The use of artificial lights at night in public locations is often seen as a benefit to public safety. However, for humans and many other species, light pollution can have damaging effects on individual health and reproductive success. Many studies have documented negative impacts of light pollution on multiple species (e.g., Rodríguez et al. 2014; LeTallec et al. 2015; Raap et al. 2016). When artificial lights impact populations of declining, threatened, and endangered species, it becomes important to monitor and mitigate the effects.

One group of animals with a well-documented sensitivity to artificial lights is sea turtles. The primary method of sea-finding in nesting and hatchling sea turtles is through visual cues offered by the brightest location along the horizon (Mrosovsky and Shettleworth 1968; Witherington and Bjorndal 1991a; Witherington 1992). Light pollution poses a threat to turtles' reproductive success along developed beachfronts when hatchlings crawl toward the artificial light rather than the ocean (McFarlane 1963; Philibosian 1976). However, not all lights are viewed equally for sea turtles: certain wavelengths do not appear to attract sea turtles and, therefore, do not negatively affect a turtle's crawl to the ocean. In general, loggerhead sea turtles (Caretta caretta) are not attracted to red-colored lights above 600 nm on the wavelength spectrum (Witherington and Bjorndal 1991b; Levenson et al. 2004).

Because sea turtles are protected under the United States Endangered Species Act (1973), multiple counties along the Gulf and East coasts of the United States have enacted legislation requiring beach-visible lighting to be sea turtle friendly (e.g., Glynn County, Georgia, Code Ordinances ch. 2–23, art. I, 1997; Georgetown County, South Carolina, Code Ordinances ch. 5.5, art. III, 2016; Brevard County, Florida, Code Ordinances ch. 46, art. III, 2017; Broward County, Florida, Code Ordinances ch. 39, art. IX, 2017). Regulations often specify minimum wavelengths of light, height requirements for fixtures, and shielding of bulbs for exterior fixtures such that they are not directly visible from the beach. Interior room lighting is generally not covered by beach lighting regulations and can be controlled through closing curtains or blinds or turning off interior lights when not in use. Owners of beachfront structures (e.g., homeowners, business owners) typically become familiar over time with their community's lighting regulations. On the other hand, out-of-town visitors are often unfamiliar with local lighting ordinances.

In these communities, various forms of environmental education and interpretive materials may be required to: 1) inform visitors of sea turtle–friendly beach behaviors, and 2) encourage visitors to participate in these behaviors. A particular concern for nesting sea turtles is the light emanating at night from large beachfront hotel rooms (Kaska et al. 2010; Oliver de Esperanza et al. 2017). Despite requests to minimize artificial light to protect nesting turtles, many guests keep lights in their rooms on and visible well into the night (M. Dodd, pers. comm., May 2016; pers. obs.). Our study sought to identify the magnitude of this problem and explore potential mitigation strategies.

Research in the fields of environmental education and interpretation highlights 2 primary methods of message delivery that may impact human behavior, both with unique advantages and disadvantages (Marion and Reid 2007; Sharp et al. 2012). Interpersonal communication, or face-to-face interactions, allow for 2-way communication between an interpreter and visitors (Knudson et al. 2003). This approach typically is viewed as desirable by site managers, in part because it allows the interpreter to respond to audience reactions, needs, and questions (Wearing and Neil 1999; Munro et al. 2008). However, interpersonal education is limited in its ability to reach large numbers of visitors and is often difficult to implement owing to the high cost of training and staffing.

Alternatively, nonpersonal communication strategies (e.g., brochures, signage) enable educators to communicate with a broader audience without relying heavily on staff resources. This approach is relatively inflexible and requires visitors to access and interpret information on their own. For individuals with limited interest in, or experience with, a topic (e.g., nesting sea turtles), such drawbacks may limit message consumption and efficacy. The static nature of nonpersonal education makes it less desirable to many natural resource managers (Hughes 2004); yet, because of its comparatively low costs, nonpersonal forms of communication typically dominate visitors' experiences (Knudson et al. 2003).

Important questions for natural resource managers therefore become: Which type of educational materials should be used in different situations? Do the benefits of interpersonal communication justify the costs, or are cost-efficient and wide-reaching nonpersonal strategies equally effective at changing visitor behavior? As a result, outcome assessment is needed to identify best practices (Silverman and Barrie 2000; Marion and Reid 2007; Washburn 2007).

Although many studies have examined the impact of different forms of interpretation on visitor knowledge and attitudes (Ham 1992; Widner and Roggenbuck 2000; Henker and Brown 2011; Sharp et al. 2012), fewer have considered impacts on overt conservation behaviors (Munro et al. 2008; Ardoin et al. 2013). The few studies that have researched the effects of education on conservation actions have found that interpersonal efforts often produce positive conservation behavior outcomes, while nonpersonal education strategies have mixed results (e.g., Taylor et al. 2007; Sharp et al. 2012; Fielding et al. 2013; Kidd et al. 2015; Warren et al. 2018). In one case, research has shown that in-room messaging promoting water conservation at hotels (e.g., encouraging visitors to hang and re-use towels) can be effective (Morgan and Chompreeda 2015). We sought to explore the efficacy of one education strategy (an in-room, nonpersonal informational rack card) on a particular conservation behavior: beachfront hotel guests' decision to engage in sea turtle–friendly actions such as turning off lights and/or closing their blinds at night.

We chose to explore this question on Jekyll Island, Georgia, a popular tourist destination where sea turtles are a prominent attraction. Jekyll Island is also a place where coastal development could potentially disrupt sea turtle nesting patterns. In the past, lights from Jekyll's beachfront structures have attracted hatchlings landward rather than to the ocean (Bagwell 2001). Additionally, the presence of tourists on the beach at night can alter sea turtle nesting behavior (Hailman and Elowson 1992). Therefore, sea turtle conservation is a high priority on Jekyll Island. Our study centered on 2 objectives. First, we wanted to understand when visible light ratios (the proportion of occupied hotel rooms with open blinds) were highest and identify factors associated with those peaks. Second, we wanted to evaluate the impact of a persuasive, nonpersonal communication tool (rack card) on hotel visitors' lighting choices. We hypothesized that we would see a significantly lower proportion of occupied guest rooms with visible lights when rack cards were present compared with nights when the cards were absent.

METHODS

Study Site. — Jekyll Island State Park (31.0704°N, 81.4193°W, WGS 84) is one of 14 barrier islands in the state of Georgia. Between May and August, sea turtles use Jekyll's beaches to nest, with hatching season occurring from July to October. Loggerhead sea turtles are the predominant nesting species; however, green (Chelonia mydas) and leatherback (Dermochelys coriacea) sea turtle nests have been documented nesting on the island (Ondich and Andrews 2013).

Jekyll Island has approximately 15.3 km of beach available on which sea turtles could nest. Approximately 8.3 km of beachfront is undeveloped, although about 5 km of that nesting habitat is located along eroding shorelines. Another 5 km of beach has a rock wall revetment that does not offer suitable nesting habitat (Ondich and Andrews 2013; Georgia Sea Turtle Center [GSTC], unpubl. data, 2018). A large portion of this rock-wall beach coincides with beachfront residential homes and hotel development. An additional 2 km of suitable nesting habitat south of the rock wall is developed by hotels, restaurants, shops, a convention center, and popular public beach access locations (Fig. 1). Because development introduces artificial lights, Glynn County, Georgia, has legislation regulating lights visible from nesting beaches. Further, the Jekyll Island State Park Authority (JIA) passed its own ordinance in 1981 and updated it in 2008 to include stricter lighting regulations than those mandated by Glynn County (Jekyll Island Auth., Georgia, Code Ordinances ch. 10, art. IV, 2015).

View Figure

• Download as Powerpoint
• Download Figure

Educational Treatment. — During the time of the study (2015–2016), Jekyll Island had 5 beachfront hotels in operation, with all but 1 having guest rooms directly visible from the beach. The tallest of these hotels stands at 5 stories. Three of the building's sides can be seen from the adjacent beach (the north, east, and south faces). When guest room blinds are left open, the light from beach-facing rooms can be seen for at least 1 km in each direction (north/south) along the beach. The light seen from hotel guest rooms includes unshielded fixtures whose bulb or filament can be seen directly on the beach and shielded fixtures whose cumulative light may illuminate the beach.

In anticipation of the possible impact this might have on nesting and hatching sea turtles, the hotel's local management team approached the JIA's GSTC (a local sea turtle hospital, education facility, and research center) to assist them in developing an effective method to educate guests to close their blinds at night. A nonpersonal education method was chosen owing to the benefits it prescribes regarding cost, time, and audience reach. An educational plaque affixed to the wall near guest room blinds was discussed initially, but corporate-level requirements dictating that rooms should look similar across all hotel locations made this idea infeasible. Because brochures do not have to be approved by corporate-level management, the decision was made to place a two-sided “Protect the Night, Hide the Light” rack card in each guest room (hereafter referred to as “rack card”). These rack cards are designed to catch the eye with colorful imagery and limited text on the front, with the back providing more-detailed information on sea turtles' sensitivity to lights and how people can modify their behavior to remedy the problem (Fig. 2).

View Figure

• Download as Powerpoint
• Download Figure

To determine whether these cards were effective at educating guests, rack card distribution was carefully monitored in regard to study design. From May to October 2015, each guest room received a rack card (treatment group). Cards were placed in rooms regardless of whether the room could be seen from the beach to simplify the instruction to the housekeeping staff and to reduce confusion about which rooms received the cards. However, only beachfront rooms were surveyed and counted as part of the analysis. Rack cards were placed on the bed for maximum visibility and were restocked by housekeeping staff when new guests arrived or if one was removed or damaged. Through weekly spot checks of rooms and conversations with guests, we confirmed that rack cards were placed as we had requested. In 2016, guest rooms did not receive rack cards (control group).

Data Collection. — Observational counts of lights in hotel rooms were conducted by the same observer 3 times weekly in 2015 and 2016 (unless inclement weather prevented a survey from occurring to ensure surveyor safety). Observations coincided with the months during which sea turtles were nesting or hatching and thus when Jekyll Island's lighting ordinance was in effect (May–October). Observation sessions were randomly chosen through a number generator (Microsoft Excel, v. 2013) with at least 1 session per week occurring during peak visitation times on a Friday, Saturday, or Sunday. To reduce the effect that day-of-week variation may have on visitation rates (higher on weekends), observation session nights in 2016 were matched to occur on the same nights of the week on which they were conducted in 2015. For example, if an observation session in 2015 occurred on the Monday of the 27th week of the year, then the 2016 observation occasion was set to also be on the Monday of the 27th week of the year. We systematically chose the nights in 2016 (the control year) to match those in 2015 (the treatment year) to avoid a potential skew toward high-occupancy weekends, or vice versa, and to minimize that potential variation across years. Finally, as hotel guests on vacation often reside in the same hotel room over multiple consecutive nights, we included in our analysis counts only from those observation occasions separated by at least 4 nights to minimize bias associated with potential independence between observations (“sampling nights”).

During each observation session, all windows on beach-facing sides of the hotel were counted as either having “no light visible” (lights off and/or blinds completely closed) or having a “light visible” (lights on and blinds fully or partially open; Fig. 3). Any light emerging through blinds (even partial light) could impact turtle behavior and would suggest that visitors were not complying with the rack card instructions. In hopes of having a maximum number of hotel guests present in their rooms and still awake in the hours after sunset, surveys began at approximately 2100 hrs. As it was not completely dark outside during the summer months before 2100 hrs, we also selected this time to ensure we were collecting observations during a period where emanating light was visible and potentially impactful to the nocturnal turtles. Last, because environmental conditions could impact a guest leaving their lights on and/or blinds open, data on environmental conditions were collected including whether it was raining (yes/no), estimated percentage of cloud cover (0%–24%, 25%–49%, 50%–74%, or 75%–100%), and air temperature.

View Figure

• Download as Powerpoint
• Download Figure

Analysis Methods. — To account for hotel occupancy rates, the estimated number of beachfront rooms occupied (i.e., occupancy rate) at the time of an observation was determined by multiplying the number of available beachfront rooms (n = 151) by the corresponding monthly occupancy rate (0%–100%) of the hotel (note: daily occupancy rates could not be obtained). The final step in estimating the proportion of occupied beachfront rooms with visible lights was to divide the number of rooms counted with lights visible by the estimated number of beachfront rooms occupied at the time of sampling (as calculated above using monthly occupancy rates). This method of adjusting the proportion of visible lights for occupancy rates likely led to an underestimation of occupied rooms during peak visitation days (e.g., Saturday nights), therefore resulting in an overestimation of the proportion of visible lights; the inverse of this pattern also held true (i.e., nights with lower-than-monthly occupancy rates likely resulted in an underestimated proportion of visible lights).

All statistical tests were conducted using SPSS Version 21.0. We used a 2 × 7 × 6 × 3 factorial analysis of covariance (ANCOVA) with one continuous covariate (temperature) to evaluate how our independent variables influenced the mean proportion of occupied rooms with visible lights. Independent variables included in the model were treatment group (i.e., year), day of the week, month, and cloud cover class (0%–24%, 25%–74%, and 75%–100%). Interaction effects between treatment and other variables in the model were additionally examined. Interactions that were not statistically significant were dismissed and not included in our final model. We conducted preliminary checks to ensure the assumptions of the ANCOVA model were not violated; these included 1) the covariate was linearly related to the dependent variable, 2) homogeneity of regression slopes, 3) an approximately normal distribution of the dependent variable for each level of the independent variables, 4) homoscedasticity, 5) homogeneity of variances, and 6) no significant outliers present. Where a variable was found to be a significant predictor in the model, a post hoc Bonferroni multiple comparisons test was used to further examine the results. A significance threshold of p = 0.05 was set for all analyses unless otherwise stated.

RESULTS

We had a combined total of 137 observation sessions for the 2 yrs (73 nights in 2015 and 64 nights in 2016). From the observation sessions, 58 sampling nights were used in analysis (29 nights from each year). In our treatment year (2015), the mean proportion of occupied guest rooms with visible light was 0.61, with a standard deviation (SD) of 0.21 (n = 29). Surprisingly, the mean proportion of occupied rooms with visible light during the control year was lower at 0.53 (n = 29), with an SD of 0.11. By month within the treatment year, the mean proportion of visible light varied from a low of 0.36 (August) to a high of 0.91 (May). In the control year, monthly mean proportions ranged from 0.49 (August) to 0.59 (June; Fig. 4).

View Figure

• Download as Powerpoint
• Download Figure

Controlling for temperature as a covariate, the ANCOVA model found a statistically significant effect of treatment (F_1,37 = 14.30 p = 0.001, η² = 0.28), day of week (F_6,37 = 2.45, p = 0.04, η² = 0.28), month (F_5,37 = 8.00, p < 0.001, η² = 0.52), and an interaction effect between treatment and month (F_5,37 = 4.49, p = 0.003, η² = 0.38; Table 1).

Table 1. Factorial analysis of covariance (ANCOVA) examining main effects and interactions of treatment (rack card vs. no rack card) and temporal/environmental variables on the proportion of visible lights after controlling for the covariate temperature (n = 58).

View Fullscreen

Pairwise comparisons between days of the week indicated that Saturdays had a significantly higher proportion of rooms with visible light than both Mondays (p = 0.001) and Wednesdays (p = 0.005). Because the interaction term between month and treatment was significant, an analysis of simple main effects was performed, with statistical significance receiving a Bonferroni adjustment, resulting in acceptance at the p < 0.025 level. A significant difference in the mean proportion of occupied rooms with visible light was found between treatment types for the months of May (F_1,37 = 20.14, p < 0.001, partial η² = 0.35), June (F_1,37 = 6.99, p = 0.01, partial η² = 0.16), and July (F_1,37 = 7.20, p = 0.01, partial η² = 0.16), but not for August through October. May, June, and July each had a significantly lower mean proportion of visible lights in the control year than in the treatment year.

Finally, the GSTC had an educator present in the hotel on Fridays in June and July 2016 (the control year) during peak check-in hours. Despite this, a 1-way ANOVA showed that there were no significant differences between months in the proportion of occupied hotel rooms with visible lights that summer (F_5,22 = 0.77, p = 0.58). We therefore concluded that these relatively infrequent interpersonal education opportunities in 2016 did not significantly alter lighting choices and behavior within our control group.

DISCUSSION

Our results suggest that 1) fewer than half of all hotel guests in beachfront rooms abided by the recommended lighting regulations to protect nesting sea turtles, and 2) the educational treatment (rack card) failed to have a positive impact on visitors' sea turtle–friendly lighting choices and behaviors.

Temporal variations were also observed with respect to these behaviors. We found that May through July had significantly higher proportions of visible lights when the rack card was present in 2015. This is probably a function of the hotel's grand opening in April of that year. We believe that the hotel's staff was learning how to best operate in those initial months, resulting in fewer staff members being aware that 1) sea turtles were nesting on the beach adjacent to the hotel beginning in May, and 2) that hotel guests' lights could pose a threat to sea turtles. As lighting issues were brought to the hotel's attention, staff was likely briefed on how to encourage guests to practice sea turtle–friendly lighting behaviors, leading to nonsignificant differences between the treatment group in the latter half of the nesting season (August through October) and the control group during that same period the following year. Although the nonpersonal rack card was ultimately ineffective at altering behaviors for many hotel guests, personal education–oriented interactions between hotel staff and guests might have gradually influenced visitors' behaviors as time progressed.

We found that the highest proportions of rooms with visible lights were observed during peak visitation times of Saturday evenings, and the lowest proportion of visible lights were observed on Monday and Wednesday nights when fewer guests were staying at the hotel. These findings could be attributed, in part, to our use of monthly hotel occupancy rates (rather than daily rates) to estimate how many beachfront rooms were occupied on any given night. But they could also reflect larger patterns in visitors' lighting choices and behaviors.

Regardless of the type of education intervention (nonpersonal vs. interpersonal) selected for conservation education, our findings suggest what one might already suspect from a management standpoint: the most important times to address beachfront lighting is during peak visitation days (e.g., weekends) and months (e.g., late spring–early summer) when more guests are present and a higher percentage of guests in beachfront rooms leave lights on with blinds open. A targeted management regime based on peak visible light ratios could prevent potential negative impacts on nesting sea turtles in the future and likely aid in the population recovery of the endangered animals. It could ultimately also cost hotel and wildlife managers less time and money in mitigating the issue. How, then, might managers effectively accomplish these goals to reduce this light pollution problem?

Our results appear to suggest that the rack cards, although relatively easy to produce and distribute, might not be the best solution. According to the elaboration likelihood model (ELM) of persuasion, our results might not be surprising. When people have the motivation, experience, ability, and predisposition to assimilate information about relevant issues, then persuasive communication can generate lasting outcomes (Cacioppo and Petty 1984; Ballantyne and Packer 2011). We presume, however, that many hotel guests on Jekyll Island do not possess inherent interest in or experience with sea turtles (K.A.M., unpubl. data, 2016). This would make them more likely to engage with information via peripheral pathways in the ELM, and it would make them less likely to respond voluntarily to nonpersonal messages that are not also accompanied by more-direct personal experience (Gore et al. 2008). Research has shown that visitors to natural areas are most likely to alter their behavior and minimize ecological impacts following direct personal contact with guides or site staff (Sharp et al. 2012; Kidd et al. 2015). The enhanced educational value of personal and authentic animal encounters (relative to other forms of information distribution) for Jekyll Island visitors has been demonstrated to influence risk perceptions with other reptile species (e.g., alligators; Skupien et al. 2016) and could lead to deeper message processing and subsequent action with respect to turtles as well. For instance, Disney's Vero Beach Resort in Florida integrates sea turtles into the resort's theme using both nonpersonal and interpersonal methods. Sea turtles are therefore part of the broader guest experience, thus keeping thoughts of them more prominent in guests' minds while in their hotel rooms (B. Witherington, pers. comm., May 2017).

Further, other nonpersonal approaches might yield better results. Some studies have shown that distribution method matters. Flyers obtained independently by park visitors at a brochure stand often have a weaker impact on visitor knowledge, attitudes, and behaviors than do flyers that are personally delivered by staff members (Watson et al. 1985; Moscardo 1999). Another study showed that the type of messaging was important: descriptive (what “everyone else” is doing), injunctive (what “should” be done), and economic messages influence guests differently based on demographics (Morgan and Chompreeda 2015). The rack card was largely injunctive in its messaging; perhaps a switch to descriptive messaging would add the element of peer pressure needed for guests to respond with sea turtle–friendly behavior. A shift to a more-coercive, regulatory, or incentive-based message might also have a greater impact on visitor behavior. A similar recommendation for incentive-based interventions was put forth in a study by Dolcinar et al. (2017) in Slovenia after stickers failed to influence behavior change in the reduction of towel and electricity use.

Although many tourists are eager to absorb and apply new information about wildlife via free-choice learning opportunities during their trips (Ballantyne et al. 2009), few tourists may view hotel rooms as a setting where this type of learning can occur. Therefore, accessible placement of information within the room may be critical, generating subtle associations that impact visitor choices. For example, a placard near the light switch or window shades in hotel rooms (vs. on the bed) would serve as an omnipresent reminder for guests to think about the implications of their decisions with respect to lights. However, as mentioned above, this type of permanent message positioning may not be approved by corporate-level hotel managers concerned with brand and consistency. If beachfront hotels were to obtain corporate approval for a sea turtle–friendly lighting plaque, its use should be tested prior to full implementation to determine its effectiveness.

While our study suggests that rack cards are not effective in modifying hotel guest behavior, it also highlights the need for more research that addresses limitations of the current study and assesses the efficacy of various communication and outreach tools on specific visitor behaviors. Estimation of behavioral change linked to educational interventions is inherently difficult (Hungerford and Volk 1990; Kollmuss and Agyeman 2002), particularly when behaviors are measured at aggregate scales (i.e., block of hotel guests vs. individual visitors). In this case, the only hotel occupancy rates available to us were grouped by month rather than day and did not differentiate beachfront rooms (which are typically filled first) from others at the hotel. This average affected our ability to precisely control for daily variations in occupancy that might have influenced our visible light count, but relative differences across similar sampling days likely remained comparable.

Finally, another major limitation that we encountered was that in June and July 2016 (the control year), the hotel requested to have a GSTC educator present during peak check-in hours on Fridays to educate guests about sea turtles, the lighting issues they face, and to promote a field-based education program. These personal interactions could have introduced additional educational content (at least for some guests) into our control group. However, ANOVAs checking for differences in visible light counts across months in 2016 suggested this impact might be negligible.

Despite these limitations, our study adopted the relatively novel approach of assessing conservation education outcomes by assessing an overt behavior (in this case, lighting choices and behaviors of hotel guests) rather than by relying exclusively on potentially biased, self-reported measures (Heimlich and Ardoin 2008; Munro et al. 2008). Future studies could examine impacts on the actions of specific visitors—offering a more-precise and potentially more-accurate scale of resolution. Regardless of analytical scale, it is likely that a single message or platform will not lead to impactful behavior change. The most effective conservation campaigns typically connect with visitors in multiple ways at multiple times, reinforcing key messages (Ballantyne et al. 2011; Ardoin and Heimlich 2013). Rather than spending money on communication strategies and hoping that they work, conservation managers and partners should work together to develop solutions for a sustainable coexistence between humans and wildlife at popular and ecologically sensitive tourist destinations. Once implemented, evaluation should take place to determine whether any impacts to people's conservation attitudes and behavior actually result in tangible biological outcomes (Veríssimo et al. 2017).

Considering our findings and the current literature, we believe that a diverse toolkit is likely needed for delivering important messages that inspire conservation behavior. With recent technological advances and cultural transitions, the way in which news and information is being consumed has changed (Choi 2015). This shift makes it difficult for managers to be assured that messages broadcast via conventional media sources will reach target audiences. In the digital age, conservation managers must be able to identify centralized, broad-reaching tactics through which they can succinctly disseminate important information (Marion and Reid 2007). Furthermore, these messages must resonate with people from diverse backgrounds, with unique interests and passions, and with different learning styles (Cassidy 2004).

Collaborations between tourism industry partners (e.g., hotel chains), conservation managers, and tourists are an important piece of this puzzle, helping to ensure an appropriate and consistent communication of information that helps to maintain enjoyable visitor experiences while achieving conservation goals (Ballantyne et al. 2009). Interpersonal interactions may be an effective means of accomplishing this goal, but managers cannot rely exclusively on these strategies because they reach only a fraction of the target audience. For example, studies suggest that formal sea turtle programming on Jekyll Island has been successful at conferring changes in participants' behaviors regarding use of sea turtle–friendly lights as well as other behaviors that can negatively impact nesting sea turtles such as littering (K.A.M., unpubl. data, 2016). However, this type of formal programming is an intense resource investment (e.g., personnel time and costs) and may prohibit certain nesting beaches from implementing similar programs in the hopes of better protecting sea turtles from light pollution. Additionally, beachfront hotels might consider investments in technological fixes to conservation problems (Heberlein 2012) that accomplish goals irrespective of visitor behavior. For example, by tinting beachfront windows, the impacts of artificial lights on turtles could be minimized even if guests and maintenance personnel did not know or comply with lighting regulations.

When dealing with situations involving endangered or threatened species such as the loggerhead sea turtle, conservation managers and their industry partners must be able to communicate quickly and effectively with the public; if they fail to do so, it is possible that the negative consequences conferred by (often unintentional) behaviors could result in regulatory violations and further decline of the species population. Conservation managers must be able to find cost-effective and time-efficient methods through which they can broadcast their messages.