Apparent changes in the trophic composition of world marine harvests: the perspective from the FAO capture database

Abstract Almost 50 yr of global multispecies harvests are represented in the FAO capture production database, and offer a broad perspective on events that underlie the major observed changes in global marine harvests. The likely relevance of top–down and bottom–up trophic interactions, versus the impacts of changes in fishing technology and markets on the trophic level of landings, are discussed on a regional basis. Despite the low resolution of this global data set, several common features emerge. Rapid increases in fleet size and technological advance, and imperfect fisheries management measures, are probably responsible for declines in peak multispecies production in many areas since the 1970s, and internal evidence suggests that peak production is not far away in the remainder. Staggered dates of peak landings in different world areas seem to reflect the spread of industrial fishing fleets from `core areas’ to the rest of the world's oceans, which largely took place from the 1960s to 1990s. A general move to higher exploitation of piscivores in global landings is implied in some areas. The hypothesis that top–down removal of predators is affecting lower trophic production is however only one explanation for declining mean trophic levels, and is more likely to emerge from local food web studies. In some regions an increased proportion of short-lived invertebrates in harvests later in the time series supports a move downwards in trophic level targeting. In the North Atlantic and some other areas, fishing down marine food webs may be a likely cause of the increase in landings of shelf planktivores. A shift from depleted apical resources to species lower in the food web may have been made on economic grounds however, independent of possible effects of a release of predatory pressure. Sharp increases in planktivores later in the time series show up in the Eastern Central Atlantic and in the Southeast Pacific and do not appear to be primarily related to depletion of predators. They seem to reflect intermittent strength of upwelling systems, and hence bottom–up effects on food web production, together with changes in harvest technology. For some areas, piscivore landings increased later in the time series than those of planktivores. In the Mediterranean, this seems mainly a bottom–up response to increased marine productivity associated with land run-off and consequent improvements in predatory foraging. In some tropical areas, later increases in piscivorous landings mainly result from expansion of distant water tuna fleets, and are probably unrelated to exploitation of forage fishes. The technological revolution of the 1950s and 1960s involved rapid application of synthetic fibres to improved gear. This led to large-scale mid-water trawling and purse seining by industrial fleets and has especially increased vulnerability of small pelagic stocks over the last few decades. This appears mainly responsible for the apparent decline in mean trophic level of harvests in areas with large stocks of these resources. From an analysis of variances of the sample data set, `Punctuated equilibrium', involving actual changes in ecosystems, rather than just continuous change in the relative harvest rates of species in a given ecosystem, is suggested as an important phenomenon, reflecting both ecological change and changing exploitation strategies.