commentary Turning back the harmful red tide

Harmful algal blooms are a serious and increasing problem in marine waters, yet scientists and funding agencies have been slow to investigate possible control strategies.

recent proliferation of harmful blooms1. Donald M. Anderson There is debate about the nature and causes Each holiday season I await the visit of one of this expansion. Some call it a global epi- relative with trepidation. Years ago he asked demic linked to pollution and human 2 whether I had “stopped that red tide problem changes to coastal ecosystems . Others argue S. FRANKS, SIO J. P. yet?” — a simple question from one con- that the expansion is in part an artefact vinced that science solves problems directed reflecting increases in the number of scien- to a so-called expert on the destructive and tists, advances in toxin detection, and the often visible ‘blooms’ of phytoplankton that proliferation of aquaculture and other activ- kill fish, make shellfish poisonous and cause ities requiring product monitoring1,3. numerous other problems in coastal waters. One thing is certain — there is a growing I explained that we did not understand the global problem at a time when human causative organisms, their ecology or reliance on coastal zones for food, recreation oceanography well enough to propose con- and commerce is rapidly expanding. Never- trol strategies, but that one day we would. theless, there is practically no exploration of Although temporarily satisfied with my direct control of marine blooms — attempt- argument, each year thereafter my brother- ing to kill or remove the cells or reduce their in-law repeated the question, and each year toxicity. At an international conference on my answer was the same. Whatever progress harmful algae held in Vigo, Spain, in June, had been made, there were new questions to only one contribution of more than 400 be addressed. Eventually, he concluded that I abstracts from 58 countries addressed direct did not want to solve the problem, as that control of marine blooms. Imagine the dif- would end my research programme. He is ference if the conference had been on agri- wrong, of course, but the explanation is far cultural pests or on algal blooms in fresh more complex than he would think, and is in water, where control efforts are common. Some red tides, such as this non-toxic bloom of part the subject of this article. Research efforts on mitigation strategies Noctiluca off California, cover huge areas, Throughout history, blooms of micro- such as shellfish-monitoring and aquacul- making it difficult to foresee environmentally scopic algae have had a major impact on fish, ture site management are critically impor- benign bloom-control strategies (see also birds, mammals and other organisms in the tant, but they treat the symptoms without http://www.redtide.whoi.edu/hab/). marine food web. These ‘red tides’ (now attacking the problem. Government officials termed harmful algal blooms) take many and the public want to know what is being To replace dwindling natural fishery forms and have many effects. Some toxic done, or what can be done, in terms of direct resources, we are turning coastal waters and algae kill wild and farmed fish. Others pro- intervention. So far, we have little to offer wetlands into marine farms at an extraordi- duce potent neurotoxins that accumulate in other than tentative predictions of bloom nary rate. Whether by fish or shrimp farms filter-feeding shellfish and poison human reductions decades from now if nutrient (which have been likened to small cities with consumers. Algal toxins can alter the struc- loadings are reduced. respect to their production of organic matter ture and function of marine ecosystems, I believe that some harmful algal blooms as pollutants), or by shellfish or seaweed cul- affecting fecundity and survival at all levels. can be controlled or managed, not 20 years ture (which strip plankton and nutrients Even non-toxic algae can be harmful from now, but in the near future, economi- from the water), we are altering near-shore when they accumulate in sufficient numbers cally and without disastrous environmental waters significantly. Coastal ecosystems are — sometimes millions of cells per litre — to consequences. This belief may brand me as a no longer pristine and will not revert to their discolour the water, shade submerged vege- heretic among my colleagues, some of whom ‘natural’ state without intervention. tation, disrupt food-web dynamics and fear that the ocean will be further despoiled cause oxygen depletion. At the other by inept human attempts to manipulate From land to ocean extreme, toxic algae can be a tiny fraction of ecosystems we do not understand. Distrust of our ability to control pests and the total phytoplankton population and still At the heart of this negativism is a convic- diseases seems to be based more on pes- be dangerous. Diarrhetic shellfish poison- tion that mankind does not possess the skills, simism than on fact. When biological con- ing, for example, has been reported with knowledge or right to manipulate the marine trol is discussed, for example, some are quick Dinophysis concentrations of a few hundred environment on any significant scale. We to point out failures such as the introduction cells per litre. are, however, already doing exactly that. By of the mongoose to oceanic islands or the The scale and timing of harmful algal polluting coastal waters, we change the giant toad to Australia4. Obscured by these blooms is highly variable. Some are local- abundance and relative amounts of critical failures is a multitude of successes in terres- ized, occurring in bays or estuaries; others plant nutrients, which in turn alters the trial biocontrol of weeds and pests4. Overall, are massive, covering thousands of square species composition of planktonic ecosys- 165 insect pests and 35 weed species have kilometres. Some occur at the same time and tems. Indeed, this may be why there is an been controlled. Less than 2 per cent of the place each year; others strike in random fash- increasing number of harmful algal blooms. introductions became pests themselves, and ion. Some last a few weeks, others years. We are harvesting fish and shellfish at an many of those were ‘generalist’ predators — Harmful algal blooms are not new phe- alarming rate, removing components of the an approach that is no longer practised. nomena. Red tides are recorded in the Bible food-web with little knowledge of how such Other examples of terrestrial manage- and in the fossil record. What is new is the enormous manipulations affect other levels. ment strategies include integrated pest

NATURE VOL 388 7 AUGUST 1997 513 | | Nature © Macmillan Publishers Ltd 1997 commentary management, which combines biological trial biological control. It is not clear whether Viruses are abundant in marine systems, control with chemical agents such as this new programme will support research replicate rapidly and tend to be host- narrow-spectrum pesticides, and ecological- on control of harmful algal blooms. specific, suggesting that a single algal species ly based pest management, which attempts During a red tide in Florida 40 years ago, could be targeted11. Parasites12 also have to work with ecosystem processes in the copper sulphate was applied to 10,000 acres potential to control algal bloom species, but management effort5. The conceptual frame- of shoreline using crop-dusting planes7. The specificity is largely unknown. There are work for pest management on land is far treatment was initially effective, but blooms numerous examples of bacterial strains13 advanced, and should be a valuable resource reappeared within weeks. Copper was exhibiting strong and specific algicidal activ- in planning the management of marine sys- deemed too expensive and non-specific to be ity, although no field applications have yet tems. Instead it is largely ignored and misun- used other than for short-term, small-scale been attempted. derstood. bloom control. In another study, 4,700 chem- Extrapolation from land to the ocean will icals were screened for use against Florida’s Prognosis for the future admittedly be difficult, as marine and terres- red-tide alga, but none was sufficiently I have mentioned only a few of many poten- trial systems differ in scale, complexity and potent in natural sea water without also hav- tial control strategies. We must cautiously dynamics6. Application of a control agent to ing adverse effects on other organisms. explore all possible approaches, but this a single crop on a parcel of land is certainly Since then, chemical control options have requires funding at the scale needed to pro- simpler than the marine equivalent, where received little attention, and no significant vide data to support informed decisions and water motions will change the distribution bloom-control research has been undertak- override our preconceptions. We also need and abundance of target organisms and en in the United States. Japan, China and to establish guidelines for acceptable marine applied control agents. Control of an out- Korea are exploring control strategies treatments. break at one site may have no effect on because they ‘farm’ their coastal waters heav- In one sense, the problems we face with blooms in later years at the same location6. ily through aquaculture. Faced with signifi- harmful algal blooms are similar to those Another difference is that the community of cant economic losses from red tides, Japan encountered in agriculture or medicine, organisms in marine ecosystems is more initiated a broad evaluation of bloom- fields in which control of pests and diseases is diverse and complex than that in single-crop control strategies in the mid-1970s. Much of a practical reality. The marine environment agricultural systems. our knowledge of possible approaches is admittedly different, but our hesitancy Yet another factor that has stalled comes from this outdated but useful series of reflects a de facto acceptance that the prob- progress is the tendency to generalize that all studies9, which continues to this day, but at a lems are insurmountable. I believe they are blooms are massive. One colleague argues much-reduced level of effort. not, and that we can make progress if new that blooms, like tornadoes or hurricanes, One promising strategy treats blooms resources are made available and if we learn can be tracked and their movements predict- with flocculents such as clay that scavenge from the mistakes and successes of more ed, but never controlled. He may be right particles, including algal cells, from sea water than 100 years of experience controlling ter- about the larger blooms, but many are small and carry them to bottom sediments. restrial pests. or localized, either permanently or during Removal efficiencies of 95–99 per cent have There is a thin line to walk here — to argue key stages of development. For example, been achieved in laboratory cultures using that we have been too cautious and that suc- destructive brown tides in Texas or New York clay, and small-scale field trials near fish cess is possible, without promising more are thought to begin in certain bays and then farms have also been successful, though than we can deliver, unrealistically raising the to spread to adjacent waters. A widespread expensive8. New clays that are an order of expectations of the public and politicians. I coastal bloom might be localized and acces- magnitude more efficient in cell removal also worry that funds needed for ecological sible at an earlier stage. have been tested in China; this capacity can studies might be diverted to control. I see the Another constraint is that most countries be further increased using coagulants such as risks, but I also see the chance that this article have no official policy for funding marine polyhydroxy aluminium chloride9. may initiate a debate that will ultimately pest management. In the United States, the Applications in China have been limited direct scientific thought and resources Department of Agriculture puts extensive to tests in shrimp aquaculture ponds but, in towards practical solutions. My brother-in- resources into terrestrial pest management. 1996, 60,000 tons of clay were used in Korea law would no doubt approve. By contrast, the equivalent agency responsi- to control a Cochlodinium red tide threaten- Donald M. Anderson is at the Woods Hole ble for the oceans, the National Oceanic and ing near-shore fish farms. About 100 km2 Oceanographic Institution, Woods Hole, Atmospheric Administration, has no marine were treated over several weeks, and nearly Massachusetts 02543, USA. pest-control programmes. Lacking strong 100,000 tonnes of clay are now stockpiled in e-mail: [email protected] leadership or targeted funding programmes anticipation of the next bloom. 1. Anderson, D. M. in Red Tides: Biology, Environmental Science at the national level, scientists opt for Clay is a non-toxic, naturally occurring and Toxicology (eds Okaichi, T., Anderson, D. M. & Nemoto, research on fundamental ecological or material. Fish and bottom-dwelling organ- T.) 11–16 (Elsevier, New York, 1989). 2. Smayda, T. J. in Novel Phytoplankton Blooms (eds Cosper, E. M. oceanographic issues less likely to be rejected isms are unaffected by extremely high clay et al.) 213–228 (Springer, New York, 1989). 10 during peer-review. loadings near pottery industries . The 3. Hallegraeff, G. M. Phycologia 32, 79–99 (1993). prospects look good, but considerable 4. Greathead, D. J. in Biological Control: Benefits and Risks (eds Control options research is needed before large-scale field Hokkanen, H. M. T. & Lynch, J. M.) (Cambridge Univ. Press, 1995). But there are signs of change in at least some applications can be attempted. Obvious con- 5. Ecologically based Pest Management: New Solutions for a New parts of the world. South Korea has estab- cerns are the fate and effects of sedimented Century (National Academy, Washington, DC, 1996). lished a harmful algal bloom engineering cells and toxins on bottom-dwelling ani- 6. Lafferty, K. D. & Kuris, A. M. Ecology 77, 1989–2000 (1996). division at its National Fisheries Research mals, and the collateral mortality of co- 7. Rounsefell, G. A. & Evans, J. E. US Fish Wildlife Serv. Spec. Sci. Rep. 270 (1958). and Development Institute. And Australia’s occurring planktonic organisms. 8. Shirota, A. Int. J. Aquacult. Fisheries Technol. 1, 195 (1989). Commonwealth Scientific and Industrial Biological control of harmful algal 9. Yu, Z., Zou, J. Z. & Ma, X. Oceanologia et Limnologia Sinica 25, Research Organization has established a blooms also has potential. Zooplankton that 226–232 (1994). Centre for Research on Introduced Marine graze on bloom species have been proposed 10.Howell, B. R. & Shelton, R. G. J. J. Mar. Biol. Ass. UK 50, Pests, which plans a proactive approach to as control agents8 but remain untested 593–607 (1970). 11.Milligan, K. L. D. & Cosper, E. M. Science 266, 805 (1994). marine pest management consistent with because of the impracticality of growing and 12.Taylor, F. J. R. J. Fish. Res. Board. Can. 25, 2241–2245 (1968). the country’s aggressive reliance on terres- maintaining predators in sufficient quantity. 13.Yoshinaga, I., Kawai, T. & Ishida, Y. Fish. Sci. 63, 94 (1997).

NATURE VOL 388 7 AUGUST 1997 514 Nature © Macmillan Publishers Ltd 1997 | |