Deer Crashes: Does topography play a factor?

        

Deer Crashes: Does topography play a factor?  

 

 

Introduction  

Driving down the Wisconsin roads can be beautiful and relaxing until a deer jumps out in front of a moving vehicle. This happens quite often as Wisconsin ranks 7th in the United States when it comes to deer crashes, and Wisconsin residents have a 1 in 57 chance of hitting a deer (State Farm). In 2019, there were over 18,414 deer crashes across the state of Wisconsin (Community Maps,2021). The average cost of a deer crash was $3,362 (Taschler, 2019). Some of these crashes seem to have occurred with similar topographies. This project will investigate if certain types of topography increase the chance of a deer crash. 

 

Literature Review  

This article explains why deer crashes seem to happen more in some areas than others.  Troy Jerman, a formal official with the Iowa DOT.  He said, “Be most alert for the presence of deer at locations where three factors converge: Food (corn fields or recently harvested corn fields),Shelter (woods),Water (streams, culverts and river crossings).” 

Jerman , T. 2010. Iowa DOT News Release. Iowa Department of Transportation. https://www.news.iowadot.gov/newsandinfo/2010/10/increased-deer-movement-means-extra-caution-needed-on-iowa-roadways.html (last accessed 1 March 2021). 

This one was a study conducted of deer crashes of deer crashes in the Czech Republic. The researchers were looking for deer crash hot spots on railroads. This also shows that in Europe deer crashes are also a problem. The project showed a similar trend as to what happens in the USA with deer crashes. This article also helped me to understand deer behavior better.  The project found out that woods and streams were more common than crops and woodlands.  

Nezval, V., and M. Bíl. 2020. Spatial analysis of wildlife-train collisions on the Czech rail network. Science Direct. https://www-sciencedirect-com.proxy.uwec.edu/science/article/pii/S0143622820304835 (Last accessed 2 March 2021). 

Some Iowa State University professors did a research project on deer crashes in the Upper Midwest. They asked some questions that will narrow down the search as well. For example, they asked states about what their undercount deer crashes are. Most states responded by saying they do not even try to get the real count. Sometimes knowing how is more important than knowing what. This article helps more with “the what.”  

Iowa State. 2004. DVC Data Collection Survey. Regional Survey of Deer Population, Vehicle Travel, and Deer-Vehicle Crash Information Collection and Management. https://iowaltap.iastate.edu/dvc-data-collection-survey/ (Last accessed 2 March 2021). 

This project was completed by researchers from the University of Georgia. They spelled out their methods of how they collect where deer crashes are happening.  They were investigating deer crashes in South Carolina in the 1990s. They noticed that vegetation influenced where deer crashes happened. They confirmed that there were certain topographical features that did increase the number of deer crashes.  

Malhotra, R., P. E. Johns, M. Madden, and G. Wein. Deer-Vehicle Collisions: Is There a Pattern? https://proceedings.esri.com/library/userconf/proc00/professional/papers/PAP446/p446.htm (last accessed 2 March 2021). 

An interview that Aaron Dekker conducted with Jeff Pritzl, a deer specialist from Wisconsin DNR, also shed some light on deer behavior. His response was that “Deer are creatures of habit. You can see this with deer paths in the woods. Their average range is about one square mile.” Pritzl, J. 2021. Deer Behavior. (last accessed 8 May 2021). 

Background 

This project was inspired because of a deer crash that happened in June 2020. After that, a hypothesis formed that the topography around the crash site had something to do with the crash. Figure B1 shows what the crash area looked like.  

 

Figure B1: 

 

After reviewing this map, the hypothesis was formed that the topography played a role. Then after reading the article by Troy Jerman, it created the formal hypothesis based off what he said: "Be most alert for the presence of deer at locations where three factors converge: Food (corn fields or recently harvested corn fields), Shelter (woods), Water (streams, culverts and river crossings).” (Jerman,2010) 

From this article, the hypothesis that was formed: 

When Fields (or food), Woods (or shelter), and Water combine, does this enhance the chance of hitting a deer?  

 

Methods: 

Study area:  

  

Figure M1: 

 

Figure M1: shows the study area which is western Wisconsin. 

Figure M2: 

 

Figure M2: Shows the simplified version of what happened. 

Just some quick definitions of what had to count as topography. To be listed as water, it had to show up on Google earth or be visibly there during a field visit. The woods had to be at least an acre. The fields could either be grass or crops, But the grass immediately next to the roads does not count. All these features had to be within a half mile of the crash site.  

To calculate which one was the biggest factor; woods, water, and fields were broken down into separate categories and calculated by percentage. After calculating percentages, to be considered as a factor they had to be over 60 percent in cases. If over 80 percent of the crashes have occurred near a certain topographic feature, then it is a major factor.  

2019 was picked as the year to check, since the 2020 COVID lockdown dropped the number of deer crashes. In Community Maps, it showed that statewide there was a drop of 2000 crashes. It was decided that the crash toll would be too low to get enough points to answer the question of topography.  

Step 1 was to select three Wisconsin counties. Criteria were: 

1. Must be a Wisconsin county.  

2. Must have a four-lane highway running through it. 

3. Must be in different deer zones.  

Figure M3. 

 

Figure M3: Shows the Wisconsin DNR deer hunting zones in 2020 (Wisconsin DNR). 

The website used to find deer crashes was Community Maps. The website is a project between UW-Madison and the Wisconsin Police Departments’ crash database. The spreadsheet of crashes was not available, so each location had to be mapped manually.  

Figure M4 

 

Figure M4: Shows the Counties that were selected.   

• Washburn County was selected because of four lane highway 53 and Northern Deer Zone 1.     

•    Jackson County has I-94 and is in Central Farmland Zone 2 and Central Forest Zone 2. 

•  Iowa County has four lane highway 151 and is southern Farmland Zone 2.   

 

Figure M5: 

 

Figure M5: Showed what the county maps looked like after selection was completed. 

Figure M6: 

 

        Figure M6: Shows what the spreadsheet partially looked like. 

 

The data was then mapped on Arc GIS pro.  

 

Results  

The results are as follows: 

Iowa County  

Figure R1: 

  

Figure R1 Shows Iowa County’s total count by terrain. 

 

Figure R2 

 

 

Figure R2: Shows the percentage of cases where fields were nearby. 

 

Figure R3 

 

Figure R3: Shows what percentage of crashes had water nearby.  

 

Figure R4: 

 

Figure 4: Shows the percentage of crashes that had woods nearby. 

 

Figure R5: 

Figure R5 This figure shows the satellite image of crashes of Iowa County with the half mile radius. 

Figure R6 

Figure R6: Shows the road map in Iowa County with crashes overlaid. 

Jackson County  

Figure R7 

  

Figure R7 shows the total number of  the topography of crashes in Jackson County. 

Figure R8 

 

Figure R8: Shows the percentage of crashes that had fields nearby. 

Figure R9: 

 

Figure R9: This figure shows the percentage of cases when water was nearby.  

Figure R10: 

 

Figure R10: This figure shows the percentage of cases when woods were nearby. 

Figure R11 

 

Figure R11: This figure shows the satellite image of crashes in Jackson County with a half mile radius buffer. 

Figure R12 

Figure R12: Shows the road map in Jackson County with crashes overlaid. 

  

Figure R13 

 

Figure R13: This figure shows the numerical combination of topography in Washburn County that was nearby the most crashes. 

Figure R14: 

 

Figure R14: This figure shows the percentage of cases when fields were nearby. 

Figure R15: 

 

Figure R15: This figure shows the percentage of cases when water was nearby. 

Figure R16 

 

Figure R16: This figure shows the percentage of cases when woods were nearby. 

Figure R17 

Figure R17: Shows the satellite image of crashes in Washburn County with a half mile radius. 

Figure R18 

 

Figure R18: Shows the road map in Washburn County with crashes overlaid.  

Totals  

Figure R19 

 

Figure R19: Shows the combined totals from all three counties.  

Figure R20: 

 

Figure R20: This figure shows the percentages when all counties’ topographical combinations were set to percentages.  

 

Figure R21: 

 

Figure R21: Shows when broken down by which factor was the biggest by percentage. 

 

Discussion  

There are multiple notes and answers that need to be addressed, and they will be here. The first notes will be about routes of roads and how that plays another role. Then the results of each county will be broken down individually. 

The route of roads also matters as to the topography. For example, I-94 in Jackson County follows or crosses a body of water for most of its length. The topography varies greatly along the routes, especially north and west of Black River Falls. Highway 151 ran mostly through fields. While Highway 53 mostly ran through woods. This will change with factor will be the leading topography that leads the way for increasing likelihood of a deer crash. 

For Iowa County, there are some notes that need to be addressed. Figure R1 showed that for Iowa County, field and field/woods had the biggest total number of crashes. While Figure R2 shows Fields cleared the 80 percent threshold easily and was just shy of hitting 100 percent. Figure R3 shows that Water cleared 60 percent of cases with no water. Figure R4 shows that Woods did not clear 60 percent. Based on the grading scale, fields were clearly the most common factor in Iowa County. Woods was not really a factor since it did not clear 60 percent. Since Water cleared 60 percent in the negative direction, this would indicate that water was not a factor in Iowa County. Even though fields and fields/woods were the most common factors in crashes when totaled, but when separated, fields were the clear winner of biggest factor of crashes. Figure R5 and Figure R6 showed that most of the crashes in Iowa County happened on Highway 151, which was built mostly through fields. The other noticeable stretch was just north of Dodgeville near Governor Dodge State Park, which was near both fields and woods. The northern stretch is more mixed topographically wise.  

Jackson County saw a completely different pattern. Figure R7 showed that Field/woods/water combined to form most of the crashes. When broken down to individual rankings, a different pattern showed. Figure R8 showed that Fields cleared the 60% threshold, meaning that it was a factor. Figure R9 had Water clear the 60 percent threshold, also being a factor. Figure R10 showed Woods easily cleared the 80 percent threshold, proving that woods were the most common factor in Jackson County. Overall, this showed that all three factors in Jackson County increased deer crash likelihood. But Woods showed the be the greatest factor. Figure R11 and Figure R12 showed that most of the crashes happened on I-94, as expected. Remember that I-94 terrain varies greatly along its path through Jackson County. The northwest side starts out as mixed where it can be woods then fields then water and maybe mixed then change back again. While the southeast side was mostly forested. I-94 also follows or crosses a body of water with quite frequency in Jackson County. Combine these factors and it paints the picture of why all three played a role in increasing the probability of a crash.  

Washburn county has some patterns that need to be explained. Figure R13 showed that Field/Woods combination was the most common. But when broken down into individual categories, the trend differed. Figure 14 shows that Fields cleared the 60 percent threshold showing that woods was a factor in creating deer crash likelihood. Figure R15 showed that Water did not clear 60 percent in either direction. Figure R16 showed that Woods easily cleared the 80 percent threshold and even cleared the 90 percent threshold. This shows that woods were the biggest factor in Washburn County. This shows that water is neither a factor nor not a factor in increasing deer crash likelihood. But when combined, fields and woods were the stronger factors. Figure R17 and Figure R18 showed that most crashes happened either on highway 53 or on highway 63. Both roads run through a mix of fields and woods.  

When all three counties are combined, the pattern sees some changes. Figure R19 showed Field/Water/Woods combination had the most crashes with Field/Woods combination being close. Figure R20 showed that by percentage, none hit the 60 percent threshold, but Field/Water/Woods had the highest percentage out of the combinations. When broken down by the three characteristics, as seen in Figure R21, Fields and Woods easily cleared the 60 percent threshold, proving that Fields and Woods were the big factors. Water just barely cleared 60 percent on the same graph. This showed water was not as big of a factor as was predicted. The biggest surprise was that only 24 percent of cases had no Fields nearby. Only 25 percent of cases had no Woods nearby. Since less than 30 percent of deer crash cases had no Fields and Woods; this would indicate that Fields and Woods are important for increasing deer crash likelihoods.  

Limitations 

There are a few limitations. The first limitation would be the accuracy of reporting of deer crashes. Sometimes these locations are not reported correctly. The caller will report that the crash happened at one location, but it really happened at another location. What also happens is someone will hit a deer at xyz, but they will then drive a mile before they must stop to report the crash so now it is at yzx. 

The other limitation was the underreporting of deer crashes. Jeff Pritzl estimated that for every crash that is reported, two are not reported. (Pritzl,2021). If every crash was reported, then the results would have given a clearer picture in either direction. Crashes are usually not reported because the crash did not cause enough damage, or the person was driving while intoxicated. 

Acknowledgments  

Jeff Pritzl, Wisconsin DNR Deer Specialist, for various answers on deer behavior. Dr. Haffner for helping fix latitude longitude problems. Community maps for publishing deer crashes and having free access. The deer I hit in June 2020 for inspiration for the project, I hope it is still doing well as it ran off after the crash. Dr. Running for various help along the way. University of Wisconsin-Eau Claire for providing GIS Software. The University of Wisconsin-Eau Claire Geography Department in general, especially Yvonne Plomedahl. Luke Freeman for keeping me on task. Martin Goettl for GIS help. My parents for motivating me and encouraging me to continue.  

Future Projects.  

The first project would be to find hotspots. Since it seemed like crashes were in hotspots. It would be interesting to see if deer crashes are clustered or just random for the same counties. Knowing that topography helps but knowing where hot spots would be more helpful.  

The second thing noticed was the sheer number of crashes of some two-lane highways versus four lane highways. The next project would be is to see how traffic counts played a role in the topography. Some topography along the four lane highways showed different results of crashes. The next step involving traffic counts would be to get the number of crashes and divide that by the number of traffic to get the probability of hitting a deer.  

The other thing noted in the crash reports was the month of the crash. It would be interesting to see what time of day and month would be the most common to hit a deer.  

Since one county swayed the results one way or the other it would be telling to see if the other 69 counties in Wisconsin show similar or quite different results. This project could also be expanded to see how other states in the United States compare. Since an article was also from Europe, that would be another case study to look at other countries. 

During class time, someone asked about other animals, like bears and raccoons, if that could be another project. That could be done as other animals might have similar topographical patterns. The downside here is that finding data on that would be hard as some of those accidents are not even reported. 

 Checking different years would either confirm or deny if there are common topographical features. This could also be done on Community Maps. Instead of comparing different years, it could be comparing years in terms of a time frame, not just one year. 

Conclusion 

In conclusion, it appeared that the topography did play a factor in crashes. Each county’s topographical leader was different. The project both confirmed and disproved the hypothesis. It is partially rejected because water was not the main factor as predicted. In the end, when woods, fields, and water combine, it does increase the likelihood of a deer crash. But when separated, fields and woods are the biggest contributors to increase the chance of a deer crash. Even though Wisconsin residents love fresh grilled venison, insurance companies do not want to pay $3362 for the fresh venison. Therefore, drivers should always be on the lookout for deer.  

 

References Cited 

1. Community Maps - Crash. https://transportal.cee.wisc.edu/partners/community-maps/crash/search/BasicSearch.do;jsessionid=00556FBC56703952046F7E82DA0AE5A2 (last accessed 1 May 2021). 

2. Gomez, B., and J. P. Jones. 2010. In Research methods in geography: a critical introduction, 20–21. Malden, MA: Wiley-Blackwell. 

3. Haffner, M., 2021. Convert DMS to DD ($2110450) · Snippets · Snippets. [online] GitLab. Available at: <https://gitlab.com/-/snippets/2110450> [Accessed 27 April 2021]. 

4. Jerman, T. 2010. Iowa DOT News Release. Iowa Department of Transportation. https://www.news.iowadot.gov/newsandinfo/2010/10/increased-deer-movement-means-extra-caution-needed-on-iowa-roadways.html (last accessed 1 March 2021). 

5. Pritzl, J. 2021. Deer Behavior. 

6. Malhotra, R., P. E. Johns, M. Madden, and G. Wein. Deer-Vehicle Collisions: Is There a Pattern?https://proceedings.esri.com/library/userconf/proc00/professional/papers/PAP446/p446.htm (last accessed 2 March 2021). 

7. State Farm.S. F. S. 2020. Where are Animal Collisions Most Likely? - State Farm®. State Farm. https://www.statefarm.com/simple-insights/auto-and-vehicles/how-likely-are-you-to-have-an-animal-collision (last accessed 11 May 2021). 

8. Taschler, J. 2018. Every day, a car hits a deer in Wisconsin. This week is among the most dangerous on our roads. Milwaukee Journal Sentinel. https://www.jsonline.com/story/money/business/2018/11/08/deer-crashes-wisconsin-peak-november/1847638002/ (last accessed 4 May 2021). 

9. Wisconsin DNR Deer Zone Map. 2020. https://dnr.wisconsin.gov. https://dnr.wisconsin.gov/sites/default/files/topic/images/dmz%20%281%29.png (last accessed 4 May 2021) 

Further readings 

Deer Crashes: Are there hotspots?

           Deer Crash: Jackson County Survey  

Disposition 

This project includes GIS work and advanced GIS approaches. It connects geography to public safety and biology disciplines. This project also helps people to visualize possible danger. 

Responsibilities/EDI 

This project informs drivers to know to be aware of topographical features that may increase the chance of a deer crash. To the best of my knowledge, nothing in this project is unethical. The data was all collected in ethical methods.  

Equality: This project helps all drivers to know where deer crashes could be likely.  Crashes can be costly; some drivers cannot afford an unexpected $3362 bill.  

Diversity: Deer do not care who the driver is.  

Inclusivity: Passengers can also know where to watch for deer, so their driver can hopefully avoid a potential deer crash. 

Paradigms 

The paradigm here is “Spatial science”. Since topography is spatial, this goes under the spatial science paradigm. (Gomez and Jones, 2010)