Tempering Expectations of Recovery for Previously Exploited Populations in a Fully ProtectedMarine Reserve

Centuries of resource extraction have impacted coral reef ecosystems worldwide. In response, area and fishery closures are often enacted to restore previously exploited populations and reestablish diminished ecosystem function. During the 19th and 20th centuries, monk seals, pearl oysters, and two lobster species were overharvested in the Northwestern Hawaiian Islands, now managed as the Papahānaumokuākea Marine National Monument, one of the largest conservation areas in the world. Despite years of protection, these taxa have failed to recover. Here, we review each case, discussing possible factors that limit population growth, including: Allee effects, interspecific interactions, and time lags. Additionally, large-scale climate changes may have altered the overall productivity of the system. We conclude that overfishing of coral reef fauna may have broad and lasting results; once lost, valuable resources and services do not quickly rebound to pre-exploitation levels. In such instances, management options may be limited to difficult choices: waiting hundreds of years for recovery, actively restoring populations, or accepting the new, often less desirable, alternate state.


Introduction
For centuries, coral reef ecosystems have been viewed as renewable resources, from which countless organisms have been harvested to serve dependent human communities [1,2].There is an underlying assumption that marine species are highly resilient to large population reductions and that abundance may be quickly restored by area or fishery closures [3][4][5].The capacity for self-repair, however, is dependent upon the extent of human impact and system resilience [6,7].Excessive and prolonged anthropogenic exploitation may reduce resilience, that is, the capacity of a system to persist while maintaining key relationships and absorbing disturbances [8].For example, 15 years after fishery closures, there has been little recovery of several marine fish stocks that were previously reduced 45%-99% [9,10].Thrush and Dayton [11] review several examples of marine ecosystem ratchets, in which recovery to pre-exploitation levels is no longer possible [12].In such instances, managers are forced to accept the new, often less desirable, alternate ecosystem state.
The Northwestern Hawaiian Islands (NWHI) span the center of the Hawaiian-Emperor seamount chain, the world's most geographically isolated archipelago (Figure 1) [13].As of 2006, these small islands, atolls, reefs, and surrounding areas comprise the Papahānaumokuākea Marine National Monument, one of the largest marine protected areas in the world.Fishing is not permitted, and the human population is small (i.e., managers, researchers, and a small crew to staff an emergency runway at Midway Atoll).Thus, the NWHI coral reef ecosystem is widely regarded to be "nearly pristine," hosting predator-dominated communities, large fish populations, numerous coral species, and a healthy algal assemblage [14].It has not, however, escaped human impact   [15], and modern species assemblages may only appear to be intact because baseline levels of abundance prior to exploitation are unknown (i.e., shifting baseline theory) [16].

Objectives
Here, we review anthropogenic impacts on the NWHI coral reef ecosystem, focusing on four of the best-studied species: Hawaiian monk seal (Monachus schauinslandi), black-lipped pearl oyster (Pinctada margaritifera), Hawaiian spiny lobster (Panulirus marginatus), and scaly slipper lobster (Scyllarides squammosus).The lack of recovery demonstrated by these taxa may be a result of Allee effects, inter-specific competition, and time lags.In addition, large-scale climate processes may have altered the carrying capacity of the entire system.We consider the implications of these losses and provide recommendations for ecosystem-based management in the NWHI, now protected as the Papahānaumokuākea Marine National Monument.

A Brief History of Anthropogenic Impacts in the NWHI Coral Reef Ecosystem
The NWHI ecosystem has been significantly altered by humans, beginning with early Native Hawaiian settlers (1000-1700 AD) and Western explorers of the late 18th century [17].Habitat disturbances include: guano mining at Laysan; modification of essential beach habitat by military actions at Midway Island, French Frigate Shoals, and Kure Atoll; creation of channels at Midway and FFS; marine debris accumulation; and pollution [18][19][20].In addition, a myriad of species occupying several trophic levels were harvested for subsistence, food markets, bait, feathers, and shells (Table 1).Between 1842 and 1915, a minimum of 1.3 million seabirds was harvested at Laysan, Lisianski, and Midway Islands [18,21,22].Hunted for human consumption and to be used as shark bait, Hawaiian monk seals (Monachus schauinsalndi) and green sea turtles (Chelonia mydas) were nearly extirpated by the 20th century [19,23].Sharks were heavily fished during the late 19th century [24][25][26] and as recently as 2000, when 990 individuals were taken in a single 21-day exploratory shark-fishing trip [27].Bottomfish stocks have been fished since the 1930s [28].Harvest of pelagic and nearshore fish species, including scrombrids and trevally, began after World War II and continued until the early 1990s.Approximately 150,000 black-lipped pearl oysters (Pinctada margaritifera) were harvested at Pearl and Hermes Reef between 1928 and 1930, severely depleting the population [29].Precious coral beds (Corallium spp.) have been exploited since 1965 [30].In the mid-1970s, a commercial lobster fishery targeting Hawaiian spiny lobsters (Panulirus marginatus) and scaly slipper lobsters (Scyllarides squammosus) became the state's most valuable demersal Hunted to near-extinction for meat, skins, and bait for shark fishery [23,35].For details, see Dedicated research surveys have since documented large populations of seabirds and sea turtles in the NWHI; however, four species have not rebounded to pre-exploitation levels: the Hawaiian monk seal, black-lipped pearl oyster, Hawaiian spiny lobster, and scaly slipper lobster.The lack of population growth in these taxa may be a result of speciesspecific factors, including Allee effects (mechanisms that lead to inverse density dependence at small population sizes [33]), redefined inter-specific interactions, or simply an inadequate amount of recovery time.Alternatively, large-scale climate processes may have altered productivity and trophic function of the entire ecosystem.Here, we explore how various factors may have prevented the recovery of seals, oysters, and lobsters in the NWHI and discuss implications for management of the Papahānaumokuākea Marine National Monument.

Hawaiian Monk Seal
In 1805, Russian explorer Yuri Lisianski provided the first written account of the Hawaiian monk seal [34].Early on, population abundance was described in historical documents as "numerous," "many," and "considerable" [35].In the next 80 years, hundreds to thousands of seals were killed for their meat, skins, and oil by sealers and shipwrecked sailors [23].By the turn of the century, few seals were observed over 1-14 months, as reported by several vessels traveling throughout the NWHI (Table 2) [35].Commercial hunting ceased due to a lack of profitability, and the species was thought to be extinct.
Observations of monk seals in the early 20th century indicated that at least 23 individuals must have survived at the nadir of the bottleneck [36].By 1958, 916 nonpup seals (i.e., seals over the age of one) were counted in the first systematic beach survey [37,38].From this count, total population size was estimated to be 1350-2962 individuals [38,39].The species was thought to be on the path to recovery, similar to other previously exploited pinniped populations.
By 1978, however, low juvenile survival had resulted in a 50% population decline (Table 2).Despite active protection and management, overall species abundance continues to decline at ∼4.1% per year as a result of low reproductive rates and high juvenile mortality, attributed to emaciation, shark predation, and entanglement in marine debris [40,41].The most recent estimate of total population size is 1,146 individuals [41].The International Union for Conservation of Nature (IUCN) has listed the species as Critically Endangered, reflecting its high risk of extinction in the wild [42].
The stunted recovery and subsequent decline of the Hawaiian monk seal may be attributed to Allee effects.Video footage reveals that female seals defend their preweaned pups against predation by Galapagos sharks.A larger number of nursing females would provide increased vigilance (e.g., observance and barking) and aggression (e.g., biting).Inbreeding depression (i.e., diminished fitness of inbred individuals) is a common Allee effect that often results in high juvenile mortality rates and low reproductive rates, both of which are exhibited by the species.Further research is required to determine the impact of inbreeding depression on the population [36].Genetic diversity has been virtually lost throughout the genome, likely as a result of long-term population reduction [36,[43][44][45][46]. Though genetic diversity is essential for population persistence and to mount an effective immune response [47], its contribution to the current decline of the species remains unknown.
Another explanation for the current decline in monk seals is inter-specific competition and reduced prey abundance.Emaciation and poor survival of juvenile seals in the NWHI suggests food limitation and stress in the forage base [48].Dietary overlap has been documented between seals and large predatory fish in the region [49][50][51][52][53], and overall prey scarcity intensifies competition among these top-level predators.Cameras mounted on the backs of seals show jacks, snappers, and sharks attempting to steal prey that seals have flushed from cover; these predatory fish also bully the seal for prey in its mouth [53].Juvenile seals appear to be most vulnerable to the impacts of competition.Furthermore, pre-weaned pups are increasingly preyed upon by sharks in the NWHI [40].In contrast, there is high pup and juvenile survivorship in the main Hawaiian Islands [54], where the biomass of large predatory fish is an order of magnitude lower [55].Without time-series data on shark, jack, and prey populations in the NWHI, we cannot determine whether predatory fish have assumed the niche once held by seals, nor whether there is increased competition due to diminished prey abundance.
Compared to other apex predators, seals are unique in their ability to forage across a wide depth range and to  [41] access cryptic prey found under rocks and buried in the sand.As central-place foragers, they undertake extensive movements to feed at neighboring banks, across habitat types and down to subphotic depths [56].They are generalists and exert top-down pressure that can structure their diverse prey community, including demersal and benthic species [49,52].Patterns of prey depletion persist in the slow-growing fish communities at subphotic depths close to the seal colonies [57].Therefore, changes in monk seal abundance could have major impacts at lower trophic levels.

Black-Lipped Pearl Oyster
The black-lipped pearl oyster is harvested and cultivated throughout its range in the Indian and Pacific Oceans for pearls and the nacre of the shell, or "mother of pearl."Though the species is uncommon throughout the NWHI, there was once a large population at Pearl and Hermes Reef [58].Between 1928 and 1930, approximately 150,000 pearl oysters were taken, eliciting a ban on harvest by the Hawai'i Territorial government with the expectation that the fishery would reopen in five years [59].In 1930, Galtsoff [29] conducted a six-week abundance survey of the Pearl and Hermes Reef stock; he found 480 individuals and concluded that the population was too depleted to sustain further harvesting.A 1993 study indicated that pearl oyster densities had not increased after 60 years of protection [60].
In 2003, Keenan et al. [59] conducted an in-depth survey of pearl oyster abundance and found a lower average density than that of similar areas in 1930 (Table 3).Additional surveys (2004)(2005)(2006) confirmed that the Pearl and Hermes Reef population remains in its depleted state after 80 years of protection [58].As described by Keenan et al. [59], Allee effects may have prevented recovery of the black-lipped pearl oyster at Pearl and Hermes Reef.Depletion of the adult population likely resulted in reduced reproductive potential of these broadcast spawners, as a result of an overall decrease in the abundance of gametes and a reduced chance of fertilization [62].Additionally, pearl oysters are protandrous hermaphrodites, starting life as males with a proportion of the population becoming female when larger.The removal of the largest adults would have removed a high percentage of the females, thereby skewing the sex ratio and decreasing the overall fecundity of the population [62].Because the planktonic larvae settle gregariously on the shells of adult oysters, the loss of settlement cues would also diminish recruitment [59].The importance of these processes is highlighted by the fact that the Pearl and Hermes Reef pearl oyster population has not increased for nearly 80 years after the cessation of harvest.
Though the last harvest of NWHI pearl oysters occurred in 1930, population recovery may require additional time.Unlike the monk seal, pearl oysters are not endemic to Hawai'i, and their recovery may be aided by external recruitment.Pearl oysters are sessile as adults, and the pelagic larvae settle 20-23 days after fertilization [63].Given the extreme isolation of the Hawaiian Archipelago and the rarity of the species throughout Hawai'i, external recruitment is likely uncommon, but the chance of such an occurrence increases with time.Also, it is possible that environmental conditions are rarely favorable to produce or support large populations in the NWHI, accounting for their scarcity throughout the archipelago.Therefore, recovery may require evolutionary time scales (i.e., hundreds of years or more) rather than management time scales (i.e., years or decades).
As filter feeders, pearl oyster populations have the potential to greatly affect levels of phytoplankton in the surrounding water column [62].The black-lipped pearl oyster has a large gill size and a high pumping rate, which makes it viable even in nutrient poor waters [64].It can graze plankton down to extremely low levels [65].For this reason, oysters are a mechanism for strong benthic-pelagic coupling, transferring planktonic energy to the benthic environment [66].The organics and minerals from the water column make up the shell and soft body tissue of the oyster, and the unassimilated portion of the filtered plankton (discarded as feces) enriches the seafloor [65].These filter-feeders also influence the ecosystem by decreasing the particle load in the water column, which allows light penetration to depths and improves conditions for photosynthesis.Therefore, the depletion of the large population at Pearl and Hermes Reef likely had a strong impact on ecosystem function, as a result of the loss of the benthic/pelagic coupling and filtering capacity.

Hawaiian Spiny Lobster and Scaly Slipper Lobster
In 1975, the National Marine Fisheries Service, the U.S. Fish and Wildlife Service, and the Hawai'i Division of Fish and Game conducted a five-year survey and assessment of the marine resources of the NWHI.Populations of the endemic Hawaiian spiny lobster were found throughout the NWHI, with particularly high concentrations at Necker Island and Maro Reef [67].Survey catch rates "showed excellent potential for development" [68] and were followed by industrialized commercial exploitation.Fishing efforts remained relatively low (<10 vessels, 400 traps hauled/vessel day) until 1984 when the introduction of new traps resulted in a fourfold increase in effort in just two years (16 vessels, 1000 traps hauled/vessel day).In addition, the new traps had a smaller mesh size that increased catch of the scaly slipper lobster, also found throughout the NWHI (Figure 2) [69,70].The NWHI lobster fishery was unregulated and unmonitored until 1983.Despite several management measures, the fishery suffered severe declines in catch per unit effort (CPUE) as determined in near-annual stock assessments from 1985 to 2000 (Figure 3).Archipelago-wide spiny lobster trap CPUE, considered an adequate proxy for density [71,72], fell from 3.41 in 1983 to 1.0 in 1986.In particular, the historically high density Necker Island population CPUE fell from 4.0 in 1983 to 1.2 in 1987.Archipelago-wide slipper lobster CPUE also declined quickly from a high of 1.21 in 1985 to 0.17 in 1991.The declining CPUE of both species resulted in a complete fishery closure in 1993, a shortened (8 week) fishery in 1994, and a one vessel exploratory fishery in 1995.The fishery was indefinitely closed in 2000 because of increasing uncertainty in the population models used to assess stock status [73].An extensive tagging program, designed to examine lobster dynamics, demonstrated that NWHI lobster abundances have not increased appreciably from 1999 levels.Spiny lobster CPUE in 2008 was 0.  Allee effects could be delaying the recovery of lobsters in the NWHI.Spiny and slipper lobster egg production was significantly reduced by the removal of females in past decades.The loss of larger females was particularly damaging, because they produce a disproportionately greater number of eggs [75,76].As lobster densities decreased due to fishing, Necker Island spiny lobster size-specific fecundity increased, a negative density-dependent, or compensatory, response [76,77].Although the long pelagic duration stage typically decouples recruitment success from fecundity [78], the lack of recovery indicates that the compensatory response may not have been enough to counter survival component Allee effects, such as the "dilution effect" (as prey groups get smaller or prey populations sparser, individual prey vulnerability increases [79]), during the pelagic or early benthic stage.The gregarious nature of palinurids and scyllarids also leaves them vulnerable to Allee effects.While it is unclear exactly which cues pelagic-phase NWHI lobsters (phyllosoma, puerulus, and nisto) utilize to orient towards their benthic settlement habitat, laboratory and field research of P. argus suggest that conspecific cues play an important role [80].The gregarious behavior is also generally thought to be a defense mechanism via group vigilance, aggression, and the dilution effect [81,82].At minimum, the "guide" effect reduces individual exposure to predation by allowing conspecifics to locate nearby, high-quality shelter [83,84].
The recovery of lobster populations may be only a matter of time.At the time of the fishery closure, it was assumed that lobster populations would rebound quickly and commercial fishing would commence in a few years.Palinurids are resilient to high fishing mortality and have the potential to recover quickly from low abundances.Pollock [85] attributes this to density-dependent factors, such as labile size-at-maturity, juvenile growth, and survival rates, which increase egg production at lower population densities.Several highly exploited species rebounded in less than a decade, including New Zealand and Australian rock lobsters (Jasus edwardsii and P. cygnus), the ornate spiny lobster (P.ornatus) of the Torres Strait and Papua New Guinea, and six species of Palinurus in Japan [86][87][88][89].Therefore, it is somewhat surprising that NWHI lobster populations have not recovered in the past decade, especially since fishing has ceased, egg production should be increasing, and oceanographic conditions favorable to lobster recruitment occurred in 1988-2002 and 2007-2009 (http://jisao.washington.edu/pdo/PDO.latest).
Severe declines in lobster populations likely impacted ecosystem function.Lobsters and other macroheterotrophs feed on the detrital and planktonic energy that is stored in the benthos as infauna (micro molluscs, annalids, etc.).Field evaluations of lobsters have found that their nutritional condition can vary over space and time, presumably in relation to prey availability [90].Gleaning energy from the benthos, populations of macro-heterotrophs comprise a pathway to higher trophic levels, serving as prey items for sharks, jacks, and seals [49,51,91,92].A sizable reduction in macro-hetereotrophs has the potential for reducing the flux of energy through the benthos and up the food web.Therefore, the loss of lobster populations could result in a benthic energy sink, constraining the flow of energy to upper trophic levels.

Oceanographic Processes
Large-scale oceanographic processes have the potential to alter trophic interactions throughout a food web.Such regime shifts result in profoundly different ecosystem structure.There is some evidence for large-scale ecosystem changes in the NWHI, linked to productivity.Of the six taxa for which long-term, in-depth population data is available, only two have recovered to significant abundance: seabirds and sea turtles.These species differ from seals, oysters, and lobsters in that they utilize the NWHI for reproductive purposes only and not to forage.Turtles and seabirds receive little energy input from the NWHI and are not trophically dependent upon this coral reef ecosystem.
Alternatively, species dependent on the NWHI for nutrient uptake have not rebounded to pre-exploitation levels.The Hawaiian monk seal, black-lipped pearl oyster, Hawaiian spiny lobster, and scaly slipper lobster represent various trophic levels in a coral reef ecosystem.The monk seal is a high trophic level species that exerts top-down predation pressure in shallow and deep-water reefs.Lobsters are low trophic level heterotrophic benthos carnivores that transfer benthic productivity up the food chain.Pearl oysters harvest planktonic productivity from surrounding waters, such that these nutrients become incorporated into coral reef food webs.Though these taxa have different life histories, population dynamics, and ecological niches, all have failed to recover from exploitation.What mechanisms could account for diminished productivity across such varied trophic levels?Marine ecosystems in the Hawaiian Archipelago may be impacted by three large-scale climate processes operating on different time scales: the inter-annual El Niño Southern Oscillation (ENSO), the Pacific Decadal Oscillation (PDO), and long-term, global climate change.At any time, all three of these climate signals may be impacting Hawai'i with varying intensities.
The best evidence of a link between climate and an ecosystem response in the NWHI occurs at the northern atolls.In some winters, often during El Niño years, the vertically mixed high-surface chlorophyll transition zone, termed the Transition Zone Chlorophyll Front (TZCF), shifts sufficiently far south that the three northern atolls (i.e., Kure Atoll, Midway Island, and Pearl and Hermes Reef) lie within its productive waters (Figure 4) [93].The temporal dynamics of monk seal populations at these northern atolls is characterized by large interannual variation in pup survival, rather than the long-term declining trend observed for the largest population at French Frigate Shoals [94].Further, monk seal pup survival increases two years after the TZCF shifts southward.The two-year lag between the position of the TZCF and monk seal pup survival is thought to represent a temporal lag between changes at the base of the food web and monk seal prey items several trophic steps removed [94].This hypothesis is supported by an ocean-plankton model indicating a strong physical, chemical, and biological gradient that impacts lower trophic level productivity [95].The relationship appears only to impact the northern atolls, so it does not account for overall monk seal dynamics or the trends observed in lobster species at the southern atolls.It may, however, have a profound impact on pearl oysters in the NWHI, which have only ever occurred in large numbers at Pearl and Hermes Reef.
Changes in the PDO (i.e., the interdecadal shift between warm and cool surface waters in the North Pacific) will also result in atmospheric and oceanic changes that could impact productivity and circulation in Hawaiian waters, potentially resulting in ecosystem regime shifts.The intensification of the Aleutian Low Pressure System during 1977-1988, characteristic of the positive phase of the PDO, may have resulted in increased deep mixing, leading to more nutrients in the mixed layer and ultimately enhanced ecosystem productivity [96].After this event ended, productivity in the NWHI diminished.In the late 1980s, 30%-50% declines were documented for seabird reproductive success, monk seal pup survival and spiny lobster recruitment [96].Between the mid-1990s and 2010, the PDO has gone through another cycle of positive and negative modes, but similar shifts in productivity have not been repeated.This lack of repeatability may disprove the existence of a relationship between PDO and productivity, or the relationship may be more complicated, as a result of time lags and complex ecosystem interactions.
ENSO and PDO oscillate between two well-defined alternate states or modes; however, the global climate change signal is expressed as an increasing temporal trend.One impact attributed to this trend is increased vertical stratification in the subtropical gyre.This leads to a reduction in nutrients mixed into the euphotic zone and reduced primary production.Based on satellite remotely sensed estimates of surface chlorophyll, the least productive areas of major oceans (i.e., the core of the subtropical gyres) have expanded over the past decade [97].Though the time series of SeaWiFS chlorophyll data are limited to the past 12 years, this trend could be responsible for long-term reduction of carrying capacity in the NWHI ecosystem.Further research is required to determine whether the change in productivity is responsible for the lack of recovery in seals, oysters, and lobsters.

Implications and Recommendations for Ecosystem-Based Management
Marine ecosystem-based management (EBM) is designed to protect ecosystem integrity and to achieve long-term sustainability [98][99][100][101].Specific objectives of EBM often include the maintenance of viable populations and the restoration of depleted resources [102,103].Marine mammals, bivalves, and lobsters are often considered keystone species [104,105] and may be essential in maintaining ecosystem integrity [106,107], as the removal of apex predators, suspension feeders, and grazers has diminished the capacity for resilience in other systems [1].Even the loss of a single species within these functional groups reduces overall biodiversity and may be detrimental to natural ecological dynamics [6,108].Therefore, while single species are not the target of protection, they remain important to management considerations [109].
Management interventions, such as fishery or area closures, are often enacted to recover depleted taxa.We have shown that these actions do not automatically result in immediate or rapid population growth.Overexploitation can result in widespread impacts that resonate through the system for decades, especially if intra-and interspecies dynamics have changed, or if energy input has been reduced.In such instances, managers generally have three options: provide additional time for recovery, actively try to restore populations, or accept that irreversible change has occurred and adapt to the new, altered system [6].We will consider each of these options as it pertains to the NWHI (which since 2006 have been protected as the Papahānaumokuākea Marine National Monument) and provide suggestions for research and management.Though it appears that adequate time has passed for seal, oyster, and lobster populations to recover from previous exploitation, we may need to alter our expectations of population growth following excessive take.For example, Hawaiian lobster and monk seal dynamics may be explained by metapopulation theory [110], in which survival and growth rates vary spatially among metapopulations, connected by gene flow [74,111,112].Extinction and recolonization events may occur at broad spatial and temporal scales, most of which are beyond the scope of contemporary research.These natural troughs and peaks occur even in the absence of anthropogenic stressors but may be exacerbated by over-exploitation.Another possibility is that reef species have low-frequency recruitment pulses, perhaps linked to stochastic local environmental processes.Oyster and lobster populations, for example, may have very low annual recruitment, with the exception that once every 50-100 years, they experience a period of high recruitment.Such theories would suggest that it is difficult to sustain single-species fisheries in coral reef ecosystems.The Papahānaumokuākea Marine National Monument is one of the largest marine protected areas in the world.As no recreational or commercial extraction occurs within its waters, it provides a natural laboratory in which to evaluate the amount of time required for population recovery when anthropogenic impacts are controlled (with the exception of global pollution and climate change).Within a protected area such as this, research becomes the main management activity.Thus, we recommend focus on the following as topics for future study: (i) long-term monitoring of multiple species to detect shifting population trends, (ii) research on inter-specific interactions to determine how systems shift over time after a perturbation, (iii) research on metapopulation dynamics to establish baseline levels of recruitment, connectivity and spatial/temporal shifts in abundance.
Another option is active management and restoration.At French Frigate Shoals, translocation of at-risk Hawaiian monk seals and removal of ten Galapagos sharks over four years reduced predation on pre-weaned and recently weaned pups from 31 in 1997 to 11 in 2003 [40].The continuation and expansion of such initiatives should be considered.Monk seals and pearl oysters may benefit from captive breeding and eventual release, though the risks associated with such activities are high and must be thoroughly researched.Aquaculture of black-lipped pearl oysters from Pearl and Hermes Reef could aid in the recovery of the population [113].Because bivalves settle in the presence of conspecifics [114], discarded shells left on the beach at Pearl and Hermes could be placed on the reef to encourage settlement.In addition, live adult oysters could be moved closer together to increase fertilization success as well as induce the settlement of larvae.These interventions could alleviate certain Allee effects, but other factors may contribute to the lack of population growth.If oceanographic processes influence carrying capacity, such management actions may be unsuccessful.Furthermore, well-intentioned attempts at restoration could have unintended and unpredictable ecological impacts of a magnitude comparable to those of overextraction.Therefore, this option requires: (i) historical data on preimpact abundance (including traditional Hawaiian knowledge in the form of orally transmitted histories, chants, and genealogies), (ii) knowledge of climate patterns, (iii) flexibility (i.e., adaptive management) to mitigate unforeseen complications.
Managers may be limited to the third option: lowering their expectations of productivity in the altered system.Coral reefs have been described as oases of biodiversity in oligotrophic, oceanic deserts [49].They may foster intense competition among species within the same functional group, such that when one species is depleted via exploitation, another fills its niche and prevents subsequent recovery.This paradigm of coral reef dynamics suggests low resilience of any one species but high functional group resilience; thus, focus on the preservation of trophic guilds is appropriate.However, catastrophic phase shifts (e.g., over-fishing of herbivorous fish followed by a sea urchin disease epidemic led to algal-dominated reefs in the Caribbean) stress the need to maintain highly diverse functional groups in order to sustain healthy ecosystems [115].The NWHI can serve as a natural laboratory to examine hypotheses regarding factors thought to confer resistance or resilience (e.g., species diversity, genetic diversity, diversity within trophic guilds, and functional redundancy) without a plethora of confounding local impacts and stressors.As such, it provides a "control" for comparison with reefs elsewhere in Hawai'i and throughout the world.Therefore, appropriate objectives for EMB are: (i) maintenance of food-web dynamics and intricacies, (ii) investigation of ecosystem function and energy flow, including research at and across all trophic levels to quantify productivity and address redundancy, (iii) assessment of capacity for resilience against potential anthropogenic disturbances, such as climate change, pollution, or invasive species.

Conclusions
Though far from pristine, the NWHI coral reef ecosystem is not beyond repair.We recommend that managers continue to provide time for recovery while permitting minor interventions for research and restoration purposes.The examples discussed in this paper require that we change the way we think about the recovery of previously exploited populations.Time should be considered in the context of the slow gradual changes and sudden shifts characteristic of natural processes.Ecosystems do not operate at a single stable state and may not recover to the original wilderness condition after a disturbance.Populations that fail to recover should not be considered isolated events, but rather interpreted as the loss of essential ecosystem components that possibly reflect broad-scale changes.The NWHI coral reef ecosystem provides an important cautionary lesson: a perturbed ecosystem does not automatically revert to its previous state when stressors are removed.This message, utilized as an educational tool, will be important in informing marine management in the Papahānaumokuākea Marine National Monument and throughout the world.

Figure 1 :
Figure 1: The Northwestern Hawaiian Islands, protected as the Papahānaumokuākea Marine National Monument since 2006 (NOAA map).

Figure 3 :
Figure 3: Commercial catch per-unit effort (CPUE; number of lobsters captured per trap haul) for Hawaiian spiny lobsters (Panulirus marginatus) and scaly slipper lobster (Scyllarides squammosus) in the NWHI from 1983 to 1999.The fishery was closed in 1993.

Table 1 :
A brief history of extraction in the NWHI: Kure Atoll (Kur), Midway Island (Mid), Pearl and Hermes Reef (PHR), Lisianski Island (Lis), Laysan Island (Lay), and French Frigate Shoals (FFS).This list is not exhaustive and does not represent all extraction occurring in the NWHI.

Table 3 :
Harvested numbers and survey counts of black-lipped pearl oysters at Pearl and Hermes Reef.