Red Tides

Oil Spills

The highest profile of all potential impacts on rocky shores is achieved by those short-term ‘environmental disasters’ with highly visible and media-friendly effects. Such factors may not necessarily be those most important in the long term but where impacts are dramatic, shores may take years to recover.

Red tides are blooms of toxic dinoflagellates and other algae which can cause the widespread mortality of marine and littoral animals. Although red tides occur naturally, there is increasing evidence that they can also result from human activities. Eutrophication of coastal waters may have caused the severe blooms of Gyrodinium aureolum, Chrysochromulina polyepis and Craetium spp. which devastated North Sea shores in 1988 (Smayda, 1989, 1990; Lindahl, 1993). The main rocky shore animals affected by these events were barnacles, dogwhelks, mussels and limpets which, as space occupiers, grazers and predators, have a strong effect on the community ecology (e.g. Southgate et al., 1984).

In recent years, British shores have been affected by two major oil spills. The Braer spilled 85,000 tons of oil near Shetland in 1995. The Sea Empress ran aground off South Wales in February 1996, releasing 72,000 tons of oil which had disastrous effects on ecosystems along more than 100km of coastline. A more severe impact was caused by the Torrey Canyon oil spill in March 1967. Of the 119,000 tons of Kuwait crude on board, 100,000 tons were spilt and 40,000 tons came ashore. 14,000 tons, borne on the highest spring tides for half a century, were stranded on the Cornish coastline and 10,000 tons of dispersant were used in the clean-up operation. The effects of this dispersant on marine life were little understood at the time. However, it later became apparent that the effects of the dispersant were more severe than those of the oil itself. The Torrey Canyon oil spill provides an excellent case study of the effects of acute pollution on the rocky littoral thanks to the long term observations on affected shores which are reported in Southward and Southward (1978, Hawkins et al. (1983) and Hawkins and Southward (1992), and summarised in Raffaelli and Hawkins (1996).

The first-generation dispersants available in 1967 proved highly toxic to marine life. Subsequent laboratory studies showed that the concentration required to kill 50% of intertidal organisms in 24 hours of exposure was between 5 and 100 ppm, with limpets in the genus Patella being highly susceptible. These lethal concentrations were much lower than those needed to disperse the oil, and consequently all animals and many algae were killed in areas of the shore close to dispersant spraying.

The effects on rocky shores of the removal of much of the biota were profound and long-lasting. The loss of Patella had a special significance in this respect, since this grazer is a key species in the north-east Atlantic, responsible for structuring midshore communities on moderately exposed and exposed rocky shores.

The time course of recolonization of rocky shores in Cornwall, expressed in years from the date of the Torrey Canyon disaster, March 1967 (extracted from Hawkins and Southward, 1992).

The table above, based on Southward and Southward (1978), summarizes information on the patterns of recolonization of various rocky shores between 1967 and 1977. The similarity of the overall pattern allows a generalized account of the course of recolonization on the midshore region. Following the death of grazing animals, a dense flush of ephemeral green algae (Enteromorpha, Blidingia, Ulva) appeared which lasted up to one year. After six months or so, large brown fucoid algae (mainly Fucus vesiculosus and F. serratus) began to colonize the shores. Very few animals were present under these dense growths of algae. Any surviving barnacles were overgrown and eventually died, whilst the dense canopy prevented subsequent recruitment of barnacles by the sweeping action of the Fucus fronds and larval barrier effects. This was probably reinforced by the dense population of the predatory dogwhelk Nucella which built up under the canopy so that barnacles declined to a minimum on most shores between 1969 and 1971.

The limpet Patella vulgata first recolonized the shores during the early winter of 1967-8 and survived well in the damp conditions under the extensive Fucus canopy, preventing subsequent recruitment of Fucus by their grazing activities. As the plants aged, grazing of the holdfasts reduced Fucus further and between 1971 and 1975 the shore became very bare with even fewer algae than before the spill.

With the disappearance of Fucus, the abnormally dense population of limpets abandoned their normal homing habit and migrated in fronts across the shore. Barnacle numbers increased on all shores once the fucoids declined. Following the bare period between 1974 and 1978, the shore went through a phase of increased Fucus cover, although overall cover never exceeded 40%. Limpet density increased with some fluctuations from 1975 before dropping in the early 1980s and then rising again, probably reflecting normal levels of spatial and temporal variation typical of limpet populations. During the period of dominance by the initial colonizing cohorts recruitment of limpets was low (e.g., 1969 - 1972) - presumably due to intense inter-age-class competition. Subsequently, recruitment improved, and after 1988 the population generally had 60-70% juvenile limpets under the length of 15mm.

After the initial decline following the spill, the barnacle population slowly built up at all shore levels. At a shore level equivalent to high water of neap tides, all counts from 1979 onwards were within one standard deviation of the pre-spill mean, although only one value (in 1990) exceeded the mean. At mid-tide level, a similarly irregular pattern was observed, with exceptionally high counts just after the bare phase in 1976. Lower on the shore, there was much greater fluctuation which probably reflected the greater influence of biological interactions, including predation and competition with spasmodic settlements of Mytilus, and this may account for the drop observed in the late 1980s.

Before recovery can be assessed, the unaffected condition must be considered. The eulittoral of exposed shores of Devon and Cornwall is normally characterised by small-scale spatial and temporal fluctuations in the major components of fucoids, barnacles and limpets. Isolated patches of Fucus occur, but they are never more than clumps of a few plants and total cover rarely exceeds 10 to 20%. The patchiness and fluctuations are partly generated by variation in recruitment and small-scale differences in microhabitat, predation and physical disturbance. Therefore, "recovery" can be defined as a return to levels of spatial and temporal variation seen on unaffected shores.

After the initial massive increase in Fucus, and a similar but aphasic increase in the key herbivore Patella, subsequent fluctuations have been much smaller. Fucus cover was clearly abnormal for the first 11 years, and was perhaps slightly elevated in the early 1980's, before fluctuating around normal levels after 15 years or so. The abundance and population structure of Patella vulgata were clearly abnormal for at least 10 years, probably 13.

The time scale for recovery at these sites seems to be at least 10 years. If limpet population structure and barnacle densities are used as criteria then 15 years may be more realistic. These time scales are not surprising when the long life spans of the main organisms are considered: Fucus 4-5 years, Patella up to 20 years, but usually < 10 years, and Chthamalus at least 5 years and possibly 20. If we estimate that the average life span of a limpet is 7-10 years it is highly likely that population structure will take 15 years or so to stabilize.

The time scale for recovery is clearly much longer than was thought by many ecologists in the early 1970s. Dense growths of seaweeds were seen as a sign of recovery, rather than of a highly disturbed community, and there were suggestions that the system had returned to normal within two years. Even the more pessimistic considered that only a few more years were needed for a complete return to normal, but Southward and Southward (1978) rightly dismissed optimistic forecasts and the myth of rapid recovery. At that time, they could only assert that some shores heavily treated with dispersants had not returned to normal after 10 years, whilst many had taken at least 5-8 years. It is now clear that it may take 15 years or so for the worst affected shores to recover. In contrast, recovery at the only shore where oil was substantially untreated because of its proximity to a seal breeding area (Godrevy) was rapid and almost complete within three years.

It was very quickly learnt from this incident that large-scale use of dispersants causes acute toxic effects. In the few weeks taken for the oil to cross the English Channel, a very different approach was adopted by the French and the Channel Islanders - manual removal or the use of suction devices with dispersants applied sparingly. These lessons were absorbed by those in charge of responding to the Santa Barbara blow-out in 1969 and subsequent spills such as the Amoco Cadiz in 1978, the Braer in 1995 and the Sea Empress in 1996. Impetus was also given to the development of less toxic dispersants (NAS 1989) and to physical dispersal and mechanical collection. In most instances, manual methods (whether removal of oiled plants or use of absorbent materials) seem to cause less disturbance than that associated with the trampling and movement of equipment, vehicles and vessels during mechanized operations.

During recent spills, there have been trials of methods which enhance the microbial breakdown of oil (see Swannell et al., 1996 and Mohammed et al., 1996 for reviews). These methods have proved successful although operational limits on their use have not been fully defined. Considerable hope has been raised by the possibility of enhancing natural biodegradation "bioremediation" by adding nutrients which limit bacterial growth to the oil, particularly in oleophilic media. Laboratory tests and field trials have been encouraging and we should be in a better position to judge once their effectiveness during the Exxon Valdez and their limited use in the 1991 Gulf War clean-ups has been fully evaluated in the long term (see Hoff, 1993). Current opinion on the effectiveness of bioremediation varies from over optimism to extreme scepticism. The results that are emerging from the applications in Alaska after the Exxon Valdez spill are often conflicting. Concerns have also been expressed about eutrophication caused by bioremediation and toxic side effects of some preparations. However, with further field trials and experience of appropriate use bioremediation still has considerable promise for the future. Evidence has accumulated from other coastal systems of the length of time needed for recovery from oil spills. Work in Panama in a deep mud associated with fringing mangroves (Burns et al, 1993) has suggested that 20 years or more is required due to the long term persistence of oil trapped in anoxic sediments and subsequent release into the water column.

Clearly, the Torrey Canyon incident was an early example of a major ecological disaster made much worse by an inappropriate response. A more considered approach to spills has emerged and more sensible procedures have evolved which were implemented during the Exxon Valdez spill. The least expensive and most ecologically sound option for restoring exposed and moderately exposed shores covered by oil is probably to do nothing, but, as Foster et al. (1990) point out, during an environmental crisis social pressure to "do something", and the political need to be seen to be doing something, often outweigh ecological considerations.

Next Section                         References