The logical question that follows from all of the research into this ecologically tremendous event is, "What will happen next?" Opinions vary on what will or can happen to the coral reefs in the next few years, but there is a consensus that the reefs will not return to their previous condition without human intervention (Hughes 1994). Whether they can, or even should, be restored with human intervention is up for debate. Some researchers (e.g., Knowlton 1992) believe that the current paucity of D. antillarum is a stable alternative to the formerly high density, and that these small populations are not likely to recover to any significant fraction of their former numbers.

Even if one takes for granted that human intervention can help to restore some of the former condition of the Caribbean coral reefs, there remains a question of what conservationist strategies can produce the most positive change. Clearly, any attempt to return to the former algal composition of the reefs will require a substantial increase in herbivory from current levels, especially the consumption of turf and macroalgae (Hughes 1994). More than a decade after the mass mortality of D. antillarum, no other herbivorous urchin or fish has had a competitive release to take over the level of herbivory formerly sustained by Diadema (Lessios 1995). Therefore, restoration of D. antillarum to levels of significant herbivory seems necessary as a part of reef recovery.

However, full recovery to pre-mortality population densities may not be a feasible, or even desirable, part of reef recovery. First of all, the pre-mortality population density has already been cited as a contributing cause of the very mortality that is requiring this recovery in the first place. For this reason, it seems better to restore the herbivory of D. antillarum without restoring its former numbers.

How is it possible to limit a population like this? The answer is simple: predator cropping. Predation on urchins, either by carnivorous fish or by humans, will be necessary to prevent exploitative competition by this voracious grazer (Hughes 1994; Hunte et al. 1986). As has been established earlier in this thesis, carnivorous fish populations are already in trouble, so they cannot be expected to recover on their own if human fishing pressure continues at recent levels.

It may be possible to restore carnivorous fish populations simply by generally restricting fishing and allowing the fish to reproduce to fill the niche they formerly occupied. McClanahan et al. (1994) demonstrated that fishing restrictions can help prevent sea urchin dominance in some coral reef habitats. It therefore seems reasonable that the same strategy could be effective in maintaining a stable herbivore community structure in Caribbean reefs.

Alternatively, these carnivorous fish could be bred and raised in captivity or in protected reef habitats, then released into less protected areas. The captive-breeding strategy should be used only if absolutely necessary in the early stages, as it would incur obvious costs for the artificial maintenance of reeflike conditions. The use of more and less protected reefs could be effective only if the protected reefs are sufficiently separated from unprotected reefs to prevent all the wandering fish from being caught as they pass through the unprotected zone. Such were the conditions among protected and unprotected reefs in McClanahan et al.'s Kenyan lagoon experiment (1994).

If the restoration of predatory fish is impossible or impractical, humans may have to intervene more directly to control D. antillarum population densities. Specifically, people may have to eliminate a certain proportion of living individuals in dense Diadema populations. As Diamond (1992) noted in the deer living in national parks, a herbivore whose natural predators have been removed by human activity must be controlled by human hunters in order to prevent overpopulation of the herbivore. Of course, the restoration of triggerfishes faces far less emotional opposition than the restoration of wolves and grizzly bears, but there are significant economic obstacles that will impede extensive fishing restrictions. It will be difficult to pass and to enforce fishing regulations sufficient to restore predatory fish populations to their former (before this century) levels, so human harvesting of urchins may be necessary to supplement the actions of nonhuman predators.

Presumably, native fishermen could find some market for these urchins, perhaps for food or decorative purposes, if the need arises to suppress growing urchin crowds. The Pacific urchin Strongylocentrotus franciscanus is already a major food resource for Japan (Tegner & Dayton 1977), so it is reasonable to expect that D. antillarum could also become an economically important resource. This could even provide an alternative source of income to fishermen who currently harvest such a dangerously high proportion of the carnivorous fishes in the Caribbean. Of course, some restrictions may have to be placed on the harvesting of these urchins, in order to prevent the excessive kills that occur in the Pacific (Tegner & Dayton 1977).

The past several paragraphs have rested on the premise that Diadema antillarum can be an effective herbivore in less dense populations than it had before the 1983 mortality, but it may not be clear how a smaller number of urchins can maintain a sufficient level of herbivory to keep algal abundance and species composition in check. When one recalls D. antillarum's ability to change its body size, either positively or negatively, in response to population density, it seems clear that individuals in a properly cropped population will increase their body size in order to maintain the same level of herbivory and even reproduction as a greater number of smaller individuals.

Another premise that has been taken for granted in much of this chapter is that Diadema antillarum populations will eventually recover in their numbers to a sufficient degree that overpopulation will again become a reasonable possibility to be feared. Some researchers, who see the current condition of sparse D. antillarum populations as a stable alternative to the excessive crowding that existed before the mass mortality (e.g., Knowlton 1992), see little likelihood that this urchin will ever recover past a certain threshold population density, below which the populations will repeatedly sink to their current, low levels. Clearly, it will be an uphill battle to restore Diadema populations, or at least a long one. There is no clear consensus on how much human intervention will be necessary to allow this recovery, but the answer seems to lie somewhere between simple fishing restrictions and captive breeding.

While captive breeding may seem excessively manipulative and artificial to some, it has the added problem of being predictably expensive. The research programs that investigate the condition of the Caribbean reefs may not ever receive sufficient financial support to allow a project as bold as nursery to the world's population of an entire species. Laboratory experiments (e.g., Bauer 1982) have already demonstrated that it is possible to raise D. antillarum in captivity, although it requires careful attention to chemical, light, and temperature conditions of the water. Furthermore, even Bauer's carefully arranged captive-breeding apparatus drew water directly from the ocean, which makes even captive individuals vulnerable to the potential of another mass mortality by waterborne pathogen (Miller & Colodey 1983).

Depending on natural recovery (with minimal intervention, including fishing restrictions) has the disadvantage of taking a tremendously long time, if it ever succeeds. In the time it takes for populations to recover that way, the governments and institutions that allowed the restrictions in the first place may run out of money and make unwise changes in the plan in order to pursue short-term interests. Such abandonment of the long-term plan could destroy the entire process. Furthermore, the longer these reefs remain dominated by space-monopolizing macroalgae, the more corals will be choked out, and the less capable the corals will be of recovery (Hughes 1994). This recovery of living coral organisms is necessary even to balance out the continual erosion and destruction of reefs, and greater recovery is required if the reefs are to rebuild themselves to anything resembling their former magnitude.

What happens if complete Diadema recovery proves impossible? Already, there have been other mortality events, even in the limited populations that survived through the 1983 mass mortality. In 1991, for example, a mortality event reduced already small populations by 97% in some parts of the Florida Keys (Forcucci 1994). Events such as this one suggest that D. antillarum may not be able to recover even to levels of effective herbivory, let alone to pre-mortality densities. At the same time, however, it is important to note that this mortality did not spread beyond the Florida Keys, thus giving weight to the argument that lower population densities are less vulnerable to the spread of a contagious pathogen.

So, if D. antillarum can't be restored, what about the possibility of increasing herbivory by increasing numbers of other reef herbivores? As has already been discussed, there has been no competitive release by other herbivorous urchins or fishes on the reefs, and there are several good reasons for this.

First of all, the fish are all being caught. As has been discussed before, the herbivorous fishes, along with the carnivorous ones, are being caught in such quantities that their populations are becoming almost ecologically insignificant (Hughes 1994). Fishing restrictions, as suggested for the restoration of carnivorous fish populations, may be a prerequisite of any substantial recovery in the populations of herbivorous fishes as well. These fishing restrictions may turn out to have harsh short-term effects on certain local fishermen, but the long-term effects may help preserve and restore the reefs which otherwise seem doomed (Hughes 1994). Furthermore, as has been mentioned earlier, there are not that many fish living on Caribbean reefs right now, so the short-term loss of fish harvest due to fishing restrictions may not even be that great.

Again, there is also the possibility of restricting fishing in some reef areas, while allowing it at others (McClanahan et al. 1994). The problem with unrestricted fishing in an ecosystem is that fishermen will deplete even the last refuges of fish, thus destroying the renewability of the fish resource. This can be prevented by restricting fishing only in certain areas. In fact, it may be far more effective and enforcible to prohibit any fishing in a few limited areas than to place moderate restrictions on a wider geographical range. As Bohnsack (1993) points out, the selection pressure applied by intense fishing removes the largest and most desirable individuals from the breeding stock, thus further weakening the genetic viability of the population with each catch. Also, the fishing and tourist industries are currently competing for incompatible uses of the same space. If certain areas are set aside as marine fishery preserves, with a complete prohibition of fishing, the economic benefits from increased tourism can equal or exceed the cost of lost fishing revenue (Bohnsack 1993). Furthermore, once a reserve becomes well established, it will export enough healthy animals that the waters outside the reserve's border will support a stronger fishery than before.

The second reason why herbivorous fishes haven't taken over the level of herbivory formerly maintained by urchins is that the fish don't eat all the same algae that the urchins do. Coral reef algae produce a number of chemical defenses against herbivory, and several of these protect certain reef algae from being grazed by fish (Hay et al. 1987). The results of these different algal preferences among different herbivores is also demonstrated by experiments with herbivore exclusion fences and algal transplants. In these experiments, areas that were protected from urchin grazing exhibited an increase in erect and filamentous algae, which were then shown to be resistant to fish grazing (Morrison 1988). Erect and filamentous algae are the same types of algae that are currently monopolizing space on reefs and growing without sufficient control (Morrison 1988). If the problematic algae are the preferred food of urchins, and are resistant to fish grazing, then it seems that only urchin recovery can suppress these algae to safe levels. For this reason, the increase of urchin populations, however difficult the task may be, could be an absolute necessity for even the mere stabilization of coral reefs.