Dominant attitudes and values toward wildlife and the environment in coastal Alabama

Surveys assessing attitudes and values about the environment can help predict human behavior toward wildlife and develop effective conservation goals along-side local communities. Coastal Alabama, in the southeastern United States, is a hotspot for biodiversity and endemism and needs protection. Land and wildlife management practices in Alabama have moved from indigenous-led, which is more in harmony with the environment, to larger-scale exploitative uses for agri-culture and plantations. We therefore predicted that a large proportion of the population has a dominant view of the environment in which land and wildlife are primarily for human benefit. To test this hypothesis, we surveyed over 1300 residents in Mobile and Baldwin counties — the two southernmost counties in Alabama — to assess attitudes toward local vertebrate wildlife, knowledge of the


| INTRODUCTION
People have different reasons for how they feel about wildlife, including utilitarian value or symbolic meaning, religious or spiritual significance, as a source of fear or attraction, or as a barometer for measuring one's concern over environmental sustainability (Manfredo, 2008). Understanding attitudes toward wildlife is important for predicting human behavior and conservation outcomes (Bath et al., 2022;Jacobs et al., 2014;Kontsiotis et al., 2021;Manfredo et al., 2009). Attitudinal surveys have been a powerful tool in conservation (Ajzen, 1991), helping institutions formulate policies and management actions that incorporate local opinions and the likelihood for acceptance and compliance, a key part of conservation success (Karanth et al., 2008). Negative attitudes against wildlife can undermine conservation initiatives and wildlife sustainability (Mogomotsi et al., 2020).
In addition to attitudinal surveys, quantitative assessment of values about wildlife are increasingly common (Bennett et al., 2017;Fulton et al., 1996;Manfredo et al., 2017;Manfredo et al., 2018;Purdy & Decker, 1989). Whereas attitudes are fast-forming and can be adapted to different situations, values represent underlying beliefs that serve as a foundation of actions (Homer & Kahle, 1988). Understanding one's values, in combination with attitudes, can help conservationists decide the best methods of intervening in a conflict situation. Values about wildlife are more difficult to assess than attitudes because the former cannot be measured directly, but rather must be inferred from statements of belief and expressions of opinion (Purdy & Decker, 1989).
Previous studies have quantitatively measured values (e.g., Fulton et al., 1996) and assessed their predictive potential (Jacobs et al., 2014) through the development of scales of value orientations. In the United States, it has become easier to assess both the attitudes and values of Americans quantitatively through a project called "America's Wildlife Values" , in which an individual's wildlife value orientation is determined by their score on mutualism and domination scales (Teel & Manfredo, 2010). Domination beliefs center on hunting and use of wildlife, while mutualist beliefs encompass caring and social affiliation. Mutualists tend to see wildlife and the environment as part of their social network and aim to live in harmony with it, while dominant individuals believe that the environment and wildlife should be used and managed for the benefit of humans .
Alabama is one of the states in the United States where the environmental outlook of inhabitants has been heavily shaped by European colonialism (Whyte, 2018). The Indian Removal Act of 1830 forcibly removed nations across the southeast from their ancestral homelands (Doran, 1975), leading to a large influx of white settlers into the Mississippi River Valley over a very short period. Indigenous people do not see themselves as separate from the land, but rather as a part of it, in turn shaping their land management strategies (Garnett et al., 2018;Schuster et al., 2019;Whyte, 2018). The swift demographic change that occurred in Alabama during the European colonization rapidly and fundamentally shifted the wildlife value orientation from mutualism to domination . In a recent survey, 14% of individuals from Alabama indicated that they have hunted in the last 12 months, 8% said they have hunted not in the last 12 months, and 35% indicated that they have not hunted before but plan to in the future . Nationwide, Alabama ranks 10th among states with the most active hunters . Although hunting in Alabama has been shown to be undertaken for many reasons including connecting to nature, socializing, managing deer populations, and as a source of local meat (Birdsong et al., 2022;Mehmood et al., 2003), it is firmly considered a dominant activity.
In this study, we test the prediction that Alabamians should exhibit a greater domination mindset over the environment compared to the national average, and that hunters in particular would be less willing to support actions to protect it. We carried out a survey in two populous counties-Mobile and Baldwin counties (Figure 1), which form the Gulf coast of Alabama-to assess knowledge, values, and attitudes toward the area and its wildlife. We analyze survey results according to demographic data, hunter status (hunter vs. non-hunter), and possession of pets, as we note that people who have pets may view wildlife as a threat to those pets and act differently (Bowes et al., 2015;Frank, 2016;Lute et al., 2016). The study area is particularly rich in biodiversity (Jenkins et al., 2015), especially around the Mobile-Tensaw Delta, harboring several endemic species (e.g., Buhlmann et al., 2009;Moreno et al., 2022). Climate change and increased urbanization are predicted to strongly affect coastal areas along the Gulf of Mexico (Anderson et al., 2013;Mulholland et al., 1997;Rabalais et al., 2007;Scavia et al., 2002), with important consequences for native flora and fauna.
Despite the incredible diversity harbored in southern Alabama, conservation management aimed at protecting native species and their habitats (Falk & Millar, 2016;Scavia et al., 2002) is often hampered by individual actions and attitudes (Hare et al., 2021;Marshall et al., 2007). Therefore, successful conservation actions cannot occur without a proper evaluation of stakeholders' attitudes and values (Fox & Bekoff, 2011;Heneghan & Morse, 2018;Lee, 2017;Manfredo et al., 2021). Our work indicates how much people in southern Alabama know of and value the diversity of the area in which they live and their general attitude toward local animals. These data will help determine whether individuals would be supportive, and therefore increase the success, of conservation actions (Jordan et al., 2020;Kansky et al., 2016;Marshall et al., 2007) in the area and whether education about the natural environment and its value should be developed for outreach activities and in the classroom to improve literacy about human-wildlife interactions.

| Data collection
We developed a 12 question survey, some of which contained sub-questions (Supplementary Materials). The goal was to assess whether people are familiar with and value the biodiversity of the area where they live/work, how much they would be willing to invest in its protection, and their attitudes toward wildlife. Specifically, the survey focused on attitudes toward non-marine wild vertebrates (hereafter referred to as "vertebrates"), which includes both terrestrial and some freshwater vertebrates (fish were not included in our survey). An initial question (Q1) asked if individuals had previously taken the survey, in which case the survey was retained only if they answered "no." The survey included four categories of questions. Category 1 (Q2-7 and Q9-10): information about the participants and their familiarity with vertebrates in their area.
This included the vertebrates that someone would normally see in their daily life and how many vertebrates they see on average during a given week. We also inquired if the person has any pets or if they do any outdoor activities, as this may increase the chance to have wildlife encounters and influence an individual's attitude toward them. Category 2 (Q8): hunting-related questions, including if the participant hunts, what they hunt, and what they use to hunt. Category 3 (Q12): demographic questions including zip code, gender, age, annual income, highest level of education, and school (if a child). We note our shortcoming in only identifying two genders-females and males-although people could decide to not respond to this question if they identified with neither of the two. Category 4 (Q11): questions assessing the attitudes of people toward wild vertebrates and the value they give to the area of study and its protection. This included types of responses to encounters with vertebrate wildlife, if the response changed depending on which animal was encountered, how well they know the biodiversity of the area, and how much the individual would be willing to monetarily invest to increase protection of nature in Mobile/Baldwin counties in Alabama. Questions were developed based on experiences and conversations that some study authors had with students and citizens about how much they knew about the biodiversity of the area in which they lived, how interested they were in preserving it, and whether some of them regularly hunt for food.
Surveys were distributed as paper copies at schools, cafes, supermarkets, flea markets, gas stations, gyms, shops, hospitals, natural parks and areas, and the University of South Alabama campus in Mobile, Alabama. We F I G U R E 1 Map of the study area. The survey was carried out in Mobile and Baldwin counties, the two southernmost counties in Alabama, which border the Gulf of Mexico. targeted different neighborhoods, age groups, and areas to obtain a sample that reflected the diversity of people living in Mobile and Baldwin counties. Paper copies were never left unattended and personally handed to everyone taking the survey. The survey was also distributed electronically through a website hosted by the University of South Alabama and accessible to everyone including those not affiliated with the university. Surveys were carried out between September 2017 and July 2018. Following survey collection, demographic responses were then compared to demographics from the 2020 U.S. Census for the state of Alabama, and Mobile and Baldwin counties (U.S. Census Bureau, 2020). All data collected from this survey are available as Supplementary Materials.

| Data analysis
All surveys were reviewed to ensure that responses were real by flagging individuals who always checked the same option (e.g., always first or last responses), took the survey multiple times, stopped taking the survey halfway, did not answer more than half of the questions, or wrote nonsense (e.g., they regularly see dinosaurs) in some of the open answers.
Analyses were run to investigate the relationship between hunter status (Q8.1) and the attitude toward different wild animals (Q11). We first built two-way contingency tables for each wild animal to show the distribution of counts for the two categorical variables of hunter status and reaction. We used Chi-square and Fisher's exact tests in R v4.1.2 (R Core Team, 2021) to test whether the distributions of reactions were significantly different between hunters and non-hunters. We also ran the analysis using gender, demographic information, pet ownership, and whether individuals spent time doing outdoor activities (regardless of hunter status) as factors that could influence the response. The analyses were repeated using a reduced model where reaction levels with less than five entries were removed from the tests. We also tested if hunters differed from non-hunters in terms of knowledge of the biodiversity of the area (Q11.29), if they think more should be invested in protecting the Mobile-Tensaw Delta (Q11.30), and how much they would invest to protect this unique area (Q11.31). As results from the Chi-square and Fisher's exact tests were always in agreement, we report only results based on the Fisher's test for the full model and the Chi-square test for the reduced model, as the Fisher's test performs better for the full model in the presence of small counts (N < 5). To test how well attitudes can be used to predict values, we use the multinom function from the nnet package (Venables & Ripley, 2002) in R to run multinomial logistic regression to test whether participants whose attitudes favored "killing" or "killing to eat" wildlife would have values oriented toward knowing and caring less about the environment and its protection. For Q11.31, responses of willingness to pay were combined, where $20 or less was treated as less supportive, while more than $20 was treated as more supportive.

| Data visualization
Data visualization was performed using Tableau Desktop software, version 2022.1.8. No data manipulation or transformation was performed for visualization purposes except for pivoting and recoding the original dataset. Data are visualized across 11 tabs, with each tab featuring a dashboard with a title, one or multiple charts, and large "call out" numbers. Descriptive subtitles are often provided to assist the audience in interpreting the visuals. Hovering over the charts provides additional context. Some dashboards include one or multiple drop-down menus, allowing the dashboard user to make selections and customize the view.

| Data and demographic information
All data collected in this survey can be visualized in an interactive tool developed for this study (https:// public.tableau.com/app/profile/ylenia.chiari/viz/Weber Herteletal__16599728420070/1Demographics-AtGlance) and are also fully available as Supplementary Materials. Our final dataset consisted of 1307 survey entries. The total combined population of Mobile and Baldwin counties according to the 2020 U.S. Census is 636,444 (U.S. Census Bureau, 2020). Therefore, our survey represents 0.2% of the combined population of those counties. The majority (66%) of individuals taking our survey identified as female. The most common age ranges of people who responded were from 13 to 19 (33%) and 20 to 30 (25%) years old. According to the U.S. census, females make up 52% of residents on average in Mobile and Baldwin counties, with 53% of residents between the ages of 18 and 65 years (U.S. Census Bureau, 2020). This shows that our survey demographics had a slightly higher representation of females than males, but similar representation of individuals 18-20 to 60-65 in comparison to the U.S. Census.
Most respondents had a college degree (38%) or a high school diploma (24%), while 86% of participants had a high school diploma or higher and 53% had a bachelor's degree or higher. The U.S. Census indicates that 89% of residents over the age of 25 in Mobile and Baldwin counties hold a high school diploma or higher, and 28% of residents hold a bachelor's degree or higher (U.S. Census Bureau, 2020). Thus, our survey respondent demographics were similar to the U.S. Census for those who have at least a high school diploma but are overrepresented for those with a bachelor's degree or higher.
The proportion of respondents was almost equally distributed across the following three annual household income categories (Q12.6): <$30K (15%), $30-70K (18%), and >$90 K (15%); a lower number of the participants (5.5%) were in the $70-90K category and 24% did not answer this question. In our survey, 15% of the participants are considered to live in poverty, earning less than $30K per year.
Among survey respondents, 1000 (77%) identified as non-hunters, 265 (20%) as hunters, and the remainder (3%) did not respond to this question (Q8.1). Hunters were almost equally represented by males (53%) and females (46%). Most hunters were in the 13-19 (40%) and 20-30 (26%) age ranges, which is a similar breakdown to the age stratification in the survey as a whole. Thirty-six percent of hunters in our survey had a 4-year college degree, while a smaller percentage had either a middle or junior high school degree (24%) or a high school degree (22%). Finally, although 30% of hunters did not respond to the question about income, those who answered had very different incomes, with the two largest annual income categories being below $30K (13%) or above $90K (17%). Pet ownership status was very similar in terms of the type of pet owned between hunters and non-hunters, although a higher percentage of hunters than nonhunters had at least one pet (85% vs. 75%, respectively), and overall, more hunters had dogs than non-hunters (80% vs. 64%, respectively). Data visualization tabs 1-4 show the complete demographic break down.

| Attitude toward wildlife
When people were asked how they would react to seeing different wild animals (Q11.1-Q11.28), there was a significant difference in the reaction between hunters and nonhunters for all animals ( p < .05) except for salamanders (Table 1, Figure 2, and Data visualization tabs 5-7). Although both hunters and non-hunters would tend to either ignore or observe most encountered animals, a larger percentage of hunters than non-hunters answered "kill" (65% vs. 17%, respectively; χ 2 = 217.7, df = 1, p < .05) or "kill to eat" (71% vs. 5.7%, respectively; χ 2 = 530.38, df = 1, p < .05) to at least one type of animal ( Figure 2, Data visualization tab 7). Furthermore, when the encountered animal is considered "dangerous" such as an alligator, a bear, a cougar, a boar, or a snake, most non-hunters would run away, while the majority of hunters would "observe" it (alligator and bear), "run away" (cougar), "kill" (snake), or "kill to eat" (boar) (Data visualization tabs 5 and 6). The number of hunters responding that they would "kill" or "kill to eat" strongly depended on the type of animal encountered. Responses of "kill" and "kill to eat" were around 3% for lizards, salamanders, owls, turtles, heron, and egrets, but reached greater than 30% for animals that are commonly eaten (e.g., dear, boar) and snakes (Data visualization tab 5). Overall, the percentage of non-hunters that would "kill" or "kill to eat" an animal was less variable, generally being 5% or lower depending on the encountered animal; however, this percentage increases to 14% of non-hunters when the animal is a snake (Data visualization tab 5). Correspondingly, 46% of total participants (hunters and non-hunters) indicated that they would "kill" (44%) or "kill to eat" (1%) a snake (Data visualization tab 5). Finally, 27% of all respondents answered "kill" as an attitude toward animals at least once, 13% answered "kill" more than once, 19% answered "kill to eat" at least once, and 14% answered "kill to eat" more than once. Therefore around 46% of the participants, regardless of hunting status, answered that they would kill an animal (to eat or just to kill) at least once (Data visualization tab 7), and 27% of all responded that they would "kill" or "kill to eat" more than once.
When attitudes toward wildlife were analyzed considering demographic information and participation in outdoor activities regardless of hunting status, we found that gender, age group, education, income, and type of activity performed outdoors were significant ( p < 0.05) factors influencing the reaction to all wild animals except for responses to owls and seagulls by gender and armadillos by level of education (Table 2). Across demographic categories and animals, the most common response was to "observe"; however, there were differences among demographic groups when responding "kill" and/or "kill to eat". More males than females would "kill" (35% vs. 23%, respectively) or "kill to eat" (27% vs. 14%, respectively). Across every income category, more than 20% of people answered "kill" at least once; but answering "kill to eat" was more prominent for incomes >$70K (21% for income between $70 and 90K and 19% for income >$90K). Finally, 36% of respondents in middle or junior high school would also "kill" and "kill to eat," which was higher than any of the other educational categories. Further details and breakdown of responses by demographic categories and animals can be seen in Data visualization tab 6.

| Knowledge and value of the area
Hunters and non-hunters did not differ in the number of vertebrates observed on average per week (Q7.1 and Q8.1) (p = .33) but did differ in their knowledge of the area (Chi-square test, p = .0002) (Q11.29, Data visualization tab 8). Although many hunters and non-hunters responded that Mobile/Baldwin counties have a similar level of biodiversity (i.e., number of wild animals) as other places in the United States (Q11.29; 56% hunters, 42% non-hunters), more non-hunters (38%) than hunters (26%) answered that they did not know. A lower percentage of individuals in both groups (20% non-hunters vs. 17% hunters) recognized this area as having a different level of biodiversity (higher or lower) than the rest of the United States (Data visualization tab 8).
When individuals were asked whether the number of protected areas in Mobile/Baldwin counties should increase, decrease, or stay the same (Q11.30), significantly (Chi-square, p < .0001) more non-hunters (56%) compared to hunters (43%) favored increasing protected areas, while more hunters (5%) than non-hunters (0.7%) would opt to decrease protected areas. Lastly, more hunters (33%) than non-hunters (22%) preferred to keep protected areas the same (Data visualization tab 10). When people were asked how much they would be willing to pay per year to maintain or increase protected areas (Q11.31), there was no significant difference T A B L E 1 Comparison of attitudes toward wildlife (Qs 11.1-11.28) between hunters and non-hunters (Q8.1).

Animal
Chi-square df p-Value between hunters and non-hunters (p = .05). The majority of both groups (55% of hunters vs. 51% of nonhunters) responded either that they would not be willing to invest money or would invest very little (maximum $10 per year) to protect these areas (Data visualization tab 10). Females and males tended to observe similar numbers of animals in a week (Q7.1 and Q12.1, p = .06) (Data visualization tab 9). When people were asked about the level of biodiversity of the area compared to other places in the United States (Q11.29), we found a significant ( p = .003) difference in response between females and males. More females (39%) than males (28%) answered that they did not know, while less females (42%) than males (50%) responded "yes" to this question. We also found that genders differed in the number of protected areas desired, with 56% of females responding "increase" versus 47% of males (p = .0001, Q11.30), with females willing to pay more to maintain or increase these areas than males ( p = .03, Q11.31) (Data visualization tab 11).
The age range of people taking the survey (Q12.2) influenced the number of vertebrates seen on average per week (Q7.1) (p = .001) and what they know about the area (Q11.29, p = .02). Within each age category, around 80% of participants responded that the study area has a similar number of wild animals to the rest of the United States or that they did not know (Data visualization tab 9). Age also significantly influenced how people feel about protected areas in Mobile/Baldwin counties (Q11.30, p = .001) and how much they are willing to invest for its protection (Q11.31, p = .0001) (Data visualization tab 11).
We found that people with different levels of education (Q12.5) also saw different numbers of vertebrate animals per week on average (Q7.1) ( p = .008) and differed in their knowledge of the diversity of the study area compared to the rest of the United States (Q11.29, p = .0002) (Data visualization tab 9). Across all educational categories, ≥60% of participants responded either that the level of biodiversity in the study area is the same as in other places in the United States or that they did not know, although a higher proportion of people with a PhD (33%) answered that the level of biodiversity was different. Education also influenced the amount of desired protected areas (p = .001, Q11.30). Although the majority in each educational category indicated that they would increase protected areas, the percentage was higher for people in graduate school (>60%) compared to other categories. Education also influenced how much participants are willing to invest for these protected areas ( p = .002, Q11.31) (Data visualization tab 11).
Finally, income significantly affected how many animals people saw on average per week (Q7.1, p = .0004), how much they know about the area (Q11.29, p = .03), how much protected area they prefer (Q11.30, p = .0001), and how much they would be willing to invest in its protection (Q11.31, p = .0001) (Data visualization tab 9). F I G U R E 2 Summary of responses to encountering terrestrial vertebrates based on hunting status. Percentages are calculated as the frequencies of hunters and non-hunters that chose a specific response type (e.g., "call animal control") at least once to any animal out of the total number of participants that answered that question. Hunters are indicated in green and non-hunters in brown.
T A B L E 2 Comparison of attitudes toward wildlife (Qs 11.1-11.28) as a function of demographic information. Within each income category, around 80% of participants indicated that the study area is similar in biodiversity to the rest of the United States or that they did not know. Furthermore, while ≥50% of participants within each income category suggested to increase protected areas, a higher proportion of individuals with income ≥$91K (27%) desired to keep it the same. A higher proportion of participants earning ≥$70K (26%-27%) also indicated that they would be willing to spend nothing for protection (vs. approximately 20% in other income categories).

| Relationship between attitude toward wildlife and knowledge and value of the area
We tested whether a more dominant attitude toward wildlife (Qs 11.1-11.28) was associated with knowing and caring less about the environment and its protection (Qs 11.29-11.31). We found no relationship between those answering "kill" or "kill to eat" and knowledge of their area's level of biodiversity (Q11.29, p = .75). This result does not change when hunting status was considered (p = .34). However, answering "kill" or "kill to eat" tended to increase the probability of answering "decrease" to the amount of desired protected areas (Q11.30, p = .03). Furthermore, when hunting status was considered, non-hunters were 13% less likely to answer "decrease" than hunters (p = .0004). We found that answering "kill" or "kill to eat" (Qs 11.1-11.28) did not influence how much the participants would be willing to invest to protect the area (Q11.31, p = .55), and this result did not change when analyzing hunters and nonhunters separately ( p = .9).

| DISCUSSION
We collected information on attitudes and values about local wildlife from 1307 individuals in Mobile and Baldwin counties in southern Alabama, which represents approximately 0.2% of the combined population of these counties. We had more than double the number of participants than a statewide study of wildlife value orientations in Alabama , indicating the strength of our sample. The demographics of participants in our survey largely reflected those reported in other studies. In general, females were more likely to respond than males (Curtin et al., 2000;Moore & Tarnai, 2002;Singer et al., 2000), younger people were more likely to respond than older people (Goyder, 1986;Moore & Tarnai, 2002), and more educated and more affluent individuals were more likely to respond than less educated and less affluent individuals (Curtin et al., 2000;Goyder et al., 2002;Singer et al., 2000). Our survey had a similar number of self-identified hunters compared to a previous statewide study of Alabama (Dietsch et al., 2018), and the proportion of male and female hunters and non-hunters in our survey reflects what has been found at the state level.

| Dominant attitudes toward wildlife
We based our interpretation of attitudes and values toward and knowledge of wildlife and the environment on the mutualism versus domination framework of Teel and Manfredo (2010) and Manfredo et al. (2018). Specifically, according to this framework, social affiliation (the desire to be in and around nature) and caring for the environment are identified more as mutualism, while hunting and use of wildlife are characteristics of domination (Teel & Manfredo, 2010). Our working hypothesis was that, due to its colonization history, we would find a more dominant attitude for the study area compared to the rest of the United States, similar to what has been observed for Alabama as a whole based on America's Wildlife Values .
Hunting is strongly associated with a domination orientation (Teel & Manfredo, 2010). In Alabama, hunters make up approximately 22% of the population  and 20% of respondents in our study. Despite the more urban landscape of Mobile and surrounding areas, the proportion of hunters in our sample was similar to the rest of Alabama and higher than the majority of states in the United States. Beyond hunting status, nearly half of all individuals in our study answered "kill" or "kill to eat" for at least one animal, providing further evidence of widespread dominant attitudes in southern Alabama regardless of hunting status. Responding "kill" or "kill to eat" was highest among the youngest ages (5-19 years old) and the most affluent economic groups (>$70K) and did not change with increasing levels of education except individuals with an advanced degree (e.g., Ph.D.).
Given the large proportion of participants, including non-hunters, that answered "kill" or "kill to eat" for at least one animal, we wanted to further understand the drivers of this attitude. Lethal removal of an animal is an example of taking wildlife that is not considered hunting. We surveyed responses to encountering animals that are considered potentially harmful to livestock, humans, or pets (i.e., coyote, bear, snake). Of the respondents who said they would either "kill" or "kill to eat" a coyote, 31% were hunters and 2.7% were non-hunters, indicating that hunters also shown a more dominant attitude toward lethal removal. We also asked respondents if they owned pets, which could affect their attitude toward a certain species . According to our survey, hunters and non-hunters had a similar percentage of pet ownership (85% of hunters and 75% of non-hunters), suggesting that having a pet does not explain the large difference observed for lethal responses to coyotes between these two demographic categories. Rather, the domination orientation of hunters is probably a better explanation for this difference, even when killing is not necessarily as a source of meat. Similar results were also obtained when asking about bears.
Conversely, snakes appear to trigger a dominant attitude from both hunters and non-hunters alike as 46% of respondents answered "kill" or "kill to eat" for snakes. While people are more likely to kill a venomous than a non-venomous snake, they cannot identify the snake species in question more than half of the time (Vaughn et al., 2022). Although the percentage of respondents that would kill a snake was higher for hunters than nonhunters in our study (32% vs. 13.6%, respectively), our findings reflect a general lethal attitude toward snakes that matches previous research. Attitudes toward snakes have been shown to be independent of the type of snake-venomous or non-venomous-and seems to be driven by general fear or disgust for snakes (Coelho et al., 2021;Crawford & Andrews, 2016;Onyishi et al., 2021). There are over 40 species of snakes that occur in Alabama, and all but six species are non-venomous. Many non-venomous species are often mistaken for venomous species by the general public; one example is the seven species of harmless water snakes (Nerodia spp.), which are often mistaken for venomous Cottonmouths (Agkistrodon piscivorus) due to their similar color and banded pattern. Additionally, there are nine snake species that are rare or endangered in Alabama and illegal to capture or kill (Alabama Division of Wildlife and Freshwater Fisheries, 2021). The public's behavior toward snakes clearly has major conservation implications, as snakes are important species for ecosystem function (Willson & Winne, 2016).
Based on our survey, we found that dominant attitudes toward wildlife are heavily skewed toward hunters. While snakes draw lethal attitudes from both hunters and non-hunters, once this group of animals is accounted for, hunters responded in much greater percentages compared to non-hunters that they are willing to kill even when it is not as a source of food such as a coyote or a bear. These results fit with the statewide study of "America's Wildlife Values" , which showed that only 23% of hunters or anglers surveyed support hunting because it is a source of local food, suggesting that hunting has other utilities or sources of meaning for self-identified hunters.

| Knowledge and value of the environment
While much of our survey addresses attitudes toward wildlife, we also asked more value-oriented questions dealing with knowledge of the area and value given to protected areas. These questions asked respondents about their perceptions of the amount of wildlife in Mobile/ Baldwin counties in comparison to other places in the United States, whether they thought that counties should increase, decrease or keep the same amount of protected areas, and how much they would be willing to pay per year. Traditionally, the sale of hunting and angling licenses has funded conservation and protection of habitats, but wildlife viewing is a newer activity that has been studied more recently and does not come with a cost (Sinkular et al., 2022). This may influence how much people are willing to spend to protect habitats they are viewing but not actively hunting or fishing in.
Survey responses indicated that most participants are not familiar with the uniqueness of the biodiversity of the area compared to the rest of the United States, especially young and middle-aged individuals below 60 years old. In addition, most participants across demographic groups are also not willing to invest much (in general nothing or less than $30) to protect it, although our data did show that more non-hunters than hunters favored increasing protected areas. Individuals with at least a 4-year college degree also favored increasing protected areas. Counterintuitively, individuals in higher income groups are less willing to spend money to increase protected areas. Previous studies have not definitively concluded a relationship between wealth and concern for the environment. In climate change research, the wealthy are responsible for the majority of carbon emissions despite their higher level of concern about the environment and willingness to pay to protect it (Nauges et al., 2021;Oswald et al., 2020), while lower income individuals have not contributed nearly as much to environmental crises, but may not say that they are very concerned about the environment or are unable to pay more to protect it (Fairbrother, 2013;Franzen & Vogl, 2013).
Our analyses indicate that the dominant attitudes of "kill" or "kill to eat" an animal are not predictors of the knowledge a participant had of the area and of how willing the person would be to invest for its protection. However, the more often an individual responded "kill" or "kill to eat" to animals, the less protected areas they would want. This result was independent of hunting status and suggests a strong relationship between a dominant attitude and behavior toward animals and value given to the environment for human use.

| Conservation implications
Overall, we found that most individuals who were supportive of killing animals self-identified as hunters. This dominant attitude goes beyond hunting for meat and includes lethal control of animals or killing for other reasons. Non-hunters, on the other hand, only appear to support lethal removal of snakes, but not other animals. Increased formal education does not appear to decrease the dominant attitude of hunters or non-hunters. Since hunting for food can have important social, health, and economic benefits, a clear conservation goal arising from our study should be targeted education to prevent killing animals for reasons other than as a source of meat. This education should happen at an early age (K-12), as our data show that it is the youngest groups (middle/high schoolers) that are the most supportive of killing animals either for food or just to kill. Educational programs in schools or camps that bring in federal or state agencies or NGOs to talk about conservation status of different species, their role in ecosystems, and responses to wildlife encounters could go a long way to improving unnecessary lethal removal of animals. This targeted education would be especially beneficial if it teaches young people to identify and respond appropriately to snakes, coyotes, bears or other wildlife perceived as harmful.
One of our surprising findings was that respondents were overwhelmingly unaware of the unique biodiversity of their region, which could influence their actions toward species as well as their underlying value system. Southern Alabama, including the Mobile-Tensaw Delta, and neighboring areas are among the most biodiverse in the United States, containing assemblages of species that rival other biodiversity hotspots around the world. Given that most federally protected areas are in the western United States (Jenkins et al., 2015), this lack of knowledge fits with a general shortage of appreciation and protection accorded to the biodiversity of the southeastern United States. Communicating to the local population about the types of wildlife that inhabit local spaces and the benefits these species provide can have a positive effect on outcomes when encountering wildlife (e.g., Ballouard et al., 2013;Pinheiro et al., 2016).
While environmental education, especially targeted toward younger ages, could help change behavior toward local wildlife, it may be difficult to influence hunters, especially due to the politics surrounding environmentalism in the United States (Blumstein & Saylan, 2007;Dunlap et al., 2001). In Alabama, hunters are already required by law to sign off on and carry each refuge's hunt brochure to legally hunt in that area, which includes which species may be hunted and by what methods (e.g., brochures located on the Outdoor Alabama website (Wildlife Management Areas, 2021)). If hunters are not abiding by the rules outlined by law, it is unlikely that environmental education would affect their choices. This suggests that conservationists could focus additional efforts on protecting species that have been identified as "kill" (rather than "kill to eat") in our survey, which could represent instances of illegal hunting. The top animals in our survey most frequently marked "kill" were snake, coyote, hog/boar, armadillo, opossum, salamanders, lizards, and alligators. For some of these species, a lethal response may be more out of fear than anything else. While education about these species may help somewhat in changing attitudes of fear to indifference or coexistence, fear can largely come from cognitive and social bias and negative media coverage in addition to lack of education (Lambertucci et al., 2021). Therefore, conservationists should also consider alternative approaches to protecting these species.
In order to protect certain species, agencies may need to turn to a financial incentive or financial/legal penalty system. Compensation schemes for landowners to allow species of wildlife to live on their land, for example, have been shown to lead to less lethal control measures of species (Dickman et al., 2011;Kontsiotis et al., 2021;Morzillo & Needham, 2015). The Endangered Species Act (United States, 1983) provides legal and financial consequences should a person kill an endangered species and could be relevant in Alabama. For example, the endangered whooping crane (Grus americana) now spends winters along the Tennessee River in northern Alabama (Lessard et al., 2018). Two individuals were recently fined $85,000 and sentenced to serve 360 h of community service for killing two whooping cranes in Louisianahopefully a severe enough sentence to deter further shootings in the area (Associated Press, 2020). Substantial financial and/or legal consequences for harming wildlife, in combination with financial incentives for coexisting with wildlife, can serve as additional tools to preserve the biodiversity of southern Alabama.
In summary, the proportion of hunters in Mobile and Baldwin counties is similar to the rest of Alabama and higher than most other states. Like Alabama as a whole, hunters are more supportive of killing animals, even when it is not as a source of food. Support for lethal responses to animals appeared to be highest among younger age groups. Both hunters and non-hunters are particularly apt to kill snakes. We also found that individuals who are likely to kill animals are less supportive of maintaining or increasing protected areas, and most respondents, regardless of hunting status would give little or no money to improve wildlife protection. Our most surprising results was how little recognition individuals of all demographics (except those with a PhD) have of the tremendous biodiversity of southern Alabama, which harbors many endemic and endangered species. Many of these attitudes can be improved through increased education about local biodiversity and animal encounters, especially among early age groups. The only exception was the large support by hunters of killing animals for reasons other than as a source of food; in these cases, improved conservation management and enforcement may be needed.