Factors associated with deer vehicle collisions in South Carolina (SC), USA

Factors associated with deer vehicle collisions in South Carolina (SC), USA | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Factors associated with deer vehicle collisions in South Carolina (SC), USA Sanjeev Sharma, Puskar Khanal This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4504927/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Every year in the United States, approximately 1.5 million deer–vehicle collisions (DVCs) occur, resulting in >200 human fatalities, >29,000 human injuries, 1.3 million deer fatalities, and >1 billion dollars’ worth of property damage. However, there was a lack of studies implementing machine learning techniques from the state level to evaluate the factors affecting DVCs. Data on DVCs on roads are valuable to reduce the occurrence of DVCs and to assist in planning. We utilized the data from 2018 to 2021 provided by Department of Transportation and Safety. The finding suggests that DVCs occurred more frequently near the developed areas, cultivated land and woody wetland and in October, from 8:00 PM to 10:00 PM and 6:00 AM to 8:00 AM. The accuracy scores 0.56 and 0.63 were obtained from machine learning and artificial neural network, opening the door for future research on more factors that affect DVCs. Machine Learning GIS Logistic Regression Land Cover Wildlife Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Fatalities and Injuries due to traffic accidents with wildlife have a great impact on the wildlife ecology and society. Wildlife-vehicle collisions (WVCs) incur significant financial costs, often with unknown actual numbers. The WVC rates are also estimated to be significantly high, for instance, 1.5 million collisions with deer in the United States (Sullivan 2011 ), four million in Belgium (Morelle et al. 2013 ), 1.5 million collisions with deer in the United States (Sullivan 2011 ). However, the collisions are mostly counted in small regions, nationwide or international numbers are estimated to be several times higher than the reported collision numbers (Gkritza et al., 2010 ; Steiner et al., 2014 ). Deer-Vehicle Collisions (DVCs) may result in significant risk to human safety, deer mortality, and expensive vehicle damage (Finder et al. 1999 ). Conover et al. ( 1973 ) found that 92% of deer hit by a vehicle die. In the United States, estimates suggest that annually > 1 million drivers are involved in DVCs, with more than 29,000 human injuries and 200 human fatalities (Conover 1997 ) resulting in > $ 1 billion in vehicle damage (Conover 1997 ). Studies have explored human perceptions and attitudes with respect to deer-related vehicle accidents (Stout et al., 1973 ;Marcoux & Riley, 2010 ). A quantitative human dimensions study, conducted within the city of Winnipeg, Canada, investigated resident opinions and tolerances toward the urban deer population; it identified DVCs as Winnipeg residents’ top deer-related concern (McCance 2010 ). Numerous studies have investigated factors correlated with DVCs. The incidence of DVCs has been attributed to deer density (Widenmaier & Fahrig, 2005 ;Sudharsan et al., 2009 ); season (Sudharsan et al. 2009 ); time of day (Marcoux et al. 2005 ); habitat type near roadways (Hussain et al. 2007 ); number of buildings (i.e., degree of development) near roadways (McShea et al. 2008); traffic volume (Sudharsan et al. 2009 ); and roadway speed limits (Sudharsan et al. 2009 ). Several management techniques have been suggested to mitigate, with varying success, the frequency of DVCs. Some of these techniques, aimed at reducing the occurrence of white-tailed deer on roadways, include deer population reduction (Brown et al., 2000 ; Riley et al., 2003 ); fencing (Clevenger & Waltho 2000 ); underpasses and overpasses (Clevenger & Waltho 2000 ); intercept feeding and whistles or repellents and reflectors. Other techniques have been aimed at improving a driver’s ability to respond to deer on roadways. These techniques encompass measures like reduced speed limits (Allen & Mccullough 1976), habitat modification, improved lighting, and warning signs (Putmam 1997 ). Prediction models are few till now for WVCs. Santos et al. ( 2018 ) developed a Bayesian hierarchical occupancy model that estimated WVC risk, especially in agricultural, open habitats, and within four-lane road sections. Visintin et al. ( 2017 ) predicted WVCs with the main determinants being traffic volume, traffic speed, and species occurrence. These studies explicitly named the advantage of predictive studies, as the analysis decimate the need for a broad data collection. Bíl et al. ( 2016 ) applied hot spot analyses such as the kernel density distribution and calculated the density of accidents along certain road section lengths. However, the need for the entire state is lacking to prioritize based upon County level. Malo et al. ( 2004 ) use a Poisson distribution, Valero et al. ( 2015 ) applied a nearest-neighbor hierarchical clustering, and Tanner et al. ( 2017 ) used a generalized linear mixed model. Furthermore, (Liu et al., 2018 ; Seo et al., 2015 ) applied regression analysis to determine the influence of environmental factors such as the roadside land use or seasonal effects. Other types of traffic accidents used data mining techniques such as machine learning (ML) for identifying the accident risk. Yang et al. ( 2021 ) investigated the severity of road accident injuries in Alabama, United States. A neural network was trained including variables such as light conditions and traffic speed to detect safer driving patterns, which may reduce fatalities and injuries by up to 40%. Komol et al. ( 2021 ) chose machine learning based classification approaches for modeling injury severity of the vulnerable road using k-nearest neighbor, support vector machine, and random forest. They found that motorcyclists have an especially high crash severity. Chen & Wu ( 2014 )used random parameters bivariate ordered probit model, they showed correlations between two drivers’ injuries such as driver age, gender, vehicle, airbag or seat belt use or traffic flow. The applications and the comparisons of diverse machine learning techniques to model traffic accidents show that these approaches are suitable for accident risk prediction. However, for a risk prediction, it is not decisive to know the impact of an individual factor, but to develop a decision model considering the environmental factors and the learning characteristics of machine learning techniques. Machine learning may also enlarge the knowledge about the accidents with wildlife, as already applied for other road accident types. Logistic regression was applied to the urban areas(Found & Boyce 2011 ) for model building, however the application for state level is lacking. Several studies were done previously for the factors associated with DVCs, most of were for county level or district level of the country. Therefore, there is a lack of studies on a large scale such as state level. The major objectives of this study were to 1) identify the factors affecting deer using machine learning and neural network approach, and 2) identify the area having higher density of DVCs based upon kernel density. Methods Study Area The study was conducted in South Carolina (SC), USA (see Fig. 1 ). SC is a state in the coastal southeastern region of the United State. In the summer, SC is hot and humid, with daytime temperatures averaging between 30–34°C in most of the state and overnight lows averaging 21–24°C on the coast and from 19–23°C inland. Winter temperatures are much less uniform in SC. Coastal areas of the state have very mild winters, with high temperatures approaching an average of 16°C and overnight lows around 5–8°C. It has an average elevation of 106.68m. It has an area of 77,856.9 km 2 . Data Collection DVCs data from 2018 to 2021 (see Table 1 ) were downloaded from the department of transportation and highway safety website. Shapefiles of the State, road, water and water area of USA were downloaded from https://www.census.gov/cgi-bin/geo/shapefiles/index.php , and National Land Cover Dataset (NLCD ) 2021 was added from the living atlas in ArcGIS Pro. These all data were imported to the ArcGIS Pro. Table 1 Data downloaded for analysis. Data Website NLCD 2021 Living Atlas inside ArcGIS Pro (which is inbuilt inside ArcGIS) County Tiger line/Shapefile Road Tiger line/Shapefile DVCs Points Department of Transportation and Highway Safety Water (Creek or streams) Tiger line/Shapefile Water Area (Pond or Lake) Tiger line/ Shapefile Railroad Tiger line/Shapefile Data Analysis and Machine Learning The data was analyzed using ArcGIS Pro, Python 3.11 and R (v. 3.6.1 R Core Team 2023) to find factors associated with DVCs. Different factors were analyzed (see Table 1 ). Different data (see Table 1 ) were imported in ArcGIS Pro and converted into common coordinate system NAD 1983 UTM Zone 17N. The SC state was exported from state shapefile of USA obtained from TigerLine shapefile website. The roads network shapefile of each county of SC was imported in the ArcGIS. They were combined and later dissolved to obtain the single layer of road network. Similarly, the water network shapefiles of each county were combined and dissolve to form a single water layer. The shapefile of rail, water, and water area obtained from the TigerLine were directly imported in ArcGIS pro. The point data of DVCs were buffered to 30 and 100m to extract the landcover information from each point. The extract by mask tool was used to extract the landcover information. For comparison with the DVC points, 7204 random points were generated using create random points tool in ArcGIS Pro and buffered to 30 and 100m (Found & Boyce 2011 ). These random points were created along the roads of the entire SC state to reduce the variation between DVCs and random points. Those buffers from DVCs points and random points were compared for landcover to find the effect of landcover on the collision(Found & Boyce 2011 ). The shapefiles of rail, water, and water area were vector, so were converted to raster by the Euclidean distance tool to measure the nearest distance from the DVCs points and random points. Extract multi-value to points tool was used to extract the landcover from the DVCs point and randomly generated points along with distance to water, distance to water area, and distance to rail. Kernel Density tool was used to know the density of road network and DVCs points. Correlation test was conducted between the DVCs points and road network for each county. Spatial join tool was used to join the roads and DVCS. Roads within 5m distance from the point of collusion was summarize the maximum collision on specific roads. Machine Learning Python 3.11 version was used for the data analysis and sublime text was used as a code editor. After the data obtained from extract multi-values to points, DVCs points were given value of 1 and randomly generated points were given the value of 0. The overlapping points were removed, and the analysis was done. Missing values were handled by giving the mean value of specific column. Random Forest (RF) and Artificial Neural network (ANN) were applied. For neural network batch size of 32 and number of epochs10 was included in the model and for random forest number of estimator was 100 and random state was 42. Two of the dependent variable’s road type and landcover were categorical so dummy variables were created to minimize the fixed effect. To evaluate the results, accuracy score and ROC curve were generated to validate the model. Result and Discussion DVCs data from 2018 to 2021 of SC was analyzed and we found that the DVC was high in the month of October, between the time of 8:00PM to 10:00PM and 6:00AM to 8:00AM (see Fig. 2 ), and in main roads compared to other roads. Higher collision (see Fig. 2 ) in the October month compared to the other months was found from the data collected between 2018 to 2021 of SC from department of transportation and vehicle safety. This finding is consistent with the months that deer are actively breeding (Goulden 1981 ). During fall, males are travelling a lot to find females, and males looking to breed (Beier and Mccullough 1990 ). More findings regarding DVCs occurrences during the fall months were documented in Iowa by (Hubbard et al. 2000 ) and in Edmonton, Alberta, Canada, by (Ng et al. 2008 ). Maximum collision was found (see Fig. 3 ) at 8:00PM to 10:00PM and 6:00AM to 8:00AM compared to other time of the day. This is not surprising given that white-tail deer are most active during dawn and dusk (Beier & Mccullough, 1990 ). Our results are consistent with the earlier works of (Hubbard et al. 2000 ; Nielsen et al. 2003 ). Higher collision (see Fig. 4 ) was found on the main roads and secondary roads compared to interstate highway, urban roads, country roads and other roads. This tells that while driving on main highway driver should be extra careful of collision due to deer. The traffic is also high on main roads and these roads are mainly close to developed areas on which deer prefer to go. Proper management around these roads will help in reducing the DVCs. Raster cells counts were summarized for all landcover within 30m and 100m from DVCs and randomly generated points. These different buffers for the DVCs and randomly generated points were used to compare the landcover (see Table 2 ). The difference of the raster counts on the DVCs and random points within 30m and 100m showed that higher DVCs occurred in developed areas, woody wetland and cultivated land which is like the result found by (Sudharsan et al. 2009 ; Farrell & Tappe 2007). Sudharsan et al. ( 2009 ) found that a higher probability of DVCs occurs in developed areas and crop or grasslands. DVCs were associated with grasslands and wooded areas (Myers et al. 2008 ). Researchers have found a higher occurrence of DVCs near woodlots (Hussain et al. 2007 ). Table 2 Total surrounding land cover of DVC points and randomly generated points buffered at 30m. Class Name DVC_Buffer30 Random_Buffer_30 Open Water 12 29 Developed Open Space 6935 8578 Developed Low Intensity 6697 5586 Developed Medium Intensity 2443 1580 Developed High Intensity 468 338 Barren Land 37 53 Deciduous Forest 295 670 Evergreen Forest 1374 1907 Mixed Forest 193 319 Shrub/Scrub 255 303 Grassland/Herbaceous 204 376 Pasture/Hay 911 1100 Cultivated Crops 912 882 Woody Wetlands 1018 740 Emergent Herbaceous Wetlands 73 43 Kernel density (see Fig. 5 ) was applied to see the effect of roads network on DVCs. Counties with denser road network has more DVCs compared to the lighter roads. Clustered distribution of DVCs was observed compared to random. Correlation was performed between the road network and DVCs points among counties, (r = 0.806, p < 0.05) was obtained from the correlation test. From the machine learning and neural network approach, we obtained an accuracy score of 0.63 from ANN and 0.56 from the RF. The ROC curve was also generated for the validation (see Fig. 6 ). Conclusion We found that the developed areas have higher DVCs compared to other land cover types. It may be due to the flow of traffic is high in the developed areas. We estimated October month, the time between 8pm to 10pm and 6am to 8am, and main roads are main cause for the DVCs. There is a positive correlation between road network and DVCs among counties. Our results obtained from machine learning using RF and ANN showed an accuracy of 56% from RF and 63% from ANN. Which tells we need a more variable other than water, water area, rai, landcover, and road type. To the best of my knowledge, this model is the first for predicting DVCs across the entire state. This model will provide the baseline for predicting DVCs based on the variables used. Proper management of landcover types and the protective major around the developed areas are the key factor for reducing the DVCs. Fencing around developed areas or speed limit around the major DVCs area may help in mitigation DVCs. Management strategies should prioritize the time of day and month when a higher number of DVCs occur. Localized management near key habitat types that are associated with higher numbers of DVCs should be considered. Building curvier road, speed limit, fencing around the prime habitat of DVCs may help in reducing DVCs. Further data on the temperature, time of a day and other factors may help in predicting better with higher accuracy. Declarations Author Contribution Sanjeev Sharma reports was provided by Clemson University College of Agriculture Forestry and Life Sciences. Sanjeev Sharma reports a relationship with Clemson University College of Agriculture Forestry and Life Sciences that includes: graduate student. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper References Beier P, Mccullough DR (1990) Factors Influencing. White-Tailed Deer Activity Patterns and Habitat Use Bíl M, Andrášik R, Svoboda T, Sedoník J (2016) The KDE + software: a tool for effective identification and ranking of animal-vehicle collision hotspots along networks. Landsc Ecol 31:231–237. https://doi.org/10.1007/s10980-015-0265-6 Brown TL, Decker DJ, Riley SJ et al (2000) The future of hunting as a mechanism to control white-tailed deer populations. Wildl Soc Bull 28:797–807 Chen X, Wu S (2014) Examining Patterns of Animal–Vehicle Collisions in Alabama, USA. Human-Wildlife Interact 8:235–244. https://doi.org/10.26077/p18k-1089 Clevenger AP, Waltho N (2000) Factors influencing the effectiveness of wildlife underpasses in Banff National Park, Alberta, Canada. Conserv Biol 14:47–56. https://doi.org/10.1046/j.1523-1739.2000.00099-085.x Conover MR (1997) Monetary and intangible valuation of deer in the United States. Wildl Soc Bull 25:298–305 Conover MR, Pitt WC, Kessler KK et al (1973) Review of Human Injuries, Illnesses, and Economic Losses Caused by Wildlife in the United States Finder RA, Roseberry JL, Woolf A (1999) Site and landscape conditions at white-tailed deer/vehicle collision locations in Illinois. Landsc Urban Plan 44:77–85. https://doi.org/https://doi.org/10.1016/S0169-2046(99)00006-7 Found R, Boyce MS (2011) Predicting deer–vehicle collisions in an urban area. J Environ Manage 92:2486–2493. https://doi.org/https://doi.org/10.1016/j.jenvman.2011.05.010 Gkritza K, Baird M, Hans ZN (2010) Deer-vehicle collisions, deer density, and land use in Iowa’s urban deer herd management zones. Accid Anal Prev 42:1916–1925. https://doi.org/https://doi.org/10.1016/j.aap.2010.05.013 Goulden HD (1981) The white-tailed deer in Manitoba. Winnipeg, Man.: Manitoba Department of Natural Resources, 1981 Hubbard MW, Danielson BJ, Schmitz RA (2000) Factors influencing the location of deer-vehicle accidents in Iowa. J Wildl Manage 64:707–713 Hussain A, Armstrong JB, Brown DB, Hogland J (2007) Land-Use Pattern, Urbanization, and Deer–Vehicle Collisions in Alabama. Human–Wildlife Interact 1:20 Komol MMR, Hasan MM, Elhenawy M et al (2021) Crash severity analysis of vulnerable road users using machine learning. PLoS ONE 16. https://doi.org/10.1371/journal.pone.0255828 Liu Y, Nieuwenhuis M, McCullagh A (2018) The effect of roadside land-use on the occurrence of deer vehicle collisions. 75:8–25 Malo JE, Suárez F, Díez A (2004) Can we mitigate animal-vehicle accidents using predictive models? J Appl Ecol 41:701–710. https://doi.org/10.1111/j.0021-8901.2004.00929.x Marcoux A, Hickling G, Riley S, Winterstein S (2005) Situational and Driver Characteristics Associated With Deervehicle Collisions in Southeastern Michigan. Proceedings, Wildlife Damage Management Conference Marcoux A, Riley SJ (2010) Driver knowledge, beliefs, and attitudes about deer-vehicle collisions in southern Michigan. Source: Human-Wildlife Interactions 4:47–55. https://doi.org/10.2307/24864502 McCance E (2010) Resident opinions concerning urban deer management in the Greater Winnipeg Area, Manitoba, Canada. Library and Archives Canada = Bibliothèque et Archives Canada Morelle K, Lehaire F, Lejeune P (2013) Spatio-temporal patterns of wildlife-vehicle collisions Spatio-temporal patterns of wildlife-vehicle collisions in a region with a high-density road network. Nat Conserv 5:53–73. https://doi.org/10.3897/natureconservation.5.4634 Myers WL, Chang WY, Germaine S et al (2008) An Analysis of Deer and Elk. -Vehicle Collision Sites along State Highways in Washington State Ng J, Nielson C, St. Clair C (2008) Landscape and Traffic Factors Influencing Deer–Vehicle Collisions in an Urban Enviroment. https://doi.org/10.26077/nxw2-yn81 . Human-Wildlife Conflicts 2: Nielsen CK, Anderson RG, Grund MD (2003) Landscape Influences on Deer-. Vehicle Accident Areas in an Urban Environment Putmam RJ (1997) Deer and Road Traffic Accidents: Options for Management. J Environ Manage 51:43–57. https://doi.org/https://doi.org/10.1006/jema.1997.0135 Riley S, Enck J, Curtis P et al (2003) Deer populations up, hunter populations down: Implications of interdependence of deer and hunter population dynamics on management1. Ecoscience 10:455–461. https://doi.org/10.1080/11956860.2003.11682793 Santos RAL, Mota-Ferreira M, Aguiar LMS, Ascensão F (2018) Predicting wildlife road-crossing probability from roadkill data using occupancy-detection models. Sci Total Environ 642:629–637. https://doi.org/https://doi.org/10.1016/j.scitotenv.2018.06.107 Seo C, Thorne JH, Choi T et al (2015) Disentangling roadkill: the influence of landscape and season on cumulative vertebrate mortality in South Korea. Landscape Ecol Eng 11:87–99. https://doi.org/10.1007/s11355-013-0239-2 Steiner W, Leisch F, Hackländer K (2014) A review on the temporal pattern of deer–vehicle accidents: Impact of seasonal, diurnal and lunar effects in cervids. Accid Anal Prev 66:168–181. https://doi.org/https://doi.org/10.1016/j.aap.2014.01.020 Stout RJ, Stedman RC, Decker DJ, Knuth BA (1973) Perceptions of Risk from Deer-Related Vehicle Accidents. Implications for Public Preferences for Deer Herd Size Sudharsan K, Riley SJ, Campa HI (2009) Relative Risks of Deer–Vehicle Collisions Along Road Types in Southeast Michigan. Hum Dimensions Wildl 14:341–352 Sullivan JM (2011) Trends and characteristics of animal-vehicle collisions in the United States. J Saf Res 42:9–16. https://doi.org/https://doi.org/10.1016/j.jsr.2010.11.002 Tanner A, Leroux S, Saunders P (2017) Road Characteristics Best Predict the Probability of Vehicle Collisions with a Non-Native Ungulate. Écoscience 24:1–11. https://doi.org/10.1080/11956860.2017.1292100 Valero E, Picos J, Álvarez X (2015) Road and traffic factors correlated to wildlife–vehicle collisions in Galicia (Spain). Wildl Res 42:25–34 Visintin C, van der Ree R, Mccarthy M (2017) Consistent patterns of vehicle collision risk for six mammal species. J Environ Manage 201:397–406. https://doi.org/10.1016/j.jenvman.2017.05.071 Widenmaier KJ, Fahrig L (2005) Inferring white-tailed deer (Odocoileus virginianus) population dynamics from wildlife. collisions in the City of Ottawa Yang D, Wu Y, Sun F et al (2021) Freeway accident detection and classification based on the multi-vehicle trajectory data and deep learning model. Transp Res Part C Emerg Technol 130:103303. https://doi.org/https://doi.org/10.1016/j.trc.2021.103303 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4504927","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":311194824,"identity":"d9511458-46aa-4b31-92dd-1f686665a7b7","order_by":0,"name":"Sanjeev Sharma","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2klEQVRIiWNgGAWjYJACAyCWY2CH8xOI02LMwEyKFhBIbCBai3z72QMFHyoOp/c3Mx/d+KPiDgM/e44BfkedyUswnHHmcO6Mw2xpt3nOPGOQ7HlDQAtDjoExb1ta7gZmHrPbjG2HGQxuELBFvv+NgfHff2npBsz8327+BGqxJ6SFAajAmLHBJsGAmYftBi/IFglCfrnxxsCw55iNIdAvZkC/HOaROPOsgIDDcswMftRIyPO3Nz+7+aPisBx/e/IG/A5jYGBDcQcPIeUgwPyAGFWjYBSMglEwggEAxaJFwLafh0QAAAAASUVORK5CYII=","orcid":"","institution":"Clemson University","correspondingAuthor":true,"prefix":"","firstName":"Sanjeev","middleName":"","lastName":"Sharma","suffix":""},{"id":311194825,"identity":"6e44b1d7-1bc8-45e2-8659-8a1182c80861","order_by":1,"name":"Puskar Khanal","email":"","orcid":"","institution":"Clemson University","correspondingAuthor":false,"prefix":"","firstName":"Puskar","middleName":"","lastName":"Khanal","suffix":""}],"badges":[],"createdAt":"2024-05-30 19:57:56","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4504927/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4504927/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":58555911,"identity":"870f605d-e184-4570-aad1-efba79e7c00c","added_by":"auto","created_at":"2024-06-18 08:01:01","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":511102,"visible":true,"origin":"","legend":"\u003cp\u003eSouth Carolina (SC) with DVCs points and randomly generated points from 2018 to 2021.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-4504927/v1/b8935a21445da4b218fe34c7.png"},{"id":58555437,"identity":"1d252782-1881-48c8-9f5b-76040d350519","added_by":"auto","created_at":"2024-06-18 07:53:01","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":64933,"visible":true,"origin":"","legend":"\u003cp\u003eDVCs in South Carolina State from 2018 to 2023.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-4504927/v1/1b2f7165735a91f3b2bdaaf5.png"},{"id":58555912,"identity":"e176cdb8-259d-44de-8ac4-bacbaf3ae06f","added_by":"auto","created_at":"2024-06-18 08:01:01","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":75324,"visible":true,"origin":"","legend":"\u003cp\u003eDVCs at different time of the day.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4504927/v1/8622929965982a39c9784c78.png"},{"id":58555910,"identity":"e0869b78-1079-4e02-b42c-8e5e9a417647","added_by":"auto","created_at":"2024-06-18 08:01:01","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":30133,"visible":true,"origin":"","legend":"\u003cp\u003eRoad types within 5m around the collusion points.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-4504927/v1/78a581ace5229f896c49c7e4.png"},{"id":58555436,"identity":"c5f0efbb-1556-4966-92aa-e71eb7135cf7","added_by":"auto","created_at":"2024-06-18 07:53:01","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":328493,"visible":true,"origin":"","legend":"\u003cp\u003eKernel density of the DVCs on the top and road network on the bottom.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-4504927/v1/ba23f71eb8854d9924a35fdb.png"},{"id":58555433,"identity":"0f134c3d-d373-4855-81a0-f2eb47c42a7b","added_by":"auto","created_at":"2024-06-18 07:53:01","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":42832,"visible":true,"origin":"","legend":"\u003cp\u003eThe ROC curves obtained from the RF and ANN to predict the deer-vehicle collisions.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-4504927/v1/72abf83d102fb2a75c088e5b.png"},{"id":76701867,"identity":"18ce97d6-c862-4b91-82e8-d80a76aeb8f0","added_by":"auto","created_at":"2025-02-19 22:01:29","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1384370,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4504927/v1/53ea68a2-22fe-4ebb-a20f-3e50d59f6842.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Factors associated with deer vehicle collisions in South Carolina (SC), USA","fulltext":[{"header":"Introduction","content":"\u003cp\u003eFatalities and Injuries due to traffic accidents with wildlife have a great impact on the wildlife ecology and society. Wildlife-vehicle collisions (WVCs) incur significant financial costs, often with unknown actual numbers. The WVC rates are also estimated to be significantly high, for instance, 1.5\u0026nbsp;million collisions with deer in the United States (Sullivan \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), four million in Belgium (Morelle et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), 1.5\u0026nbsp;million collisions with deer in the United States (Sullivan \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). However, the collisions are mostly counted in small regions, nationwide or international numbers are estimated to be several times higher than the reported collision numbers (Gkritza et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Steiner et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Deer-Vehicle Collisions (DVCs) may result in significant risk to human safety, deer mortality, and expensive vehicle damage (Finder et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e1999\u003c/span\u003e). Conover et al. (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e1973\u003c/span\u003e) found that 92% of deer hit by a vehicle die. In the United States, estimates suggest that annually\u0026thinsp;\u0026gt;\u0026thinsp;1\u0026nbsp;million drivers are involved in DVCs, with more than 29,000 human injuries and 200 human fatalities (Conover \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1997\u003c/span\u003e) resulting in \u0026gt;\u003cspan\u003e$\u003c/span\u003e1\u0026nbsp;billion in vehicle damage (Conover \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1997\u003c/span\u003e). Studies have explored human perceptions and attitudes with respect to deer-related vehicle accidents (Stout et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e1973\u003c/span\u003e;Marcoux \u0026amp; Riley, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). A quantitative human dimensions study, conducted within the city of Winnipeg, Canada, investigated resident opinions and tolerances toward the urban deer population; it identified DVCs as Winnipeg residents\u0026rsquo; top deer-related concern (McCance \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2010\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eNumerous studies have investigated factors correlated with DVCs. The incidence of DVCs has been attributed to deer density (Widenmaier \u0026amp; Fahrig, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2005\u003c/span\u003e;Sudharsan et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2009\u003c/span\u003e); season (Sudharsan et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2009\u003c/span\u003e); time of day (Marcoux et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2005\u003c/span\u003e); habitat type near roadways (Hussain et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2007\u003c/span\u003e); number of buildings (i.e., degree of development) near roadways (McShea et al. 2008); traffic volume (Sudharsan et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2009\u003c/span\u003e); and roadway speed limits (Sudharsan et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). Several management techniques have been suggested to mitigate, with varying success, the frequency of DVCs. Some of these techniques, aimed at reducing the occurrence of white-tailed deer on roadways, include deer population reduction (Brown et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Riley et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2003\u003c/span\u003e); fencing (Clevenger \u0026amp; Waltho \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2000\u003c/span\u003e); underpasses and overpasses (Clevenger \u0026amp; Waltho \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2000\u003c/span\u003e); intercept feeding and whistles or repellents and reflectors. Other techniques have been aimed at improving a driver\u0026rsquo;s ability to respond to deer on roadways. These techniques encompass measures like reduced speed limits (Allen \u0026amp; Mccullough 1976), habitat modification, improved lighting, and warning signs (Putmam \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e1997\u003c/span\u003e). Prediction models are few till now for WVCs. Santos et al. (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) developed a Bayesian hierarchical occupancy model that estimated WVC risk, especially in agricultural, open habitats, and within four-lane road sections.\u003c/p\u003e \u003cp\u003eVisintin et al. (\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) predicted WVCs with the main determinants being traffic volume, traffic speed, and species occurrence. These studies explicitly named the advantage of predictive studies, as the analysis decimate the need for a broad data collection. B\u0026iacute;l et al. (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) applied hot spot analyses such as the kernel density distribution and calculated the density of accidents along certain road section lengths. However, the need for the entire state is lacking to prioritize based upon County level. Malo et al. (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2004\u003c/span\u003e) use a Poisson distribution, Valero et al. (\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) applied a nearest-neighbor hierarchical clustering, and Tanner et al. (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) used a generalized linear mixed model. Furthermore, (Liu et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Seo et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) applied regression analysis to determine the influence of environmental factors such as the roadside land use or seasonal effects.\u003c/p\u003e \u003cp\u003eOther types of traffic accidents used data mining techniques such as machine learning (ML) for identifying the accident risk. Yang et al. (\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) investigated the severity of road accident injuries in Alabama, United States. A neural network was trained including variables such as light conditions and traffic speed to detect safer driving patterns, which may reduce fatalities and injuries by up to 40%. Komol et al. (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) chose machine learning based classification approaches for modeling injury severity of the vulnerable road using k-nearest neighbor, support vector machine, and random forest. They found that motorcyclists have an especially high crash severity. Chen \u0026amp; Wu (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2014\u003c/span\u003e)used random parameters bivariate ordered probit model, they showed correlations between two drivers\u0026rsquo; injuries such as driver age, gender, vehicle, airbag or seat belt use or traffic flow.\u003c/p\u003e \u003cp\u003eThe applications and the comparisons of diverse machine learning techniques to model traffic accidents show that these approaches are suitable for accident risk prediction. However, for a risk prediction, it is not decisive to know the impact of an individual factor, but to develop a decision model considering the environmental factors and the learning characteristics of machine learning techniques. Machine learning may also enlarge the knowledge about the accidents with wildlife, as already applied for other road accident types. Logistic regression was applied to the urban areas(Found \u0026amp; Boyce \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2011\u003c/span\u003e) for model building, however the application for state level is lacking. Several studies were done previously for the factors associated with DVCs, most of were for county level or district level of the country. Therefore, there is a lack of studies on a large scale such as state level. The major objectives of this study were to 1) identify the factors affecting deer using machine learning and neural network approach, and 2) identify the area having higher density of DVCs based upon kernel density.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy Area\u003c/h2\u003e \u003cp\u003eThe study was conducted in South Carolina (SC), USA (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). SC is a state in the coastal southeastern region of the United State. In the summer, SC is hot and humid, with daytime temperatures averaging between 30\u0026ndash;34\u0026deg;C in most of the state and overnight lows averaging 21\u0026ndash;24\u0026deg;C on the coast and from 19\u0026ndash;23\u0026deg;C inland. Winter temperatures are much less uniform in SC. Coastal areas of the state have very mild winters, with high temperatures approaching an average of 16\u0026deg;C and overnight lows around 5\u0026ndash;8\u0026deg;C. It has an average elevation of 106.68m. It has an area of 77,856.9 km\u003csup\u003e2\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eData Collection\u003c/h2\u003e \u003cp\u003eDVCs data from 2018 to 2021 (see Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) were downloaded from the department of transportation and highway safety website. Shapefiles of the State, road, water and water area of USA were downloaded from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.census.gov/cgi-bin/geo/shapefiles/index.php\u003c/span\u003e\u003cspan address=\"https://www.census.gov/cgi-bin/geo/shapefiles/index.php\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, and National Land Cover Dataset (NLCD ) 2021 was added from the living atlas in ArcGIS Pro. These all data were imported to the ArcGIS Pro.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eData downloaded for analysis.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eData\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWebsite\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNLCD 2021\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLiving Atlas inside ArcGIS Pro (which is inbuilt inside ArcGIS)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCounty\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTiger line/Shapefile\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRoad\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTiger line/Shapefile\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDVCs Points\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDepartment of Transportation and Highway Safety\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater (Creek or streams)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTiger line/Shapefile\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater Area (Pond or Lake)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTiger line/ Shapefile\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRailroad\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTiger line/Shapefile\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eData Analysis and Machine Learning\u003c/h2\u003e \u003cp\u003eThe data was analyzed using ArcGIS Pro, Python 3.11 and R (v. 3.6.1 R Core Team 2023) to find factors associated with DVCs. Different factors were analyzed (see Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Different data (see Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) were imported in ArcGIS Pro and converted into common coordinate system NAD 1983 UTM Zone 17N. The SC state was exported from state shapefile of USA obtained from TigerLine shapefile website. The roads network shapefile of each county of SC was imported in the ArcGIS. They were combined and later dissolved to obtain the single layer of road network. Similarly, the water network shapefiles of each county were combined and dissolve to form a single water layer. The shapefile of rail, water, and water area obtained from the TigerLine were directly imported in ArcGIS pro. The point data of DVCs were buffered to 30 and 100m to extract the landcover information from each point. The extract by mask tool was used to extract the landcover information. For comparison with the DVC points, 7204 random points were generated using create random points tool in ArcGIS Pro and buffered to 30 and 100m (Found \u0026amp; Boyce \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). These random points were created along the roads of the entire SC state to reduce the variation between DVCs and random points. Those buffers from DVCs points and random points were compared for landcover to find the effect of landcover on the collision(Found \u0026amp; Boyce \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2011\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe shapefiles of rail, water, and water area were vector, so were converted to raster by the Euclidean distance tool to measure the nearest distance from the DVCs points and random points. Extract multi-value to points tool was used to extract the landcover from the DVCs point and randomly generated points along with distance to water, distance to water area, and distance to rail. Kernel Density tool was used to know the density of road network and DVCs points. Correlation test was conducted between the DVCs points and road network for each county. Spatial join tool was used to join the roads and DVCS. Roads within 5m distance from the point of collusion was summarize the maximum collision on specific roads.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eMachine Learning\u003c/h2\u003e \u003cp\u003ePython 3.11 version was used for the data analysis and sublime text was used as a code editor. After the data obtained from extract multi-values to points, DVCs points were given value of 1 and randomly generated points were given the value of 0. The overlapping points were removed, and the analysis was done. Missing values were handled by giving the mean value of specific column. Random Forest (RF) and Artificial Neural network (ANN) were applied. For neural network batch size of 32 and number of epochs10 was included in the model and for random forest number of estimator was 100 and random state was 42. Two of the dependent variable\u0026rsquo;s road type and landcover were categorical so dummy variables were created to minimize the fixed effect. To evaluate the results, accuracy score and ROC curve were generated to validate the model.\u003c/p\u003e \u003c/div\u003e"},{"header":"Result and Discussion","content":"\u003cp\u003eDVCs data from 2018 to 2021 of SC was analyzed and we found that the DVC was high in the month of October, between the time of 8:00PM to 10:00PM and 6:00AM to 8:00AM (see Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), and in main roads compared to other roads. Higher collision (see Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) in the October month compared to the other months was found from the data collected between 2018 to 2021 of SC from department of transportation and vehicle safety. This finding is consistent with the months that deer are actively breeding (Goulden \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e1981\u003c/span\u003e). During fall, males are travelling a lot to find females, and males looking to breed (Beier and Mccullough \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1990\u003c/span\u003e). More findings regarding DVCs occurrences during the fall months were documented in Iowa by (Hubbard et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2000\u003c/span\u003e) and in Edmonton, Alberta, Canada, by (Ng et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2008\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eMaximum collision was found (see Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) at 8:00PM to 10:00PM and 6:00AM to 8:00AM compared to other time of the day. This is not surprising given that white-tail deer are most active during dawn and dusk (Beier \u0026amp; Mccullough, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1990\u003c/span\u003e). Our results are consistent with the earlier works of (Hubbard et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Nielsen et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). Higher collision (see Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) was found on the main roads and secondary roads compared to interstate highway, urban roads, country roads and other roads. This tells that while driving on main highway driver should be extra careful of collision due to deer. The traffic is also high on main roads and these roads are mainly close to developed areas on which deer prefer to go. Proper management around these roads will help in reducing the DVCs.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eRaster cells counts were summarized for all landcover within 30m and 100m from DVCs and randomly generated points. These different buffers for the DVCs and randomly generated points were used to compare the landcover (see Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The difference of the raster counts on the DVCs and random points within 30m and 100m showed that higher DVCs occurred in developed areas, woody wetland and cultivated land which is like the result found by (Sudharsan et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Farrell \u0026amp; Tappe 2007). Sudharsan et al. (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2009\u003c/span\u003e) found that a higher probability of DVCs occurs in developed areas and crop or grasslands. DVCs were associated with grasslands and wooded areas (Myers et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Researchers have found a higher occurrence of DVCs near woodlots (Hussain et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2007\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eTotal surrounding land cover of DVC points and randomly generated points buffered at 30m.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClass Name\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDVC_Buffer30\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRandom_Buffer_30\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOpen Water\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e29\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDeveloped Open Space\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6935\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e8578\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDeveloped Low Intensity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6697\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5586\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDeveloped Medium Intensity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2443\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1580\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDeveloped High Intensity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e468\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e338\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBarren Land\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e53\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDeciduous Forest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e295\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e670\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEvergreen Forest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1374\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1907\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMixed Forest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e193\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e319\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShrub/Scrub\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e255\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e303\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrassland/Herbaceous\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e204\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e376\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePasture/Hay\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e911\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1100\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCultivated Crops\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e912\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e882\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWoody Wetlands\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1018\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e740\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEmergent Herbaceous Wetlands\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e43\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eKernel density (see Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) was applied to see the effect of roads network on DVCs. Counties with denser road network has more DVCs compared to the lighter roads. Clustered distribution of DVCs was observed compared to random. Correlation was performed between the road network and DVCs points among counties, (r\u0026thinsp;=\u0026thinsp;0.806, p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) was obtained from the correlation test. From the machine learning and neural network approach, we obtained an accuracy score of 0.63 from ANN and 0.56 from the RF. The ROC curve was also generated for the validation (see Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eWe found that the developed areas have higher DVCs compared to other land cover types. It may be due to the flow of traffic is high in the developed areas. We estimated October month, the time between 8pm to 10pm and 6am to 8am, and main roads are main cause for the DVCs. There is a positive correlation between road network and DVCs among counties. Our results obtained from machine learning using RF and ANN showed an accuracy of 56% from RF and 63% from ANN. Which tells we need a more variable other than water, water area, rai, landcover, and road type. To the best of my knowledge, this model is the first for predicting DVCs across the entire state. This model will provide the baseline for predicting DVCs based on the variables used. Proper management of landcover types and the protective major around the developed areas are the key factor for reducing the DVCs. Fencing around developed areas or speed limit around the major DVCs area may help in mitigation DVCs. Management strategies should prioritize the time of day and month when a higher number of DVCs occur. Localized management near key habitat types that are associated with higher numbers of DVCs should be considered. Building curvier road, speed limit, fencing around the prime habitat of DVCs may help in reducing DVCs. Further data on the temperature, time of a day and other factors may help in predicting better with higher accuracy.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eSanjeev Sharma reports was provided by Clemson University College of Agriculture Forestry and Life Sciences. Sanjeev Sharma reports a relationship with Clemson University College of Agriculture Forestry and Life Sciences that includes: graduate student. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBeier P, Mccullough DR (1990) Factors Influencing. White-Tailed Deer Activity Patterns and Habitat Use\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eB\u0026iacute;l M, Andr\u0026aacute;šik R, Svoboda T, Sedon\u0026iacute;k J (2016) The KDE\u0026thinsp;+\u0026thinsp;software: a tool for effective identification and ranking of animal-vehicle collision hotspots along networks. Landsc Ecol 31:231\u0026ndash;237. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10980-015-0265-6\u003c/span\u003e\u003cspan address=\"10.1007/s10980-015-0265-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrown TL, Decker DJ, Riley SJ et al (2000) The future of hunting as a mechanism to control white-tailed deer populations. Wildl Soc Bull 28:797\u0026ndash;807\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen X, Wu S (2014) Examining Patterns of Animal\u0026ndash;Vehicle Collisions in Alabama, USA. Human-Wildlife Interact 8:235\u0026ndash;244. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.26077/p18k-1089\u003c/span\u003e\u003cspan address=\"10.26077/p18k-1089\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClevenger AP, Waltho N (2000) Factors influencing the effectiveness of wildlife underpasses in Banff National Park, Alberta, Canada. Conserv Biol 14:47\u0026ndash;56. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1046/j.1523-1739.2000.00099-085.x\u003c/span\u003e\u003cspan address=\"10.1046/j.1523-1739.2000.00099-085.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConover MR (1997) Monetary and intangible valuation of deer in the United States. Wildl Soc Bull 25:298\u0026ndash;305\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConover MR, Pitt WC, Kessler KK et al (1973) Review of Human Injuries, Illnesses, and Economic Losses Caused by Wildlife in the United States\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFinder RA, Roseberry JL, Woolf A (1999) Site and landscape conditions at white-tailed deer/vehicle collision locations in Illinois. Landsc Urban Plan 44:77\u0026ndash;85. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1016/S0169-2046(99)00006-7\u003c/span\u003e\u003cspan address=\"10.1016/S0169-2046(99)00006-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFound R, Boyce MS (2011) Predicting deer\u0026ndash;vehicle collisions in an urban area. J Environ Manage 92:2486\u0026ndash;2493. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1016/j.jenvman.2011.05.010\u003c/span\u003e\u003cspan address=\"10.1016/j.jenvman.2011.05.010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGkritza K, Baird M, Hans ZN (2010) Deer-vehicle collisions, deer density, and land use in Iowa\u0026rsquo;s urban deer herd management zones. Accid Anal Prev 42:1916\u0026ndash;1925. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1016/j.aap.2010.05.013\u003c/span\u003e\u003cspan address=\"10.1016/j.aap.2010.05.013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoulden HD (1981) The white-tailed deer in Manitoba. Winnipeg, Man.: Manitoba Department of Natural Resources, 1981\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHubbard MW, Danielson BJ, Schmitz RA (2000) Factors influencing the location of deer-vehicle accidents in Iowa. J Wildl Manage 64:707\u0026ndash;713\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHussain A, Armstrong JB, Brown DB, Hogland J (2007) Land-Use Pattern, Urbanization, and Deer\u0026ndash;Vehicle Collisions in Alabama. Human\u0026ndash;Wildlife Interact 1:20\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKomol MMR, Hasan MM, Elhenawy M et al (2021) Crash severity analysis of vulnerable road users using machine learning. PLoS ONE 16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0255828\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0255828\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu Y, Nieuwenhuis M, McCullagh A (2018) The effect of roadside land-use on the occurrence of deer vehicle collisions. 75:8\u0026ndash;25\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMalo JE, Su\u0026aacute;rez F, D\u0026iacute;ez A (2004) Can we mitigate animal-vehicle accidents using predictive models? J Appl Ecol 41:701\u0026ndash;710. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.0021-8901.2004.00929.x\u003c/span\u003e\u003cspan address=\"10.1111/j.0021-8901.2004.00929.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarcoux A, Hickling G, Riley S, Winterstein S (2005) Situational and Driver Characteristics Associated With Deervehicle Collisions in Southeastern Michigan. Proceedings, Wildlife Damage Management Conference\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarcoux A, Riley SJ (2010) Driver knowledge, beliefs, and attitudes about deer-vehicle collisions in southern Michigan. Source: Human-Wildlife Interactions 4:47\u0026ndash;55. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2307/24864502\u003c/span\u003e\u003cspan address=\"10.2307/24864502\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcCance E (2010) Resident opinions concerning urban deer management in the Greater Winnipeg Area, Manitoba, Canada. Library and Archives Canada\u0026thinsp;=\u0026thinsp;Biblioth\u0026egrave;que et Archives Canada\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMorelle K, Lehaire F, Lejeune P (2013) Spatio-temporal patterns of wildlife-vehicle collisions Spatio-temporal patterns of wildlife-vehicle collisions in a region with a high-density road network. Nat Conserv 5:53\u0026ndash;73. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3897/natureconservation.5.4634\u003c/span\u003e\u003cspan address=\"10.3897/natureconservation.5.4634\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMyers WL, Chang WY, Germaine S et al (2008) An Analysis of Deer and Elk. -Vehicle Collision Sites along State Highways in Washington State\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNg J, Nielson C, St. Clair C (2008) Landscape and Traffic Factors Influencing Deer\u0026ndash;Vehicle Collisions in an Urban Enviroment. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.26077/nxw2-yn81\u003c/span\u003e\u003cspan address=\"10.26077/nxw2-yn81\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Human-Wildlife Conflicts 2:\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNielsen CK, Anderson RG, Grund MD (2003) Landscape Influences on Deer-. Vehicle Accident Areas in an Urban Environment\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePutmam RJ (1997) Deer and Road Traffic Accidents: Options for Management. J Environ Manage 51:43\u0026ndash;57. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1006/jema.1997.0135\u003c/span\u003e\u003cspan address=\"10.1006/jema.1997.0135\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRiley S, Enck J, Curtis P et al (2003) Deer populations up, hunter populations down: Implications of interdependence of deer and hunter population dynamics on management1. Ecoscience 10:455\u0026ndash;461. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/11956860.2003.11682793\u003c/span\u003e\u003cspan address=\"10.1080/11956860.2003.11682793\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSantos RAL, Mota-Ferreira M, Aguiar LMS, Ascens\u0026atilde;o F (2018) Predicting wildlife road-crossing probability from roadkill data using occupancy-detection models. Sci Total Environ 642:629\u0026ndash;637. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1016/j.scitotenv.2018.06.107\u003c/span\u003e\u003cspan address=\"10.1016/j.scitotenv.2018.06.107\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeo C, Thorne JH, Choi T et al (2015) Disentangling roadkill: the influence of landscape and season on cumulative vertebrate mortality in South Korea. Landscape Ecol Eng 11:87\u0026ndash;99. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11355-013-0239-2\u003c/span\u003e\u003cspan address=\"10.1007/s11355-013-0239-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSteiner W, Leisch F, Hackl\u0026auml;nder K (2014) A review on the temporal pattern of deer\u0026ndash;vehicle accidents: Impact of seasonal, diurnal and lunar effects in cervids. Accid Anal Prev 66:168\u0026ndash;181. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1016/j.aap.2014.01.020\u003c/span\u003e\u003cspan address=\"10.1016/j.aap.2014.01.020\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStout RJ, Stedman RC, Decker DJ, Knuth BA (1973) Perceptions of Risk from Deer-Related Vehicle Accidents. Implications for Public Preferences for Deer Herd Size\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSudharsan K, Riley SJ, Campa HI (2009) Relative Risks of Deer\u0026ndash;Vehicle Collisions Along Road Types in Southeast Michigan. Hum Dimensions Wildl 14:341\u0026ndash;352\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSullivan JM (2011) Trends and characteristics of animal-vehicle collisions in the United States. J Saf Res 42:9\u0026ndash;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1016/j.jsr.2010.11.002\u003c/span\u003e\u003cspan address=\"10.1016/j.jsr.2010.11.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTanner A, Leroux S, Saunders P (2017) Road Characteristics Best Predict the Probability of Vehicle Collisions with a Non-Native Ungulate. \u0026Eacute;coscience 24:1\u0026ndash;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/11956860.2017.1292100\u003c/span\u003e\u003cspan address=\"10.1080/11956860.2017.1292100\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eValero E, Picos J, \u0026Aacute;lvarez X (2015) Road and traffic factors correlated to wildlife\u0026ndash;vehicle collisions in Galicia (Spain). Wildl Res 42:25\u0026ndash;34\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVisintin C, van der Ree R, Mccarthy M (2017) Consistent patterns of vehicle collision risk for six mammal species. J Environ Manage 201:397\u0026ndash;406. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jenvman.2017.05.071\u003c/span\u003e\u003cspan address=\"10.1016/j.jenvman.2017.05.071\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWidenmaier KJ, Fahrig L (2005) Inferring white-tailed deer (Odocoileus virginianus) population dynamics from wildlife. collisions in the City of Ottawa\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang D, Wu Y, Sun F et al (2021) Freeway accident detection and classification based on the multi-vehicle trajectory data and deep learning model. Transp Res Part C Emerg Technol 130:103303. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1016/j.trc.2021.103303\u003c/span\u003e\u003cspan address=\"10.1016/j.trc.2021.103303\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Machine Learning, GIS, Logistic Regression, Land Cover, Wildlife","lastPublishedDoi":"10.21203/rs.3.rs-4504927/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4504927/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eEvery year in the United States, approximately 1.5 million deer–vehicle collisions (DVCs) occur, resulting in \u0026gt;200 human fatalities, \u0026gt;29,000 human injuries, 1.3 million deer fatalities, and \u0026gt;1 billion dollars’ worth of property damage. However, there was a lack of studies implementing machine learning techniques from the state level to evaluate the factors affecting DVCs. Data on DVCs on roads are valuable to reduce the occurrence of DVCs and to assist in planning. We utilized the data from 2018 to 2021 provided by Department of Transportation and Safety. The finding suggests that DVCs occurred more frequently near the developed areas, cultivated land and woody wetland and in October, from 8:00 PM to 10:00 PM and 6:00 AM to 8:00 AM. The accuracy scores 0.56 and 0.63 were obtained from machine learning and artificial neural network, opening the door for future research on more factors that affect DVCs.\u003c/p\u003e","manuscriptTitle":"Factors associated with deer vehicle collisions in South Carolina (SC), USA","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-06-18 07:52:57","doi":"10.21203/rs.3.rs-4504927/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"9d62e77b-906f-4754-81e4-f40bf93a7d08","owner":[],"postedDate":"June 18th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-02-19T21:53:15+00:00","versionOfRecord":[],"versionCreatedAt":"2024-06-18 07:52:57","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4504927","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4504927","identity":"rs-4504927","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}