Interactions Between Humans and White-Tailed Deer in Illinois: A Cross-Sectional Survey

Abstract

SARS-CoV-2 has been found in multiple species, including cervids such as wild white-tailed deer (WTD), in multiple regions in the United States, including Illinois. The virus has been shown to transmit among WTD, and across species in both directions (deer-to-humans and humans-to-deer). Cross-species transmission requires infectious contact between WTD and humans, the form and frequency of which is poorly understood. The aim of this cross-sectional survey was to understand the frequency and type of contact between the general public in the state of Illinois and WTD, and to identify human populations at highest risk for such contact. An online survey was distributed using convenience sampling from list serves, social media, and community partners or extension liaisons. Questions addressed frequency and distance of contact with WTD, encompassing live animals and bodily fluids. Standard and ordinal logistic regression were used to identify factors associated with contact. An overall risk score was calculated, and linear regression was used to identify factors associated with risk. We found that hunters and those who have deer feeding on their property are more likely to report contact with deer, and that people reporting a larger lot size and living in counties with higher proportions of potential deer habitat are more likely to report deer feeding on their property. These results will better identify people with a high likelihood of WTD contact for messaging and further research. Our survey did not distinguish between contact with live and dead WTD, thus the findings are most relevant to deer-to-human cross-species transmission than the human-to-deer direction.

Introduction and Purpose

White-tailed deer (WTD, Odocoileus virginianus) are highly susceptible to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and may serve as a reservoir for variant SARS-CoV-2 strains that may be extinct in humans (Feng et al., 2023; Pickering et al., 2022). Multiple spillbacks from WTD to humans have now been identified in North America (Feng et al., 2023). However, the pathway for spillover and potential spillback of SARS-CoV-2 between humans and WTD is unknown.

Throughout the US, monthly prevalence of SARS-CoV-2 in WTD ranged from 6 to 20% during 2021–2022 with R₀ between 1 and 2.5 in most eastern US counties where WTD occur (Bevins et al., 2023; Hewitt et al., 2024). Infection pressure in humans (measured as SARS-CoV-2-induced mortality rates) were significantly correlated with prevalence of SARS-CoV-2 in deer at the county level but the effects of human density are weak (Hewitt et al., 2024). This suggests that other factors such as human behavior or mobility could play a role in shaping infection rates in deer. Such human dimensions of wildlife management have long been recognized as important areas of study (Decker and Enck, 1996), but the focus has been on wildlife management and human-wildlife conflict and on the experiences of hunters and farmers (Stinchcomb et al., 2022). There is still much to be learned about human behavior that leads to human-wildlife transmission of wildlife-related zoonoses (Decker et al., 2010).

To address the gaps related to how deer and humans come into potentially infectious contact, we conducted surveys of the public in Illinois to understand the frequency and type of close contact between the general public in Illinois and white-tailed deer (WTD). We consider the factors that may predict human and deer interactions such as location, residence type, and occupation, with the goal of identifying individuals at high risk for spillover and spillback events. Our work is a stepping-stone towards streamlining future research into spillover and spillback transmission events, and improving the efficiency and efficacy of public health communication efforts.

Methods

This study was ruled exempt (non-human subjects research) by the University of Illinois Institutional Review Board (protocol #23422). Informed consent was obtained from all individual participants included in the study.

A cross-sectional study was designed to measure how often the general public in the state of Illinois comes into close contact with WTD. Questions in the survey addressed frequency, proximity, and type of contact with wild white-tailed deer, as well as demographic information, residence area, and participation in activities with likely WTD contact. The surveys were developed and disseminated on a secure online platform called REDCap (Research Electronic Data Capture) which is hosted by the University of Illinois Urbana-Champaign with the help of the Interdisciplinary Health Sciences Institute and Research IT—Technology Services. REDCap provides (1) an intuitive interface for validated data capture; (2) audit trails for tracking data manipulation and export procedures; (3) automated export procedures for seamless data downloads to common statistical packages; and (4) procedures for data integration and interoperability with external sources. The survey was beta-tested for logistical or technical issues by team members, and their responses were removed before the data collection process.

The survey was distributed broadly across the state of Illinois via a public link using a convenience sampling method from list serves, social media, and community partners or extension liaisons. Participation was open to all residents of Illinois who were over 18 years of age and self-reported as able to read English. All responses that did not confirm residence in Illinois, did not respond to the question about coming within 50 ft of a wild WTD in the last year, or did not complete the survey were removed prior to analysis.

Analysis was based on the causal map shown in Figure 1. We hypothesized that contact with deer would be a function of the presence of deer near a participant’s residence and activities that might bring the participant into deer habitat. The presence of deer near a residence was considered to be a function of the region of the state, the type of residence area (urban, suburban, rural, or other), and the size of the residence lot. Alternatively, one question asked if deer fed on the participant’s property, which was used as a proxy for deer presence. For activities that might increase contact, the following occupations/hobbies were assessed: biologist, veterinarian, animal control, hunter, meat processor, taxidermist, rehabilitation specialist, and other (with an opportunity to self-identify activities that might increase deer contact). Participants were also asked if they hunted wild WTD, if they hunted captive deer species, and if they owned livestock (and, if so, what type of livestock they owned).

Figure 1.

Causal map for the analysis of data from a survey of Illinois residents on their interactions with wild white-tailed deer. The outcome is in a black box, latent variables in a gray box, and predictor variables in white boxes (Color figure online).

Logistic regression was used to examine the impact of the identified exposure variables on the probability of ever being within 50 yards of WTD. Due to the potential for multicollinearity, based on the causal map, lot size was not included in this analysis. Two sets of predictor variables were considered to represent different potential aspects of the presence of deer in the respondent’s area: (1) region and residence area; (2) deer feeding on property. All models also included the identified occupations and hobbies (meat processor, biologist, and other), hunting, and the ownership of livestock. The occupations veterinarian, animal control, taxidermist, and rehabilitation specialist were not included due to low numbers, and ownership of livestock was not included in the ordinal logistic regression models for the same reason.

Cumulative odds ordinal logistic regression with proportional odds was used to examine the impact of the identified exposure variables on the frequency of being within 50 yards of WTD and the proximity of the closest WTD contact. The same two models were fit as for logistic regression.

To assess the combined risk of both proximity to wild deer and frequency of that proximity, a total risk score variable was calculated as

$$s=\sum _i\sum _j{p}_i{f}_j$$

where $p_i$ is the value associated with proximity $i$ and $f_j$ is the value associated with frequency $j$. The values of $p$ were empirically assigned as follows: touching = 10, 0–3 ft = 8, 3–6 ft = 6, 6–10 ft = 4, and > 10 ft = 2. The values of $f$ were assigned as follows: > 5 days a week = 10, 3–4 days a week = 8, 1–2 days a week = 6, < 1 day a week = 4, I do not recall = 2, and never = 0. This calculated risk score was then modeled using linear regression against factors identified using the causal map. Due to the nested nature of these factors, two models were fit: a feeding model, which included the variable “deer feeding on property” as a proxy for deer near residence, and a regional model, which included factors related to location and residence area.

All statistics were performed in Rstudio ver. 4.3.0. Data cleaning and visualization utilized the tidyverse suite of packages (Wickham et al., 2019), logistic and linear regression were performed with the glm function from the stats package (R core team, 2023), and ordinal logistic regression was performed with the polr function from the MASS package (Venables and Ripley, 2002). Model selection was conducted by AIC values, where the lowest AIC values indicated a relatively better fit and models were considered significantly better than others when their AIC values were more than 2 less than other models. In the best fit models, relationships were considered statistically significant if p < 0.05.

Results

A total of 715 participants initiated the survey, of which 13 were removed due to not being Illinois residents, 20 were removed for incomplete survey responses, and 4 were removed for failing to answer the primary question (“have you been within 50 ft of a wild white-tailed deer in the last year?”), leaving 678 participants for analysis (Table 1, Fig. 2). The datasets generated during and/or analysed during the current study are available in the Illinois Data-Bank, https://doi.org/10.13012/B2IDB-1661924_V1. Participation was highest in Champaign and McDonough counties, home to University of Illinois Urbana-Champaign and Western Illinois University, respectively. Both of these universities allowed survey advertisement in their weekly community email newsletter.

Table 1.

Participants in a cross-sectional survey of Illinois residents regarding contact with wild white-tailed deer.

Variable	Level	Have you been within 50 ft of a wild white-tailed deer in the last year?
		Yes	No	I do not recall
Number of participants (%)		600 (88.5)	72 (10.6)	6 (0.9)
Age (%)	18–20 years	34 (89.5)	3 (7.9)	1 (2.6)
	21–30 years	81 (83.5)	14 (14.4)	2 (2.1)
	31–40 years	93 (87.7)	11 (10.4)	2 (1.9)
	41–50 years	114 (91.2)	11 (8.8)	0 (0.0)
	51–60 years	120 (88.9)	14 (10.4)	1 (0.7)
	60 + years	151 (88.8)	19 (11.2)	0 (0.0)
	Prefer not to answer	4 (100.0)	0 (0.0)	0 (0.0)
Gender (%)	Male	205 (92.3)	15 (6.8)	2 (0.9)
	Female	371 (85.9)	57 (13.2)	4 (0.9)
	Other	5 (100.0)	0 (0.0)	0 (0.0)
	Prefer not to answer	15 (100.0)	0 (0.0)	0 (0.0)
Highest level of education (%)	Did not attend school	1 (100.0)	0 (0.0)	0 (0.0)
	Elementary	0 (0.0)	0 (0.0)	0 (0.0)
	Middle School/Junior High	1 (100.0)	0 (0.0)	0 (0.0)
	High School Graduate	26 (83.9)	4 (12.9)	1 (3.2)
	Some College	80 (90.9)	8 (9.1)	0 (0.0)
	2-year College (Associates Degree)	46 (83.6)	8 (14.5)	1 (1.8)
	4-year College (Bachelors degree)	152 (90.5)	16 (9.5)	0 (0.0)
	Some Graduate School	60 (87.0)	8 (11.6)	1 (1.4)
	Masters Degree	157 (87.7)	20 (11.2)	2 (1.1)
	Doctorate Degree (PhD or similar)	59 (88.1)	7 (10.4)	1 (1.5)
	Professional Degree (MD, DVM, DDS, JD, etc.)	14 (100.0)	0 (0.0)	0 (0.0)
	Other	3 (100.0)	0 (0.0)	0 (0.0)
Annual family income (%)	$0–$40,000	55 (84.6)	9 (13.8)	1 (1.5)
	$40,001–$60,000	74 (88.1)	10 (11.9)	0 (0.0)
	$60,001–$80,000	71 (88.8)	9 (11.2)	0 (0.0)
	$80,001–$100,000	83 (92.2)	7 (7.8)	0 (0.0)
	$100,001–$120,000	72 (84.7)	12 (14.1)	1 (1.2)
	$120,001–$140,000	41 (93.2)	3 (6.8)	0 (0.0)
	$140,001–$160,000	26 (83.9)	5 (16.1)	0 (0.0)
	$160,001 +	80 (87.9)	9 (9.9)	2 (2.2)
	Prefer not to answer	93 (90.3)	8 (7.8)	2 (1.9)
Region of Illinois (%)	Central	392 (88.9)	45 (10.2)	4 (0.9)
	Northern	154 (88.0)	19 (10.9)	2 (1.1)
	Southern	54 (87.1)	8 (12.9)	0 (0.0)
Type of residence area (%)	Urban (Greater than 50,000)	85 (85.9)	13 (13.1)	1 (1.0)
	Suburban (Between 2500 and 50,000)	289 (89.8)	30 (9.3)	3 (0.9)
	Rural (Less than 2500)	157 (88.7)	18 (10.2)	2 (1.1)
	Exurban (Does not live inside of an incorporated area/town)	64 (85.3)	11 (14.7)	0 (0.0)
	Prefer not to answer	2 (100.0)	0 (0.0)	0 (0.0)
Lot size of residence (%)	No Lot (apartment, condo, townhouse)	55 (87.3)	7 (11.1)	1 (1.6)
	Less than 0.25 (1/4) of an acre	143 (87.7)	17 (10.4)	3 (1.8)
	Between 0.25 (1/4) and 0.5 (1/2) of an acre	143 (90.5)	15 (9.5)	0 (0.0)
	Between 0.5 (1/2) and 1 acre	75 (88.2)	8 (9.4)	2 (2.4)
	Between 1 and 5 acres	85 (84.2)	16 (15.8)	0 (0.0)
	Between 5 and 10 acres	30 (93.8)	2 (6.2)	0 (0.0)
	Between 10 and 50 acres	33 (84.6)	6 (15.4)	0 (0.0)
	Between 50 and 100 acres	7 (100.0)	0 (0.0)	0 (0.0)
	Greater than 100 acres	27 (96.4)	1 (3.6)	0 (0.0)
Do you participate in any of these occupations/ hobbies? (%)	Biologist	36 (83.7)	6 (14.0)	1 (2.3)
	Veterinarian	6 (66.7)	2 (22.2)	1 (11.1)
	Animal control	3 (100.0)	0 (0.0)	0 (0.0)
	Hunter	104 (95.4)	5 (4.6)	0 (0.0)
	Meat processor	33 (943)	2 (5.7)	0 (0.0)
	Taxidermist	5 (83.3)	1 (16.7)	0 (0.0)
	Rehabilitation specialist	4 (80.0)	0 (0.0)	1 (20.0)
	Other*	68 (87.2)	10 (12.8)	0 (0.0)
	None of these	397 (87.8)	50 (11.1)	5 (1.1)
What types of livestock/ poultry do you own? (%)	none	538 (88.3)	65 (10.7)	6 (1.0)
	beef	19 (95.0)	1 (5.0)	0 (0.0)
	Sheep or goats	8 (88.9)	1 (11.1)	0 (0.0)
	swine	4 (100.0)	0 (0.0)	0 (0.0)
	Commercial poultry	1 (100.0)	0 (0.0)	0 (0.0)
	Backyard poultry	35 (85.4)	6 (14.6)	0 (0.0)
	deer	0 (0.0)	0 (0.0)	0 (0.0)
	Other+	12 (92.3)	1 (7.7)	0 (0.0)

^*Responses include hiking, running, camping, bird watching, farming, gardening, naturalist, and work with harvested deer (deer check stations and home processing).

⁺Responses include horses/donkeys, bison, honey bees, and llamas.

Figure 2.

Map of number of responses to a survey about human interaction with white-tailed deer in the state of Illinois, USA. Gray indicates no responses (Color figure online).

A majority of participants (600/678, 88%) reported ever being within 50 ft of a wild WTD outside, but most (461/600, 77%) reported doing so less than once a week, and few (29/600, 5%) reported touching the deer (Table 2). When asked about frequency of different proximities to deer, most reported infrequent close contact but more frequent contact at further distance (Fig. 3), although contact at any distance remained infrequent for most. It should be noted that several respondents reported confusion as to whether this question referred to live or dead WTD, a distinction that should be made clear in future studies. A total of 16% (109/678) reported direct contact with wild white-tailed deer or with bodily fluids from a wild deer in the last 2 years, including 79/111 hunters, 33/36 meat processors, 12/43 biologists, 4/6 taxidermists, 3/9 veterinarians, and 11/78 who reported another high-contact job or hobby. As respondents could select more than one occupation or hobby, some of these are counted twice; for instance, 5/6 taxidermists were also meat processors. The most common physical contact was with meat products or blood, which might be relevant for deer-to-human transmission but not human-to-deer transmission.

Table 2.

Responses of participants in a cross-sectional survey of Illinois residents regarding contact with wild white-tailed deer on their level of contact with white-tailed deer. Percents are calculated for values within a particular question.

Variable	Level	N (%)
Have you ever been in close proximity (within 50 ft) of a wild deer outside?	Yes	600 (88.5)
	No	72 (10.6)
	I do not recall	6 (0.9)
How often do you come within 50 ft of a wild deer outside?	Never	48 (7.1)
	Less than once a week	461 (68.2)
	1–2 days/week	93 (13.8)
	3–4 days/week	24 (3.6)
	5 + days/week	24 (3.6)
	I do not recall	26 (3.8)
How close have you been to a wild deer in the last 2 years in an outdoor environment?	Touching	29 (4.3)
	0–3 ft	50 (7.4)
	3–6 ft	93 (13.7)
	6–10 ft	90 (13.3)
	10–20 ft	110 (16.2)
	20–50 ft	148 (21.8)
	50–100 ft	78 (11.5)
	100–500 ft	43 (6.3)
	500 + ft	6 (0.9)
	Never	26 (3.8)
	I do not recall	6 (0.9)
How often to you see wild deer on your property?	Never	223 (33.1)
	Less than once a week	192 (28.5)
	1–2 days/week	87 (12.9)
	3–4 days/week	62 (9.2)
	5 + days/week	110 (16.3)
Do wild deer feed on your property?	Yes	347 (51.3)
	No	294 (43.4)
	I do not recall	35 (5.2)
	Prefer not to answer	1 (0.1)
Where on your property do wild deer feed?	At a bird feeder/waterer/bath	21 (6.5)
	Garden (flowers, vegetables, other)	108 (33.4)
	Wooded area	121 (37.5)
	Pasture area	51 (15.8)
	Livestock feeding area (including horses and poultry)	4 (1.2)
	Other*	18 (5.6)
What type of (physical) contact have you had with wild deer in the last 2 years? Select all that apply	Blood	82 (12.1)
	Feces	46 (6.8)
	Meat Products	92 (13.6)
	Brain, Spinal Fluid, or other Nervous Tissue	25 (3.7)
	Saliva	20 (2.9)
	Petting/Touching	37 (5.5)
	Urine	38 (5.6)
	None	566 (83.5)

^*Responses include under trees, lawn, agricultural fields, and combinations of the other options. Only one answer was allowed, and some respondents reported multiple areas.

Figure 3.

Heat map of count of responses to the question “For each of the distances listed, please check the best scenario that describes how often you may come that close to a wild deer” in a survey of residents of Illinois, USA. Colors indicate the frequency of the response and are shown on a semi-log scale for ease of visualization (Color figure online).

A summary of all models fitted is shown in Table 3 and Supplemental Table 1. Across all outcomes, the model using the variable of “feeding” (deer feed on property) is a better fit to the data than the model using region and locality type.

Table 3.

Summary of models fitted to data from a survey of residents of Illinois, USA regarding proximity to wild white-tailed deer (WTD).

Model name	Covariates	AIC
		Score1	Logistic2	Ordinal Logistic: Distance3	Ordinal Logistic: Frequency4
Feeding	Do deer feed on your property? (yes, no) High risk occupation (yes, no) Hunter (yes, no) Own livestock* (yes, no)	6544	425	1521	1171
Regional	Region (Northern, Southern, Central) Locality Type (Urban, Suburban, Rural, Other) High risk occupation (yes, no) Hunter (yes, no) Own livestock* (yes, no)	6923	463	1623	1248

¹Calculated combination of distance and frequency, range 0–300.

²Have you been within 50 ft of a wild WTD in the last year? Answers: yes, no.

³How close have you come to a wild WTD? Answers: < 10 ft, 10–50 ft, 50–500 ft, > 500 ft.

⁴How frequently do you come within 50 ft of a wild WTD? Answers: Daily, Weekly, < Weekly, Never.

^*Not included in ordinal logistic regression models due to rank deficiency.

The only variable consistently associated with ever being within 50 ft of a wild WTD outside across all models was identifying as a hunter (Fig. 4); hunters were approximately 3 times as likely to report being within 50 ft of a wild WTD as non-hunters (95% CI: 1.2, 8.6 in the feeding model; 1.4, 10.0 in the region model).

Figure 4.

Odds ratios (points) and 95% confidence intervals (lines) from logistic regression models for three possible response variables: ever within 50 ft of a wild WTD outside (black), the closest proximity to a wild WTD in the last 2 years (green), and the typical frequency of being within 50 ft of a wild WTD (brown). Orange dashed line indicates no association (Color figure online).

Those who reported deer feeding on their property were 2.6 times as likely to have been within 50 ft of a wild WTD as those who did not (95% CI: 1.5, 4.6). The feeding model had a better overall fit, with an AUC of 0.653 compared to 0.599 for the region model, but neither model was a good fit, suggesting there may be other factors that we did not account for that are more important for explaining the results.

To avoid rank deficiency due to small numbers, ownership of livestock was removed from the ordinal logistic regression models. Measures of distance were simplified to within 10 ft, 10–50 ft, 50–500 ft, and over 500 ft, and measures of frequency were simplified to never, less than weekly, weekly, and daily. Similar patterns were seen as with the logistic regression model, with only hunting and reporting WTD feeding on their property showing a significant association with how close or how frequently respondents have come to a wild WTD in the last 2 years (Fig. 4). The model fit for both was improved with the feeding model compared to the region model (ΔAIC = 101 for distance, 76 for frequency).

Total risk scores ranged from 0 to 300, with a median of 24 and a heavily right-skewed distribution. A total of 43 respondents had a risk score of 0, corresponding to responding “Never” to all distances listed. A majority of respondents (n = 432) had risk scores between 0 and 50, and only 76 had risk scores above 100, indicative of more frequent close interactions. When included as the response variable in a multivariable regression, risk scores were significantly higher in rural areas, among hunters, and among those reporting deer feeding on their property (p < 0.001) (Fig. 5). Again, model fit was improved with the feeding model compared to the region model (ΔAIC = 379).

Figure 5.

Model coefficients (points) and 95% confidence intervals (lines) from linear regression models for the outcome of total risk score for contact with wild white-tailed deer. Orange dashed line indicates no association (Color figure online).

Examining the demographics of those in the 90th percentile of risk score or higher (values of 108 or higher), there are a higher proportion of younger individuals (51.4% under the age of 40 compared to 34% of the lower risk group). High risk individuals are also significantly more likely to live in a rural area (40.8% vs 24.5% in the lower risk group), to be a hunter (38% vs 13.7%), meat processor (18.3% vs 3.8%), taxidermist (4.2% vs 0.5%), or biologist (15.5% vs 5.2%). High risk individuals were also more likely to report deer feeding on their property (65.7% vs 49.6% of the lower risk group).

As the presence of deer feeding on a participant’s property was consistently important to the best fit models, we conducted a post-hoc logistic regression analysis of factors related to that variable. We considered lot size, residence area type, region, and the proportion of the county that was considered deer habitat to be predictor variables. The latter was calculated by calculating the percentage of pixels within the county that were labeled deer habitat by the USGS Gap Analysis Project (Aycrigg et al., 2013). This analysis found that larger lot sizes (1–10 acres, OR = 2.6, or > 10 acres, OR = 3.6, compared to no lot) and a higher % of deer habitat in the county (OR = 3.2) were strongly associated with a higher probability of deer feeding on the participant’s property (Fig. 6, Supplemental Table 2). Suburban and other residence areas were also significantly associated with higher probabilities of deer feeding, although the effect was much smaller (OR = 1.2 and 1.7, respectively).

Figure 6.

Odds ratios (points) and 95% confidence intervals (lines) from logistic regression models for the outcome of having wild WTD feed on their property. Orange dashed line indicates no association (Color figure online).

Discussion

Our findings suggest that Illinois residents surrounded by deer habitats and those that hunt will have higher risk of close contact with wild white-tailed deer (either dead or alive). Our findings demonstrate the utility of human dimensions data for better understanding factors that determine where and in which human groups deer-human disease transmission might be highest and provide useful background information for conducting this type of study more broadly.

Illinois residents that lived in rural areas had a higher overall risk score, but residence area type was not significantly associated with either proximity or frequency of deer interaction. Deer feeding on a property was a better predictor of deer interaction, which is understandable; rural areas are associated with more deer habitat, but do not guarantee the presence of deer on a property or contact with these deer. Wild white-tailed deer were mostly reported in the comments to feed on vegetation or in gardens (as has been reported before (Cornicelli et al., n.d.; Hadidian, 1991), which could be a mechanism of indirect contact if humans frequently tend the gardens; gardens are common but not ubiquitous in rural and suburban areas alike. The post-hoc model with deer feeding as an outcome found a stronger association with deer habitat at a county level, as well as lot size and residence area. Interestingly, suburban and exurban areas were more associated with deer feeding on properties than rural areas, although this may be due to a perception bias related to damage to gardens.

In Illinois, there is a high reproductive rate in WTD 85.8% of adult deer and 20.5% of fawns are pregnant with a litter size that increases with age (Green et al., 2017), and high deer densities are common in suburban areas, and human interaction with these deer have been studied with regards to deer management (Urbanek and Nielsen, 2012). One previous study failed to find WTD in suburban yards in North Carolina, despite showing them in nearby woodlands (Kays and Parsons, 2014). However, this was based on 5 spring/summer months of camera trapping, which may have failed to capture winter visitations (Anderson et al., 2011) and may not be representative of all suburban areas. Other research has identified WTD living in suburban woodlots, with smaller home ranges than neighboring rural areas and generally crepuscular activity (Cornicelli et al., 1996); several respondents to this survey voluntarily reported seeing deer in this time period. The current survey did not ask when deer fed on the property; future surveys may benefit from adding this question—several respondents commented that there were specific times of year in which they saw deer more, and that answers to frequency questions were seasonally variable.

Although we did not include lot size in the primary models due to potential multicollinearity, we suspect that the disconnect between rurality and deer feeding may be due to an overlap with lot size, as 49/75 respondents with lot sizes of > 10 acres and 47/134 with lot sizes of 1–10 acres were in rural areas; yard size is known to correlate with wild herbivore visitation (Johansson and DeGregorio, 2023). Of the 59 respondents from rural areas who reported no deer feeding on their property, 43 had < 1 acre of property and 6 had none. In addition, rural respondents came from counties with a wide range of deer habitat distributions, from 4 to 86%, with 27/177 coming from counties with < 20% deer habitat by area. Therefore, rurality may not be as good a measure of potential deer contact as lot size and amount of local deer habitat. In addition, previous studies have found that deer behavior may change in suburban and urban areas due to differences in management, resulting in deer with small flight distances and potentially leading to closer contact distances for people in these areas (Honda et al., 2018). In addition, the distribution of houses may affect deer habitat, as WTD in Illinois exurban areas have been found to prefer forest edges and therefore be more common near linearly arranged houses than clumped residence areas (Anderson et al., 2011).

One caveat to the analyses regarding home property is that many respondents commented that they interacted with deer away from home, which may limit the value of the data on their home environment. Many referred to seeing deer while walking in parks, nature preserves, and cemeteries, or while visiting family or friends, and hitting deer with their vehicles was a common report. In particular, a large number of respondents commented on the presence of a herd of deer on the Western Illinois University campus, including fawns being hidden in a high-traffic area of campus. However, most of these locations of deer contact were in areas near to the respondents’ homes, which may explain why the presence of deer feeding on respondents’ property was strongly associated with deer contact. In addition, people commented on aspects of their property that may impact deer presence, either decreasing through fencing for exclusion or increasing due to larger neighboring properties or nearby creeks, parks, and wooded areas, as has been noted previously (Fehlmann et al., 2020). Thus, future surveys would benefit from additional questions regarding where contact with deer occurs and assessing other factors about the place of contact that may affect deer presence. One other aspect of deer contact frequently mentioned in the comments but not well-addressed in the survey is interaction with scat—many referred to deer feces in their yard, and several commented on their dogs regularly eating deer feces.

Occupations that involve deer exposure will have a higher risk of close contact with wild white-tailed deer. Occupations listed in the survey were biologist, veterinarian, animal control, taxidermist, meat processor, hunter, and rehabilitation specialist. Due to the special nature of hunting as a category, we separated hunters from other risky occupations and hobbies in our models. From our findings, hunters had a greater probability of contact with wild WTD than any other occupation listed, but this may be due to sample size; hunting was the most commonly reported of the occupations and hobbies. We saw that our highest risk group was also more likely to report several of the other hypothesized high-risk activities, including meat processing, taxidermy, and working as a biologist. Importantly, many of these occupations will have close contact primarily, if not entirely, with dead deer. While this may be a risk for deer-to-human transmission, these contacts are not pathways of human- to-deer transmission. Future studies would benefit from carefully differentiating how different types of human-deer interactions might contribute to each direction of cross-species transmission.

Outside the context of high-risk occupations and hobbies, close contact with deer was rare, and increasingly infrequent at close range (Fig. 3). For most individuals in this survey, potential exposure to wild alive WTD is likely to be a rare event. Other work in this region has found that these rare encounters are considered memorable, and are highly valued as a way of connecting with nature (Hicks, 2017). However, their rarity means that the risk of zoonotic transmission might be low for these individuals. Messaging around WTD safety might therefore be targeted at higher-risk individuals, such as hunters and those likely feeding deer or to have deer feeding on their property (large lot sizes in areas with high amounts of deer habitat). However, some respondents did comment that deer in certain areas were less fearful of people than in others, and that the startle distance of these deer was noticeably decreasing over time. Therefore, even low-risk individuals may be found to have close, high-risk contact with deer in areas of high deer density. It is unclear what impact messaging about zoonotic disease transmission risk might have in these different populations with regards to behaviors and attitudes towards wild deer (Decker et al., 2012); further research is required to design appropriate and meaningful risk communication approaches to decrease cross-species transmission potential.

Some caveats must be highlighted with these results. First, the participants were recruited through convenience sampling, and so may not be fully representative of the population. In particular, the respondent group was dominated by two university towns (University of Illinois Urbana-Champaign and Western Illinois University), which may have resulted in over-representation in a particular demographic. In addition, all responses were self-reported; the high value that people place on interactions with wildlife (Hicks, 2017) may result in some amount of recall bias and over-reporting of contact. Lastly, it is clear that not all contact is equivalent in terms of zoonotic transmission risk. No questions were asked about the duration of contact—processing a hunted carcass and moving the leg of a coyote kill, for example, would have very different risk profiles. Future research would benefit from careful delineation of various risky activities, with close attention to the direction of potential transmission (deer to human vs. human to deer).

Conclusion

In this study, we identified risk factors of contact with wild white-tailed deer by exploring human and deer interactions in Illinois. Transmission of SARS-CoV-2 between human and deer has been identified in the United States, but the exact mechanism and transmission pathway(s) has not been identified. The findings in this survey suggest that direct human-to-deer transmission is likely to be rarer than the reverse in many contexts.

Supplementary Material

Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1007/s10393-024-01694-7.