Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Long-term assessment of whale shark population demography and connectivity using photo-identification in the Western Atlantic Ocean

  • Jennifer A. McKinney ,

    Roles Conceptualization, Formal analysis, Investigation, Methodology, Project administration, Resources, Visualization, Writing – original draft, Writing – review & editing

    jennifershark@gmail.com

    Affiliation Louisiana Department of Wildlife and Fisheries, New Orleans, LA, United States of America

  • Eric R. Hoffmayer,

    Roles Conceptualization, Formal analysis, Investigation, Methodology, Project administration, Resources, Visualization, Writing – original draft, Writing – review & editing

    Affiliation National Marine Fisheries Service, Southeast Fisheries Science Center, Mississippi Laboratories, Pascagoula, MS, United States of America

  • Jason Holmberg,

    Roles Conceptualization, Data curation, Investigation, Project administration, Resources

    Affiliation Wild Me, Portland, OR, United States of America

  • Rachel T. Graham,

    Roles Conceptualization, Investigation, Methodology, Project administration, Resources, Visualization, Writing – original draft, Writing – review & editing

    Affiliation MarAlliance, San Pedro, Ambergris Caye, Belize

  • William B. Driggers III,

    Roles Conceptualization, Investigation, Visualization, Writing – original draft, Writing – review & editing

    Affiliation National Marine Fisheries Service, Southeast Fisheries Science Center, Mississippi Laboratories, Pascagoula, MS, United States of America

  • Rafael de la Parra-Venegas,

    Roles Conceptualization, Investigation, Project administration, Resources

    Affiliation Ch'ooj Ajauil AC, Cancún, Quintana Roo, México

  • Beatriz E. Galván-Pastoriza,

    Roles Conceptualization, Investigation, Resources

    Affiliation Ch'ooj Ajauil AC, Cancún, Quintana Roo, México

  • Steve Fox,

    Roles Conceptualization, Investigation, Project administration, Resources

    Affiliation Utila Whale Shark Research, Utila, Bay Islands, Honduras

  • Simon J. Pierce,

    Roles Conceptualization, Investigation, Methodology, Resources, Visualization, Writing – original draft, Writing – review & editing

    Affiliations Wild Me, Portland, OR, United States of America, Marine Megafauna Foundation, Truckee, CA, United States of America

  • Alistair D. M. Dove

    Roles Conceptualization, Investigation, Project administration, Resources

    Affiliation Georgia Aquarium, Atlanta, GA, United States of America

Abstract

The predictable occurrence of whale sharks, Rhincodon typus, has been well documented in several areas. However, information relating to their migratory patterns, residency times and connectivity across broad spatial scales is limited. In the present study photo-identification data is used to describe whale shark population structure and connectivity among known aggregation sites within the Western Central Atlantic Ocean (WCA). From 1999 to 2015, 1,361 individuals were identified from four distinct areas: the Yucatan Peninsula, Mexico (n = 1,115); Honduras (n = 146); northern Gulf of Mexico, United States (n = 112), and Belize (n = 49). Seasonal patterns in whale shark occurrence were evident with encounters occurring in the western Caribbean Sea earlier in the year than in the GOM. There was also a significant sex bias with 2.6 times more males present than females. Seventy sharks were observed in more than one area and the highest degree of connectivity occurred among three aggregation sites along the Mesoamerican Reef. Despite this, the majority of resightings occurred in the area where the respective sharks were first identified. This was true for the WCA as a whole, with the exception of Belize. Site fidelity was highest in Mexico. Maximum likelihood modelling resulted in a population estimate of 2,167 (95% c.i. 1585.21–2909.86) sharks throughout the entire region. This study is the first attempt to provide a broad, regional population estimate using photo-identification data from multiple whale shark aggregations. Our aim is to provide population metrics, along with the description of region-scale connectivity, that will help guide conservation action in the WCA. At a global level, rapidly growing photographic databases are allowing for researchers to look beyond the description of single aggregation sites and into the ocean-scale ecology of this pelagic species.

Introduction

The whale shark (Rhincodon typus) is the world’s largest extant fish and has a cosmopolitan distribution in tropical and warm temperate marine ecosystems [1]. Since its description by Smith in 1828 [2], the whale shark has remained an enigmatic species. Many aspects of its life history, such as age at maturity, longevity, and reproductive cycle are still poorly understood [1, 36]. This lack of knowledge has hindered conservation efforts. Based on > 50% reduction in the global population over the last 75 years, the IUCN Red List of Threatened Species recently changed the conservation status of the whale shark from “Vulnerable” to “Endangered” [7]. Additionally, this species benefits from several broad protection plans, including being listed under Appendix II of the Convention on International Trade in Endangered species (CITES). Despite these efforts, a more targeted conservation action requires a better understanding of the distribution, population structure and movement patterns [8].

While the whale sharks’ range is well established, information on the global population structure and movement patterns remains limited. Analyses of mitochondrial [9] and nuclear [10] DNA sequence data revealed shared haplotypes for whale sharks in the Atlantic, Indian and Pacific oceans and demonstrated historical gene flow among these ocean basins. These findings are supported by satellite tracking data which have documented large-scale movements across broad, oceanic expanses [1113]. Conversely, genetic data also reveal significant differences in haplotype frequencies between the Indo-Pacific and Atlantic regions [9,10] suggesting whale sharks within the Atlantic Ocean, particularly the Gulf of Mexico (GOM) [14], are relatively isolated from conspecifics in the Indian and Pacific oceans.

Elucidating large-scale movement patterns of whale sharks has been a challenging, yet growing field of research. The advent of satellite tracking technology has shed light on whale shark short-term movements but examining long-term patterns (>1 year) has been hampered by a number of factors, such as poor retention and non-reporting of tags [1,15]. Movements of aquatic animals have been monitored through the use of external tags since at least 1653 [16]. Typically, conventional tags are imprinted with a unique identifier/code and contact information (e.g. phone number) to allow the observer to relay resighting data to the investigator. While this method has been successful for tracking movements of many aquatic organisms, in whale sharks this approach has been shown ineffective for population studies, due to tag loss, lack of reporting, and inaccurate reporting [17,18]. However, the use of natural spot patterns has been proven useful for identifying individuals, having utility for movement and population studies.

Natural marks (e.g. fin shape, scars, spot patterns) have been widely used to identify individuals from a wide-range of taxa [19], including several shark species [2024]. Photo-identification of whale sharks based on skin spot-patterns provides a non-invasive method in which individuals can be monitored throughout their lifetime [25] and thus has utility for movement studies, as well as population modelling [18, 2629]. The applicability of photo-identification in studying the movements of marine organisms was first demonstrated by Würsig and Würsig [30] using individual-specific markings and/or fin shapes to identify and monitor cetaceans. More recently, photo-based analytical techniques, originally developed for investigating social structure in marine mammals [31], have been successfully applied to whale shark sighting data and used to calculate various population parameters [28,29,32,33].

Whale sharks form predictable aggregations at specific sites throughout their range. In the Western Central Atlantic Ocean (WCA), these aggregations occur off the coasts of Belize, Honduras, Mexico and the United States [18, 32, 3438]. The ability to reliably encounter whale sharks in close proximity to shore has resulted in the development of ecotourism industries in Isla Holbox, Isla Mujeres, Cancun (Mexico), Gladden Spit (Belize) and Utila (Honduras) [34]. While it has been suggested that ecotourism can adversely impact shark populations [4, 18, 26, 39, 40], it also creates an opportunity for the collection of whale shark data at an unprecedented level. Rather than relying on a few dedicated researchers, every tourist with a camera represents a potential citizen scientist collecting photographic data. Holmberg et al. [26] utilized photographic data collected from a number of sources, including citizen scientists, to estimate survival rate for whale sharks at Ningaloo Marine Park off Western Australia. These efforts led to the creation of online repositories for whale shark photographs obtained by researchers and citizen scientists around the world [25, 26, 41]. The largest of these repositories is the Wildbook for Whale Sharks photo-identification library (www.whaleshark.org), established in 2003, which has over 33,000 whale shark encounters, including over 4,000 photos taken in the WCA region. The objective of this study is to utilize these data to generate population parameters, and summarize long-term movements of whale sharks in the WCA.

Materials and methods

Study area

The GOM and Caribbean Sea are large, adjacent, and semi-enclosed seas within the WCA (Fig 1). The Caribbean Current, the principal current in the Caribbean Sea, enters the GOM through the Yucatan Channel and forms the Loop Current and the Yucatan Current [42]. These currents serve as major transport mechanisms allowing for regional connectivity for many marine organisms, including invertebrates [43], reef fishes [4446], and large pelagic teleosts [47]. Within the study region is the Mesoamerican Barrier Reef System, the second largest reef system in the world, which runs from the Honduran Bay Islands along the Belizean coast to the Yucatan Peninsula.

thumbnail
Fig 1. Spatial distribution of sightings data collected through Wildbook for Whale Sharks in the Western Central Atlantic Ocean during 1999–2015.

Movements between whale shark sightings within the Gulf of Mexico and Caribbean, including Honduras (HN), Belize (BZ), Mexico (MX), United States of America (US), and the greater Caribbean region (CRB) based on photo-identification data.

https://doi.org/10.1371/journal.pone.0180495.g001

Standardized photo-identification

Whale shark images used in this study were collected by researchers and recreational divers at various aggregation sites within the WCA, including Utila, Honduras, Gladden Spit, Belize, Yucatan Peninsula, Mexico, in the northern GOM, United States (US), and beyond (Fig 1). Due to the limited number of submissions overall throughout the Bahamian and Caribbean Island chains, this data was omitted from statistical analysis. Photographic identification relies on a clear image of the spot patterning behind the gills (also referred to as the “fingerprint”), on either the left or right sides. When evident in images or specifically noted by person submitting the image, sex was assigned based on the presence (male) or absence (female) of claspers. In addition to images, ancillary data is collected such as sighting date, time, location, and estimated total length. Total length was typically estimated using approximate lengths of nearby boats and divers for scale [18]. Identification of individual sharks was performed using the techniques and computer-assisted matching tools described in Arzoumanian et al. [25]. Individual identifications were assigned to sharks based on having a high-quality spot “fingerprint” coupled by low matching scores with existing sharks and visual assessment by trained researchers. By centralizing the data online, it is possible to make identifications across datasets and geographically distinct research efforts, providing a more cohesive look at spatial and temporal linkages. All whale shark sightings data used in this study are third-party data publicly available on www.whaleshark.org. Data was collected and used according to the Wildbook for Whale Sharks Terms and Conditions, with appropriate permissions. The authors confirm that the data are owned by a third party and that they had no special access privileges to the data sets.

All identification and sighting locations were imported into ArcMap 10.2 (Environmental Systems Research Institute, Inc., Redlands, CA, USA) and projected using the global WGS 1984 PDC Mercator. The minimum straight-line distance (km) between sighting locations was calculated and summed per individual to determine total distance traveled. This was considered to be the minimum distance traveled by each shark, which is an underestimation of true movement. Therefore, spatial error created by paths that may have crossed land was considered insignificant.

Regional connectivity and population size

The total number of uniquely identified individuals was determined for each area and summed across the entire study region. Since whale sharks are known to aggregate in large numbers, repeated encounters with the same individual within a single calendar month were consolidated to a single occurrence when investigating seasonality of occurrence. A discovery curve was constructed to visualize the number of newly identified individuals per year to determine if abundance estimates can be applied to the dataset [48]. A Student’s t-test was used to determine if there was a significant difference in length between sexes [49]. Differences in sex ratios were also investigated using a chi-square test with Yates’ correction to determine if there was any significant deviation from the assumed 1:1 male to female ratio [49]. These statistical tests were applied to the entire dataset and considered significant at alpha = 0.05. Mean values are reported with standard error throughout.

The residence times for individuals within the study area were investigated using the “movement” module in the compiled version of SOCPROG 2.7 [31]. Individual sighting data were used to estimate Lagged Identification Rate (LIR), the probability that the animal will be sighted again after a variable time lag from its first sighting. This approach, a modification of maximum likelihood methods, has been specifically designed to account for situations where individual identification data have been obtained on an opportunistic basis, and allows for non-random distribution of sampling effort (which, in this case, is focused on known feeding aggregations) [31, 50]. All individuals in the database were included with a location code corresponding to one of the four areas referenced above. Whole-site and within-between site LIRs were estimated and fitted to a series of open and closed population movement scenarios. The model with the lowest quasi-Akaike information criterion (QAIC) value, accounting for over-dispersion of the data, was considered to be the best description of shark residency patterns. The most parsimonious model with 100 bootstraps was used to generate parameter estimates and confidence intervals. Similarly, a movement model was also used to test the probabilities of travel between the four sites within the region, with an additional “outside” option to incorporate sharks that moved away from the primary study areas. In this case, the LIR represented the probability of the individual being sighted again in either the same area or moving to a different area after a specified time lag. In this case, the two most parsimonious model scenarios, fully-mixed and migration with full interchange, were fitted to the data with 100 bootstraps to calculate movement parameters with 95% confidence intervals [50].

Results

Western Central Atlantic Whale Shark demographics

From 1999 to 2015, 4,298 encounters from 1,361 individually identified whale sharks were available for analysis from the WCA (Fig 1). Over 300 people contributed photographic data for analysis, with approximately 82.1% (n = 3,572) of photographs being provided by researchers and 17.9% (n = 777) provided by recreational divers or ‘citizen scientists’. The largest number of identified sharks were obtained from Mexico from 2001–2015 (1,115 sharks during 3,669 encounters), followed by Honduras from 1999–2015 (146 sharks during 337 encounters), United States from 2003–2015 (112 sharks during 123 encounters), Belize from 1999–2015 (49 sharks during 144 encounters), and the greater Caribbean Sea from 2003–2015 (20 sharks during 25 encounters). The number of newly-identified whale sharks per year exhibited no asymptote throughout the study; however, the percentage of resighted individuals remained constant from 2011–2015 at 75.0% (Table 1, Fig 2). Half (n = 683) of the 1,361 sharks identified in this region were reported only once in the database. Of the 678 resighted sharks, 90% (n = 608) were only observed in the area where initially encountered. Only 10% (n = 70) of resighted individuals were observed in more than one area.

thumbnail
Fig 2. Discovery curve for newly identified whale sharks from the Western Central Atlantic Ocean (1999–2015).

https://doi.org/10.1371/journal.pone.0180495.g002

thumbnail
Table 1. Whale shark photo-identification records for Western Central Atlantic Ocean from 1999–2015 with details of new records and resightings from previous years.

https://doi.org/10.1371/journal.pone.0180495.t001

A seasonal pattern of whale shark occurrence was evident within the WCA, with encounters occurring off Honduras and Belize earlier in the year than off Mexico and the United States (Fig 3). The highest percentage of encounters occurred during spring off Honduras (57.1%, March to May) and Belize (86.6%, April to June), whereas off Mexico the peak time was primarily during summer (91.7%, June to Sept) (Fig 3A–3C). In US waters, the peak time for whale shark encounters was more protracted, occurring during summer and fall (78.0%, June to October) (Fig 3D).

thumbnail
Fig 3.

Monthly distribution of whale shark encounters in a) Honduras, b) Belize, c) Mexico, and d) US waters.

https://doi.org/10.1371/journal.pone.0180495.g003

There was a significant sex bias with 2.6 times more males present than females (χ2 = 135.2, n = 707, p < 0.001). Estimated total length for whale sharks ranged from 1.8 to 12.1 m, with males ranging from 1.8 to 11.0 m (6.8 ± 0.1 m, n = 169) and females ranging from 2.4 to 10.6 m (6.7 ± 0.2 m, n = 61). There was no significant difference in estimated total length between sexes (t228 = 0.148, p = 0.883). Approximately 89% of the whale sharks were estimated to be smaller than 8.0 m (Fig 4A).

thumbnail
Fig 4.

Length frequency distribution for a) all identified individuals and b) individuals that exhibited movement (n = 70) within the Western Central Atlantic region from 1999–2015.

https://doi.org/10.1371/journal.pone.0180495.g004

Regional connectivity and population size

The highest degree of connectivity occurred among the three aggregation sites along the Mesoamerican Reef, with the majority of the movements documented between Honduras and Belize (n = 40), Honduras and Mexico (n = 39), and Belize and Mexico (n = 18) (Fig 1). The time at liberty for the resighted sharks ranged from 0.1 to 16.5 years and peaked between 2–4 years (1.8 ± 2.6 years) (Fig 5). Sixteen sharks had sighting records longer than 10 years (Table 2). The longest time between sightings of the same individual was 16.5 years (BZ-008), which was first observed in Belize in 1999, resighted in Belize again in 2002, then observed in Mexico in 2008, 2011, 2013, 2014, and 2015. In some cases, sharks were observed repeatedly in the same area for several days to months, and in other cases, many years passed between encounters (Table 2). There was also a time lag evident for sharks observed during the same year among regions along the Mesoamerican Reef. The mean lag time between sharks observed in Belize and Honduras was 26 ± 4.7 days (3–35 days), whereas the lag time between Belize/Honduras and Mexico was 4.4 ± 0.3 months (2–6 months). In almost every case, when sharks were observed in both Belize/Honduras and Mexico in the same year, they were observed in Belize or Honduras during spring and in Mexico during summer. There were 13 movements observed between the United States and Mexican waters, with no apparent seasonal pattern between the movements. In the same year, some sharks moved from Mexico to the United States (MX-030; 2008), whereas others made the reverse transition (MX-343; 2010).

thumbnail
Fig 5. Time-at-liberty (years) distribution of whale sharks in the Wildbook for Whale Sharks photo-identification library from the Western Central Atlantic Ocean.

https://doi.org/10.1371/journal.pone.0180495.g005

thumbnail
Table 2. Summary statistics of whale sharks tracked for over 10 years via photo-identification records in the Western Central Atlantic Ocean from 1999–2015.

Lag refers to the time period between sightings. Country abbreviations include: Honduras (HN), Belize (BZ), Mexico (MX), and United States of America (US).

https://doi.org/10.1371/journal.pone.0180495.t002

Modelled LIR declined rapidly during the first 200 days following initial sighting (Fig 6) There was a slight increase in probability of reidentification at one year (Fig 6), suggesting an annual periodicity of whale shark presence in the study area. The most open model accounting for emigration, reimmigration, and mortality (Model G) best fit the empirical data based on the QAIC (Table 3). In this modeled scenario, there were 57.4 ± 62.6 S. E. (95% c.i. 28.4–277.2) sharks likely to be observed on any given day. The mean residency was 0.8 ± 7.2 S.E. (95% c.i. 0.5–29.7) days within the study area followed by 6.9 ± 8.8 S.E. (95% c.i. 5.1–36.1) days outside of the surveyed area.

thumbnail
Fig 6. Probability of resighting an individual whale shark over time (LIR; mean +- S.D.) within the Western Central Atlantic compared to the best fitting model (dark line).

https://doi.org/10.1371/journal.pone.0180495.g006

thumbnail
Table 3. Model comparisons for lagged identification rate of whale sharks throughout the entire study area (A-H) and within/between areas (I-L).

https://doi.org/10.1371/journal.pone.0180495.t003

The LIR was also used to investigate movement within and between the four focal areas in this study, with the addition of a hypothetical “outside” location to account for movement into unmonitored areas. The fully mixed model (Model I; Table 3) best fit the empirical data and generated a population estimate of 2,167 (± 378) (95% c.i. 1695.91–3207.76) sharks. The LIR for resightings in the area in which the shark was originally identified, was much greater than the probability of the shark being resighted within a different area. Resighting probabilities showed the same rapid decline over the first year after sighting with a slight increase at close to a year (Fig 7). After two years, annual LIR showed a slight increase for a probability for resighting in a different area and a steady decrease for the same area. Animals originally observed in Mexico had a 90.8% probability of resighting at the same location; whereas in the other areas, the probability of resighting was approximately 50% (Table 4). In fact, the probability of an individual moving outside of the areas modeled was greater (55.3%) than the resighting probabilities for Honduras (43.4%), Belize (54.5%), and the US (49.2%). Movement between Honduras and Belize was 11% and 12% from these locations to the US. Besides being seen again in the same area, moving to an “outside” area had the highest transition probability for the US (26.8%), Honduras (16.4%), and Mexico (6%). Whale sharks coming from the “outside” have the highest probability of transition to Mexico (19.4%) or the US (11.2%).

thumbnail
Fig 7. Probability of resighting an individual whale shark over time (LIR; mean ± S.D.) within the same location (green circle) or a different location (red asterisk) in the Western Central Atlantic compared to the best fitting model (dark line).

https://doi.org/10.1371/journal.pone.0180495.g007

thumbnail
Table 4. Transition probabilities amongst aggregations sites in the Western Central Atlantic.

https://doi.org/10.1371/journal.pone.0180495.t004

Discussion

Whale sharks within the Atlantic Ocean appear to be functionally separate from other populations worldwide [14]. We provide the first WCA regional population estimate for this species of 2,167 (± 345) sharks using maximum likelihood models. Based on our findings, a significant proportion of the WCA whale shark population has been sampled with photo-identification. Although the discovery curve did not reach an asymptote, the resighting rate remained over 70% for the last five years of the study, which is typically indicative of high site fidelity and a relatively small population size [51]. Yet for whale sharks, we are reporting one of the largest population sizes in the literature. In the Indian Ocean, resightings rates were considerably lower than reported here (10–32%) and inversely correlated with population size [52]. The number of individuals identified in this study represented 63% of the modelled population estimate and thus reasonably support this conclusion. The results of this study demonstrate that whale sharks within the WCA represent an open population with significant emigration and re-immigration (incorporating permanent emigration), which is not surprising considering their large size and movement capabilities of this species [28, 32, 33, 40]. Previous studies have shown open population models are the most appropriate for population estimates of whale sharks from photo-identification data [41, 53] and the lack of asymptote in the discovery curve, as we observed, indicates that new animals are entering the region [28, 29]. Given the vast available habitat for whale sharks in the WCA, it is entirely plausible that additional aggregation sites occur within the area that have not been identified by researchers or tourism. Mexico and the U.S. had higher percentages of connectivity with the hypothetical “outside” area than Honduras or Belize, indicating that sharks from the broader region are more likely to be identified in the northern Caribbean and GOM than in the southern Mesoamerican Reef area.

Our modelled population estimate is comparable to previous site-specific studies inside [18, 32, 40] and outside [26, 29, 54] the WCA. For example, mark-recapture models at Holbox Island, Mexico predicted an overall population of 521–809 sharks with annual periodicity [40]. At Gladden Spit, Belize, 106 individuals were identified with a mean sightings rate of 4–6 sharks per day but mark-recapture population estimates could not be calculated [18]. In Utila, Honduras, daily estimates of 4.6 sharks were reported, but a total population estimate was not provided [32]. In comparison, the modelled whale shark population estimate (2,837 sharks ± 1,243.9) for the Arabian Gulf was similar to our study, although the proportion of the population sampled was far smaller (15%) during the four years of the study [29]. Smaller population estimates are reported for Ningaloo Reef in Western Australia (320–440 individuals) [54], and for the Seychelles (348–488 individuals) [17].

Population estimates generated from volunteer sightings data should be taken with a bit of caution for several reasons. First, the uneven distribution in effort throughout the range and the high resighting rate, particularly in Mexico, is indicative of sampling biases that can impact the overall estimate. Due to the anecdotal nature of the dataset, it is impossible to account for heterogeneity of capture probability. Whale shark photo-identification data collection is facilitated by the ecotourism industry, and therefore these data are inherently biased by the seasonality of tourism, weather, and distance from shore. Tourism biases are most likely evident in the Yucatan Peninsula, which supports the largest whale shark ecotourism industry in the world [55] and comprised the largest number of encounters in the present study. There is no doubt that the coast of Mexico supports one of the largest aggregations in the world; however, the tourism industry sponsors (almost daily survey) trips by dedicated researchers. Other areas cannot match this level of survey effort. While imagery provided by “citizen scientists” have been shown to be effective for use in mark-recapture modeling for population estimates [53], our study indicates that dedicated initiatives at each aggregation site are needed to spread awareness, encourage participation, and often collect/submit photos to the database.

Year-round monitoring, outside of peak aggregation times and tourism season, is likely confounded by whale sharks inhabiting waters at greater distances from shore and sea conditions that prevent underwater photography. Cagua et al. [56] demonstrated with acoustic telemetry that sightings data could miss year-round residency due to individuals utilizing subsurface waters, or moving further from shore. Honduras is the only country in the WCA to consistently report whale shark sightings for every month of the year; however, anecdotal accounts of year-round whale shark presence are prevalent in all four known aggregation sites in the WCA. Photo-identification data fails to recognize important habitat areas that do not have tourism or directed research effort. For example, satellite telemetry has shown the Bay of Campeche to be a high use area by whale sharks [13], yet no photo-captures were available from that area. Also, Belize is underrepresented in the current study (n = 47 sharks), as earlier work identified 106 individual sharks [18]; however, equipment failure resulted in the loss of digital images prior to the inclusion in the Wildbook system (R. Graham pers. comm.). It is therefore possible that the results reported herein are an underrepresentation of true population size. Despite these caveats, photo-identification has proven useful for providing the first regional population estimate with the best available data.

Adequate conservation measures rely on an understanding of the full reproductive stock, which is typically unavailable at individual aggregation sites. The vast majority of whale shark data available is derived from juvenile male-dominated aggregation sites, with few exceptions [1]. Even the sexually integrated population in the Red Sea does not represent a reproductive population because the animals are immature [28]. This is also true for the entire WCA, which is dominated by males and 89% of the sampled individuals were smaller than size of maturity estimates (7.0–8.0 m) [5, 40]. A recent study using stereo-video photogrammetry concluded visual size-estimates of whale sharks tends to underestimate larger individuals and overestimate smaller ones [57]. The ecology described here could change as more accurate methods of measuring marine megafauna become prevalent. Regardless, both the smallest and largest whale sharks tend to be absent from most aggregations, including those in the WCA. The scattered reports of neonatal and mature whale sharks in the Atlantic tend to be associated with mid-oceanic islands [13, 5862] which receive little research or tourism. Until those demographics can be more reliably surveyed, scientific understanding and management of the species will be biased towards aggregating sub-adults.

Connectivity among whale shark aggregations in the WCA has been well-demonstrated using satellite telemetry, conventional tags, and photo-identification [13, 18, 63]. Yet, movement studies of whale sharks are inherently limited by current methods. Deployment times for satellite tags rarely exceeds one year [6466], and sightings data is limited by visibility bias and constraints on survey effort [56]. In the present study, the repeated long-distance movements of at least 1,600 km (between Honduras/Belize and the US Gulf Coast), were documented in two individuals. Additionally, movements between Mexico and the northern Gulf of Mexico aggregation site, were documented thirteen times. Other studies have reported multinational connectivity, but usually on a smaller spatial scale [29, 67]. It is difficult to compare movements from photo-identification to telemetry studies because, like conventional tags, only the distance between two locations is known. However, this methodology is most useful for long temporal scales, where electronic tracking technology is limited. With the current technology, it is clear that international cooperation is needed for effective research and management of whale shark populations. Information gathering can only improve as new methods and technologies are developed and researchers at individual aggregation sites continue to work collaboratively.

Conclusion

We provided the first regional population estimate and provided further support for the high degree of connectivity among whale shark aggregations sites in four countries within the WCA over a 16-year period. Broad-scale movements within this region have been corroborated by the available short-term tagging data (conventional and satellite); however, multi-year periodicity of whale shark occurrence can only currently be elucidated using long-term photo-identification. Despite the inherent biases of opportunistic, observational data, photo-identification has proven to be a useful tool to determine demography, movement, site fidelity, and habitat use of whale sharks over large spatial and temporal scales. Whale shark ecotourism has bolstered the photo-identification process allowing both “citizen scientists” and researchers to contribute to a global database. This paper demonstrates the usefulness of multinational participation and collaboration with regard to research and conservation efforts for whale sharks in the WCA.

Acknowledgments

We would like to thank The UK Darwin Initiative, the Summit Foundation, the Comisión de Áreas Naturales Protegidas (CONANP), International Foundation for Animal Welfare (IFAW), On Wings of Care, Aqua-Firma, Shark Foundation, and Mission Blue for providing logistical support for this work. We are also indebted to all of the contributors to the whale shark photo-identification database and the tour operators from Isla Mujeres, Cancún, Holbox, Belize, and Utila. We wish to thank two anonymous reviewers whose comments helped to greatly improve this manuscript.

This research has made use of data and software tools provided by Wildbook for Whale Sharks (http://www.whaleshark.org), an online photo-identification mark-recapture database operated by the non-profit scientific organization Wild Me with funding derived from Qatar Whale Shark Research Project, Siren Fleet, and public donations.

The authors’ work on this study was supported by their affiliated organization. With the exception of SJP who was supported by Aqua-Firma, Shark Foundation and two private trusts.

References

  1. 1. Rowat D, Brooks KS. A review of the biology, fisheries and conservation of the whale shark Rhincodon typus. Journal of Fish Biology. 2012 Apr 1;80(5):1019–56. pmid:22497372
  2. 2. Smith AN. Descriptions of new, or imperfectly known objects of the animal kingdom, found in the south of Africa. South African Commercial Advertiser. 1828 Nov;3(2).
  3. 3. Wintner SP. Preliminary study of vertebral growth rings in the whale shark, Rhincodon typus, from the east coast of South Africa. Environmental Biology of Fishes. 2000 Dec 1;59(4):441–51.
  4. 4. Martin RA. A review of behavioural ecology of whale sharks (Rhincodon typus). Fisheries Research. 2007 Mar 31;84(1):10–6.
  5. 5. Norman BM, Stevens JD. Size and maturity status of the whale shark (Rhincodon typus) at Ningaloo Reef in Western Australia. Fisheries Research. 2007 Mar 31;84(1):81–6.
  6. 6. Hsu HH, Joung SJ, Liu KM. Fisheries, management and conservation of the whale shark Rhincodon typus in Taiwan. Journal of Fish Biology. 2012 Apr 1;80(5):1595–607. pmid:22497399
  7. 7. Pierce, S.J. & Norman, B. Rhincodon typus. The IUCN Red List of Threatened Species 2016: eT19488A2365291. http://dx.doi.org/10.2305/IUCN.UK.2016-1.RLTS.T19488A2365291.en. Downloaded on 15 September 2016.
  8. 8. Fromentin JM, Powers JE. Atlantic bluefin tuna: population dynamics, ecology, fisheries and management. Fish and Fisheries. 2005 Dec 1;6(4):281–306.
  9. 9. Castro AL, Stewart BS, Wilson SG, Hueter RE, Meekan MG, Motta PJ, Bowen BW, Karl SA. Population genetic structure of Earth's largest fish, the whale shark (Rhincodon typus). Molecular Ecology. 2007 Dec 1;16(24):5183–92. pmid:18092992
  10. 10. Schmidt JV, Schmidt CL, Ozer F, Ernst RE, Feldheim KA, Ashley MV, Levine M. Low genetic differentiation across three major ocean populations of the whale shark, Rhincodon typus. PloS One. 2009 Apr 7;4(4): e4988. pmid:19352489
  11. 11. Hearn AR, Green J, Román MH, Acuña-Marrero D, Espinoza E, Klimley AP. Adult female whale sharks make long-distance movements past Darwin Island (Galapagos, Ecuador) in the Eastern Tropical Pacific. Marine Biology. 2016 Oct 1;163(10):214.
  12. 12. Brunnschweiler JM, Baensch H, Pierce SJ, Sims DW. Deep-diving behavior of a whale shark Rhincodon typus during long-distance movement in the western Indian Ocean. Journal of Fish Biology. 2009 74: 706–714. pmid:20735591
  13. 13. Hueter RE, Tyminski JP, de la Parra R. Horizontal movements, migration patterns, and population structure of whale sharks in the Gulf of Mexico and northwestern Caribbean Sea. PloS One. 2013 8(8), e71883. pmid:23991000
  14. 14. Vignaud TM, Maynard JA, Leblois R, Meekan MG, Vázquez‐Juárez R, Ramírez‐Macías D, Pierce SJ, Rowat D, Berumen ML, Beeravolu C, Baksay S. Genetic structure of populations of whale sharks among ocean basins and evidence for their historic rise and recent decline. Molecular Ecology. 2014 May 1;23(10):2590–601. pmid:24750370
  15. 15. Wilson SG, Polovina JJ, Stewart BS, Meekan MG. Movements of whale sharks (Rhincodon typus) tagged at Ningaloo Reef, Western Australia. Marine Biology. 2006 Mar 1;148(5):1157–66.
  16. 16. McFarlane GA, Wydoski RS, Prince ED. External tags and marks. In Historical review of the development of external tags and marks. American Fisheries Society Symposium. 1990 (Vol. 7, pp. 9–29).
  17. 17. Rowat D, Speed CW, Meekan MG, Gore MA, Bradshaw CJ. Population abundance and apparent survival of the Vulnerable whale shark Rhincodon typus in the Seychelles aggregation. Oryx. 2009 Oct 1;43(04):591–8.
  18. 18. Graham RT, Roberts CM. Assessing the size, growth rate and structure of a seasonal population of whale sharks (Rhincodon typus Smith 1828) using conventional tagging and photo identification. Fisheries Research. 2007 Mar 31;84(1):71–80.
  19. 19. Speed CW, Meekan MG, Bradshaw CJ. Spot the match–wildlife photo-identification using information theory. Frontiers in Zoology. 2007 Jan 16;4(1):2.
  20. 20. Myrberg AA Jr, Gruber SH. The behavior of the bonnethead shark, Sphyrna tiburo. Copeia. 1974 Jun 13:358–74.
  21. 21. Van Tienhoven AM, Den Hartog JE, Reijns RA, Peddemors VM. A computer‐aided program for pattern‐matching of natural marks on the spotted raggedtooth shark Carcharias taurus. Journal of Applied Ecology. 2007 Apr 1;44(2):273–80.
  22. 22. Castro AL, Rosa RS. Use of natural marks on population estimates of the nurse shark, Ginglymostoma cirratum, at Atol das Rocas Biological Reserve, Brazil. Environmental Biology of Fishes. 2005 Feb 1;72(2):213–21.
  23. 23. Anderson SD, Goldman KJ. Photographic evidence of white shark movements in California waters. California Fish and Game. 1996 Sep 1;82(4):182–6.
  24. 24. Taylor G. Whale sharks of Ningaloo Reef, Western Australia: a preliminary study. Western Australian Naturalist. 1989;18(1):7–12.
  25. 25. Arzoumanian Z, Holmberg J, Norman B. An astronomical pattern‐matching algorithm for computer‐aided identification of whale sharks Rhincodon typus. Journal of Applied Ecology. 2005 Dec 1;42(6):999–1011.
  26. 26. Holmberg J, Norman B, Arzoumanian Z. Robust, comparable population metrics through collaborative photo‐monitoring of whale sharks Rhincodon typus. Ecological Applications. 2008 Jan 1;18(1):222–33. pmid:18372568
  27. 27. Brooks K, Rowat D, Pierce SJ, Jouannet D, Vely M. Seeing spots: photo-identification as a regional tool for whale shark identification. Western Indian Ocean Journal of Marine Science. 2010;9(2):185–94.
  28. 28. Cochran JE, Hardenstine RS, Braun CD, Skomal GB, Thorrold SR, Xu K, Genton MG, Berumen ML. Population structure of a whale shark Rhincodon typus aggregation in the Red Sea. Journal of Fish Biology. 2016 Sep 1;89(3):1570–82. pmid:27401632
  29. 29. Robinson DP, Jaidah MY, Bach S, Lee K, Jabado RW, Rohner CA, March A, Caprodossi S, Henderson AC, Mair JM, Ormond R. Population Structure, Abundance and Movement of Whale Sharks in the Arabian Gulf and the Gulf of Oman. PloS One. 2016 Jun 30;11(6):e0158593. pmid:27362839
  30. 30. Würsig B, Würsig M. The photographic determination of group size, composition, and stability of coastal porpoises (Tursiops truncatus). Science. 1977 Nov 18;198(4318):755–6.
  31. 31. Whitehead H. SOCPROG programs: analysing animal social structures. Behavioral Ecology and Sociobiology. 2009 Mar 1;63(5):765–78.
  32. 32. Fox S, Foisy I, De La Parra Venegas R, Galván Pastoriza BE, Graham RT, Hoffmayer ER, Holmberg J, Pierce SJ. Population structure and residency of whale sharks Rhincodon typus at Utila, Bay Islands, Honduras. Journal of Fish Biology. 2013 Sep 1;83(3):574–87. pmid:23991875
  33. 33. Araujo G, Snow S, So CL, Labaja J, Murray R, Colucci A, Ponzo A. Population structure, residency patterns and movements of whale sharks in Southern Leyte, Philippines: results from dedicated photo‐ID and citizen science. Aquatic Conservation: Marine and Freshwater Ecosystems. 2016 Jan 1
  34. 34. Graham RT. Whale sharks of the western Caribbean: an overview of current research and conservation efforts and future needs for effective management of the species. Gulf and Caribbean Research. 2007;19(2):149–59.
  35. 35. Hoffmayer ER, Franks JS, Driggers III WB, Oswald KJ, Quattro JM. Observations of a feeding aggregation of whale sharks, Rhincodon typus, in the north central Gulf of Mexico. Gulf and Caribbean Research. 2007;19(2):69–73.
  36. 36. de la Parra Venegas R, Hueter R, Cano JG, Tyminski J, Remolina JG, Maslanka M, Ormos A, Weigt L, Carlson B, Dove A. An unprecedented aggregation of whale sharks, Rhincodon typus, in Mexican coastal waters of the Caribbean Sea. PloS One. 2011 Apr 29;6(4):e18994. pmid:21559508
  37. 37. Cárdenas‐Palomo N, Herrera‐Silveira J, Velázquez‐Abunader I, Reyes O, Ordonez U. Distribution and feeding habitat characterization of whale sharks Rhincodon typus in a protected area in the north Caribbean Sea. Journal of Fish Biology. 2015 Feb 1;86(2):668–86.
  38. 38. Hacohen-Domené A, Martínez-Rincón RO, Galván-Magaña F, Cárdenas-Palomo N, de la Parra-Venegas R, Galván-Pastoriza B, Dove AD. Habitat suitability and environmental factors affecting whale shark (Rhincodon typus) aggregations in the Mexican Caribbean. Environmental Biology of Fishes. 2015 Aug 1;98(8):1953–64.
  39. 39. Hammerschlag N, Gallagher AJ, Wester J, Luo J, Ault JS. Don’t bite the hand that feeds: assessing ecological impacts of provisioning ecotourism on an apex marine predator. Functional Ecology. 2012 Jun 1;26(3):567–76.
  40. 40. Ramírez‐Macías D, Meekan M, La Parra‐Venegas D, Remolina‐Suárez F, Trigo‐Mendoza M, Vázquez‐Juárez R. Patterns in composition, abundance and scarring of whale sharks Rhincodon typus near Holbox Island, Mexico. Journal of Fish Biology. 2012 Apr 1;80(5):1401–16. pmid:22497390
  41. 41. Holmberg J, Norman B, Arzoumanian Z. Estimating population size, structure, and residency time for whale sharks Rhincodon typus through collaborative photo-identification. Endangered Species Research. 2009 Apr 8;7(1):39–53.
  42. 42. Murphy SJ, Hurlburt HE, O'Brien JJ. The connectivity of eddy variability in the Caribbean Sea, the Gulf of Mexico, and the Atlantic Ocean. Journal of Geophysical Research: Oceans. 1999 Jan 15;104(C1):1431–53.
  43. 43. Johnson DR, Perry HM, Graham WM. Using nowcast model currents to explore transport of non-indigenous jellyfish into the Gulf of Mexico. Marine Ecology Progress Series. 2005 Dec 23;305:139–46.
  44. 44. Cowen RK, Paris CB, Srinivasan A. Scaling of connectivity in marine populations. Science. 2006 Jan 27;311(5760):522–7. pmid:16357224
  45. 45. Johnston MW, Purkis SJ. Spatial analysis of the invasion of lionfish in the western Atlantic and Caribbean. Marine Pollution Bulletin. 2011 Jun 30;62(6):1218–26. pmid:21511308
  46. 46. Hines A, Acero P, Ortí G, Wilbur AE, Freshwater DW. Reconstructing the lionfish invasion: insights into Greater Caribbean biogeography. Journal of Biogeography. 2011 Jul 1;38(7):1281–93.
  47. 47. Kraus RT, Wells RD, Rooker JR. Horizontal movements of Atlantic blue marlin (Makaira nigricans) in the Gulf of Mexico. Marine Biology. 2011 Mar 1;158(3):699–713.
  48. 48. Baker JD, Harting AL, Johanos TC. Use of discovery curves to assess abundance of Hawaiian monk seals. Marine Mammal Science. 2006 Oct 1;22(4):847–61.
  49. 49. Zar JH. Biostatistical analysis 4th edn. New Jersey. 1999.
  50. 50. Whitehead H. Selection of models of lagged identification rates and lagged association rates using AIC and QAIC. Communications in Statistics—Simulation and Computation. 2007 Nov 5;36(6):1233–46.
  51. 51. Baird RW, Webster DL, Mahaffy SD, McSweeney DJ, Schorr GS, Ligon AD. Site fidelity and association patterns in a deep‐water dolphin: Rough‐toothed dolphins (Steno bredanensis) in the Hawaiian Archipelago. Marine Mammal Science. 2008 Jul 1;24(3):535–53.
  52. 52. Andrzejaczek S, Meeuwig J, Rowat D, Pierce S, Davies T, Fisher R, Meekan M. The ecological connectivity of whale shark aggregations in the Indian Ocean: a photo-identification approach. Royal Society Open Science. 2016 Nov 1;3(11):160455. pmid:28018629
  53. 53. Davies TK, Stevens G, Meekan MG, Struve J, Rowcliffe JM. Can citizen science monitor whale-shark aggregations? Investigating bias in mark–recapture modelling using identification photographs sourced from the public. Wildlife Research. 2013 Jan 11;39(8):696–704.
  54. 54. Meekan M, Bradshaw C, Press M, McLean C, Richards A, Quasnichka S, Taylor J. Population size and structure of whale sharks (Rhincodon typus) at Ningaloo Reef Western Australia. Marine Ecology Progress Series. 2006 Aug 18; 319: 275–285.
  55. 55. Gallagher AJ, Hammerschlag N. Global shark currency: the distribution, frequency, and economic value of shark ecotourism. Current Issues in Tourism. 2011 Nov 1;14(8):797–812.
  56. 56. Cagua EF, Cochran JE, Rohner CA, Prebble CE, Sinclair-Taylor TH, Pierce SJ, Berumen ML. Acoustic telemetry reveals cryptic residency of whale sharks. Biology Letters. 2015 Apr 1;11(4):20150092. pmid:25832816
  57. 57. Sequeira AM, Thums M, Brooks K, Meekan MG. Error and bias in size estimates of whale sharks: implications for understanding demography. Royal Society Open Science. 2016 Mar 1;3(3):150668. pmid:27069656
  58. 58. Macena BC, Hazin FH. Whale Shark (Rhincodon typus) Seasonal Occurrence, Abundance and Demographic Structure in the Mid-Equatorial Atlantic Ocean. PloS One. 2016 Oct 26;11(10):e0164440. pmid:27783634
  59. 59. Hazin FH, Vaske Júnior T, Oliveira PG, Macena BC, Carvalho F. Occurrences of whale shark (Rhincodon typus Smith, 1828) in the Saint Peter and Saint Paul archipelago, Brazil. Brazilian Journal of Biology. 2008 May;68(2):385–9.
  60. 60. Ketchum JT, Galván-Magaña F, Klimley AP. Segregation and foraging ecology of whale sharks, Rhincodon typus, in the southwestern Gulf of California. Environmental Biology of Fishes. 2013 Jun 1;96(6):779–95.
  61. 61. Acuña-Marrero D, Jiménez J, Smith F, Doherty PF Jr, Hearn A, Green JR, Paredes-Jarrín J, Salinas-de-León P. Whale shark (Rhincodon typus) seasonal presence, residence time and habitat use at Darwin Island, Galapagos Marine Reserve. PLoS One. 2014 Dec 31;9(12):e115946. pmid:25551553
  62. 62. Clingham E, Brown J, Henry L, Beard A, Dove AD. Evidence that St. Helena island is an important multi-use habitat for whale sharks, Rhincodon typus, with the first description of putative mating in this species. PeerJ Preprints; 2016 Mar 21.
  63. 63. Gifford A, Compagno LJ, Levine M, Antoniou A. Satellite tracking of whale sharks using tethered tags. Fisheries Research. 2007 Mar 31;84(1):17–24.
  64. 64. Hammerschlag N, Gallagher AJ, Lazarre DM. A review of shark satellite tagging studies. Journal of Experimental Marine Biology and Ecology. 2011 Feb 28;398(1):1–8.
  65. 65. Lea JS, Wetherbee BM, Queiroz N, Burnie N, Aming C, Sousa LL, Mucientes GR, Humphries NE, Harvey GM, Sims DW, Shivji MS. Repeated, long-distance migrations by a philopatric predator targeting highly contrasting ecosystems. Scientific Reports. 2015 Jun 9;5:11202. pmid:26057337
  66. 66. Norman BM, Reynolds S, Morgan DL. Does the whale shark aggregate along the Western Australian coastline beyond Ningaloo Reef? Pacific Conservation Biology. 2016 Apr 22;22(1):72–80.
  67. 67. Berumen ML, Braun CD, Cochran JE, Skomal GB, Thorrold SR. Movement patterns of juvenile whale sharks tagged at an aggregation site in the Red Sea. PLoS One. 2014 Jul 30;9(7):e103536. pmid:25076407