The major predators of fish (including non-commercial species) are the fish themselves. Predation by man is substantially less, roughly the same level as other mammals. Predation represents an important process in the regulation of fish populations, but predator- prey interactions and their effects on fish resources are very diverse and complex and require in-depth study. The situation is further complicated by the non-equilibrium nature of both the local and ecosystem environments in which predation and competition occur. In reality, the world's fisheries are targeted at both predator and prey assemblages. Attaining most beneficial use requires knowledge of the interactions, the ecosystem effects, and bio-economic modelling in the multi-species context.
This paper examines predator-prey interactions as a component of the regulation of fisheries resources. It includes observations on the impact and scale of predation in marine and freshwater ecosystems, approaches to including predation effects in mathematical modelling of single- and multi-species fisheries, and provides examples of how these effects have been incorporated in management advice and in harvesting strategies.
Predation can have a significant impact on marine and freshwater ecosystems. The reduction of the large baleen whales in the Antarctic meant a greater abundance of krill in the short term through reduced mortalities from predation. Subsequently there was population growth among the remaining baleen whales, seals, seabirds and squid. Similar effects can be observed in freshwater ecosystems, in seal predation on marine organisms, and in predation on eggs and larvae. Some of the more dramatic effects of predation at community level occur when species are introduced into freshwater lakes. An example is the destruction of the haplochromine cichlids of Lake Victoria after the Nile perch was introduced.
The harvesting of marine mammals is a special case, in view of the aesthetic and moral constraints being imposed by public opinion. The continuing recovery of most marine mammal populations, however, has created the dilemma of whether to continue the near total ban on harvesting, or to permit limited harvesting under carefully controlled conditions. Much more study is required to understand the consequences of these alternatives to the maintenance of balanced ecosystems. The public needs to be better informed of the extent of recovery of these populations and the ecosystem effects, in order to permit a more reasoned discussion of the costs and benefits (aesthetic, social and economic) of the alternative approaches.
The scale of predation is reported for the following: cod, tuna, marine walleye pollock, mammals, cephalopods, and birds. The magnitude of shrimp predation in the waters of Greenland is presented as an example.
The annual consumption of northern shrimp by Pacific halibut was estimated as 1 300 tonnes for 1990 and 1 100 tonnes for 1991. The consumption of shrimp by redfish was estimated to be 33 600 tonnes in 1990 and 8 700 tonnes in 1991. The nominal catch of shrimp in the two years was about 52 000 tonnes and 58 000 tonnes respectively.
Attempts have been made to incorporate the effects of predation in fisheries models. Extensions to single species assessments take into account the effects of cannibalism and the repercussions for stock size, recruitment, yield and management strategies. Single species models can also be extended into simple multi-species models. In an example of the interactions between Norway lobster and cod in the Irish sea, where these are the two most valuable species and where cod accounts for an estimated 88% of all predation on the lobster, it was concluded that cod exploitation should be kept as high as possible (while taking care to avoid declining recruitment) to reduce predation and increase lobster yield.
Other multi-species models are more complex. The ICES approach consists of two models. The first is MSVPA (Multi-Species Virtual Population Analysis), which is used to estimate the past stock numbers, the past predation mortality coefficients and the parameters to compute them, and the past fishing mortality coefficients. It takes as its primary inputs the total numbers caught by age group by all fishing fleets, the food consumption rates and prey preferences of the predatory fish, and individual weights by age for all the species. The second model is MSFOR (Multi-Species Forecast), which is used to predict future yields, stock numbers and biomass, and future predation mortality coefficients. Its inputs are the outputs from the MSVPA (stock numbers, predation and other fixed parameters), estimated or assumed recruitment, and assumed future fishing mortalities.
Two case studies of the application of multi-species models are given. In the example of hake and anchovy stock in the northwest Mediterranean, the conclusion favoured management that maximized yields of the predator, hake. This is largely due to the higher price that hake commands. Moreover the incorporation of a predator-prey interaction into the assessment did not provide any further insights beyond that available from a single species assessment. The important results were from the single species VPA applied to hake, particularly in respect to quantifying the extent of potential benefit from having an age at first capture of 4 years (equivalent to a length of about 40 cm).
Owing to the complexity of multi-species ecosystems, harvesting strategies should not be oversimplified. The idea that the total production of a system can be increased by fishing down the predators, and then harvesting the more abundant prey lower down in the food web, does not seem to work in practice.
As fishing increases on the apex predator, its role as a regulating agent on the abundance of the lower trophic species is replaced, at least in part, by the next predator down; the older individuals of which are able to move into a higher trophic level than formerly. The extent to which this happens is increased when there are many competing predators on the same group of prey. Furthermore, as the natural effect of predators is to regulate the abundance of the prey, the substantive and selective removal of predators will be de- stabilizing, as reflected by increased variation in the abundance of the prey. Selective removal of the prey, on the other hand, will be detrimental to the predators, although less so if alternative prey are available.
The economics of predator-prey harvesting also need to be assessed. Flaaten (1989) combined a simple biological model involving a single predator and a single prey from May et al. (1979), with economically independent harvesting sectors, one for each of the two species. As expected, harvesting the predator in this very simple situation increased the harvestable stock of the prey, and harvesting the prey reduced the harvestable stock of the predator.
In the case where the prey is inexpensive-to-catch and valuable, and the predator is expensive-to-catch trash, he concluded that it may be economically justified to subsidise the harvesting of the predator.
In the reverse, when the predator is inexpensive-to-catch and valuable, and the prey is expensive-to-catch and trash, the increase in the harvestable stock of the prey resulting from the harvesting of the predator, may make it economic to harvest the prey when previously it was not. This would occur, however, at the expense of the fishery on the predator.
Examples are provided of predation effects being incorporated into management advice for marine fisheries. In the Northeast and Northwest Atlantic, capelin catches are restricted to leave enough as food for the higher valued cod. In contrast, anchovy and pilchard continue to be heavily exploited in the Southeast Atlantic even though they are valuable forage fish for the higher value hake and other predators. Other examples include harvesting of the South African fur seal, and culling of grey seals in Scotland.
The policy and institutional implications of multi-species management need to be examined. Three problems, in particular, are highlighted: fishermen restricted (as through licensing) to exploiting certain species may become disadvantaged; fishery resources are often shared with other apical predators (whales, seals and birds); and institutional problems arise when species come under the jurisdiction of several independent agencies.
Nevertheless, there is certain to be an increasing trend toward studying and managing fish stocks in a multi-species context, and in some cases toward managing the ecosystem where they belong. In multi- species fisheries, the search for economic viability will continue to be the dominant influence on the strategies used to harvest predator- prey stocks.
When the predators have a much higher per unit price than the prey, as is usual, the harvesting strategies will give preference to sustainable exploitation of the predators, with the yields from the prey stocks being of lesser importance.
In conclusion, research to support multi-species management is necessary, but will not be equally relevant across all fisheries. Research will depend on how well established species interactions are, its cost effectiveness in complex ecosystems, and appropriateness in cases where single species assessments might be adequate. Whatever emphasis is given to the research, it will be essential to model the effects in its full bio-economic context. At least some minimum understanding of the species interactions and the abiotic factors will be needed. As appropriate, this should include a fuller understanding of predator-prey relationships, natural climatic processes, ecological responses, and the impacts of exploitation and management.