Using Snake Roadkill Patterns to Indicate Effects of Climate Change on Snakes in Taiwan

Abstract

This study investigates the impacts of climate change on snake behavior and distribution in Taiwan by analyzing roadkill data from the Taiwan Roadkill Observation Network (TaiRON). Focusing on data from 2012 to 2019, the analysis reveals temporal and spatial changes in snake roadkill patterns, shedding light on the ecological effects of a warming climate. From 2012 to 2019, the number of snake roadkill events exhibited a rising trend, particularly during peak activity months from May to October, which accounted for over 70% of annual cases. However, a notable increase was observed in January, traditionally a low-activity period, with roadkill numbers rising 14.9-fold and proportions increasing nearly 6-fold over the study period. This shift suggests that warmer winters are extending the active period for snakes, potentially altering their seasonal behaviors. Spatially, snake roadkill numbers showed a northward and upward migration, reflecting a response to rising temperatures and habitat shifts to higher-altitude regions. These migratory trends, while adaptive, expose snakes to heightened roadkill risks in newly occupied habitats. The findings underscore the potential of roadkill data as a robust ecological monitoring tool for understanding species responses to climate change. By integrating citizen science with ecological and spatial analyses, this research highlights the critical role of environmental changes in driving snake activity and distribution shifts. This study emphasizes the need for climate-adaptive conservation strategies, including road design improvements and biodiversity-focused policies, to mitigate roadkill risks and safeguard snake populations. These insights contribute to broader efforts in ecological conservation and the formulation of evidence-based policies to address the impacts of climate change on cold-blooded animals.

1. Introduction

Climate change is one of the most challenging environmental issues globally, and its impacts on ecosystems and biodiversity are becoming increasingly evident (Pecl et al., 2017 [1]; Weiskopf et al., 2020 [2]). Rising temperatures, changing rainfall patterns, and the frequency of extreme weather events have led to changes in the distribution ranges, breeding cycles, and activity patterns of many species. For instance, under warm winter conditions, many ectothermic animals (such as amphibians and reptiles) experience extended active periods, further affecting their population structures and ecological interactions (Rutschmann et al., 2024 [3]; Williams et al., 2015 [4]). However, existing research on these phenomena largely relies on meteorological data and ecological models, which can reveal macro trends but lack direct evidence regarding species behavior and activity details.

In recent years, roadkill data (wildlife–vehicle collisions, WVCs) have emerged as a new ecological monitoring tool and have gradually gained attention (Schwartz et al., 2020 [5]). Such data not only reflect human activities’ interference with nature but also provide important clues for observing seasonal behaviors and spatial distribution changes in wildlife. For ectothermic animals (such as snakes), whose physiological activities are highly sensitive to temperature changes, roadkill data can accurately capture how their behavior patterns adjust due to changes in environmental conditions. For example, studies by Goldenberg et al. (2024) [6] and Nori et al. (2014) [7] have indicated that the timing and frequency of snake roadkill events are highly influenced by climate change drivers. Therefore, analyzing snake roadkill data can serve as a unique and low-cost method to explore the profound impacts of climate change on ectothermic animals.

In addition, human activities such as religious practices can further influence the patterns observed in roadkill data (Mo & Mo, 2022 [8]; Viviano et al., 2023 [9]). For example, some organizations in the Changbai Mountains have been reported to release snakes during Buddhist ceremonies, aiming to promote spiritual merit (Wang et al., 2022 [10]). However, these practices have inadvertently contributed to an increase in the number of road-killed snakes in the surrounding areas. This highlights the complex interplay between cultural traditions and ecological dynamics, where well-meaning actions can unintentionally disrupt local ecosystems and exacerbate wildlife mortality.

The release of snakes not only amplifies roadkill rates but also raises concerns about potential ecological impacts, particularly when the released animals include non-native species. These cases underscore the importance of roadkill data as a tool for detecting such events and monitoring the spread of potentially invasive species (Ham et al., 2022 [11]). By capturing the spatiotemporal patterns of snake activity, roadkill records provide valuable insights into how both natural and anthropogenic factors, including religious practices, influence wildlife populations under changing climatic conditions.

1.1. The Potential of Roadkill Data in Climate Change Research

Roadkill data, characterized by low costs, extensive coverage, and long-term accumulation, have emerged as a powerful tool for ecological monitoring (Shilling et al., 2015 [12]). They provide critical insights into the interaction between wildlife and road environments and reflect behavioral responses to climate change (Capula et al., 2014 [13]). By capturing temporal and spatial patterns of animal activity, roadkill records can effectively track shifts in behavior and distribution, particularly for ectothermic animals sensitive to environmental changes. Leveraging citizen science platforms enhances data collection, enabling researchers to explore these patterns with increased precision and efficiency (Oddone Aquino & Nkomo, 2021; Périquet et al., 2018; Schwartz et al., 2020; Swinnen et al., 2022 [5,14,15,16]). The temperature trends highlighted in this study are sourced from TCCIP, which directly links these changes to climate change. This provides a strong contextual basis for discussing the observed ecological impacts as a response to broader climate dynamics.

Taiwan, as a biodiversity hotspot, has established platforms such as the “Taiwan Roadkill Observation Network” (TaiRON) to collect roadkill events of wildlife across the island. These data not only provide a solid foundation for studying the ecological behaviors of ectothermic animals like snakes but also address the shortcomings of traditional climate change research in observing micro-level behaviors. However, current research on snake roadkill is relatively scarce, especially in the context of climate change, and the potential behavioral adjustment mechanisms and ecological effects have not been fully revealed.

1.2. The Impact of Climate Change on Snake Distribution and Roadkill Events

As ectothermic animals, snakes’ metabolism and behavior are easily affected by temperature changes. Relevant studies have shown that warm winter conditions can shorten the hibernation period of snakes and even lead to earlier activity. For example, research by Capula et al. (2014) [13] in the Mediterranean region indicated that the peak timing of snake roadkill was advanced by about a month due to warm winters. This behavioral change may stem from the restructuring of foraging, breeding, and migration patterns influenced by climate change. At the same time, Nori et al. (2014) [7] pointed out that the distribution patterns of venomous snakes in South America have largely changed due to rising temperatures, with habitat suitability at lower elevations declining while higher elevations have become more suitable, further reflecting the driving effects of climate on snake distribution ranges.

Based on the above research findings, the impacts of climate change on snakes can be summarized as follows:

(1)

Extended activity periods and seasonal changes: When winter temperatures rise, snakes may shorten their hibernation periods and even become active during previously less active months.

(2)

Northward distribution shifts and high-altitude expansion: Global warming may prompt snakes to migrate to cooler high-altitude or northern regions to adapt to new environmental conditions.

(3)

Increased frequency of road interactions: As activity ranges expand or active times lengthen, snakes’ opportunities for contact with roads also increase, leading to more roadkill events.

This study aims to utilize snake roadkill data provided by the TaiRON platform to analyze the effects of climate change on snake distribution and behavior patterns in Taiwan. By examining the spatiotemporal patterns of snake roadkill, we aim to reveal their response mechanisms to climate change and provide scientific evidence for future conservation.

2. Introduction to the Taiwan Roadkill Observation Network (TaiRON)

The Taiwan Roadkill Observation Network (TaiRON) is Taiwan’s first citizen science project focused on recording and analyzing wildlife deaths caused by road development. Established in 2011 by zoologists from the Taiwan Endemic Species Research Institute, the project aims to collect and analyze roadkill data from across Taiwan through citizen participation, providing strong support for ecological conservation research (TaiRON, 2024 [17]).

TaiRON’s data mainly come from volunteers across the country, who contribute by uploading information about roadkill incidents via a mobile app or website. Each report typically includes a photo of the animal, along with the time and GPS coordinates of the event, and basic species identification. Since its inception, TaiRON has attracted over 19,000 registered members, with around 5000 active contributors who regularly upload data. Volunteers simply need to enable GPS, take a photo of the roadkill, and upload it to complete the entry. If the carcass is in good condition, volunteers may also send specimens to the Endemic Species Research Institute for further analysis (TaiRON, 2024 [17]).

All observation data are reviewed by the TaiRON project management team to ensure the accuracy of time, location, and species classification. Verified data are marked as “confirmed”, while unverified data are marked as “unconfirmed” (TaiRON, 2024 [17]). These filtered data are widely used in various fields, including the formulation of ecological conservation policies, disease monitoring, and environmental education. For example, TaiRON’s data revealed the impact of roads on snake mortality, prompting the government to take effective measures to reduce roadkill hotspots. Additionally, through the analysis of white-nose-syndrome specimens collected by TaiRON, researchers discovered that rabies had entered Taiwan as early as 2010, raising government awareness of cross-species disease transmission.

Since its establishment, TaiRON has gradually become a key tool for biodiversity conservation in Taiwan, playing an important role in enhancing the coordination between road planning and ecological protection. With TaiRON’s data, researchers can establish accurate roadkill risk models, predict the frequency of wildlife–vehicle collisions, and identify high-risk areas. These research findings not only provide strong scientific evidence for biodiversity conservation in Taiwan but also prompt the government to adjust policies, such as reducing the use of rodenticides in agricultural fields, thereby more effectively reducing wildlife mortality risks and promoting ecological protection.

3. Research Methods

3.1. Snake Roadkill Data

Snake roadkill data refer to records of incidents where snakes are killed due to vehicle collisions. This type of data is of research value for understanding the distribution and behavior patterns of snakes and their interactions with road environments. With the increase in road construction, the risk of snakes crossing roads has also risen. Since 2011, TaiRON, 2024 [17] has established a large database that collects records of wildlife roadkill incidents from across Taiwan, including detailed data on snakes. These data not only aid in studying the interactions between snakes and roads but also serve as a basis for assessing the impacts of climate change on snake distribution and behavior.

TaiRON’s data collection relies on citizen science participation, where volunteers upload details such as snake photos, time of death, GPS coordinates, and species classification through a mobile application or website. The transparency and openness of the data allow TaiRON to efficiently collect roadkill records from all over Taiwan, covering various types of wildlife.

This study utilized snake roadkill data collected from the Taiwan Roadkill Observation Network (TaiRON), a citizen science platform that records wildlife–vehicle collision incidents. Volunteers contribute details such as species identification, GPS location, and photographs via mobile apps or websites. To improve data usability and accuracy, artificial intelligence (AI) techniques and web scraping tools (Instant Data Scraper, 1.2.1) were employed to extract, clean, and organize large datasets. These technologies streamline the removal of duplicate entries, correct species misclassifications, and enhance geospatial data precision.

The cleaned data were converted into a format readable by GIS and analyzed spatially using tools like ArcGIS 10.2. The GPS coordinates of each record were visualized as geographic points on a map, generating distribution maps of snake mortality hotspots, which can also display changes over different years. These hotspot analyses helped identify potential correlations between specific areas and road design, habitat conditions, or climate factors, providing empirical foundations for further research.

The combination of AI and GIS technologies not only enhanced the efficiency and accuracy of processing the snake roadkill data but also supported the establishment of snake mortality risk models, allowing for in-depth analysis of environmental variables (such as road density or climate conditions) on snake roadkill. These research findings are of reference value for formulating ecological protection policies and improving road design.

3.2. Climate Data

The climate data for this study were sourced from the Taiwan Climate Change Projection and Adaptation Knowledge Platform (TCCIP) (TCCIP, 2024 [18]). This platform provides climate change projection data for Taiwan, covering various climate indicators (such as temperature, precipitation, wind speed, etc.) and supports detailed regional analysis. The core indicators of particular interest in this study were the annual time series data of the maximum, average, and minimum temperatures across Taiwan and its four major regions: Northern, Central, Southern, and Eastern. These indicators were crucial for studying the environmental impacts on snake behavior and mortality patterns.

The activity range, breeding season, and survival conditions of snakes are highly influenced by temperature; therefore, temperature changes were a key variable in this study. The historical climate data provided by TCCIP can reveal long-term temperature change trends, aiding in exploring how climate change affects the ecological needs and mortality risks of snakes.

In terms of data processing, this study extracted historical temperature data from the TCCIP platform, covering the annual average, maximum, and minimum temperatures for Taiwan and its four major regions. After extraction, missing values were filled, and the data were standardized to ensure completeness and consistency. All data were organized and classified by year to support the time series comparative analysis with snake roadkill data.

Additionally, this study utilized visualization tools to generate climate time series graphs to display temperature change trends in different regions. These visual results were compared with the spatial distribution patterns of snake mortality events. For example, an increasing temperature trend in certain areas may correlate with an increase in snake mortality events, thereby revealing the potential impacts of climate change on snake behavior and mortality patterns.

The analysis of these climate data provided important support for studying the relationship between snake ecology and climate change as well as offering scientific evidence for future ecological protection policies and biodiversity management.

3.3. Comparative Analysis Methods

This study employed two main methods for analysis: spatial distribution analysis and temporal trend comparison, aimed at exploring the link between snake roadkill patterns and climate change. As an exploratory study, it avoided complex statistical models, focusing instead on identifying potential trends through straightforward spatial and temporal comparisons.

First, GIS technology was used to plot snake roadkill GPS data onto a map of Taiwan, enabling spatial distribution analysis. These data, processed with AI and converted into GIS-compatible formats, generated annual or seasonal maps that highlighted snake mortality hotspots and their changes over time. This visualization helped analyze the distribution of roadkill incidents and their connection to environmental factors like road density and urbanization.

Second, the distribution maps of snake roadkill events were compared with climate data (such as temperature extremes) to examine potential correlations between climate variations and mortality trends. These comparisons, which used maps and time series graphs, allowed the researchers to assess whether shifts in climate may influence snake activity and survival. For example, an increase in roadkill in warmer areas may indicate that temperature changes affect snake behavior.

In summary, this study integrated spatial and temporal analysis to investigate the connection between snake roadkill hotspots and climate change. It provides essential insights into the impacts of climate change on species distribution and ecology, laying the groundwork for future research and policy development.

4. Research Results

We collected data on snake roadkill events from 2011 to 2024. Since TaiRON was established in 2011, the data for that year are relatively incomplete, and the roadkill phenomenon fluctuated during the period from 2020 to 2023 due to the impact of the COVID-19 pandemic. Therefore, we only selected data from 2012 to 2019 for analysis.

According to the analysis in

Table 1. Snake roadkill counts (year by month). Data source: TaiRON (2024) [17].

Year	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec
2012	9	6	28	87	160	162	188	130	124	194	164	36
2013	10	19	58	148	293	366	390	276	350	340	207	64
2014	42	32	69	215	336	508	554	358	377	424	247	46
2015	41	31	98	266	418	515	387	275	440	366	269	114
2016	48	25	43	290	410	482	443	277	254	397	318	247
2017	126	45	54	196	407	440	611	573	452	482	288	116
2018	82	35	116	310	534	326	579	420	440	652	531	238
2019	134	63	117	382	279	368	471	329	334	542	188	60

and

Table 2. Monthly proportion of roadkill incidents relative to the total annual roadkill count. Data source: TaiRON (2024) [17].

Data Source: TaiRON (2024) Year	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec
2012	0.007	0.005	0.022	0.068	0.124	0.126	0.146	0.101	0.096	0.151	0.127	0.028
2013	0.004	0.007	0.023	0.058	0.114	0.142	0.152	0.107	0.136	0.132	0.081	0.025
2014	0.013	0.010	0.022	0.067	0.105	0.158	0.173	0.112	0.118	0.132	0.077	0.014
2015	0.013	0.010	0.030	0.083	0.130	0.160	0.120	0.085	0.137	0.114	0.084	0.035
2016	0.015	0.008	0.013	0.090	0.127	0.149	0.137	0.086	0.079	0.123	0.098	0.076
2017	0.033	0.012	0.014	0.052	0.107	0.116	0.161	0.151	0.119	0.127	0.076	0.031
2018	0.019	0.008	0.027	0.073	0.125	0.076	0.136	0.099	0.103	0.153	0.125	0.056
2019	0.041	0.019	0.036	0.117	0.085	0.113	0.144	0.101	0.102	0.166	0.058	0.018

and Figure 1 and Figure 2, the number of snake roadkill events showed an overall increasing trend from 2012 to 2019. Snake roadkill incidents exhibited a notable upward trend, particularly during the May–October activity peak, which accounted for over 70% of annual cases. Winter roadkill events, traditionally rare, also greatly increased. For example, January cases surged from 9 in 2012 to 134 in 2019, reflecting a 14.9-fold increase. This shift suggests that warmer winters are extending snake activity periods, altering their seasonal behaviors and increasing road mortality risks.

Inter-monthly variations showed fluctuations in the data from May to October, especially in the summer months of July and August, where the number of snake roadkill events largely increased, maintaining a relatively stable proportion at a high level (around 15%). This suggests that environmental conditions such as high temperatures and rainfall may promote more frequent snake activity, thereby increasing the chances of crossing roads and raising the risk of roadkill. In contrast, non-peak months like January and October also showed a gradually increasing trend, indicating that the seasonal distribution of snake activity may be changing with environmental shifts. Notably, in January, the number and proportion of snake roadkill events greatly increased, indicating that the low activity period in winter is being affected by climate change, leading to an extension of the snakes’ active period.

The changes in roadkill numbers and proportions in January exhibit characteristics reflecting the potential impact of long-term environmental changes on snake behavior. First, from the perspective of the increase in roadkill numbers, the growth in January is particularly notable, rising from 9 cases in 2012 to 134 cases in 2019, an increase of 14.9 times, indicating a rise in winter snake roadkill events. Secondly, the proportion of roadkill events in January increased from 0.007 in 2012 to 0.041 in 2019, a growth of 5.86 times, suggesting that the activity level of snakes in winter is gradually strengthening. This indicates that although winter is typically a low activity period for snakes, the ecological characteristics of January seem to be becoming more pronounced. Compared to other months, the increase in January was at a higher level in both quantity and proportion. For example, the increase in numbers during peak activity months like July and August was about 2.5 times, with proportions remaining relatively stable or only slightly fluctuating. This growth pattern suggests that the changes in the data during the winter months (especially January) were greater, potentially reflecting underlying changes in snake behavior during winter. Additionally, the changes in the January roadkill data may also be related to climate change, as global warming could extend or broaden the activity time and range of snakes in winter, thereby increasing winter roadkill events. Overall, the roadkill data for January reflect changes in seasonal activity and the impact of long-term ecological environmental changes on snake behavior, particularly indicating that the activity patterns of snakes in winter may be becoming more frequent, a phenomenon that warrants further ecological and climate-linked research.

The spatial distribution of snake roadkill events showed a clear trend of northward and upward migration to higher mountainous areas, which may be closely related to climate change. As the climate warms, temperatures in low-altitude areas may exceed the suitable range for many snake species, forcing them to gradually migrate to cooler, higher-altitude regions. At the same time, climate change has made previously unsuitable high-altitude and northern areas increasingly habitable (see Figure 3, Figure 4 and Figure 5), providing new habitats for snakes. However, this migration process increases the frequency of snakes crossing roads, thereby raising the risk of roadkill in high-altitude and northern areas. Therefore, the changes in the spatial distribution of snakes not only reflect the impact of climate change but also increase the threat of road traffic to their survival, especially in these newly migrated-to regions.

5. Conclusions

This study underscores the utility of roadkill data as a robust ecological monitoring tool, revealing impacts of climate change on snake behavior and distribution in Taiwan. Key findings include the extension of active periods into winter months, northward and upward habitat shifts, and increased roadkill risks in these newly occupied areas. By integrating citizen science with advanced analytical methods, this research highlights the need for adaptive conservation strategies, such as wildlife-friendly road designs and targeted habitat protection. While this study focused on temperature trends, it is important to note that these trends were derived from TCCIP data, which attribute them to climate change. This supports the use of the term “climate change” in our discussion. Future research could expand on other aspects of climate change, such as altered precipitation or extreme weather patterns, to complement these findings.

Climate change has had an impact on the distribution range and activity patterns of snakes. Not only has it changed the breeding seasons and foraging behaviors of snakes, but it has also led to their migration to northern and high-altitude areas. These behavioral changes are directly related to the increased frequency of snake encounters with roads, further exacerbating roadkill incidents. Particularly under warm winter conditions, snakes become active earlier, increasing their risk on the roads.

The research findings highlight the long-term impacts of climate change on the ecological behaviors of snakes, especially the changes in their winter activity. As temperatures rise, the hibernation period of snakes may shorten, leading to more frequent activity during previously low-activity months. These changes need to be considered in future ecological conservation plans and pose new challenges for road design and policies aimed at reducing roadkill risks.

In summary, this study not only demonstrates how climate change affects the activity and mortality patterns of snakes but also indicates the importance of roadkill data in monitoring species behavioral changes, advancing ecological conservation, and formulating relevant policies. In the future, further integration of roadkill data with climate models should be utilized to explore the impacts of climate change on other cold-blooded animals and provide more specific scientific guidance for enhancing biodiversity conservation.

6. Research Limitations

This study aimed to reveal the impact of climate change on the distribution and behavioral patterns of snakes in Taiwan through the analysis of snake roadkill data. However, there were some issues and limitations encountered during the execution of the research that need to be addressed or further explored in future studies.

• Incompleteness of data: Although the TaiRON platform has established a relatively complete roadkill database, the reliability and scope of the data may fluctuate due to its reliance on citizen science. In particular, data from certain regions may be insufficient, affecting the representativeness of the overall analysis results. For example, roadkill incidents in remote areas are harder to document, leading to potential biases in the research findings.
• Limitations of climate data: While this study used climate data provided by TCCIP to explore the effects of climate change on snake behavior, there are certain limitations in the analysis of climate data. For instance, the impacts of climate change are often reflected in long-term trends, but changes in snake behavior may be influenced by shorter-term or localized climate fluctuations, which may not be fully captured in the data.
• Complexity of snake physiology and behavior: As ectothermic animals, snakes’ activity patterns are influenced by various environmental factors such as temperature, humidity, and food availability. Therefore, climate change may have complex interactions with snake behavior and breeding seasons, which have not been thoroughly explored in this study. For example, snakes’ adaptation to environmental changes may depend not only on temperature but also on ecological factors such as food chains and predator activities.
• Simplification of models and analytical methods: This study employed spatial distribution analysis and temporal trend comparison methods to visually explore snake roadkill incidents. Although these methods are intuitive and have good practical applicability, they lack more precise statistical modeling, such as regression analysis or time series analysis, which could further reveal the quantitative relationships between environmental variables and snake behavior.
• Impact of human disturbance factors: In addition to climate change, snake activity and roadkill incidents may also be influenced by human factors such as road construction, traffic flow, and agricultural activities. These anthropogenic disturbance factors may obscure or confound the effects of climate change on snake behavior, and future research needs to consider these factors more deeply.
• The COVID-19 lockdown: The unique characteristics of reduced human and vehicle movement during the pandemic complicate direct comparisons with earlier data. We emphasize that these differences highlight the need for future research to systematically analyze the impacts of traffic volume on roadkill patterns under varying conditions.

In conclusion, this study provides preliminary evidence regarding the impact of climate change on snake behavior, but it also faces issues such as incomplete data, temporal and spatial variability, and simplification of analytical methods. Future research can improve upon the data coverage and analytical methods and consider other ecological factors to more accurately capture the effects of climate change on snakes and other cold-blooded animals.