Individual animals do not necessarily exhibit largescale movements (to the extent of their occurrence) on a regular basis, but instead constrain their movements to smaller areas where they forage, rest, and avoid predators (Powell 2000). These smaller areas are considered home ranges (Burt 1943) and are a fundamental aspect of spatial biology. Understanding animal home ranges can help researchers develop conservation methods (Safi et al. 2007), understand individual movements (van Dam and Diez 1998), locate invasive species (Harrington and Macdonald 2008), and protect populations (Kramer and Chapman 1999). Deriving robust estimates of home ranges typically require animal sightings or locations that are usually obtained via biotelemetry (Gubbins 2002; Cooke 2008). These telemetry data are usually collected directly by scientists; however, some researchers have used citizenscientists (individuals of the public who are opportunistically involved in scientific data collection [Bonney et al. 2009; Silverton 2009]) to help log animal sightings (Aguilar-Perera et al. 2012).

Animal sightings collected by citizen-scientists may help researchers understand the geographic extent of invasive species or assist in spatial decision-making processes for the protection of animal habitats. Whereas some studies have utilized hand-written global positioning system (GPS) coordinates (Lunney et al. 2000; Aguilar-Perera et al. 2012), other studies such as those by Azzurro et al. (2013) and van der Wal et al. (2015) utilized webbased data collection methods, allowing citizen-scientists to upload their own information on animal sightings. These studies have, for example, helped researchers understand how to manage invasive species and increase conservation efforts (Knapp and Owens 2005; Cooke 2008; Anderson et al. 2017). Web-based interactive maps allow users to query layers and are useful sources for gathering information from citizen-scientists for use in wildlife management decision-making. In another project, researchers devised a method for identifying individual sea turtles over periods of months and years through a web map of citizen-based sea turtle sightings data (Baumbach and Dunbar 2017). This trend of collecting data online or through smartphone devices is quickly becoming a popular method of data gathering that facilitates public contributions to scientific endeavors and, at the same time, educates the public about the importance of ongoing research (Newman et al. 2012; Land-Zandstra et al. 2016).

Few studies provide tourists with the opportunity to directly record animal sightings into a geographic information system (GIS). This is largely due to the public's lack of knowledge about how to work with GIS technologies used by researchers and conservation managers to map data and reveal spatial patterns (Rodgers and Carr 1998). Additionally, the process of analyzing data for spatial analyses can be complex and further complicated by postcollection processes such as data storage, quality control, and deciding which analysis methods to implement (Urbano et al. 2010). Although generally unknown to the public at large, several GIS analytical tools are available (Rodgers and Carr 1998; Calenge 2006; Halpin et al. 2009; Fortin et al. 2012; Fujioka et al. 2014), yet these often require high-cost software and/or significant training, which preclude their use by nonscientists. Nevertheless, researchers have requested aid from the public to gather animal sightings on a more continual basis. Public involvement in field research is a growing opportunity to collect information on animals when researchers are unable to be present in the field.

Recently, researchers have been engaging citizenscientists to help collect data using smartphones (Teacher et al. 2013). Yet of the thousands of software applications available from both major application vendors (Apple AppStore^© and Google Play Store^©), as of 2015 only 33 were designed for nature-related, citizen-based data collection (Jepson and Ladle 2015). Such applications use built-in GPS capabilities in smartphones (despite low accuracy) to record approximate sightings locations, which are then uploaded to the cloud or a server for storage. Citizen-science projects using smartphones can provide data that are useful for projects such as mapping marine litter (Jambeck and Johnsen 2015), mapping invasive species (Adriaens et al. 2015; Wallace et al. 2016), and tracking animal sightings (Sequeira et al. 2014; Stelle and King 2014). Such studies utilize both web maps and smartphone applications to maximize the number of methods through which citizen-scientists can upload data.

Several sea turtle species provide opportunities to identify home ranges using citizen-based sightings due to their fidelity to specific foraging sites that are accessible to snorkelers and self-contained underwater breathing apparatus (SCUBA) divers (León and Diez 1999; Schofield et al. 2010). Sea turtles have been reported in almost every ocean and sea around the globe. Until biotelemetric advances in animal tracking, sea turtles were difficult to track unless flipper-tagged at nesting beaches and visually sighted by fishermen or community members in distant locations (Carr 1984). Examples of biotelemetric techniques include, but are not limited to, the use of satellite (Schofield et al. 2010; Hawkes et al. 2011; Gaos et al. 2012), acoustic (Taquet et al. 2006; Fuentes et al. 2019), and radiotracking (Berube et al. 2012).

Gaos et al. (2012) elucidated internesting movements and foraging home ranges using satellite telemetry for critically endangered eastern Pacific hawksbills (Eretmochelys imbricata). Schofield et al. (2010) and Hawkes et al. (2011) used satellite telemetry to monitor loggerhead (Caretta caretta) foraging home ranges over multiple seasons and years to determine how habitat use changed over these temporal scales. However, Cuevas et al. (2008) determined that satellite telemetry may not be the most appropriate method to determine small-scale movements and should be replaced with radiotelemetry to more accurately determine home range sizes. Acoustic tracking to monitor sea turtles at small spatial scales, typically of less than 1–2 km², has successfully been used in several studies (van Dam and Diez 1998; Scales et al. 2011; Carrión-Cortez et al. 2013). In contrast, in a study using both sonic and radio transmitters applied to green turtles in the Gulf of California, Mexico, Seminoff et al. (2002) found when food items are less concentrated in regions where resident turtles are found, home ranges may be larger because turtles travel greater distances to obtain preferred food items than when these items are highly abundant. In the case of hawksbills, sponge distribution has been found to influence home range sizes (van Dam and Diez 1998; Scales et al. 2011) and, therefore, investigating abundance of preferred sponge prey species within an area may improve our understanding of habitat use within home ranges. Using radio transmitters alone to analyze the home ranges of juvenile hawksbills in the Port Royal region of Roatán, Honduras, Berube et al. (2012) found that these turtles had a home range of less than 1 km² and likewise suggested abundant sponge prey was likely to influence home range size.

In many areas, sea turtles use habitats that overlap areas used by tourists (Marcovaldi and Dei Marcovaldi 1999; Campbell and Smith 2006; Hayes et al. 2016), offering excellent opportunities to harness citizen-based sea turtle information. Baumbach and Dunbar (2017) used a web map to aid citizen-scientist dive tourists in logging sea turtle sightings data along with photographs and dive log information through their web-based map of the island of Roatán, Honduras. However, they recognized that not all dive tourists have access to computers immediately after diving, which hindered their involvement with data provision. In response, the authors created a smartphone application to facilitate logging post dive sea turtle sightings, an advancement that has helped increase citizen-based data provision.

Here we describe the development of a web-based map system with smartphone data inputs, and we interpret data generated from citizen-scientists to estimate sea turtle home ranges. We used hawksbill sea turtles as a case study to evaluate whether sightings collected by citizen-scientists during a 3-yr period could provide comparable home ranges to previous studies. We also investigated whether the abundance of sponge prey species influenced hawksbill home range sizes within a marine protected area (MPA). Through this study we promote new methods for citizenscientists to log data and upload photos of sea turtles using GIS mapping and smartphone technologies.

METHODS

Study Site.—The island of Roatán lies approximately 60 km north of mainland Honduras with the protected Sandy Bay West End Marine Reserve (SBWEMR), located along 13 km of coastline on the western end of the island. In an ongoing study, the SBWEMR has been divided into the 3 general zones of West Bay (2.97 km²), West End (2.99 km²), and Sandy Bay (2.83 km²) (Wright et al. 2017; Fig. 1) and is surrounded by a barrier reef immediately offshore that extends out approximately 92 m to the reef crest at a depth of approximately 18 m. Following the reef crest is a gradual descent for 2.2 km before an abrupt drop to depths greater than 130 m (Hayes et al. 2016). However, benthic variability exists in specific areas of the West Bay and West End zones where sandy areas lacking coral shelf and crest, slope to an abrupt dropoff. The reef crest comprises a wide variety of substrates, including hard corals, soft corals, and sponges, which host a wide variety of marine life. This abundant biodiversity attracts divers from around the world, with most dive tourism occurring in the West Bay and West End zones primarily from January to April and reaching peak dive tourism in March (Hayes et al. 2016).

View Figure

• Download as Powerpoint
• Download Figure

In-Water Sponge Counts. — We conducted in-water transects to determine the abundance of barrel sponges (Geodia neptuni) during approximately 70-min dives in each of the 3 zones within the SBWEMR from June to September of 2017 and from June to July of 2019. On reaching the reef crest, the dive team, consisting of 2–4 divers, spread out in a line perpendicular to the reef slope. Divers then swam in a straight line parallel to the slope line (Fig. 2). During the first half of each dive (on the outbound portion), 1 of 3 researchers counted each sighting of G. neptuni along a straight-line, unmarked swim transect to the farthest extent of each dive. We chose to survey this sponge because it is a main diet item for juvenile hawksbills in the area of Roatán (Berube et al. 2012; Baumbach et al. 2015). To prevent double counts, we did not count sponges during the inbound portion of the dive. In some cases, sites had multiple counts done over multiple dives, although swim line (position of the counter in the dive line) and extent of swim transects were random. We averaged sponge counts from sites with multiple dives to provide an estimation of the number of sponges for that dive site. Depths of transects were variable depending on dive direction and bathymetry. Final counts were tallied in Microsoft Excel (v. 16.11.1). Zone areas in square kilometers were calculated in ArcMap by measuring along the boundaries of each zone.

View Figure

• Download as Powerpoint
• Download Figure

Web-Based Map Development and Application. — We worked with 12 dive shops in Roatán that offer dive expeditions for tourists multiple times each day, providing many opportunities for sighting resident juvenile green (Chelonia mydas) and hawksbill turtles due to their residence within the SBWEMR. We provided sightings, species ID, web map, and smartphone app training to dive instructors who then, in turn, assisted dive tourists in identifying sea turtle species and recording sightings. Tourists could also receive additional species information within our interactive web map. We previously designed a web-based map for the island of Roatán (https://arcg.is/1CamGy) during the summer of 2015 with the aim of collecting sea turtle sightings from citizen-scientist dive tourists from each of the 12 participating dive shops (Baumbach and Dunbar 2017). A more detailed description of both the development and data input fields of the Roatán web-based map can be found in Baumbach and Dunbar (2017). Briefly, these include name, sea turtle species, dive site, and depth sighted (for a full list see Table 1). To input data immediately after diving, citizenscientists were asked to open the map hosted on the Protective Turtle Ecology Center for Training, Outreach, and Research, Inc. (ProTECTOR, Inc.) web site (www.turtleprotector.org), scroll to the dive site they visited, choose the turtle species, and place an icon at the dive site where the turtle was sighted. Once the turtle icon was placed, a pop-up box appeared asking for metadata such as name, e-mail, date, weather conditions, and turtle gender. We specified how to upload photos in the instructions on the map and explained that photos were required for us to verify turtle species. To determine an average distance error for citizen-based turtle sightings, we also measured the straight-line distance between 20 different dive sites within the West Bay and West End zones and calculated a mean and standard deviation.

Table 1. A comparison of data fields between the Roatán web map and the TURT smartphone application.

View Fullscreen

Smartphone Application Development. — We developed a smartphone application, called Turtles Uniting Researchers and Tourists (TURT), using Esri's AppStudio for Desktop (ver. 3.3.110), accompanied by Qt Creator (v. 5.12.1) to provide citizen-scientists access to a global map on which they are able to log sea turtle sightings anywhere in the world. In collaboration with the AppStudio for ArcGIS developer team, we created a quick report application on the AppStudio for ArcGIS web site that provided the basic layout and structure of the application, and then we imported it into AppStudio for Desktop. TURT opens to a welcome page that prompts citizenscientists to log a new turtle sighting and then goes through a series of pages where citizen-scientists select the species of turtle seen as indicated by different colored sea turtle icons, upload multiple photos or videos of a single turtle, choose the dive site location where they saw the turtle, and enter data into specific data fields. Fields for name, e-mail, turtle record type (in-water sighting or turtle product), country, date, and photo are required before citizen-scientists are able to upload the turtle sighting record. For citizen-scientists who may be logging sightings in remote locations, we developed TURT to save reports in an offline mode which can then be uploaded once connected to Wi-Fi or satellites. TURT also has an information button where citizen-scientists can view sample photos of sea turtles for identification as well as providing specifications about what data to upload within the application. We custom-coded TURT with Qt Creator to provide links in the confirmation page where citizenscientists view all currently logged turtle sightings based on record type.

Sea Turtle Identification and Home Range Analysis. — We collected sea turtle sightings from the Roatán web-based map and TURT for individual turtles in the SBWEMR that had 10 or more sightings. This minimum number is appropriate as a low sighting number when working with endangered species (Muths 2003; Silva et al. 2008). Individual turtles were previously tagged with Inconel flipper tags (Style 681, National Brand and Tag Company, USA), which were then used along with photo ID to positively identify hawksbills (Dunbar et al. 2014). Photos and scute patterns were analyzed in a computerized photo identification program as described in Dunbar et al. (2017) to identify individual turtles from a collection of photos in the feature layers of each map. We then plotted sightings data in ArcGIS Online on individual web mapping applications, with each individual hawksbill identified by its flipper-tag numbers. Individual hawksbill sightings were then plotted on the map in 2-wk intervals from July 2014 to December 2017 to determine distribution by dive site in the SBWEMR over time. Time series maps of each turtle may be viewed at https://arcg.is/zvTqe (BBQ150), https://arcg.is/09nWGe (BBQ260), https://arcg.is/055fyK (BBQ205), and https://arcg.is/8yq9n (BBQ346).

We used ArcMap (ver. 10.5) to map the same individual turtles at different dive sites within the SBWEMR, and then measured the distance between the two farthest sightings along with the area of sightings for each turtle. Maximum distance (km) between the two farthest sightings was determined with the standard measure tool in ArcMap. Home range area (km²) for each hawksbill was calculated using minimum bounding geometry with convex hull in ArcGIS Pro (ver. 2.1, Esri, Redlands, CA). When home range area covered portions of land, vertices were manually moved away from land and followed the shoreline to more adequately depict a potential area where hawksbills may be sighted. Finally, we overlaid all individual home ranges to determine if there was any overlap among turtles. We exported home range maps in a JPEG format.

Statistical Analyses. — Because of sample distribution and nonnormal data, we analyzed sponge counts in the SBWEMR using a Kruskal-Wallis H test in SPSS (IBM 2015) to determine if the number of sponges differed among the West Bay, West End, and Sandy Bay sampling areas. We then used a Mann-Whitney U-test for post hoc pairwise comparisons. Additionally, we conducted Spearman correlations to determine if there were any significant associations between home range size and several variables such as sponge numbers, sea turtle size and weight, and total data collection duration. Finally, we used Wilcoxon signed rank tests to determine if there was any significant variation among individual hawksbills in both size and weight. Values are presented as means ± standard deviations (SD).

RESULTS

We conducted a total of 42 transect sponge count dives across all 3 zones (see Table 2 for dive counts per zone). Zones were approximately equal to each other in area (Table 2). We calculated the average distance error between dive sites to be 170 ± 88 m. Sponge counts differed significantly among the 3 areas sampled (x²₂ = 10.16, p = 0.006, η² = 0.25). Counts during individual transects yielded from 3 to 66 sponges for West Bay (n = 15; 36 ± 19), 1 to 65 sponges for West End (n = 19; 24 ± 19), and 2 to 23 sponges for Sandy Bay (n = 8; 9 ± 7) (Table 2). Post hoc pairwise comparisons showed that West Bay (x²₁ = 9.04, p = 0.003, η² = 0.41) and West End (x²₁ = 4.30, p = 0.038, η² = 0.17) had significantly higher sponge counts compared with Sandy Bay, although West Bay and West End did not differ significantly from each other (x²₁ = 2.89, p = 0.089, η² = 0.088).

Table 2. Sponge counts of Geodia for the 3 zones within the Sandy Bay West End Marine Reserve. There are significantly more Geodia within the West Bay and West End zones than in the Sandy Bay zone. A Mann-Whitney U-test was used to calculate p-values for differences in sponge counts among the 3 zones

View Fullscreen

We identified 4 individual hawksbills (RMP T047, RMP T048, RMP T077, RMP T078) that fit our mapping criteria of 10 or more sightings from 2014 to 2017. No turtle sightings were recorded in 2015. Turtle RMP T077 had 10 sightings records, meeting the minimum requirement for this study, whereas RMP T047 had the most individual sightings with 20 records. RMP T048 and RMP T078 fell between these values with 13 and 19 records, respectively (Table 3; Fig. 3). Turtles within our study ranged in size from 48.9 to 62.4 cm minimum curved carapace length (CCLmin) and weights ranged from 14.2 to 28.6 kg (Table 3). Wilcoxon signed tests showed that there was no difference in size or weight among hawksbills (CCLmin = 57.68 ± 6.35 cm, p = 0.12; weight = 22.18 ± 7.38 kg, p = 0.12).

Table 3. A comparison of the maximum distance traveled and home range calculated through multiple convex hull for 4 individual hawksbills within the Sandy Bay West End Marine Reserve. These data were calculated using turtle sightings uploaded by citizenscientists to either the Roatán web map or through the TURT smartphone application. CCLmin = minimum curved carapace length.

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

Of the 4 turtle home ranges analyzed using citizenbased sea turtle sightings, RMP T047 had the longest maximum distance and largest area (max distance = 6.90 km, area = 1.44 km²), whereas RMP T048 showed the shortest maximum distance (max distance = 1.87 km) among all reported turtles. Although RMP T047 was sighted at both ends of the SBWEMR, all other turtles had a maximum sighting distance of less than 3.5 km. Home ranges calculated from minimum bounding convex hulls ranged from 0.22 km² (RMP T048) to 1.44 km² (RMP T047) with a mean of 0.68 ± 0.55 km² (Table 3; Fig. 3). Spearman correlation showed no difference between home range size and sponge numbers within each hawksbill's home range (ρ = 0.9, p = 0.083). Similarly, Spearman rank tests showed no significant differences between home range size and hawksbill size and weight (ρ = 0.8, p = 0.33; ρ = 0.6, p = 0.41). Turtle sightings occurred over periods ranging from 450 to 1123 d (Appendix 1), yet Spearman correlation showed no difference between home range size and total data collection duration (ρ = 0.6, p = 0.41). Our analyses showed that all hawksbill home ranges overlapped with each other (Fig. 4), with RMP T047′s home range encompassing all other home ranges and extending into the Sandy Bay zone. Therefore, RMP T047 represented the largest percent overlap (100%) with other turtles, whereas RMP T048 represented the smallest percent overlap (15%) compared with RMP T047.

View Figure

• Download as Powerpoint
• Download Figure

DISCUSSION

In this study, we demonstrated the successful use of both web map and smartphone applications to estimate hawksbill home ranges using citizen-based sightings data. Our results indicated that hawksbills in the SBWEMR typically had home ranges of less than 1 km², with the exception of one turtle that was seen once outside of the West Bay and West End zones. These results are consistent with previous studies that also found hawksbill home ranges to be confined spatially (van Dam and Diez 1998; Scales et al. 2011; Berube et al. 2012). The current study demonstrates that SBWEMR boundaries are sufficient for the protection of the individual turtles studied. However, we suggest some caution in the interpretation of these results because dive tourists do not typically dive outside the SBWEMR and, thus, are less likely to provide turtle sightings from outside the protected area. This conclusion is supported by similar studies in marine protected areas, such as those by van Dam and Diez (1998) and Scales et al. (2011), which identified juvenile hawksbill home ranges that were within marine protected areas. In contrast, an established marine protected area at Punta Coyote, Costa Rica, was found to be insufficient by not encompassing the entire extent of hawksbill home ranges, which typically covered 0.6 km² in that study (Carrión-Cortez et al. 2013). Further population level home range analyses may demonstrate either the sufficiency or insufficiency of the SBWEMR for hawksbill protection. If insufficient, marine reserve boundaries may need to be reassessed by reserve conservation managers. Similarly, Carrión-Cortez et al. (2013) recommended that the protected area in Punta Coyote either be extended to cover the reef areas where hawksbills aggregated or that a new marine protected area be created that encompassed the identified home ranges for those turtles.

We discovered that the 4 hawksbill home ranges within our study overlapped with each other to varying extents within Zones 1 and 2 within the SBWEMR. This overlap among individual hawksbills provides evidence that the West Bay and West End zones provide abundant and adequate sponge food items to accommodate shared resources among hawksbills. RMP T047 utilized Zone 3 as a part of its home range, yet no other hawksbill home ranges overlapped within Zone 3, which may in part be due to a lower abundance of adequate sponges in this area of the reserve. Although utilizing different data collection and analysis tools, several studies have also previously reported a high degree of overlap in hawksbill home ranges at each study site (van Dam and Diez 1998; Scales et al. 2011; Berube et al. 2012; Carrión-Cortez et al. 2013), suggesting that these sites have adequate food items and are able to accommodate relatively large populations of developing juvenile hawksbills (Scales et al. 2011). Our results suggest that the small home ranges of hawksbills concentrated within the West Bay and West End zones of Roatán may likewise be due to a high relative abundance of sponge food items in these zones, providing agreement with previous investigations. Juvenile hawksbills in the SBWEMR are resident (evidenced by resightings over time as shown in Appendix 1) and, thus, may be restricting their home ranges to areas with abundant quantities of sponge in order to reduce foraging energy expenditure and increase growth.

We found significant differences in the abundance of sponge prey between West Bay and Sandy Bay as well as between West End and Sandy Bay. The greater abundance of sponges in West Bay and West End may help explain the concentration of hawksbill home ranges within Zones 1 and 2. However, there was no significant difference between individual hawksbill home ranges and sponge numbers within each home range throughout the SBWEMR, although sponge numbers differed by zones when analyzed independently. Even though hawksbills with larger home ranges had access to more sponges within their range, most of these sponges occurred within West End and West Bay, with very few being counted in Sandy Bay. Yet, hawksbills with smaller home ranges still had access to large numbers of sponge within a smaller area, suggesting that West Bay and West End zones may provide optimal foraging conditions. In contrast, RMP T047 was seen once in Sandy Bay where there are few Geodia sponges and, thus, it is possible this individual turtle had a larger home range than hawksbills not seen utilizing the Sandy Bay zone in order to obtain adequate food items. We find some agreement for this conclusion in the study by Seminoff et al. (2002), who used multiple convex polygons to demonstrate that green turtles in Bahía de los Angeles, Mexico had large mean home ranges of 16 km² because algal patches were less abundant, requiring turtles to move greater distances among foraging patches.

Results from the current study suggest citizen-based sea turtle sightings may be a useful method to estimate home range areas. For example, juvenile hawksbills within the SBWEMR have home ranges that spanned from 0.22 to 1.44 km² and are similar, yet slightly larger, compared with juvenile hawksbills along the southeastern end of Roatán, which ranged from 0.15 to 0.55 km² (Berube et al. 2012). This pattern of relatively small juvenile hawksbill home ranges has also been noted elsewhere in the Caribbean such as in Puerto Rico (van Dam and Diez 1998) and Belize (Scales et al. 2011). These studies suggested that these small home ranges may have been due to an abundance of high-quality food items in their study sites. In agreement, we also suggest that small home ranges within our study site may be due to an abundance of high-quality food items, which may reduce foraging competition between individuals and allow hawksbill home ranges to overlap.

Data collected from citizen-scientists represent an untapped source of sea turtle sightings. The SBWEMR is visited by many dive tourists who typically write or digitally store their dive information from their dive computers and then date photographs of wildlife seen during dives. From these citizen-based sightings we were able to estimate the approximate home ranges of individual hawksbills that could later be verified through moreprecise biotelemetry tracking techniques, or vice versa. Some scientists have expressed concerns about the quality of data uploaded by citizen-scientists, stating that citizenscientists are untrained and lack the knowledge to upload data of similar quality to that of trained scientists (Alabri and Hunter 2010). However, Williams et al. (2015) found information from tourist dive logs to be useful and reliable, as did Goffredo et al. (2010), who found data logged by dive tourists were as adequate as those of a trained marine biologist. In this study, we relied on citizen-scientists to log sea turtle sightings within an average of 170 ± 88 m of the dive site buoy. Adequate and reliable data collected from citizen-scientists, such as simple sightings and location data, can provide information over multiple years, especially when scientists are unable to be in the field. Nevertheless, the design of web maps or smartphone applications to facilitate citizen-based data uploads should be educational and simple to use (Newman et al. 2010). We have integrated this approach by providing the necessary fields for citizen-scientists to populate, along with the presence of an information button that provides links to assist in identifying turtle species on both the smartphone application and web map.

Although we have applied a user-friendly approach to logging citizen-based sightings data, as suggested by Newman et al. (2010), user frequency for logging sea turtle sightings on the web map and smartphone application remains relatively low. One way to address this is to perhaps display banners with quick response (QR) codes in supporting dive shops that describe the uploading process to encourage more visitors to use TURT and our Roatán web map. Dive tourists are able to take photos of these banners and later log sea turtle sightings at their convenience. However, the drawback in this approach is the introduction of recollection error and, as with many citizen-based input processes, the potential for loss of interest to participate with increasing time after the sighting event.

GIS can provide citizen-scientists with the ability to aid researchers in logging data to identify movement patterns from any animal at local and international scales in order to improve conservation efforts (Lunney et al. 2000; Lee et al. 2006; Newman et al. 2010; Wood et al. 2011; Catlin-Groves 2012). Thus, GIS remains a powerful tool that can be used to identify a variety of patterns to help wildlife managers understand how to better manage protected areas locally and work internationally in cases of highly migratory species (Blumenthal et al. 2006). Although other studies have used similar approaches with whale sharks (Rhincodon typus) (Holmberg et al. 2009), manta rays (Manta alfredi) (Jaine et al. 2012), and sharks (Carcharhinus amblyrhynchos) (Vianna et al. 2014), to our knowledge, this is the first study to develop a citizenscience approach for identifying home ranges of sea turtles. We recommend citizen-based data can be integrated with new GIS technologies as a beneficial method for determining animal movements and home range sizes that may be applied in any location around the globe.