Based on counts, modelled population estimates and habitat availability, it is inferred that approximately 75% of the global Whale Shark population occurs in the Indo-Pacific, and 25% in the Atlantic. In the Indo-Pacific, a population reduction of 63% is inferred over the last three generations (75 years), and in the Atlantic a population reduction of more than 30% is inferred (see the discussion for each subpopulation below). Combining data from both regions, it is likely that the global Whale Shark population has declined by >50% over the last 75 years.

Available datasets for Whale Sharks, in particular those from coastal sites, are typically dominated by juvenile male sharks. Acknowledging this, we make the assumption that trend data in the visible fraction of the population (i.e. mainly juveniles) is representative of the adult population.

Global population size
Whale Sharks are individually identifiable based on their characteristic spot patterns (Taylor 1994, Arzoumanian et al. 2005, Marshall and Pierce 2012). A global database of Whale Shark sightings, comprising submitted photographs from both researchers and the public, is hosted online at Wildbook for Whale Sharks (www.whaleshark.org) (Wild Me 2016, Norman et al. submitted). As of February 2016, there were 7,011 individual sharks on this database, identified from images submitted between 1964 and 2016. However, 70% of sexed individuals (n = 3,420) were male, with the majority of these likely to be immature based on length estimates (Norman and Stevens 2007, Ramírez-Macías et al. 2012, Rohner et al. 2015). This dataset is assumed to not fully represent female, small juvenile or adult sharks (Norman et al. submitted). Therefore, the total represents a minimum number of sharks alive over this period.

Two global-scale genetic studies on Whale Sharks have estimated genetic effective population size – the number of breeding adults – albeit based on small sample sizes of 70 (Castro et al. 2007) and 68 sharks (Schmidt et al. 2009), respectively. Castro et al. (2007) used mitochondrial DNA to estimate current genetic effective population size to be 119,000–238,000 sharks. Schmidt et al. (2009) estimated genetic effective population size to be 103,572, with a standard error of 27,401–179,794, based on microsatellite analysis. However, lack of knowledge on species-specific mutation rates mean these estimates should not be used for management purposes (J. Schmidt pers. comm., T. Vignaud and S. Planes pers. comm).

Global population structure
While global connectivity of Whale Shark subpopulations has been postulated (Sequeira et al. 2013b), a large global genetic study using both mitochondrial DNA and microsatellite analysis has demonstrated that the Whale Shark subpopulations in the Atlantic and Indo-Pacific are functionally separate (Vignaud et al. 2014). Large-scale photo-identification studies in the Indian Ocean (Brooks et al. 2010) and western Atlantic (Hueter et al. 2013), along with Wildbook online database results (Norman et al. submitted), have shown limited connectivity exists between non-contiguous countries. Based on these results, here we have considered two separate subpopulations in this assessment.

Atlantic subpopulation trend
A decline of ≥30% in the Atlantic subpopulation is inferred over the last three generations (75 years) based on data from tuna fleet observers off a likely centre of abundance for this subpopulation. Between 1980 and 2010 there was a decline in sightings per unit effort (SPUE) off western Africa, with SPUE peaking in 1995 and declining thereafter (Sequeira et al. 2014; Table 1 in the supplementary material). In absolute terms, sightings decreased from about 500 during the 1990s to around 150 during the 2000s. Peak-month sightings also declined by approximately 50% over this time (Sequeira et al. 2014).

At Gladden Spit in Belize, Whale Shark sightings declined from a mean of 4 to 6 sharks per day between 1998 and 2001 to less than 2 per day in 2003 (Graham and Roberts 2007), with reports from diving guides indicating that numbers have remained low until 2016 (R. Graham, pers. comm.).

In the Azores, there was a significant increase in sightings in 2008 and afterwards compared to the decade before (Afonso et al. 2014; Table 1 in the supplementary material). This was strongly correlated with the location of the 22°C isotherm, indicating that this increasing sighting trend is due to environmental conditions (Afonso et al. 2014).

Atlantic subpopulation size
Regional counts of identified sharks or modelled abundance estimates are available from many of the larger known aggregation or feeding areas. Ramírez-Macías et al. (2012) photo-identified 350 individual Whale Sharks from Holbox Island in Mexico between 2005 and 2008, and estimated that 521–809 sharks participate in this aggregation. Aerial surveys from this area and the adjacent Caribbean coast have counted up to 420 sharks in a single aerial survey (de la Parra Venegas et al. 2011). The largest-known aggregation as of February 2016 occurs seasonally off the Yucatan cost of Mexico, with over 1,100 identified sharks (Norman et al. submitted). Satellite-tagged sharks from this aggregation have been tracked to the northern Gulf of Mexico (Hueter et al. 2013), where aggregations of up to 100 sharks have been reported (Hoffmayer et al. 2005), south to Belize where 106 individual sharks were identified between 1998 and 2003 (Graham and Roberts 2007), and off the island of Utila, Honduras, where 95 sharks were identified between 1999 and 2011 (Fox et al. 2013). One shark, tagged in 2007, was tracked swimming across the equator to the South Atlantic Ocean, near the Mid-Atlantic Ridge (Hueter et al. 2013). The end point of this track was 543 km southeast from the Saint Peter and Saint Paul Archipelago, where 54 Whale Shark sightings were recorded between 2000 and 2005 (Hazin et al. 2008). This individual was subsequently photo-identified back off the Yucatan coast in 2011 and 2012 (Hueter et al. 2013).

Aside from St Helena Island (A. Dove, pers. comm.), there have been few photographic records from elsewhere in the Atlantic (Wild Me 2016). There were, however, 2,297 records of Whale Sharks from the logbooks of tuna purse-seine vessels between 1980 and 2010, mostly from the eastern Atlantic (Sequeira et al. 2014) and particularly off the coast of Gabon (Capietto et al. 2014). An additional 1,449 sightings were recorded from the Azores archipelago between 1998 and 2013 (Afonso et al. 2014).

Indo-Pacific subpopulation trend
A decline of >50% (likely around 63%) in the Indo-Pacific Whale Shark subpopulation is inferred over the last three generations (75 years) based on relevant indices of abundance from Mozambique, the broader Western Indian Ocean, the Philippines, Taiwan, Thailand and the Western and Central Pacific, and actual levels of exploitation in mainland China, the Maldives, India, the Philippines and Taiwan.

A commercial fishery for Whale Shark existed in Taiwan until 2007 (Hsu et al. 2012). Information provided by fishers operating from Hongchun harbour in southern Taiwan indicated that 50–60 sharks were caught each season in the mid-1980s, declining to less than 10 per year in each of 1994 and 1995 (Chen and Phipps 2002; Table 1 in the supplementary material). Although definitive catch trends are not available, there was a significant (58%) decrease in the estimated annual catch in 1997 of 272 sharks (Chen and Phipps 2002) to a reported catch of 113 sharks over 15 months in 2001–2002 (Chen and Phipps 2002; Table 1 in the supplementary material). A decline in the mean total length of landed sharks was noted between 2002 and 2007 (Hsu et al. 2012; Table 1 in the supplementary material). A decline in the mean size of landed sharks was also noted in southern Chinese waters, from 8.27 m prior to 2004 to 6.3 m in 2008–2011 (Li et al. 2012; Table 1 in the supplementary material).

Data from observers aboard the tuna purse-seine fleet in the Western and Central Pacific noted 1,073 Whale Shark sightings, with most from the Bismark and Solomon Seas (Harley et al. 2013). The occurrence of Whale Shark in free school sets decreased by approximately 50% between 2003 (1%) and 2012 (0.5%), potentially representing a fall in abundance (Harley et al. 2013; Table 1 in the supplementary material), although a weak linear increase in occurrence probability was modelled by Sequeira et al. (2014) between 2000 and 2010 (Table 1 in the supplementary material). However, model performance for this latter dataset was poor (Sequeira et al. 2014).

The influence of environmental variability on sightings of Whale Shark can complicate the interpretation of trend data, particularly with the shortage of long-term datasets. A decline in Whale Shark sightings along the eastern coast of South Africa between 1993 and 1998 (Gifford 2001) was included in the proposal for listing the sharks on Appendix II of the Convention on International Trade in Endangered Species (CITES 2002). However, with the benefit of hindsight, the substantial variability in sighting data from this area over summer months (Cliff et al. 2007) and seasonal changes in their oceanic distribution (Sequeira et al. 2012), makes it unclear whether these earlier data are indicative of a true subpopulation decline. North of this area, in the northern Mozambique Channel and broader western Indian Ocean, a slight increase in Whale Shark sightings was noted between 1991 and 2000 based on tuna purse-seine vessel data, then a decrease from 2000 to 2007 (Sequeira et al. 2013a; Table 1 in the supplementary material). In absolute terms, 600 sightings were reported from the 1990s decreasing to ~200 across 2000–2007 (Sequeira et al. 2014). Peak monthly sightings decreased by around 50% over the study period (Sequeira et al. 2014). In Inhambane, Mozambique, in the southern Mozambique Channel, sightings declined 79% between 2005 and 2011 (Rohner et al. 2013; Table 1 in the supplementary material). This decreased rate of sightings has persisted to at least January 2016 (S. Pierce, unpubl. data). Sharks routinely move between South Africa and southern Mozambique (Rohner et al. 2015, Norman et al. submitted), but there is limited or no connectivity between these and other known coastal feeding areas to the north in Djibouti, Seychelles and Tanzania (Brooks et al. 2010, Norman et al. submitted). Prior to the species being protected in the Maldives in 1995, catches of Whale Shark declined from around 30 each year from one of the significant fishing locations up until the early 1980s to a catch of 20 or less Whale Sharks from the entire archipelago by 1993 (Anderson and Ahmed 1993; Table 1 in the supplementary material).

Two hundred and fifty-three Whale Shark sightings were recorded by a local dive charter company in the Andaman Sea, Thailand, between 1991 and 2001 (Theberge and Dearden 2006). Sightings per unit effort showed a significant decline over this period, with an overall decrease from 1.58 Whale Sharks per trip in 1992–1993 to 0.13 sharks per trip in 2000–2001 (Theberge and Dearden 2006; Table 1 in the supplementary material). A low absolute number of sightings persisted until at least the 2002–2003 season, although effort data were not recorded (Theberge and Dearden 2006). Following the conclusion of data collection for that study, shark sightings have likely increased in frequency according to reports from dive operators. However, sharks are perceived to be smaller than those sighted in the 1990s (P. Dearden, pers. comm).

Bradshaw et al. (2008) analysed tourism sightings at Ningaloo Reef, Australia, between 1995 and 2004, corrected for search effort and environmental fluctuation, and identified a 40% decline in sighting rate and a decline in mean shark length of 1.6 m over this time period (Table 1 in the supplementary material), although seasonal shifts in peak abundance to outside observation months may also have contributed to this observed decline (Mau and Wilson 2007, Holmberg et al. 2009). Analysis of individual sighting data from 1995 to 2006 identified behavioural heterogeneity in the sharks at Ningaloo, with the majority showing some site fidelity in that they were sighted in multiple seasons (Holmberg et al. 2008). This (majority) subset of the study population was increasing over the course of this work (Holmberg et al. 2008). Follow-up analyses on a slightly longer dataset (1995 to 2008) confirmed this result, with smaller individuals contributing in larger numbers to recruitment, thereby potentially explaining the observed decline in mean size (Holmberg et al. 2009; Table 1 in the supplementary material). However, a genetic study on Ningaloo sharks indicated declining genetic diversity over five consecutive years for mtDNA (2007–2012) and two (2010–2012) for microsatellites (Vignaud et al. 2014; Table 1 in the supplementary material).

Whale Sharks were fished in the Philippines prior to protection in 1998, with Whale Shark catch per unit effort (i.e., per boat) declining from 4.44 to 1.7 in Pamilacan and 10 to 3.8 in Guiwanon between two surveys conducted in 1993 and 1997 (Alava et al. 2002; Table 1 in the supplementary material).

Indo-Pacific subpopulation size
Six hundred and seventy sharks were photo-identified in southern Mozambique between 2003 and 2014 (Norman et al. submitted). Off Mafia Island in Tanzania, 131 sharks were photo-identified between 2006 and 2014 (Norman et al. submitted). Off Djibouti, 297 individuals were photo-identified between 2003 and 2010 (Rowat et al. 2011), while 443 individuals were photo-identified off Mahe in the Seychelles between 2001 and 2009 (Rowat et al. 2011). A subset of the Seychelles dataset, comprising records from 2004 to 2009, was modelled to produce an abundance estimate of 469 to 557 sharks from that area (Brooks et al. 2010). Comparisons of identities collected from the Seychelles, Djibouti, Tanzania and Mozambique, comprising a total of 1,069 individual sharks sighted up to 2009, found no matches between these countries (Brooks et al. 2010). Norman et al. (submitted) updated these figures and found three international matches (all between Mozambique and Tanzania) from 1,098 individuals. A significant aggregation of juvenile Whale Sharks also occurs in the Red Sea off the Saudi Arabian coast, with 122 individuals identified to 2013 (J. Cochran, pers. comm). Aggregations of up to 100 sharks have been noted off the Qatar coast (Robinson et al. 2013). One hundred and six sharks have been identified around South Ari Atoll in the Maldives (Riley et al. 2010, Davies et al. 2012), with a modelled abundance estimate of around 74 to 104 sharks (Davies et al. 2012). There were 1,184 Whale Shark sightings in the Indian Ocean recorded by tuna purse-seine vessel logbooks between 1991 and 2007, with almost all of these from the Western Indian Ocean (Sequeira et al. 2013a).

Eighty-two sightings were recorded off Christmas Island, Australia, between 1996 and 2008 (Hobbs et al. 2009). Whale shark photo-identifications range back to the early 1990s off Ningaloo Reef in Western Australia (Meekan et al. 2006; Holmberg et al. 2008, 2009). Three hundred and eighty-six sharks were identified here between 1995 and 2008 (Holmberg et al. 2009), with over 1,000 sharks identified by 2015 (Norman et al. submitted). Although two satellite-tagged sharks were tracked moving from Ningaloo Reef to the vicinity of Christmas Island (Sleeman et al. 2010), no exchange of photo-identified sharks has been documented to date (Wild Me 2016). One photo-identified Whale Shark was, however, recorded from both Ningaloo Reef and Borneo (B. Norman, unpubl. data).

Whale Shark abundance at Ningaloo Reef, Australia has been modelled by two studies. Meekan et al. (2006) estimated super-population size to be 319 to 436 sharks between 1992 and 2004. Holmberg et al. (2009) estimated annual abundance to vary between 86 and 143 sharks in the years 2004 to 2007, when length was used as a covariate. Whale Shark abundance in this area is correlated with the Southern Oscillation Index and several other oceanographic variables, which potentially relate to the strength of ocean currents and local productivity (Wilson et al. 2001, Sleeman et al. 2010).

Over 820 individual sharks have now been photo-identified from the Philippines (Schleimer et al. 2015), with large aggregations recorded from around Donsol (Quiros 2007), Oslob on Cebu Island (Araujo et al. 2014, Schleimer et al. 2015), and Pintuyan on Southern Leyte (Schleimer et al. 2015). Excluding the Philippines, 326 sharks have been photo-identified from Southeast Asia (here including Cambodia [3], Indonesia [64], Malaysia [32], Myanmar [4], Taiwan [7] and both coasts of Thailand [216]) and added to the global database by February 2016 (Wild Me 2016).

In the Eastern Pacific, Whale Sharks are seasonally present in small numbers around Darwin and Wolf islands in the north of the Galapagos Archipelago (Hearn et al. 2013). Acuña-Marrero et al. (2014) identified 82 individuals here from 2011–2013 and estimated an annual study population size of 695±166 individuals for the Galapagos Islands. Males and immature females are rarely observed at this location, so this estimate refers predominantly to large, apparently pregnant females (Acuña-Marrero et al. 2014). Two hundred and fifty-one individual sharks were photo-identified in the Gulf of California, Mexico, between 2003 and 2009: 129 from Bahía de Los Ángeles, 125 from Bahía de La Paz and smaller numbers from other sites (Ramírez-Macías et al. 2012). Modelled abundance estimates from this dataset were 54 in 2008 and 94 sharks in 2009 from Bahía de Los Ángeles, and between 19 and 62 through 2005–2009 at Bahía de La Paz (Ramírez-Macías et al. 2012). Outside of these areas (Galapagos Islands and Mexico), few photo-identified sharks have been added to the global database from the Eastern Pacific (Wild Me 2016). Twenty-five whale sharks were reported from tuna purse-seine vessel observers in 2014 (Clarke 2015).