, detection systems, and overpasses, particularly in forested and wetland habitats. Generally, these studies highlight the need for alignment between ecological understanding, legal frameworks, and human behavior in the development of effective roadkill mitigation strategies.

Taxon-Specific and Biodiversity-Oriented Studies (4%, 36 of 973 articles) have significantly advanced the understanding of wildlife–vehicle collisions (WVCs) by illuminating species-specific vulnerabilities and informing targeted mitigation strategies. For example, Visintin et al. (2017) developed a predictive risk modeling framework to assess WVC risks across six native terrestrial mammal species in southeastern Australia, including the koala (Phascolarctos cinereus) and the common ringtail possum (Pseudocheirus peregrinus). Their models explained up to 34.3% of the spatial variation in species occurrences and up to 19.4% in collision data, demonstrating how collision risk is influenced by both species distribution and traffic parameters. Similarly, Delgado et al. (2018) conducted a multi-species analysis of 35 amphibians, reptiles, and mammals in Andalusia, Spain, and found that roadkill distribution was not significantly associated with the presence of underpasses, suggesting that road permeability is shaped by more than just structural features such as culverts. Their findings point to the importance of considering additional ecological and behavioral factors. In Brazil’s Cerrado region, Silveira Miranda et al. (2020) identified taxon-specific roadkill aggregation patterns, noting that birds and mammals exhibited different spatial hotspots. The study cautioned against overly generalized analyses and advocated for group-specific approaches to avoid misleading conclusions. In Alberta, Canada, Paul et al. (2014) validated the reliability of citizen science data by comparing 640 volunteer-reported wildlife observations with systematically collected data, identifying three consistent roadkill hotspots, and demonstrating the value of public participation in road ecology research. Most recently, Bíl et al. (2024) analyzed a comprehensive WVC dataset in Czechia and found that collisions predominantly involved roe deer (Capreolus capreolus) and wild boar (Sus scrofa), with motorcyclists facing 1,600 times higher odds of injury compared to car occupants. These studies collectively highlight the necessity of taxonomic precision and species-level data in enhancing WVC mitigation, improving wildlife conservation, and advancing road safety interventions.

Taxonomic and conservation biases in wildlife road mortality research

The analysis of wildlife taxa in roadkill studies reveals two major research biases: a pronounced taxonomic imbalance and uneven representation of species based on conservation status. As shown in Fig. 12 and Supplementary III, mammals dominate the literature (n = 933 studies), followed by reptiles (n = 359), general wildlife categories (n = 300), amphibians (n = 208), birds (n = 203), and marsupials (n = 86). In stark contrast, arthropods—particularly insects (n = 18) and crustaceans/myriapods (n = 5)—are severely underrepresented. This pattern reflects broader conservation biases, where 75% of studies focus on species classified as “Least Concern,” while only a small fraction address taxa facing higher extinction risks.

Critically Endangered species are particularly neglected, with just 1% of studies (n = 13) investigating species such as Agassiz’s desert tortoise (Gopherus agassizii), Asiatic cheetah (Acinonyx jubatus venaticus), and Malayan tiger (Panthera tigris jacksoni). Even fewer studies (n = 5) focus on other threatened species, including the Oregon silverspot butterfly (Speyeria zerene hippolyta) and the Southwestern Pond turtle (Emys pallida). This disproportionate attention to low-risk species limits the ecological relevance of the research and weakens its potential to inform conservation strategies for vulnerable taxa.

These disparities highlight critical knowledge gaps regarding the impacts of roads on biodiversity. The limited inclusion of threatened species and invertebrates calls into question the representativeness of existing roadkill studies. A more balanced research agenda is needed to ensure that road ecology informs conservation policies for both well-known and overlooked species.

Table 6 illustrates how mitigation strategies vary across taxonomic groups. Mammals such as moose (Alces alces) and Florida panthers (Puma concolor coryi) benefit from interventions such as fencing and wildlife underpasses. Reptiles have been supported by measures such as drift fences and seasonal road closures. However, small vertebrates and invertebrates often lack targeted mitigation strategies. Cross-taxa approaches—including GIS-based hotspot modeling, citizen science platforms such as iNaturalist, and AI-powered detection systems—offer scalable solutions, but their implementation remains limited. Addressing these persistent imbalances requires inclusive, interdisciplinary frameworks to reduce wildlife mortality and protect underrepresented species effectively.

Fig. 12
figure 12

Frequency of wildlife taxa in roadkill studies from Scopus, categorized by taxonomic group and conservation status

Full size image
Table 6 Key wildlife taxa affected by road mortality and corresponding mitigation strategies proposed in the selected literature
Full size table

Discussion

This bibliometric and systematic review reveals both the rapid expansion and persistent gaps in global wildlife roadkill research. While the field has advanced significantly—particularly over the past two decades, paralleling global infrastructure growth and biodiversity threats—progress remains uneven across taxonomic groups, geographic regions, and research themes.

The pronounced geographic concentration of wildlife roadkill research in North America and Europe raises serious concerns regarding conservation efficacy in underrepresented yet biodiversity-rich regions such as Southeast Asia, Central Africa, and South America. This disparity underscores the urgent need to reorient research priorities toward areas facing the highest biodiversity threats. As Ducatez and Lefebvre (2014) observe, high-conservation-value regions—particularly in South and Central America and Eastern Asia—remain significantly under-studied compared to temperate zones, which benefit from better funding and research infrastructure. This pattern reflects broader trends in conservation science, where investments and academic focus often prioritize well-resourced areas over those experiencing the most severe biodiversity loss (Zabel et al. 2019). Bridging this gap requires targeted investment in research infrastructure, knowledge exchange, and collaborative networks in underserved regions to ensure a more equitable global conservation agenda.

Addressing this systemic imbalance requires targeted investments in capacity-building and institutional support for neglected regions. Porzecanski et al. (2022) emphasize the importance of inclusive conservation frameworks that actively involve local stakeholders and are tailored to regional contexts. Similarly, O’Connell et al. (2022) highlight the value of partnerships and mentorship in facilitating knowledge exchange and promoting place-based conservation strategies. Engaging diverse stakeholders not only fosters local resilience but also aligns grassroots conservation actions with broader global biodiversity targets, contributing to more equitable and effective outcomes (Valdez et al. 2024).

Wildlife roadkill research exhibits notable taxonomic and thematic biases. A disproportionate emphasis has been placed on large, charismatic mammals—particularly ungulates and carnivores—while ecologically critical but less conspicuous groups such as amphibians, reptiles, and invertebrates remain significantly underrepresented, especially in tropical (Teixeira et al. 2013b; Kouris et al. 2024). This taxonomic skew limits the comprehensiveness of conservation responses and overlooks the broader ecological implications of road mortality.

The conservation outlook for amphibians is particularly urgent due to the dual threats of road mortality and emerging infectious diseases such as Chytridiomycosis. Roadkill peaks during seasonal migrations to breeding habitats, particularly in rainy periods when amphibians cross roads in large numbers (Zhang et al. 2018; Bastos et al. 2019; Sur et al. 2022). The presence of nearby wetlands, which serve as critical breeding sites, further increases risk, as individuals frequently move between terrestrial and aquatic environments that intersect with roadways (Gryz and Krauze-Gryz 2008). Compounding these challenges, the global spread of the Chytrid fungus has caused dramatic population declines in many amphibian species, amplifying the long-term impact of road mortality on already vulnerable populations (Dąbrowska and Sołtysiak 2015; Reshetylo et al. 2019). These converging threats erode population resilience and increase extinction risk, particularly in fragmented landscapes (Fraga et al. 2022).

Thematically, most studies remain descriptive, focusing primarily on species detection and mortality rates. There is a dearth of research examining the demographic, behavioral, and genetic impacts of roadkill, particularly in vulnerable populations (Olgun et al. 2022; Moore et al. 2023). Although mitigation measures such as underpasses and fencing are increasingly reported, their effectiveness is rarely assessed systematically—especially for small-bodied species for whom such structures are often inadequate (Silva et al. 2020). Future research should prioritize standardized, long-term evaluations of mitigation strategies across varied ecological and socio-political settings. Moreover, inclusive approaches such as citizen science should be further utilized to identify roadkill hotspots and inform targeted interventions (Schwartz et al. 2020). A more balanced and integrative approach to road ecology is essential to ensure that conservation strategies address the needs of all taxa, including those historically overlooked in both research and policy (Santos et al. 2015).

Technological and methodological innovations have significantly influenced the evolution of wildlife roadkill research. Tools such as citizen science, camera trapping, artificial intelligence (AI), machine learning, and remote sensing have expanded both data collection and analytical capacity. Citizen science platforms and camera traps have enhanced spatial and temporal monitoring, while AI and machine learning have enabled efficient species identification and predictive modeling with high levels of accuracy (Willi et al. 2018; Bijl and Heltai 2022).

Despite their potential, the integration of these technologies into wildlife roadkill studies remains limited. Barriers include challenges in data standardization, validation, and the applicability of tools across diverse ecological contexts (Schneider et al. 2020). Remote sensing offers additional promise for monitoring wildlife movements at scale, particularly when integrated with machine learning techniques (McCarthy et al. 2024). Realizing the full potential of these technologies will require interdisciplinary collaboration among ecologists, engineers, and data scientists to bridge methodological divides and develop scalable, robust solutions (Glover-Kapfer et al. 2019). As new technologies such as drones and next-generation AI models become more widely accessible, establishing standardized protocols and validation frameworks will be essential to ensure data quality and inform evidence-based conservation strategies (Dujon et al. 2021). Ultimately, embedding these technologies within a broader conservation framework could transform wildlife roadkill research from descriptive monitoring into proactive, data-driven mitigation planning.

Policy integration and conservation implications

Although wildlife road mortality is a well-documented conservation concern, the integration of scientific findings into policy and infrastructure planning has not kept pace with the scale of road expansion and its ecological impacts. While extensive empirical evidence demonstrates the detrimental effects of roads on wildlife populations, there remains a substantial disconnect between available data and its application in environmental impact assessments (EIAs), spatial planning, and the design of protected areas—especially in regions of high biodiversity value (Silva et al. 2020). This gap is particularly troubling for already imperiled species and highlights the urgency of aligning conservation policies with global frameworks such as the Post-2020 Global Biodiversity Framework and the Sustainable Development Goals (Xu et al. 2021).

Although some progress has been made in identifying effective mitigation measures, implementation remains inconsistent and often suboptimal. For example, wildlife underpasses are frequently promoted as essential infrastructure, yet few are rigorously evaluated in terms of their efficacy across different habitats and species (Marcelino et al. 2025). Without context-sensitive assessment, such measures may fail to achieve intended conservation outcomes. Effective policy must extend beyond the mere installation of infrastructure to include long-term monitoring and the incorporation of roadkill data into broader urban and regional planning frameworks. Integrating habitat connectivity into planning processes can reduce wildlife-vehicle collisions while preserving ecological corridors (Martínez et al. 2024).

Moreover, the spatial mismatch between roadkill hotspots and the location of mitigation measures necessitates more informed decision-making. Enhanced ecological research, combined with adaptive infrastructure design, can help address this issue (Eberhardt et al. 2013; Martins et al. 2023). Citizen science offers a valuable solution to address data deficiencies, providing broad spatial coverage and improving data resolution (Pawgi et al. 2024). In addition to facilitating data collection, such community-based initiatives enhance public engagement and political support, which are crucial for long-term conservation investments (Chyn et al. 2024).

Ultimately, advancing policy integration requires a coordinated, multi-scalar approach that addresses both global conservation targets and local ecological contexts. Recognizing regional variability in roadkill patterns is essential for designing locally adapted strategies grounded in empirical evidence and cross-sector collaboration (Heigl et al. 2017; Balčiauskas et al. 2024). Establishing institutional mechanisms that embed biodiversity considerations into transportation planning is vital to ensure that future infrastructure development proceeds without undermining ecological integrity or conservation outcomes.

Limitations and methodological considerations

This bibliometric analysis presents several limitations that may influence the interpretation of its findings. The exclusive use of the Scopus database, while comprehensive, may omit relevant studies indexed elsewhere (Lemeshow et al. 2005), thereby skewing geographic and thematic coverage. Limiting the dataset to research articles further excludes grey literature and local case studies, which often contain valuable insights not captured in peer-reviewed publications (Benzies et al. 2006). Additionally, persistent taxonomic and geographic biases—such as the underrepresentation of invertebrates and research from Africa and Southeast Asia—may restrict the generalizability of the results. The reliance on citation-based metrics also tends to favor widely cited work, potentially overlooking emerging or context-specific research that holds long-term value (Coryn 2006). These constraints suggest that bibliometric analysis should be complemented with other approaches, such as systematic reviews or field-based assessments, to provide a more balanced and inclusive understanding of wildlife roadkill research.

Conclusion

This comprehensive bibliometric review underscores the substantial growth and evolving scope of global wildlife roadkill research, highlighting its increasingly critical role in biodiversity conservation amid the rapid expansion of road infrastructure. The findings reveal a consistent rise in scholarly output since the 1960s, with a marked acceleration after 2000, largely driven by interdisciplinary collaborations and significant contributions from North America and Europe. In contrast, biodiversity-rich regions such as Southeast Asia and Africa remain underrepresented. The literature predominantly focuses on mammals and reptiles, while amphibians, invertebrates, and critically endangered species receive comparatively limited attention, reflecting persistent taxonomic biases. Although mitigation strategies—such as wildlife crossings and fencing—are frequently examined, their effectiveness, particularly for small-bodied species, has not been systematically assessed. Emerging technologies, including artificial intelligence and citizen science, offer promising tools for data collection and analysis but require wider adoption and integration. This study calls for more inclusive and taxonomically comprehensive research, enhanced policy integration, and targeted conservation efforts in high-risk regions to effectively address the ecological and socio-economic consequences of wildlife road mortality.