Conservation strategies for understanding and combating the primate bushmeat trade on Bioko Island, Equatorial Guinea

Bioko Island, Equatorial Guinea is among the important places in Africa for the conservation of primates, but a cultural preference for bushmeat and a lack of effective law enforcement has encouraged commercial bushmeat hunting, threatening the survival of the remaining primate population. For over 13 years, we collected bushmeat market data in the Malabo market, recording over 35,000 primate carcasses, documenting “mardi gras” consumption patterns, seasonal carcass availability, and negative effects resulting from government intervention. We also conducted forest surveys throughout Bioko's two protected areas in order to localize and quantify primate populations and hunting pressure. Using these data, we were able to document the significant negative impact bushmeat hunting had on monkey populations, estimate which species are most vulnerable to hunting, and develop ecological niche models to approximate the distribution of each of Bioko's diurnal primate species. These results also have allowed for the identification of primate hotspots, such as the critically important southwest region of the Gran Caldera Scientific Reserve, and thus, priority areas for conservation on Bioko, leading to more comprehensive conservation recommendations. Current and future efforts now focus on bridging the gap between investigators and legislators in order to develop and effectively implement a management plan for Bioko's Gran Caldera Scientific Reserve and to develop a targeted educational campaign to reduce demand by changing consumer attitudes toward bushmeat. Using this multidisciplinary approach, informed by biological, socioeconomic, and cultural research, there may yet be a positive future for the primates of Bioko.

estimates of maximum sustainable production, most taxa hunted for bushmeat are overexploited; potentially more than six times sustainable levels (Bennett, 2002;Bennett et al., 2002;Fa & Brown, 2009).
However, not all wildlife species are equally threatened by hunting.
Factors such as ecological flexibility (e.g., broad dietary breadth, ability to exploit numerous habitats), anti-predator behavior, and life history traits can influence species' vulnerability to hunting (Linder & Oates, 2011;McGraw, 2007;Struhsaker, 1999). Diurnal primates, for example, are particularly threatened, with over 70% of species in the region thought to be hunted unsustainably (Fa & Brown, 2009), despite most species receiving at least some level of legal protection under both national and international legislation (e.g., CITES, African Convention on the Conservation of Nature and Natural Resources).
Primates play a vital role in ecosystem functioning in terms of seed dispersal, and the preservation of primate populations is critically important for the maintenance of forest structure and forest regeneration (Chapman & Onderdonk, 1998;Poulsen, Clark, & Smith, 2001;Wrangham, Chapman, & Chapman, 1994). Declines and/or losses of these species can lead to cascading negative ecological consequences, including reductions in the number of large hardwood trees, a transition toward fast-growing, low-density pioneer species, and declines in the overall tree community diversity, threatening the persistence of the ecosystems they inhabit and the people who depend on them (Abernethy, Coad, Taylor, Lee, & Maisels, 2013;Chapman & Onderdonk, 1998;Effiom et al., 2013;Laurance et al., 2012;Terborgh et al., 2008;Vanthomme, Belle, & Forget, 2010).

Hundreds of millions of dollars have been spent on conservation in
central Africa, leading to the development and implementation of numerous strategies to better understand and combat the bushmeat trade, which have been met with varied levels of success (Pailler, 2005;Pyhälä, Osuna Orozco, & Counsell, 2016). Development objectives, such as poverty alleviation, are widely utilized and have improved livelihoods of some of those dependent on forest resources, but real conservation effectiveness is rare and often not evaluated (Roe et al., 2015), and as stand-alone measures, development objectives have had minimal success in reducing bushmeat consumption (Astaras, 2009;Oates, 1999;. In the Oban Division of the Cross River National Park, for example, despite a proposed budget of 18.43 million European Currency Units over a 7-year period (most of which went to development projects and international consultants), high-intensity unregulated hunting in the park led to low mammal densities (Oates, 1999). Heavy hunting in the park has continued since Oates' account, resulting in extremely low mammal densities, and local communities in the vicinity of the park are now "somewhat antagonistic" due to unfulfilled development promises stemming from the creation of the park (Morgan et al., 2011;Morgan, Abwe, Dixson, & Astaras, 2011).
Furthermore, many development projects struggle to meet their own objectives due to limited funds, capacity, and available time (Wicander & Coad, 2015). Forest guard patrols in protected areas have shown to be successful at reducing hunting (Bruner, Gullison, Rice, & da Fonseca, 2001;Campbell, Kuehl, Diarrassouba, N'Goran, & Boesch, 2011;Corlett, 2007;de Merode & Cowlishaw, 2006;Hilborn et al., 2006;Rowcliffe, de Merode, & Cowlishaw, 2004;Tranquilli et al., 2012), but they do not fully address the problem of bushmeat demand. These patrols often lack adequate financial resources (Njuh Fuo & Memuna Semi, 2011;Oates et al., 2004), can be ineffective if improperly implemented, and, in some cases, have contributed to conflicts with local communities (Pyhälä et al., 2016). Blanket criminalization of hunting and consumption could deter hunting if properly enforced, but enforcement regimes are often ineffective or absent, and, as such, have been relatively unsuccessful in reducing the overall trade (Barnes, 1996;Biggs, Courchamp, Martin, & Possingham, 2013;Burton, 1999;Miron, 1998;Rivalan et al., 2007). Domestication of bushmeat species has been proposed as a way to alleviate demand (Cooper, 1995;Grande Vega, Carpinetti, Duarte, & Fa, 2013;Jori, Mensah, & Adjanohoun, 1995), but it has been shown to be economically inviable in the absence of enforcement, while wild meat remains essentially a free good (Brooks, Roberton, & Bell, 2010;Mockrin, Bennett, & LaBruna, 2005;Nasi et al., 2008). Despite readily available protein alternatives at cheaper prices, taste and cultural preferences for bushmeat contribute to the persistence of its high demand (Bowen-Jones & Pendry, 1999;East, Kumpel, Milner-Gulland, & Rowcliffe, 2005;Kümpel et al., 2007;Morra, Hearn, & Buck, 2009;Reid, Morra, Bohome, & Fernández, 2005;Schenck et al., 2006). Education in order to change perceptions toward wildlife and conservation may have long lasting impacts, but societal change is often a slow process, and too many species require solutions in the short-term in order to ensure their persistence. Thus, education and outreach should be critical components of any comprehensive strategy, but they do not address the inherent immediacy of the bushmeat crisis.
What is clear is that there is no panacea for the bushmeat crisis and that our understanding of how to most effectively solve the problem remains unclear, largely due to the extremely complex nature of the bushmeat trade, spanning from individual actors to national and international-level policy considerations. No single solution can stand alone in the face of such an intricate problem; rather, we need to address the bushmeat trade from as many angles as possible.
It has long been said, however, that conservation is a crisis discipline (Soulé, 1985), and some of the best laid theories often fall victim to limitations of funding, logistics, and the realities on the ground (Cronin, Libalah, Bergl, & Hearn, 2014;James, Green, & Paine, 1999). Due to these limitations, researchers in central Africa have often focused their efforts on a particular niche, for example, socioeconomics or ecology (Brashares, Golden, Weinbaum, Barrett, & Okello, 2011;Foerster et al., 2012), with the aim of contributing data to an overarching conservation effort. Recent studies (e.g., Fa, Olivero, et al. [2015], Nasi & Van Vliet [2011], Ziegler et al. [2016]) have begun to broaden the focus to provide regional understanding and scope, while still providing great specificity in the details of their findings.
However, there have been few instances where it has been possible to develop and implement a long-term multidisciplinary approach tailored to a particular site; where both research and conservation activities could be implemented in an area small enough to feasibly manage multiple projects as well as monitor progress.
Bioko Island, Equatorial Guinea provides a unique opportunity to study the bushmeat trade and its effects on primate populations in central Africa. The Bioko Biodiversity Protection Program (BBPP), an academic partnership between Drexel University and the Universidad Nacional de Guinea Ecuatorial (UNGE), has carried out research and conservation activities, and has been spearheading conservation efforts on Bioko since the program's inception in 1998. In recent years, however, the BBPP has developed and implemented a more comprehensive approach to biodiversity conservation on Bioko, leveraging the strength of its long-term conservation and monitoring programs to shift to a more results-based approach that encompasses current education, research, and planning techniques. In this review, we aim to (1) detail the multifaceted conservation framework (Figure 1 1.1 | Bioko Island: A bushmeat case study Bioko Island, Equatorial Guinea (2,017 km 2 ) is a small volcanic island in the Gulf of Guinea, just 37 km off the coast of Cameroon ( Figure 2). The island has been recognized as a hotspot for biodiversity (Myers, Mittermeier, Mittermeier, da Fonseca, & Kent, 2000;Oates et al., 2004), owing in part to its small size, location, and biogeographic history, as well as its seven diurnal primate taxa (Table 1), which make it one of the highest priority sites in Africa for the conservation of primates (Oates, 1996). Human population density varies widely on Bioko, from >100 people/km 2 in Malabo in the north to <10 people/km 2 in the south (Albrechtsen, Fa, Barry, & Macdonald, 2006). Much of the island's biodiversity occurs within two protected areas that comprise approximately 40% of the island, Pico Basilé National Park (PBNP) (330 km 2 ) and the Gran Caldera Scientific Reserve (GCSR) (510 km 2 ). Since the late 1990s, urban development surrounding Malabo has expanded greatly, but due to a combination of rugged terrain, isolation, heavy rainfall, and an island-wide ban on logging activities (Republic of Equatorial Guinea, 1991), large swaths of intact forests remain, especially within PBNP and GCSR (Zafra-Calvo et al., 2010). Despite the readily available intact habitats and biological wealth of Bioko, there are neither management plans for its protected areas, nor detailed enforcement strategies in place with which to effectively conserve its biodiversity.
Bioko Island provides a unique opportunity for the study of the bushmeat trade. Bushmeat hunting is the primary threat to the persistence of primates on Bioko. Government attempts to regulate the bushmeat trade in Equatorial Guinea have so far focused on reducing supply by regulating hunting (Republic of Equatorial Guinea, 1988), banning hunting inside protected areas (Republic of Equatorial Guinea, 2000Guinea, , 2003, and prohibiting the hunting, sale, and consumption of primates (Republic of Equatorial Guinea, 2007). Each of these legislative efforts have ultimately been toothless, however, as objectives have been too broad, unfeasible (e.g., no staff/infrastructure to enforce laws in protected area), and/or lacking detailed strategies for funding and implementation. As a result, forests and protected areas are entirely unmanaged and hunting is extensive throughout the island, both outside (legally) and inside (illegally) of protected areas (Cronin, Riaco, & Hearn, 2013;Cronin et al., 2016;Grande Vega et al., 2013;Grande-Vega, Farfán, Ondo, & Fa, 2016). This hunting is conducted nearly exclusively for profit, predominantly by commercial hunters from the mainland sector of Equatorial Guinea (Albrechtsen, Macdonald, Johnson, Castelo, & Fa, 2007;Grande Vega et al., 2013;Hearn, Morra, & Butynski, 2006;Reid et al., 2005). The market structure and taxonomic profile are relatively similar to other regional markets (Albrechtsen et al., 2007;Cronin, Woloszynek, et al., 2015;Fa, Yuste, & Castelo, 2000). The bushmeat trade on Bioko is confined to a relatively small, contained (insular) system (barring easily identifiable imports from the mainland), with simple transport routes (Fa, 2000), and consumption primarily restricted to Malabo (Albrechtsen et al., 2007). Malabo's population is not dependent on bushmeat, as alternative protein sources are readily available, and bushmeat contributes an insignificant proportion of the population's minimum protein requirement Grande Vega et al., 2013;Morra et al., 2009;Reid et al., 2005) and fulfills only a fraction of the economic needs for relatively few individuals Reid et al., 2005). Rather, it seems that consumption of bushmeat, and especially of primates, is associated with wealth and status Cronin, Woloszynek, et al., 2015;Reid et al., 2005). As a result, larger vertebrates, specifically monkeys, are in decline on Bioko (Cronin, Bocuma Meñe, et al., 2015;Cronin et al., 2010Cronin et al., , 2013Grande-Vega et al., 2016;Hearn et al., 2006).

| Bushmeat market surveys
At the time of the first contemporary conservation assessment of primates on Bioko, primate populations were relatively abundant, but researchers also documented the existence of a bushmeat market on Bioko and warned of the potential negative impacts hunting could have on the island's primates (Butynski & Koster, 1994). Subsequent studies documented the extent of the market, and demonstrated that throughout the 1990s and early 2000s, wildlife on Bioko was heavily exploited, with some species, including primates, hunted unsustainably (Albrechtsen et al., 2007;Hearn et al., 2006;Morra et al., 2009). These studies formed a critical baseline for conservation planning, but lacking a true temporal component, were only able to provide general "snapshot" details of market characteristics and trends. Cronin, Woloszynek, et al. (2015), however, conducted a comprehensive bushmeat market study using long-term data collected between October 1997 and September 2010, which allowed for the detection of short-and long-term effects of market interventions, species-specific hunting patterns within taxonomic groupings, and seasonality in hunting patterns across several years. Market data were classified into groups (e.g., primates) and analyzed relative to three distinct periods based on conservation activities, government interventions, and notable market changes using an intervention model (Box & Tiao, 1975) and suite of time series analyses (see Cronin, Woloszynek, et al. (2015) for an in-depth description of methodology).
Over 197,000 carcasses from 45 different taxa were recorded during the course of the study. More than 35,000 of these carcasses were primates, making up about 18% of the entire volume of the market  Bioko putty-nosed monkey c Cercopithecus nictitans martini Least concern Vulnerable Taxonomic classification follows Grubb et al. (2003), except for Preuss's monkey, which is allocated to the genus Allochrocebus following Grubb (2006).  (Cronin, Woloszynek, et al., 2015). The overall market grew significantly over time concurrent to a transition toward increased shotgun hunting.
The volume of primate carcasses in the market also increased gradually until October 2007 (Figure 3a), when the hunting, sale, and consumption of primates were banned by Presidential Decree (Republic of Equatorial Guinea, 2007). As a result, the primate carcass rate temporarily dropped to nearly zero carcasses/market day, but then swiftly increased to rates 3-4 higher than pre-ban, reaching a maximum of 37.42 carcasses/ market day in April 2010. Cronin, Woloszynek, et al. (2015) termed this pattern a "mardi gras" mentality in relation to attempted conservation interventions, in which bushmeat volume actually increased following implementation of the intervention largely due to a lack of enforcement as market players sought to maximize their gains before the potential effects of the legislation could take hold. Not all primate taxa were hunted equally, however, as interspecific differences revealed via trend analyses shed light on the drivers of the rapid increase in the overall primate carcass rate following the decree. Five of the seven primates occurred in the market at a significantly greater rate in the period following the primate hunting ban, but two species (Piliocolobus pennantii and Cercopithecus nictitans) did not follow the same pattern ( Figure 3b). As a result of both environmental factors and a history of unrestricted hunting, populations of these two species are restricted to the remote southern extent of Bioko within the GCSR (Figure 3c) (Butynski & Koster, 1994;Cronin et al., 2013Cronin et al., , 2016, suggesting that a combination of isolation and long-term BBPP conservation activities focused on the area, have provided at least passive protection from hunting. Furthermore, due to their limited geographic range, these species can serve as indicator species, alerting us to hunting activity in particular areas of the GCSR.

| Forest surveys
While many bushmeat studies have been conducted on Bioko, there have been fewer comprehensive field surveys for primates. Rather, the emphasis has been on maintaining a localized, yet regular, long-term monitoring and research effort (Cronin et al., 2010;Hearn et al., 2004Hearn et al., , 2006 in order to provide passive protection in key areas. Although there have been myriad conservation benefits to this program, more broadly- Belebu, on the northern boundary of the GCSR, was a village of several hundred people and served as the primary access point for the GCSR via paved roads from Luba. The area around Belebu was extensively hunted and regularly organized bushmeat transports brought offtake to Malabo (JMEE, pers. obs.). There also has been a long history of plantation agriculture around Belebu, so in addition to the loss of primary forest in the area, shotguns were widely used to both hunt bushmeat and control agricultural pests (Butynski & Koster, 1994). It should be noted that these surveys took place between January 2011 and February 2012 and, as aforementioned, direct access to Ureca and the southern beaches via the new road has changed patterns of hunting pressures in the reserve (Figure 2).
Unsurprisingly, primate abundance was negatively associated with shotgun hunting (Cronin et al., 2016). Primate encounter rates were significantly lower at Belebu than at either of the other two sites, while concurrently shotgun hunting was highest at Belebu (Table 2). Although these data indicated that hunting was adversely impacting the overall primate population, they did not explain how individual species were affected by differing levels of hunting pressure. A "hunting response index" (HRI) was developed in order to infer species-specific vulnerability ( Figure 4). An HRI has been used before (e.g., Linder and Oates [2011]) to provide an estimate of vulnerability to hunting by comparing relative differences in species' encounter rates between FIGURE 3 (a) Time series plot of the overall primate carcass rate (carcasses/market day). Vertical lines delineate breaks between periods of distinct market activity identified via intervention analysis (Box and Tiao, 1975). The second break (between "Pre-ban" and "Post-ban") coincides with the October 2007 Presidential Decree banning primate hunting (Republic of Equatorial Guinea, 2007). Adapted from Cronin, Woloszynek, et al. (2015). (b) Average change in carcass rate per month (slopes) and 95% confidence intervals for each of the diurnal primate species. Adapted from Cronin, Woloszynek, et al. (2015). (c) Species range estimation for P. pennantii and C. nictitans derived from the ecological niche models developed using Maxent, both of which are restricted to the southern extent of Bioko within the GCSR. Adapted from Cronin, Bocuma Meñe, et al. (2015).
| 5 of 16 highly and lowly hunted forests while controlling for habitat type. HRI values of less than one suggest that a species is vulnerable to hunting, values greater than one suggest the species may be resilient, and a value equal to one suggests no effect from hunting. Both Cercopithecus erythrotis and C. nictitans exhibited some resiliency to hunting, which in the case of C. nictitans supports results from Linder and Oates' (2011) study in Korup National Park in Cameroon, as well as reports of relatively high densities of C. nictitans in other heavily hunted forests throughout the region (Garcia & Mba, 1997;Matthews & Matthews, 2002;Muchaal & Ngandjui, 1999). In contrast, the other four primate species were all vulnerable to hunting, as each was encountered less in heavily hunted forests ( Figure 4). P. pennantii was most vulnerable to hunting, a trait it shares with many other highly threatened forms of red colobus across Africa (Struhsaker, 2005). This vulnerability has been attributed to its high degree of ecological specialization (e.g., limited dietary and habitat flexibility), as well as its large body and group size, and slow and ineffective anti-predator responses (González-Kirchner, 1997;McGraw, 2007;Struhsaker, 1999).

| Biomonitoring and research presence
Forest guards can be beneficial in combating the illegal harvesting of wildlife (Bruner et al., 2001;Campbell et al., 2011;Corlett, 2007;de Merode & Cowlishaw, 2006;Hilborn et al., 2006;Rowcliffe et al., 2004;Tranquilli et al., 2012), and are believed to be the most cost-effective and expeditious solution to poaching, as well as an integral part of effective long-term protected area planning (Bennett, 2011 (Figure 2), and, in doing so, converted a number of former hunters to enthusiastic conservationists. Furthermore, data collected by the monitoring teams have been vital to BBPP's efforts, contributing to a number of reports and publications advancing the cause of biodiversity conservation on Bioko (Cronin et al., 2010;Cronin et al., 2013Cronin et al., , 2016Cronin, Bocuma, Meñe, et al., 2015;Fitzgerald, Ordway, Honarvar, & Hearn, 2011;Hearn et al., 2006;Rader, Ela Mba, Morra, & Hearn, 2006), as distilling data into formats understandable to policy makers as well as the general population is key to the development of successful wildlife management plans.
The number of primate group sightings is in parentheses. of the wettest in the world with over 10,000 mm of rain annually (Font Tullot, 1951;Nosti, 1947), and its rugged terrain has left steep and deep ravines unexplored, and restricted potential areas for primate surveys, especially during the wet season, when access and mobility are extremely limited. As a result, our inferences about primate distributions could not adequately account for inaccessible areas and/or range shifts.
We have been able to overcome these hurdles by generating ecological niche models (ENMs) using the program Maxent (Phillips, Anderson, & Schapire, 2006). ENMs were developed specifically to maximize the utility of presence-only data collected in similar situations (Elith et al., 2011), p > 0.05), and reported that bushmeat was their preferred protein source (DF = 1; p > 0.05). However, differences existed among ethnic groups in regards to preferred bushmeat type; Fang respondents had a higher preference for primates (Fisher's Exact Test; p < 0.05). These findings were similar to Fa, Juste, Burn, and Broad (2002) in which they reported that in 1990-1991 the Fang also had a significant preference for primates. This suggests that certain cultural preferences may transcend significant economic growth and changes in the bushmeat market (Cronin, Woloszynek, et al., 2015), and the erosion of other seemingly entrenched societal norms related to consumption (e.g., a taboo on eating Colobus satanas) (Colell, Mate, & Fa, 1994;Kümpel et al., 2008). In a comparison of two hunting villages, one Bubi and one (3) and a continuous research and biomonitoring presence in the GCSR, to advocate for a movement toward conservation lead and supported by the government.

| Focusing conservation efforts: Primates as umbrella species
A good example of our strategy in practice is the case of Pennant's red colobus (P. pennantii), which illustrates the biological importance of the GCSR, as well as the utility of using primates as umbrella species for conservation on Bioko. P. pennantii is perennially considered among the world's most endangered primates (Cronin, Hearn, & Oates, 2014;Mittermeier et al., 2007Mittermeier et al., , 2010. Previous work has suggested that P. pennantii is the only primate species endemic to Bioko (Groves, 2007;Groves & Ting, 2013;Oates, 2011), despite high subspecific endemism among the other six species (Oates, 2011), and that it is also the most vulnerable primate to the impacts of hunting on the island (Cronin et al., 2016). Forest surveys have suggested that the population size of P.
pennantii has declined significantly since 2006 (Cronin, unpublished data), and that its geographic distribution (Cronin, Bocuma, Meñe, et al., 2015;Cronin et al., 2013Cronin et al., , 2016 is restricted entirely within the boundaries of the GCSR to an area nearly half the size of previous estimates (IUCN, 2016;Oates, 2011). Furthermore, the distribution of P.
pennantii encompasses not only the areas with the highest species richness of monkeys on Bioko, but also much of the critical sea turtle nesting habitat along the southern beaches of the GCSR (Figure 6)

| Conservation recommendations
Securing the long-term future of the GCSR will require a multifaceted approach, including (1)  A critical problem that must be addressed is the commitment of the Equatoguinean government to biodiversity conservation, namely through support for protected areas and their management. Equatorial Guinea is far from unique in this situation, as many African rain forest protected areas are underfunded by 50-80% of their necessary annual operating costs (Bruner, Gullison, & Balmford, 2004;Wilkie, Carpenter, & Zhang, 2001), and at least 75% lack a secure, long-term funding program (Struhsaker, Struhsaker, & Siex, 2005). In a comprehensive analysis, Struhsaker et al. (2005) estimated that the annual cost of operating a protected area in African rain forests was between 23 and FIGURE 5 Using forest survey data, the BBPP developed ecological niche models for each of Bioko's monkey species (Cronin, Bocuma Meñe, et al., 2015), which allowed for the identification of (a) hotspots of primate species richness. Using these primate hotspots and existing infrastructure as a guide, we designated (b) priority areas for primate conservation on Bioko, as well as a series of "Ranger Bases" at principal protected area access points and "Bushmeat Checkpoints" that would utilize existing infrastructure at key transit "choke points" to serve as bushmeat inspection points. Also designated were two remote camps, Moraka and Moaba, both long used by the BBPP, but also key sites for landing small boats, from which forests guards could monitor and protect two of most important beaches on the southern coast 208 USD/km 2 , noting however, that these levels were believed to be insufficient, and that even doubling estimates to about 400 USD/km 2 would still have left the costs significantly lower than protected areas in developed nations (James, Green, et al., 1999). Blom (2004) had similar results, estimating that the average yearly expenditure to achieve effective management at numerous protected areas across Cameroon, Equatorial Guinea, Gabon, and Central African Republic was approximately 212 USD/km 2 . Given the high abundance and diversity of species in these forests (e.g., Oates et al., 2004), this suggests that investment in African rain forest protected areas is highly cost-effective (Bruner et al., 2004).  approximately 452 million USD/year. Furthermore, James, Gaston, and Balmford (1999) estimated that roughly 10 billion USD per year are spent safeguarding the world's protected areas, from which approximately 600 billion USD/year is generated in direct in-country expenditure from visitors to these areas (Balmford et al., 2015).
Unfortunately, visitors to Equatorial Guinea are scarce, as it is considered among the least visited countries in the world (Mark, 2015), meaning generating large-scale profits from ecotourism will be difficult in the near-term. However, if some policies were amended to make it easier to visit and transit throughout the country, there is significant potential for ecotourism on Bioko due to its intact forests, expansive beaches, and abundant wildlife, including high densities of both diurnal primates and marine turtles. We recognize that these values are rough estimates and that there are inherent limitations in detailed valuations of ecosystem services or potential future tourism expenditures, but, in general, evidence suggests that increased investments in protected areas and their management could yield substantial returns.
Given the current conditions and the task ahead, government-led conservation efforts should be prioritized on Bioko in order to concentrate efforts how and where they are likely to have the maximum possible conservation benefit. Initial efforts should focus on the protection of primates and marine turtles, the taxa most threatened by current offtake patterns. The optimal scenario for the preservation of primates would be a complete ban on shotgun hunting, which accounts for over 90% of all primate carcasses (Cronin, Woloszynek, et al., 2015;Grande Vega et al., 2013). Albrechtsen, Fa, Davidsen, and Macdonald (2004) tested this policy in a model based on Bioko, which also included manipulating market prices for larger animals and increased incomes from alternative livelihoods. The gun hunting ban performed best, not just for the protection of large animals, but also for the conservation of small animals and for reducing the size of the hunting population. Guns have been confiscated previously on Bioko in 1974 (Butynski & Koster, 1994), to the benefit of wildlife, and more recently Grande Vega et al. (2013)  Grande Vega et al., 2013). Perhaps the most practical solution would be the implementation of forest guards (Bennett, 2011), which has been successful elsewhere in reducing hunting and improving the effectiveness of protected areas (Bruner et al., 2001;Campbell et al., 2011;Corlett, 2007;de Merode & Cowlishaw, 2006;Hilborn et al., 2006;Rowcliffe et al., 2004;Tranquilli et al., 2012). Intensive monitoring for bushmeat hunting could be conducted by these forest guards (or in the immediate future, the trained military personnel) stationed at "ranger bases" situated at key protected area access points, and by randomized searches of those transiting into and out of protected areas by INDEFOR-AP staff. By focusing on these natural "choke points," guards may be more effective than with a more generalized enforcement scheme. The development and implementation of a "ranger base" or restricted access point at Belebu is especially important to the long-term future of the GCSR ( Figure 5b). It provides the only vehicular access into the GCSR and to the southern extent of Bioko. The highest richness and densities of wildlife are found in this area, providing a biological incentive for protection, but the southern beaches area also a major tourist attraction, and the inability to control access and to generate profits that could then be put back into the management of the GCSR represents a significant missed economic opportunity.
A more broadly focused, systematic program of regular forest patrols covering the entirety of Bioko's protected areas should also be implemented in the long-term, but in the beginning, we recommend that patrol efforts be prioritized following the conservation prioritization framework we developed using our ecological niche model results ( Figure 5b). This framework was created to maximize conservation effectiveness based on amounts of total coverage, as well as the estimated relative investment necessary to effectively patrol the area.
We also recommend that forest guard positions eventually be staffed by people living in or alongside protected areas (e.g., Ureca, Moka, Belebu, Moeri, Basilé Fang), as they are best suited for the positions given their local knowledge of the area they will be patrolling. This will aid in the success of the guard program by attaching an economic value to the stewardship of wildlife, and by helping to empower and engage local communities in the process of conservation.
The highest priority zone is the southwestern sector of the GCSR, which should be considered a "ritical zone" in each of the following management strategies (Figure 5b). Using patrols to make this area a "no-take zone" could be enforced more easily than any offtake restrictions (Milner-Gulland & Bennett, 2003), as it is already protected passively via isolation, difficult terrain, and limited access from the ocean. This area contains all seven diurnal primate taxa at densities higher than elsewhere on Bioko, and has significantly less hunting than other sites (Cronin, Bocuma Meñe, et al., 2015;Cronin et al., 2016). It also holds the highest densities of Mandrillus leucophaeus and C. satanas are also the only safe sites for landing small boats, a tactic employed by both terrestrial and marine poachers ( Figure 5B). Thus, by concentrating anti-poaching efforts in the region, guards would be able to maximize conservation benefits at minimal cost.
The second priority zone is, in effect, an extension of the first zone to include the entire southern extent of Bioko (Figure 5b). This would be potentially more of an investment to monitor as it is a larger area, but delineation would be slightly less arbitrary than the first priority zone, and thus, may be easier to enforce. This zone is delineated with a relatively straight line from east to west across the spine of the southern highlands, encompassing the northern rim of the Gran Caldera, but passing below the pastures on the slopes of Pico Biao.
This conservation zone would contain an identical faunal species assemblage, but would likely also encompass the entire range of C.
nictitans. Protecting this zone would also conserve the entirety of the unique monsoon forest habitat type as well as afromontane formations on the two peaks.
The third priority would be to simply protect the currently delineated protected areas (Figure 5b). This is, perhaps, the best place to start, as the protected areas already legally exist, and would require no new designation. This strategy is aided, like the previous two, by the fact that they were originally created since much of the terrain they encompass was deemed inferior for agriculture and overly difficult to access and exploit. In spite of their legal status and difficult terrain, development continues to gradually progress inside Bioko's protected areas with little consideration of their status. Finally, given the extensive territorial waters of Equatorial Guinea, the commercial fishing sector represents a much underutilized resource, but also an opportunity to reduce pressure on Bioko's terrestrial wildlife. Fish availability and bushmeat demand have been shown to be directly linked (Brashares et al., 2004), thus increased availability of fresh fish may help to alleviate demand for terrestrial wildlife. Malabo consumers have a preference for fresh meat (Reid et al., 2005), but the only commercial sources of meat currently "produced locally" are bushmeat and fish , as well as the occasionally available "cebu" beef (humped cattle). The organization and improvement of the Equatoguinean national fishing fleet may reduce pressure on Bioko's terrestrial wildlife. Furthermore, increasing numbers of non-African fleets are fishing in the Gulf of Guinea, heavily exploiting stocks to the point of decline (Pauly & Zeller, 2016;Pauly et al., 2014), and forcing small scale fisheries to compete with industrial fleets (Belhabib, Sumaila, & Pauly, 2015;Pauly & Zeller, 2016). If fish stocks do not begin to be managed more effectively for the Equatoguinean population, the supply of fish will decline, likely leading to increased demand for bushmeat (Brashares et al., 2004).