.
Ciguatera (Fish Poisoning), El Niño, and Pacific Sea Surface Temperatures Simon Hales, Phil Weinstein, and Alistair Woodward Department of Public Health, Wellington School of Medicine, University of Otago, Wellington, New Zealand
ABSTRACT Ciguatera (fish poisoning) is the most frequent cause of human illness caused by ingestion of marine toxins. The toxins are ingested by herbivorous fish which feed on marine algae and are then passed up the food chain to humans eating reef fish. We have used a unique database of reported fish poisoning cases in the South Pacific islands to investigate the relationship between fish poisoning and El Niño Southern Oscillation, a periodic disruption of global climate that is associated with marked changes in rainfall, temperatures, and ocean currents. This provides an opportunity to study the effects of a strong climate signal on a sensitive ecosystem in a region that has been less subject to local human disturbance than most others. Using a mixed ecological study design, we calculated correlations between reports of fish poisoning in individual Pacific Islands, estimates of local sea
surface temperature, and the Southern Oscillation Index. Strong positive correlations between the annual incidence of fish poisoning and local warming of the sea surface were found in a group of islands which experience warming during El Niño conditions. In another group of islands, which experience cooling of the sea surface during El Niño events, there were weaker negative correlations between fish poisoning and local sea surface temperature. The results are consistent with other evidence suggesting a close interdependence of marine ecosystems and climate. Increases in ciguatera may result if the climate continues to warm as a result of the enhanced greenhouse effect. Coral reefs have been under increasing pressure from human populations in recent years; ciguatera may be a sensitive indicator of environmental disturbance in tropical marine ecosystems.
INTRODUCTION
Herbivorous reef fish become contaminated when feeding on the macroalgae. The toxins become more concentrated as they move up the food chain, as when herbivorous fish are preyed on by larger carnivores including barracuda, snapper, grouper, and moray eel (Baden et al. 1995). There has been much speculation about a role for climate factors in ciguatera and other diseases caused by harmful marine algae, but little quantitative evidence to date (Hallegraeff 1993; de Sylva 1994; Tibbetts 1998). We have used a database of reported fish poisoning cases in the Pacific Islands to investigate the relationship between fish poisoning and El Niño Southern Oscillation (ENSO). ENSO is a recurring climate pattern, centered on the tropical Pacific, that causes marked year-to-year variation
Ciguatera (fish poisoning) is the most frequent cause of human illness caused by ingestion of marine toxins (Baden et al. 1995) and has substantial adverse health impacts in Pacific Island countries that rely on fish as a major protein source (Lewis 1992). The disease is caused by the ingestion of reef fish that have been contaminated by ciguatoxins. Ciguatoxins are produced by marine dinoflagellates, (particularly Gambierdiscus toxicus) that live on the surfaces of marine macroalgae.
Address correspondence to: S. Hales, Department of Public Health, Wellington School of Medicine, University of Otago, PO Box 7343, Mein St., Wellington, New Zealand; E-mail
[email protected].
©1999 Blackwell Science, Inc.
21
in rainfall, sea surface temperature (SST), and ocean currents. ENSO is a semiregular cycle with two extremes, El Niño and La Niña. The state of the ENSO cycle is indicated by the Southern Oscillation Index (SOI); during an El Niño event, the SOI is negative, whilst during La Niña conditions the SOI is positive. El Niño is associated with strong warming of the sea surface in the eastern equatorial Pacific (Figure 1). In the years covered by this study, the Islands of the South Pacific have been relatively unaffected by local social and economic disturbances that influence disease rates. Therefore, the region may be an ideal location to detect the first biological effects of global warming (Tester 1994).
MATERIALS AND METHODS We examined correlations between reports of fish poisoning in individual South Pacific islands, estimates of local SST anomalies, and the SOI. Incidence data for fish poisoning were obtained from the South Pacific Epidemiological and Health Information Service (SPEHIS, Secretariat of the Pacific Community, Noumea, New Caledonia: 1973–
1994). A total of 22 islands reported to the SPEHIS during the period of the study. Ciguatera was rarely reported by 11 islands. In order to ensure that the islands included in the study were as similar as possible (apart from exposure to interannual climate variations) three islands in the northern hemisphere were excluded. The SOI data were obtained from NASA: http:// daac.gsfc.nasa.gov/CAMPAIGN_DOCS/FTP_ SITE/ INT_DIS/readmes/soi.html Time series of local SST anomalies were estimated using the International Research Institute for Climate Prediction (IRI) data library to access observed SST data: http://ingrid.ldgo.columbia. edu/expert/SOURCES/.CAC For each island group, appropriate latitude (Y–Y9) and longitude (X–X9) ranges were read from a map and entered into the data selection fields to the nearest degree. The time (T) grid was limited to the period January 1973 to December 1994 by the same method. The data were averaged over all resulting (X, Y) grid points, using the “filters” field, to produce monthly estimates. The resulting output was downloaded separately for each island group. Pearson correlations were calculated using January–December annual aver-
FIGURE 1. Effect of El Niño on sea surface temperature in the South Pacific. The light area is warmer than usual and shaded area cooler than usual during El Niño events. Hales et al.: Fish Poisoning and El Niño
22
soning and local SST. The effect of ENSO on SST depends upon the location. SST anomalies were negatively correlated with SOI in five islands (indicating that in these islands, warming occurs during El Niño events). In the remaining three islands, SST anomalies were positively correlated with SOI—indicating cooler temperatures during El Niño events.
TABLE 1 Pearson correlations between the Southern Oscillation Index and fish poisoning; Islands of the South Pacific, 1973–1994
20.50 20.19 20.45 20.38 0.19 20.24 20.28 0.06
Tuvalu Rarotonga Kiribati Western Samoa French Polynesia Fiji Vanuatu New Caledonia
CORRELATIONS BETWEEN FISH POISONING AND LOCAL SST ANOMALIES The results fall into two groups. In the eastern group of islands, all of which experience local warming during El Niño events, there were positive correlations between fish poisoning and local SST anomalies; the correlations were statistically significant in four islands. In each of these, there was an approximately linear relationship between fish poisoning and SST (Figure 2). The eastern group included three islands with the highest per capita rates of reported fish poisoning. In the westerly group of islands (which experience local cooling during El Niño events), there were negative correlations between fish poisoning and local SST. When we plotted time series of SST for geographical regions encompassing each group (180–220E, 0–20S and 140–180E, 10–25S), we found that maximum temperatures were higher in the former (easterly) group (Figure 3). We reran the analyses, this time including data from the previously excluded islands. This did not alter the pattern of the results.
ages. Average reporting rates were calculated based on 1994 populations (data from the Secretariat of the Pacific Community population/demography program).
RESULTS Table 1 shows correlations between average annual reports of fish poisoning and the average annual SOI. In most islands the correlation is negative, indicating a tendency for fish poisoning to occur during El Niño conditions. Table 2 shows correlations between local SST and SOI and between the incidence of fish poi-
TABLE 2 Pearson correlations between sea surface temperature, the Southern Oscillation Index, and fish poisoning; per capita reporting rates, Islands of the South Pacific, 1973–1994 Local SST vs. SOI
Tuvalu Rarotonga Kiribati Western Samoa French Polynesia Fiji Vanuatu New Caledonia
Fish Poisoning vs. Local SST ( p value)
20.76 20.51 20.93 20.53 20.81 0.70 0.74 0.78
Approximate Rate (cases/1000 population)
0.65 (0.001) 0.61 (0.003) 0.54 (0.01) 0.49 (0.02) 0.13 (0.58)
5 4 5 ,1 3
20.05 (0.81) 20.17 (0.46) 20.24 (0.27)
Ecosystem Health
Vol. 5
No. 1
,1 ,1 ,1
March 1999
23
FIGURE 2. Relationship between local sea surface temperature anomalies and ciguatera for four islands in which a statistically significant relationship was found.
DISCUSSION Annual reports of fish poisoning were found to be correlated with El Niño events. Positive correlations between the annual incidence of fish poisoning and local SST anomalies were confined to the easterly group of islands that experience warming during El Niño conditions. Here, in four out of five islands, the relationship between fish poisoning and local SST was moderately strong, statistically significant, and approximately linear. Maximum sea surface temperatures and approximate per capita rates of fish poisoning were also appreciably higher in this group. Because population figures are not available for all islands for the whole period of the study, we based our estimates of rates of fish poisoning on 1994 populations. If population trends differed between islands this could have biased these estimates. With the exception of Western Samoa, demographic trends were probably broadly similar and are unlikely to explain much of the large differences in rates of fish poisoning. Random misclassification of the data, for example from underreporting, is likely to have bi-
ased the results toward null. Conversely, serial correlation of the data may have increased the apparent statistical significance of the results. An interrelationship between marine ecosystems and climate is biologically plausible, with a range of possible mechanisms (McGowan et al. 1998; Lenton 1998). Fish poisoning is known to be related to physical disturbance of coral reefs and there is evidence that coral ecosystems are sensitive to a range of environmental factors, including increased sea temperatures (Goreau 1990; Glynn 1993; Gleeson & Strong 1995; Brown 1997). The surfaces of dead coral are suitable for colonization by macroalgae and (subsequently) ciguatoxigenic organisms (de Sylva 1994). The organisms responsible for producing the toxins are also known to be temperature sensitive. Warmer conditions lead to increases in the growth of dinoflagellates and feeding behavior in fishes, which would increase the potential for toxins to accumulate at the top of the food chain (Bomber et al. 1988; Morton et al. 1992). Several further studies are possible to clarify the mechanisms involved. It will be important to
Hales et al.: Fish Poisoning and El Niño
24
FIGURE 3. Time series of average sea surface temperature in two regions of the South Pacific. Eastern region (E): averaged for the region 0–20S 180–220E. Western region (W): averaged for the region 10–25S 160–180E. Data from http://ingrid. ldgo.columbia.edu/expert/SOURCES/.CAC
ACKNOWLEDGMENTS
investigate further the reasons for the different response to local SST during El Niño and La Niña conditions. Tuvalu, Rarotonga, and Kiribati are all relatively isolated atoll countries, with high per capita quantities of fish caught, and a high percentage of these fish usually being from the local reef (Dalzell 1992; Lewis 1992). Negative correlations between local SST and ciguatera could result from importation of fish by westerly islands such as Fiji, Vanuatu, and New Caledonia, or migration of larger carnivorous fish from east to west during El Niño events. In the westerly group, SST may not be high enough to cause an increase in ciguatera during La Niña conditions (Figure 3). The present findings suggest a role for climatic factors in human diseases caused by G. toxicus. Whilst the environmental factors affecting ciguatera are undoubtedly complex, increases in ciguatera (and perhaps other diseases caused by harmful marine algae) may result if the climate continues to warm as a result of the enhanced greenhouse effect. More broadly, ciguatera may be a sensitive indicator of environmental stress in coral reef systems. Coral reefs have been under increasing pressure from human populations in recent years (Jameson et al. 1995) and increases in ciguatera are likely if anthropogenic disturbance of these critical ecosystems is allowed to continue. Ecosystem Health
Simon Hales is supported by a Training Fellowship of the Health Research Council of New Zealand. We would like to thank Yvan Souares and the Community Health Programme, Secretariat of the Pacific Community, Noumea (the former South Pacific Commission) for access to data, NASA for distributing the SOI data, and Benno Blumenthal at the IRI for the SST data. There was no conflict of interest in this study.
REFERENCES Baden, D., Fleming, L., Bean, J. (1995) Marine toxins. In: De Wolff, F. (ed) Handbook of Clinical Neurology. Elsevier Science, Amsterdam. Bomber, J., Guillard, R., Nelson, W. (1988) Roles of temperature, salinity and light in seasonality, growth and toxicity of ciguatera-causing Gambierdiscus toxicus Adachi et Fukuyo (Dinophyceae). Journal of Experimental Marine Biology and Ecology 115, 53–65. Brown, B. (1997) Coral bleaching: Causes and consequences. Coral Reefs 16, S129–S138. Dalzell, P. (1992) Ciguatera fish poisoning and fisheries development in the South Pacific region. Bulletin de la Societe de Pathologie Exotique 85, 435–444. Gleeson, M. & Strong, A. (1995) Applying MCSST to Vol. 5
No. 1
March 1999
25
coral reef bleaching. Advanced Space Research 16, 151–154. Glynn, P. (1993) Coral reef bleaching: Ecological perspectives. Coral Reefs 12, 1–17. Goreau, T. (1990) Coral bleaching in Jamaica. Nature 343, 417. Hallegraeff, G. (1993) A review of harmful algal blooms and their apparent global increase. Phycologia 32, 79–99. Jameson, S., McManus, J., Spalding, M. (1995) State of the Reefs: Regional and Global Perspectives. International Coral Reef Initiative. Washington, D.C. Lenton, T. (1998) Gaia and natural selection. Nature 394, 439–447. Lewis, R. (1992) Ciguatera in the pacific. Bulletin de la Societe de Pathologie Exotique 85, 427–434. McGowan, J., Cayan, D., Dorman, L. (1998) Climate-
ocean variability and ecosystem response in the Northeast Pacific. Nature 281, 210–217. Morton, S., Norris, D., Bomber, J. (1992) Effect of temperature, salinity and light intensity on the growth and seasonality of toxic dinoflagellates associated with ciguatera. Journal of Experimental Marine Biology and Ecology 157, 79–90. de Sylva, D. (1994) Distribution and ecology of ciguatera fish poisoning in Florida with emphasis on the Florida Keys. Bulletin of Marine Science 54, 944–954. Tester, P. (1994) Harmful marine phytoplankton and shellfish toxicity: Potential consequences of climate change. Annals of the New York Academy of Sciences 740, 69–76. Tibbetts, J. (1998) Toxic tides. Environmental Health Perspectives 106, A326–A331.
Hales et al.: Fish Poisoning and El Niño
