Public Knowledge of Monarchs and Support for Butterﬂy Conservation

: Pollinator populations in North America are in decline, including the iconic monarch butterﬂy. In order to determine if public knowledge of monarchs informs opinions on butterﬂy conservation, we surveyed the public to assess their knowledge of monarchs. We also asked participants about their attitudes towards general butterﬂy conservation and if they believe that butterﬂy gardens contribute to conservation. Respondents generally had some knowledge of monarchs but were unaware of monarch population declines and the necessity of milkweed to their life cycle. Respondent knowledge was correlated with more positive attitudes about butterﬂy conservation. Furthermore, membership in an environmental organization increased the likelihood that the participant had prior knowledge of monarchs and cared about monarch conservation. Respondent socioeconomic factors of age and sex were also signiﬁcantly correlated with conservation attitudes—older and female participants had more positive attitudes towards general butterﬂy conservation. Interestingly, females were also less likely than males to admit having prior knowledge of monarchs, indicating that gender may also play an important role in conservation outreach efforts. Our study indicates that educational efforts need to be directed more toward individuals not already associated with an environmental organization as these individuals are predisposed to regard conservation positively.


Introduction
Social norms play an integral role in actualizing conservation programs and environmental policies since perceived social norms can alter individuals' conservation decisions. Both Milfont [1] and Primmer and Karppinen [2] showed that perceived norms can impact forester conservation decisions. Similarly, Van Dijk et al. [3] found that farmers consider social expectations when planning conservation management strategies. Social norms and subsequent public opinions can sway the enforcement and funding of conservation-oriented policies through financial appropriations and voting trends [4][5][6]. For instance, Endangered Species Act policies in the United States [7,8] are more likely to be successful with buy-in from the general public and stakeholders [9][10][11][12]. However, general apathy towards conservation, low levels of education on conservation issues, and negative associations with pro-environmental groups have been implicated as potential roadblocks for public and stakeholder support of ESA policies [13][14][15].
Education or exposure to different values may be an effective way to change the public's opinion on conservation ideas [12,[16][17][18][19] as attitudes can be based partially on social expectations and narratives [20][21][22]. For instance, during interviews with immigrant women in Los Angeles, Sustainability 2018, 10, 807 2 of 16 Lassister and Wolch [23] found that many times the participants' opinions of marine animal welfare changed after exposure to US norms on animal welfare and treatment. Public outreach efforts, particularly those with hands-on experience, have been particularly useful in this vein [24][25][26][27][28]. García-Cegarra and Pacheco [28] found that whale-watching tour participants stated being more likely to care about marine conservation after taking the tour, and a survey of insect zoo attendees revealed that respondents felt more favorably towards insects after visiting and physically interacting with insects [29]. Furthermore, Suh and Samways [30] show that the public was generally ignorant of the insects rather than "not interested" in their preservation.
The influence of public opinion and policy support is particularly critical for invertebrate (i.e., insects, spiders, etc.) conservation. However, public support for insect communities is generally lacking despite an estimated $4.5 billion in economic benefits provided to North America annually [31,32]. Surveys of the Connecticut public and children, revealed a general dislike of insects [16,33]. Additionally, when looking at the valuation placed on invertebrates versus birds, mammals, and even reptiles and fish, invertebrates have been consistently the lowest rated in terms of monetary support for conservation [20,34]. One invertebrate of note is the monarch butterfly (Danaus plexippus L. (Lepidoptera: Nymphalidae)) since it is often used as a teaching device in schools and outreach events due to its iconic orange color and its annual trek across North America [35,36]. Due to this migration, the vast majority of the United States public has the opportunity to interact with this species at some point during the year. However, due to over-wintering site habitat loss as well as the loss and contamination of food resources (milkweed) and habitat in its summer range [37][38][39][40], the decline of monarch populations has been of increasing concern [41][42][43]. Concerns over monarch decline have caused the US Fish and Wildlife to undertake an effort to evaluate its threatened status in accordance with the Endangered Species Act [44][45][46] in addition to private efforts to conserve and create suitable habitat [47,48]. Current recommendations for monarch conservation include the transition of marginal agricultural lands into milkweed plantings, particularly in the northcentral and southern regions of the US including Kentucky where our survey was conducted [49][50][51]. However, a large amount of stakeholder participation is required, which may require more targeted educational efforts.
Given the overall charisma of monarchs [52] and efforts to change the conservation status of monarchs [53], we aimed to determine if the general public's views on butterfly conservation generally were informed by prior knowledge of monarchs [54,55]. Specifically, we wanted to determine (1) if prior knowledge of monarchs (i.e., appearance, required food sources, and potential decline) affects monarch conservation attitudes. Additionally, we wanted to find the impacts of (2) participation in environmental or garden organizations and (3) socioeconomic factors on both prior knowledge of monarchs and corresponding conservation attitudes. We expected that greater levels of prior knowledge would increase the likelihood of positive attitudes towards butterfly conservation. Furthermore, we expected that participation in environmental and garden organizations would increase both the likelihood of prior knowledge and positive attitudes towards conservation but that environmental organizations would have a greater relative impact compared to gardening groups. Finally, we anticipated that socioeconomic variables would influence participation in environmental or gardening groups, prior knowledge of monarchs, and attitudes towards conservation since data have shown similar trends [56]. The outcomes of our study will inform future education and public outreach needs in order to increase the public's desire for monarch and general insect conservation [57], particularly at a local scale where prior studies have indicated the greatest need [54].

Materials and Methods
Our main objectives were achieved through a public survey. The study and survey received approval from the Institutional Review Board from the University of Kentucky (Protocol No. 16-0316-X4B), and each respondent acknowledged the terms of participation prior to beginning the survey. We conducted a survey of the public in the Commonwealth of Kentucky over 51 days from 9 May to 22 July 2016 in Fayette County, Kentucky public parks with the approval of the University of Kentucky and park managers. Prior to the survey, the questionnaire went through four rounds of focus groups. Each focus group consisted of 5-8 individuals recruited from the University of Kentucky listservs where participants were given the current rendition of the survey and allowed to ask questions and make comments on the clarity and neutrality of the questions. Following focus groups, one pilot study was performed at the UK Arboretum (50 responses) to ensure the accuracy of the information and clarify the content. For the finalized survey, a total of 34 public parks and events were visited at various times throughout the day (approximately 8 a.m.-11 a.m., 3 p.m.-5 p.m., and 6 p.m.-9 p.m.) with some parks sampled repeatedly for special events such as movie nights, outdoor concerts, and sporting events. Although we conducted our survey in and around public parks, we also selected many of these events as being unrelated to conservation in order to reduce potential sampling bias. While on site, potential respondents (any individual passing by) over the age of 18 were able to approach the survey station or were approached by a moderator about taking the survey advertised as related to parks in order to avoid biasing the sample towards only pro-environmental participants [58]. Among observable characteristics, the sample contains a greater proportion of respondents with children and a smaller proportion of more senior respondents compared to the population. While it is difficult to understand the potential direction of the bias in this specific analysis, a meta-analysis by Wiernik et al. [59] showed that age had no effect on environmental awareness or knowledge, and, more broadly, either no or negligible effects across a range of environmental sustainability variables, environmental awareness or knowledge. Although Dupont et al. find that willingness to pay for environmental improvements significantly increases with the presence of children [60], Torgler et al. find no evidence of having children on a range of environmental attitudes [61]. Of course, differences caused by demographic characteristics or other unobservable characteristics of the sample may still affect the results.
Respondents were then given an online survey with an estimated completion time of 10 min on a tablet device through the platform provided by the research firm Qualtrics (Provo, UT, USA). To single out the effect on conservation attitudes associated with monarchs being a flagship species in addition to its charismatic appearance, each respondent was randomly assigned to a survey about either monarch or viceroy butterflies, a non-threatened, monarch doppelganger (Limenitis archippus L. (Lepidoptera: Nymphalidae)) [62]. The two versions of the survey are identical except for a brief excerpt at the beginning of the survey, each with 40 questions (combination of True-False, multiple choice, and Likert-style). To ensure that respondents were relatively informed, a short excerpt about the corresponding butterfly's food resources and their native range was given followed by three True-False questions about that excerpt. Respondents in the monarch version were also informed that the monarch was under review for the US list of endangered or threatened species as well as its sole reliance on milkweed plants during its larval stage.
The remaining survey questions pertained to prior knowledge of monarchs (stated overall knowledge, food requirements, population decline, and identification), participation in environmental or gardening organizations, prior donations to environmental organizations, extent of outdoor activity, sociodemographic, and attitudes towards butterfly conservation and the use of butterfly gardens for conservation. To quantify each participant's ability to identify a monarch, each respondent was asked to identify a monarch butterfly in two steps. First, they were presented with four sets of paired, similar-looking butterflies ( Figure 1A) and asked which set contained the monarch. If they identified the correct pair, they were asked to identify which of the pair was a monarch ( Figure 1B), again followed by a question regarding their certainty. In addition, the following Likert questions were asked to estimate prior knowledge: "Have you ever heard or read about monarch/viceroy butterflies prior to this survey? How much did you know about the decline in monarch populations prior to this survey? How much did you know about the importance of milkweeds to monarch prior to this survey? How important is it to you to help conserve butterfly species? How much do you believe that installing butterfly plants can actually help butterfly conservation?" Respondents could choose one of five answers to each of the questions from "not at all" to "a great deal," with the exception of butterfly gardens, which used a slightly different scale (five answers ranging from "not important" to "very important"). In total, 789 participants (Table 1) completed the full survey and were included in the dataset for statistical analyses. Respondents could choose one of five answers to each of the questions from "not at all" to "a great deal," with the exception of butterfly gardens, which used a slightly different scale (five answers ranging from "not important" to "very important"). In total, 789 participants (Table 1) completed the full survey and were included in the dataset for statistical analyses. In order to answer our three questions, the following variables were used as the dependent variables in separate ordered logistic regressions: monarch knowledge, viceroy knowledge, knowledge of monarch declines, knowledge of milkweed importance, ability to identify a monarch, and willingness to conserve butterflies. All models included age, sex, education, income, household with children, race, outdoor recreation, member in a garden club, member of an environmental group, and previous donation to environmental cause as independent variables with the exception of willingness to conserve which had the added variable of monarch knowledge. Because of incomplete responses to these independent variables, the sample used to perform regression analysis (which relies on complete responses) decreases to the results seen in Tables 2 and 3.
All analyses were conducted in Stata 13 (StataCorp LP, College Station, TX, USA) with average marginal effects (MEs) calculated in the same program. MEs are commonly used in the social science discipline representing the change of likelihood of survey respondents be in each of the five possible answer groups (from "not at all" to "a great deal") following one unit of difference between them (e.g., male versus female, or one year older) while holding all other factors unchanged [63]. As an example, if the marginal effect for an independent indicator variable (e.g., female = 1) equals 0.2 for particular level of the dependent variable, this means that the observations with that characteristic (i.e., being a female) are 20% more likely to have selected that particular level while holding all other factors unchanged. A marginal effect of −0.05 would indicate observations with that characteristic are 5% less likely to select that level of the dependent variable. Such marginal effects can be generated for every level of the dependent variable, and the sum of all marginal effects for a given independent variable across all levels of the dependent variable must sum to zero. While ME at every level of the responses is generated in the ordered logit models, for brevity, we interpret ME only at the lowest ("not at all" or "not important") and highest ("a great deal" or "very important") extremes of the responses for only those variables that were significant or marginally significant in the corresponding model.

General Attitudes towards Monarch Conservation
When asked "how important is it to you to help conserve butterfly species," respondents had mixed responses. Very few thought it was "not important" (6.34%), while the responses among "somewhat important," "important," and "very important" were evenly divided (Figure 2), and the remaining 4.56% answered "didn't know." This last group is excluded from the logit model. Most individuals also thought that butterfly gardens contributed to butterfly conservation, with 54.5% thinking that it was "very important" and 35.2% answering that it was "important."

General Attitudes towards Monarch Conservation
When asked "how important is it to you to help conserve butterfly species," respondents had mixed responses. Very few thought it was "not important" (6.34%), while the responses among "somewhat important," "important," and "very important" were evenly divided (Figure 2), and the remaining 4.56% answered "didn't know." This last group is excluded from the logit model. Most individuals also thought that butterfly gardens contributed to butterfly conservation, with 54.5% thinking that it was "very important" and 35.2% answering that it was "important."

Prior Monarch Knowledge Correlated with Conservation Attitudes
The general public's reported knowledge about either monarchs or viceroys was lacking, but viceroy knowledge more so (Table 2). When asked about overall prior knowledge, 56.08% and 16.60% of participants reported not knowing anything about viceroys and monarchs, respectively (Figure 3).

Prior Monarch Knowledge Correlated with Conservation Attitudes
The general public's reported knowledge about either monarchs or viceroys was lacking, but viceroy knowledge more so (Table 2). When asked about overall prior knowledge, 56.08% and 16.60% of participants reported not knowing anything about viceroys and monarchs, respectively ( Figure 3). The same trend appears for those who reported knowing "a lot" or "a great deal" about viceroys (4.81% and 2.53%, respectively) and monarchs (19.26% and 15.84%, respectively). When asked about prior knowledge of monarch population declines, the majority of respondents did not know anything (40.25%) or only knew "a little" (26.58%). Respondent prior knowledge of the importance of milkweeds as a food resource to monarchs was similar to that of monarch decline. Again, the vast majority of respondents did not know anything (53.67%) or knew "a little" (15.95%) about the importance of milkweed (Figure 3). The same trend appears for those who reported knowing "a lot" or "a great deal" about viceroys (4.81% and 2.53%, respectively) and monarchs (19.26% and 15.84%, respectively). When asked about prior knowledge of monarch population declines, the majority of respondents did not know anything (40.25%) or only knew "a little" (26.58%). Respondent prior knowledge of the importance of milkweeds as a food resource to monarchs was similar to that of monarch decline. Again, the vast majority of respondents did not know anything (53.67%) or knew "a little" (15.95%) about the importance of milkweed (Figure 3). When given the opportunity to identify a monarch/viceroy pair out of a lineup (Figure 4), most respondents (78.71%) could identify the pair correctly. This proportion is significantly different than 25% (a random guess of an answer out of the four alternatives) (p < 0.01). Claiming to have prior knowledge of monarchs (p < 0.01), in general, increases the likelihood of a correct answer (ME = 7.61%; i.e., 7.61% more likely). However, only 50.89% of respondents who answered the first identification question correctly could subsequently correctly identify the monarch from the monarch/viceroy pair, (40.05% of entire sample), which was not significantly different (p = 0.75) from null proportion of 50% (corresponding to a random guess by respondents as to which butterfly shown is a monarch). This indicates that respondents have a general idea of monarchs' appearance but are not necessarily able to differentiate them from a similar looking species. The only notable predictor of ability to correctly identify a monarch in the second question was stated prior knowledge of monarchs, but only with marginal significance (p = 0.08), and the effect was rather small (ME = 2.88%). Increases in stated prior monarch knowledge also highly correlated with increased likelihood of respondents finding butterfly conservation "very important" (ME = 6.8%, Table 3) and decreased the likelihood of "not important" (ME = −2.4%). When given the opportunity to identify a monarch/viceroy pair out of a lineup (Figure 4), most respondents (78.71%) could identify the pair correctly. This proportion is significantly different than 25% (a random guess of an answer out of the four alternatives) (p < 0.01). Claiming to have prior knowledge of monarchs (p < 0.01), in general, increases the likelihood of a correct answer (ME = 7.61%; i.e., 7.61% more likely). However, only 50.89% of respondents who answered the first identification question correctly could subsequently correctly identify the monarch from the monarch/viceroy pair, (40.05% of entire sample), which was not significantly different (p = 0.75) from null proportion of 50% (corresponding to a random guess by respondents as to which butterfly shown is a monarch). This indicates that respondents have a general idea of monarchs' appearance but are not necessarily able to differentiate them from a similar looking species. The only notable predictor of ability to correctly identify a monarch in the second question was stated prior knowledge of monarchs, but only with marginal significance (p = 0.08), and the effect was rather small (ME = 2.88%). Increases in stated prior monarch knowledge also highly correlated with increased likelihood of respondents finding butterfly conservation "very important" (ME = 6.8%, Table 3) and decreased the likelihood of "not important" (ME = −2.4%). Sustainability 2018, 10, x FOR PEER REVIEW 9 of 15

Participation in Environmental Organizations Impacts Prior Knowledge and Attitudes
Participants that were members of garden clubs (p = 0.01), environmental organizations (p < 0.01), had previously donated to an environmental group (p < 0.01) were more likely to have prior knowledge of monarchs and their decline ( Table 2). Such participation increased the likelihood that the respondents knew "a great deal" about monarchs and decreased the likelihood that they know nothing at all. The same trend was true for stated knowledge of monarch declines-being part of a garden club or environmental organization decreased the chance of reporting being "not at all" familiar (ME = −23.49%, −15.16%, respectively) and increased the chance that respondents knew "a great deal" (ME = 10.29%, 4.25%, respectively). Respondents who donated to an environmental cause (p < 0.01) had a similar association to knowledge of monarch declines (ME "not at all" = −17.68%; ME "a great deal" = 4.33%). Furthermore, those who spent more time outdoors for recreation are also slightly more likely to be aware of monarch declines (p = 0.02), with one additional day of recreation per week corresponding to a smaller chance of being not familiar (ME = −1.73%) and a greater chance of knowing "a great deal" (ME = 0.37%).
Prior knowledge of milkweed's importance and ability to identify a monarch was less straightforward. Being an environmental organization member (p < 0.01, ME "not at all" = −20.43%, ME "a great deal" = 8.64%) or having donated to environmental causes (p < 0.01, ME "not at all" = −11.47%, ME "a great deal" = 4.05%) increased the likelihood that the respondent knew about the importance of milkweeds as food resources for monarchs, but being a member of a garden club did not impact prior knowledge of milkweed importance (p = 0.12). Those who did more outdoor recreation had some evidence of milkweed knowledge (p = 0.06) since they were less likely to answer "not at all" (ME = −1.5%) and more likely to answer "a great deal" (ME = 0.5%). Association with garden (p = 0.43) or environmental organizations (p = 0.53), having made a donation to an environmental cause (p = 0.93), or having prior knowledge of viceroys (p = 0.27) did not have a significant impact on the ability to find the monarch/viceroy pair.
When asked about monarch conservation, participation in environmental organizations diverged from those who participated in outdoor recreation or in garden clubs ( Table 3). Members of environmental organizations (p = 0.06) and those who had donated to environmental causes (p < 0.01) were more likely to find butterfly conservation "very important" (ME = 8.8% and 12.8%, respectively). However, members of garden clubs (p = 0.42) and those with greater levels of outdoor recreation (p = 0.29) were no more likely to find butterfly conservation important.

Participation in Environmental Organizations Impacts Prior Knowledge and Attitudes
Participants that were members of garden clubs (p = 0.01), environmental organizations (p < 0.01), had previously donated to an environmental group (p < 0.01) were more likely to have prior knowledge of monarchs and their decline ( Table 2). Such participation increased the likelihood that the respondents knew "a great deal" about monarchs and decreased the likelihood that they know nothing at all. The same trend was true for stated knowledge of monarch declines-being part of a garden club or environmental organization decreased the chance of reporting being "not at all" familiar (ME = −23.49%, −15.16%, respectively) and increased the chance that respondents knew "a great deal" (ME = 10.29%, 4.25%, respectively). Respondents who donated to an environmental cause (p < 0.01) had a similar association to knowledge of monarch declines (ME "not at all" = −17.68%; ME "a great deal" = 4.33%). Furthermore, those who spent more time outdoors for recreation are also slightly more likely to be aware of monarch declines (p = 0.02), with one additional day of recreation per week corresponding to a smaller chance of being not familiar (ME = −1.73%) and a greater chance of knowing "a great deal" (ME = 0.37%).
Prior knowledge of milkweed's importance and ability to identify a monarch was less straight-forward. Being an environmental organization member (p < 0.01, ME "not at all" = −20.43%, ME "a great deal" = 8.64%) or having donated to environmental causes (p < 0.01, ME "not at all" = −11.47%, ME "a great deal" = 4.05%) increased the likelihood that the respondent knew about the importance of milkweeds as food resources for monarchs, but being a member of a garden club did not impact prior knowledge of milkweed importance (p = 0.12). Those who did more outdoor recreation had some evidence of milkweed knowledge (p = 0.06) since they were less likely to answer "not at all" (ME = −1.5%) and more likely to answer "a great deal" (ME = 0.5%). Association with garden (p = 0.43) or environmental organizations (p = 0.53), having made a donation to an environmental cause (p = 0.93), or having prior knowledge of viceroys (p = 0.27) did not have a significant impact on the ability to find the monarch/viceroy pair.
When asked about monarch conservation, participation in environmental organizations diverged from those who participated in outdoor recreation or in garden clubs ( Table 3). Members of environmental organizations (p = 0.06) and those who had donated to environmental causes (p < 0.01) were more likely to find butterfly conservation "very important" (ME = 8.8% and 12.8%, respectively). However, members of garden clubs (p = 0.42) and those with greater levels of outdoor recreation (p = 0.29) were no more likely to find butterfly conservation important.

Socioeconomic Factors Influence Knowledge and Attitudes
Older participants were more likely to know about monarch declines but the marginal effects were often very low. For instance, the marginal effect for one year older in age (p = 0.03) on answering knowing about monarch declines with "not at all" is −0.03%. This means that all else equal, a person who is one year older than another is 0.03% less likely to answer "not at all." Conversely, the marginal effect of age is 0.06% in reporting "a great deal" of knowledge of monarch declines. Age was not a significant predictor of prior knowledge of monarchs (p = 0.84) or viceroys (p = 0.09) in general or milkweed importance (p = 0.16). However, older participants were more likely to find butterfly conservation "very important" (p < 0.01, ME = 0.7%).
Participant sex was only a significant predictor variable in regressions for prior viceroy knowledge and the importance of butterfly conservation. Females were significantly more likely than males to state that they know about viceroys "not at all" (p < 0.01, ME = 12.70%) in general but not monarchs (p = 0.90). Sex had no impact on state knowledge of monarch declines (p = 0.18) or importance of milkweed (p = 0.87). Males had greater self-reported knowledge of monarchs and especially viceroys, but males were no more likely to correctly identify the monarch/viceroy pair or monarch butterfly specifically compared to females. For the importance of butterfly conservation, sex was an important predictor (p = 0.01, Table 3). Females were more likely than males to state that conservation is "very important" (ME = 7.47%) and less likely to state that is was "not important" (ME = −2.01%).
Race was correlated with all aspects of prior knowledge and willingness to conserve butterflies. Individuals who self-identified as black were more likely than white individuals to not know about ("not at all") viceroys (p < 0.01, ME = 16.10%), monarchs (p < 0.01, ME = 11.56%), monarch declines (p < 0.01, ME = 15.30%), and the importance of milkweed (p < 0.01, ME = 16.50%). No other races were significant predictors in these models. Interestingly, Asian (p = 0.07, ME = −16.07%) and Hispanic (p = 0.08, ME = −14.27%) respondents were significantly less likely to correctly identify monarchs correctly than were white respondents. Whereas, black respondents were no more likely to correctly identify the monarch than white respondents (p = 0.14). However, race-related predictors for willingness to conserve butterflies was more similar to those for prior knowledge (Table 3)-black participants (p = 0.04) were less likely to find it "very important" (ME = −8.81%) and more likely to find it "not important" (ME = 2.91%).
Other socioeconomic factors were less important in all regressions. For both general viceroy and monarch knowledge, greater education indicated a smaller likelihood of knowing "nothing" ( Table 2). The same trend was true for knowledge of monarch declines (p < 0.01) and the importance of milkweed (p < 0.01). In both of these cases the marginal effects were stronger for knowing nothing (ME = −5.36% and ME = −5.76%, respectively) versus knowing a great deal (ME = 0.86% and ME = 1.56%, respectively). Respondent education levels did not impact their ability to identify monarchs (p = 0.30) or willingness to conserve butterflies (Table 3). Neither being in a household with children nor household income (p-values listed in same order) had an impact on prior knowledge of viceroys (Table 2), monarchs, monarch declines (p = 0.36, p = 0.14), milkweed importance (p = 0.24, p = 0.46) or monarch identification (p = 0.66, p = 0.39) and willingness to conserve butterflies (Table 3).

Discussion
Generally, the surveyed respondents stated that they had prior knowledge (from "moderate" to "a lot") of monarchs but had less knowledge of monarch declines and the importance of milkweed to the monarch life cycle (Figure 2). More people claimed knowledge of monarchs than viceroys, interesting as the state butterfly of Kentucky is, in fact, a viceroy and featured on select state license plates [64]. This interest in monarchs compared to viceroys is unsurprising given recent media coverage of monarchs in relation to other topics such as agriculture biotechnology, climate change, and gardening [65][66][67][68]. Stated prior knowledge of monarchs was correlated with both being able to identify monarchs and in willingness to conserve butterflies (concern for wildlife), which corroborates what prior studies have found [29,69,70] and lends urgency to education efforts to gain public buy-in for milkweed plantings [49,51,66].
Association with environmental groups, including having donated to an environmental cause, increased both the likelihood of monarch knowledge and positive attitudes about butterfly conservation. Participation in garden clubs also increased knowledge but did not correlate with positive attitudes about butterfly conservation, while outdoor recreation and its potential for exposure to monarch did correlate with either. The social values of environmental group members for efforts to save endangered species possibly influenced their support for monarch conservation efforts [71,72], while the garden club members did not hold these same values. In terms of conservation, increased environmental group membership might foment an increased acceptance of environmentally friendly attitudes and behaviors [73][74][75].
As prior studies have found, environmentally friendly behaviors and attitudes are not predicated on personal values alone as signaled by environmental group membership, but also by demographic and cultural factors [76]. Similarly, we found that even after taking into consideration prior knowledge of monarch and membership in environmental groups, some demographic factors played an integral role in explaining attitudes towards butterfly conservation. We found that an increase in age correlated with an increase in positive attitudes towards conservation as have some other studies [77] though, in contrast to prior studies, higher household income (that tends to increase with age) did not [78,79]. However, the marginal effects of sex were prominent, with self-identified females exhibiting more positive attitudes towards butterfly conservation than did males despite having lower reported levels of monarch knowledge. Since this is a trend seen in prior studies on environmental attitudes and environmentally friendly behaviors, sex clearly has a large impact on conservation strategies in addition to educational efforts to improve public knowledge [79][80][81][82][83][84][85].
Since it is difficult to explain the sex differences in conservation attitudes, the results of our study emphasize the need for public education and recruitment to environmental organizations [69,86,87]. In terms of educational goals, if increasing public awareness is done in order to increase interest in conservation practices, it appears that more emphasis on specific conservation needs and methods is required (i.e., planting milkweeds and other nectar plants) [88][89][90]. Future education and outreach efforts for monarch conservation should also target individuals who are not already members of environmental organizations as this group is already inclined to support butterfly conservation. This is especially important as monarch habitat is dependent in large part on private landowners (often farmers), who are likely not members of environmental organizations [91,92]. But previous studies have shown that increased education increased the likelihood of farmers perceiving and mitigating environmental concerns such as monarch habitat loss [93][94][95].

Conclusions
Conservation is not a one-sided task. Our study shows that educating the public involves more than the appearance of an animal species, particularly for insects. As there has been little research conducted on social understanding and perception toward conservation or protection of insect species, our analysis adds to this limited depository [52,62]. Based on our work, further studies can examine how individuals may be willing to contribute to protecting the entire range of monarch butterflies including Mexico and Canada or how the public may view the tradeoff between preserving milkweed versus land management such as agricultural weed control [38]. Our research strengthened the idea that identifying public attitudes related to conservation is necessary to determine what aspects of conservation are important to the general public [96].