Overall, we estimated that 8 of 12 elasmobranch species observed at Cocos Island declined significantly from 1993 to 2013. Six of these were declines in relative abundance, while the remaining 2were declines in probability of occurrence. The 4 remaining species increased in the odds of their presence and were among the larger bodied sharks at Cocos Island.
Large citizen-science collected data sets require careful scrutiny to ensure quality and consistency among observers.
Results based on the effects of divemaster-recorded environmental variables should be interpreted cautiously because these variables were not always standardized, as would be the case in a scientific survey. However, our analysis of this data set showed that individual divemasters had little influence on the number of sharks observed, and our parameter estimates for time trends were robust when using only a subset of the divemasters (Supporting Information). These results are in accordance with previous research indicating the effectiveness of using diver-collected data to assess trends in marinemegafauna (Ward-Paige and Lotze 2011; Vianna et al. 2014).
We hypothesized that large-bodied wide-ranging pelagic sharks and planktivores would experience declines, primarily as a result of overfishing.
The temporal trends for 4 of the 6 species within this category, including the iconic scalloped hammerhead, were in accordance with this hypothesis.
The scalloped hammerhead is considered endangered within the eastern tropical Pacific, where it is caught as bycatch in at least Mexico, Costa Rica, and Ecuador (Baum et al. 2007; Kyne et al. 2012). Scalloped hammerhead sharks are known to move among the major offshore islands in the region: Cocos, Galapagos, and Malpelo (Bessudo et al. 2011). Although each of these islands is designated as an MPA, scalloped hammerheads are still caught both illegally within these protected areas and legally outside them (Kyne et al. 2012). Thus, substantial declines in this species are not surprising. In addition to water temperature and seasonality (Ketchum et al. 2014), our models also revealed the importance of El Niño activity in driving the relative abundance of scalloped hammerheads at Cocos Island. During El Niños, scalloped hammerheads are thought to shift their distribution, either into deeper waters (Bessudo et al. 2011) or away from the equator (Lea and Rosenblatt 2000).
Silky shark, the other large pelagic shark that declined significantly, is the most commonly caught shark species in the eastern Pacific’s tuna purse seine fisheries (Watson et al. 2009).
Although silky sharks are listed as near threatened globally, they are considered vulnerable in the eastern tropical Pacific because of directed fishing for their fins and bycatch (Watson et al. 2009; IUCN 2014). From 1994 to 2004, capture rates of silky sharks as bycatch in purse seine fisheries in this region are estimated to have fallen by 50% (Minami et al. 2007). Although we examined silky shark presence instead of counts, our results indicate a similar dramatic decline.
Worldwide, mobula and manta rays are threatened by overfishing (Ward-Paige et al. 2013; IUCN 2014).
The population status of these species has been uncertain in the eastern tropical Pacific, but our results indicate dramatic declines in relative abundance of 78% and 89%, respectively. These declines likely stem from the multination fisheries in the eastern tropical Pacific because both tend to have a large home range and low rebound potential (Dulvy et al. 2014b).
Contrary to our initial prediction, tiger sharks showed significant increases in their odds of occurrence over time, arising from the abrupt increase observed since 2007 (Fig. 4b). It is possible that within this system of strong fishing pressure, tiger sharks have an advantage over other elasmobranch species because of their relatively high intrinsic rate of increase (Hutchings et al. 2012) and high post-hooking survival rate (Gallagher et al. 2014). Tiger shark population increases have been documented recently in the northwestern Atlantic (Baum and Blanchard 2010) and South Africa (Dudley and Simpfendorfer 2006). In the latter case these increases were attributed to competitive release. However, the abrupt increase in tiger shark observations at Cocos Island beginning in 2007 suggests that tiger sharks have simply moved to Cocos Island and established long-term residency there. Even though tiger sharks are a pelagic species capable of long migrations, recent evidence suggests that some individuals may display year-round residency at isolated reefs (Werry et al. 2014). The estimated increase should thus be interpreted cautiously because it may better reflect tiger shark movement than population trends.
Also contrary to our initial hypothesis, we observed a slight increase in the odds of occurrence for whale sharks at Cocos Island.
There is, however, large interannual variability for this species; its odds of occurrence at Cocos Island appeared to spike every 3 years (Fig. 4j). This suggests that Cocos may be a stopover for whale sharks moving to feeding or mating grounds (Hearn et al. 2013). Our results are in contrast to documented whale shark declines elsewhere in the world, which have resulted primarily from overfishing (IUCN 2014). Although whale sharks are protected under several international agreements, this species has continued to decline in many places (IUCN 2014).
We had expected that smaller sharks (whitetip reef) and bottom-feeding rays (eagle and marble rays) would experience increases in their relative abundance because of mesopredator release, but all 3 species instead declined greatly in relative abundance. This is likely due to a combination of other predators (Galapagos, tiger, and blacktip) increasing in presence, thereby changing species interactions, and illegal fishing activity within the Cocos Island MPA (Baskett et al. 2007; Arias et al. 2014).
We initially hypothesized that reef-associated sharks (blacktip, Galapagos, and silvertip), because of their high site fidelity, would be better protected by the Cocos Island MPA.
Our results are consistent with this hypothesis for both blacktip and Galapagos sharks, but silvertip sharks declined over time. Silvertip sharks may be in direct competition with blacktip and Galapagos sharks, which may explain why the latter 2 species increased at the same time as the recent severe declines in silvertip sharks occurred (Figs. 4e, f, and g). Additionally, increases in the presence of blacktip and Galapagos sharks could be due to the Cocos Island MPA working effectively for these largely reef-restricted species.
Despite substantial declines in 8 shark and ray species we documented, Cocos Island continues to be hailed as an example of a successful MPA and aworld class location for viewing large numbers of elasmobranchs (Friedlander et al. 2012; Edgar et al. 2014). This suggests a problem of shifting baselines, with recreational divers failing to recognize how much of the megafauna at Cocos Island already has been lost. Moreover, while many divers are excited by the increasing number of some larger elasmobranch species (i.e., tiger, blacktip, Galapagos, and whale sharks), these shifts reflect the changing community assemblage that has occurred at Cocos Island over the past 21 years and are not necessarily an indication of the MPA’s effectiveness (Baskett et al. 2007). It is unclear if the current dynamics of the Cocos Island elasmobranch community are simply indicative of a long transient response following creation of the MPA (White et al. 2013). Although management efforts have increased in the past decade, illegal fishing still occurs within the island’s waters (Arias et al. 2014). It is unclear if the Cocos Island MPA is even properly designed (Botsford et al. 2003) to protect such large and wide-ranging species (Hooker and Gerber 2004; Gr¨uss 2014).
Conservation efforts at Cocos Island cannot be focused simply on expanding the protected area (Arias et al. 2014); rather, efforts should be put toward increasing enforcement and management (Kelaher et al. 2015). Costa Rica’s efforts to increase their MPA coverage are admirable, but the establishment of MPAs cannot be the end point. Explicit plans and dedicated funding for monitoring and enforcement must be in place to prevent the creation of a network of paper parks. These plans need to include using both theory about MPAs and empirical data (White et al. 2011). Further, there must be stronger penalties for noncompliance with MPA rules to offset the potential gains of illegal fishing (Arias et al. 2014). We found that data collected by divemasters can be a reliable way to discern trends in relative abundance. We recommend that monitoring and enforcement of Costa Rica’s MPAs be increased substantially and that international environmental NGOs and foundations contribute to these efforts. Such efforts are urgently required if Cocos Island is to recover its elasmobranch populations and claim its status as a truly successful MPA.
