-from Epidemiological Bulletin, Vol. 21 No. 2, June 2000-

Fish Mortality in Southeastern Caribbean Countries

A fish-kill outbreak affected several southeastern Caribbean countries from July to September 1999. As thousands of demersal reef fish washed ashore on some beaches of Barbados, Grenada, Saint Vincent and the Grenadines, and Tobago, concern grew among the national authorities of the Caribbean Community and Common Market (CARICOM), particularly for the serious direct impact of the fish-kills on the fishing industry, and their indirect effects on public health and the tourism industry.

In response to the magnitude of the problem, a Regional Meeting was convened by the Secretary General of CARICOM in Barbados in October 1999. The Pan American Health Organization (PAHO), through the Office of the Caribbean Program Coordination (CPC) in Barbados, responded in a timely manner to the request for technical cooperation from the affected countries. An epidemiological questionnaire was sent to all affected and non-affected countries and a technical team from the PAHO-CPC Office visited the countries affected by the outbreak. The purpose of this note is to present a brief description of the findings and provide recommendations to respond to similar natural or man-made disasters in the future.

Fish-kill events were reported in Grenada at the end of July, in Saint Vincent and the Grenadines on September 2, in Barbados on September 16, and in Tobago in the middle of September. The problem started to subside at the beginning of October, probably due to the depletion of the fish population at risk, to the diminution of the factor(s) associated with the kill events, or to a combination of both. Other countries affected were Guyana (in July) and Venezuela (in August). No other CARICOM Country reported fish mortality.

As estimated by the National Fisheries Division of Saint Vincent and the Grenadines, the direct economic impact amounted to a US$192,000 loss due to the decrease in fish landing and a US$120,000 loss of fish exports to Martinique. Additionally there was a 75 % drop in fishing and fish vending activities in comparison to the same months in previous years. There was also an unestimated loss due to a decline in the consumption of fish in the national and tourist populations. Indirect losses associated with a decrease in the number of visitors and reservation cancellations were not assessed, but highly publicized.

The following hypotheses were formulated to explain the kill events: a) volcanic eruption, b) chemical spill and/or cruise discharges, c) algae bloom, and d) pathogenic microorganisms. Hypotheses a) and b) were excluded for the following reasons: firstly, the Seismic Unit in Trinidad & Tobago did not register any significantly different seismic activity in the affected areas during the time of the outbreak as compared to data from previous years; secondly, the hypothesis of chemical spill was unsustainable, because the dilution factor would have more likely limited the fish kill to an area close to where any chemical spill would have occurred.

The hypothesis of the algae bloom - associated with an increase of the influx of nutrient-rich water form the Orinoco and Amazon Rivers in South America, an increase of water temperature, and oxygen depletion - is supported by physical observations made by fishermen from affected countries and pilots of regional airlines, indicating the presence of greenish/brownish discoloration of the sea water before and at the time of the kill events. Satellite pictures of the Caribbean presented by a technical team from Japan at the CARICOM Meeting, compared with images from previous years, clearly showed an unusual volume of the influx from the Orinoco and Amazon Rivers into the affected areas, including a gradient increase of temperature and of the sea water nutrient content as measured by chlorophyll gradient levels. This hypothesis was not confirmed by the determination of an increase in phytoplankton in the affected areas. However, it should be noted that water samples used for determination of algae blooms and physicochemical analysis were not collected in a timely manner nor using equipment allowing to sample different layers of the water column in affected areas. The samples were not preserved with lugol solution or 4 % formalin for phytoplankton studies.

In relation to the hypothesis of association with microorganisms, the Veterinary Division Laboratory of the Barbados Ministry of Agriculture, in collaboration with Dr. H. Ferguson of the University of Scotland, isolated a bacteria named Streptococcus iniae from 3 out of 4 specimens of moribund fish washing off the affected coast of Barbados. Histopathological studies revealed vasculitis, inflammation around the gills and pericaditis and the presence of gram+ bacteria in the liver, spleen and brain. None of the fish specimens exhibited external lesions. This finding is very significant, as S. iniae has been associated with fish mortality in fresh water farming settings in various countries, and with human occupational diseases through the inoculation of the bacteria during the manipulation of fish. The Veterinary Division of the Ministry of Agriculture of Barbados has continued working on the subject and will make invaluable contributions to deal with such events in the future.

During the outbreak, the PAHO-CPC office took the following actions, together with the national authorities of all the affected countries: a) creation of inter-sectorial task forces at the national level, b) jointly with the Institute of Marine Affairs of Trinidad and Tobago, creation of an e-mail network linking the national authorities of all affected and non affected countries and international experts on the subject, including staff from PAHO’s specialized centres such as the Caribbean Epidemiology Center (CAREC), the Pan American Institute for Food Protection and Zoonoses (INPPAZ), and the Pan American Center for Sanitary Engineering and Environmental Sciences (CEPIS), c) preparation of guidelines for safe disposal of dead fish, d) dissemination of a technical bulletin alerting all national health authorities in the area to the problem and its possible direct and indirect impact on public health, and e) provision of technical expertise on algae bloom, phytoplankton and marine toxins issues to all affected countries, and preparation and dissemination of a final report.

The following recommendations were extracted from the final report prepared by PAHO-CPC:

1. A surveillance program should be established to monitor algae blooms and other physicochemical parameters predictor of algae bloom in affected areas and those areas that could be potentially affected in the future.

The program should include:

  1. Active surveillance to establish trends in the environmental explanatory/predictor variables to determine the likelihood of occurrence of similar events in the future. A surveillance program should address the following points :
  1. Sampling of water for phytoplankton and microbiological analysis to detect concentration of phytoplankton species, pathogens and contaminant indicators, i.e., total coliforms, fecal E. coli and S. faecalis.

  2. Determination of appropriate physicochemical parameters.

  3. Sampling on a bi-weekly or weekly basis. The latter is highly recommended. The sampling protocol should include:

2. A research design should be set up to systematically collect and analyze information on variables (contributors) associated with increased kills of reef fish. The following multivariate theoretical model was suggested:

FK(P) = a + b1(x1) + b2(x2) + b3(x3) + b4(x4)+ b5(x1x2)+...+e

where FK (P)= Outcome Variable or Probability of fish killing
a= intercept
b1,...,bn = Regression coefficients
x1,...,xn = Variables or contributor factors
x1= Temperature (° C)
x2= Phytoplankton (Number of cells per liter)
x3= Toxic dinoflagellates a and b (0 or 1)
x4= S. iniae (cfu per gram)
e = Error

This model provides room to determine the action and interaction of the variables. In the example above, the interaction between temperature and phytoplankton is expressed by b5(x1x2). Other interaction factors could be included in the model. At the same time, phytoplankton concentrations and types could be used as an outcome variable.

3. The educational strategy to the general population should receive special attention. A national task force should prepare messages to the public based on scientific evidence. These messages must be very clear to minimize economic consequences. For example, in a scenario where an algae bloom is affecting reef areas, the consumption of fish and shellfish must be prohibited only from those affected areas.

There was not conclusive evidence associating the fish kills in the southeastern Caribbean to one single factor. These events were more likely due to the action and interaction of several physicochemical and biological environmental factors as previously discussed. Historical evidence of seasonal nontoxic and toxic algae blooms in Venezuela and in Trinidad and Tobago, and their related public health implications suggest that a surveillance and monitoring system of the likely environmental factors associated with these events should be implemented as a valuable tool to manage similar problems in the future and minimize their economic and public health impact.

The e-mail network linking national authorities from the health and agricultural sectors was very useful to exchange information on the fish kills events, and should be maintained by the Institute of Marine Affairs. Communication to the general public must be carefully managed based on risk assessment to protect public health and minimize economic consequences. PAHO/WHO and its specialized centers, CAREC and INPPAZ, in collaboration with the PAHO office in Trinidad and Tobago, held a workshop on “Analysis of Selected Foodborne Pathogens and Marine Toxins”. Participants received field training on phytoplankton sampling and identification, and a demonstration of marine toxin bioassays was organized. Additionally, a Subregional Network for the control of red tides and marine toxins, now comprised of Cuba, the Dominican Republic, Guatemala, Panama and Venezuela, was created in 1990 with PAHO/WHO’s cooperation. The functioning of this network should be revisited and extended to the English speaking Caribbean.

(1) AOAC, 1990. Paralytic Shellfish Poison, Biological Method. In: Association of Official Analytical Chemist. Official Methods of Analysis. pp 881-2
(2) Balladin DA, Lee Lum L, Lum Kong A. 1999. Report of Fish Killing in Tobago. Institute of Marine Affairs, Trinidad and Tobago. 11p.
(3) Barbados Fisheries Division Report. 1999. September 1999: Fish kill on the South Coast of Barbados. Barbados 24 September 1999, 6 p.
(4) Bowser, PR et al. 1998. Streptococcus iniae Infection of Tilapia Oreochromis niloticus in a Recirculation Production Facility. J. World Aquaculture Society. Vol. 29 (3): 335-339.
(5) Colorni A and Diamant A. 1999. Pathology in Mariculture: Specific Problems and Research in Israel. In: Abstracts: Japanese-Israeli Symposium on Aquaculture.
(6) De Campo M. 2000. Personal communication.
(7) Ferguson H, Saint John S, Roach J. October 1999. Report presented at the CARICOM Meeting on Fish Kills. Barbados.
(8) Texas Parks and Wildlife (1999) - Fishing: Frequently asked questions about red tide. http://www.tpwd.state.tx.us/fish/recreat/redtide.htm
(9) Garcia, G. 1999. Report No. 2. Fish Kills Episodes in the Caribbean Subregion. 28 September 1999, 2 p.
(10) Institute of Marine Affairs. 1999. IMA Investigate Fish Kill in Tobago. 24 September 1999. 1p
(11) Institute of Medicine. 1991. Seafood Safety. National Academy Press. Washington, D.C., 20418. 432 p.
(12) La Barbera, A. Personal Communication. 1999.
(13) Morris, K et al. 1999. Report of Fish Killing in Saint Vincent and the Grenadines.
(14) Seafood Network Information Center - University of California, Davis. Natural Marine Toxins. http:/seafood.ucdavis.edu/pubs/natural.htm
(15) Philippeaux H. 24 September 1999. Report No.1 Situation Report on Fish Kills Episodes in the Caribbean Subregion. 3 p.
(16) Siung Chang A. Personal Communication. Presentation made at the CARICOM Meeting, Barbados October 1999.
(17) Botany Department - University of the Western Cape, South Africa. Red tide and shellfish poisoning. http://www.botany.uwc.ac.za/Envfacts/redtides
(18) Spencer D. 1999. Personal communication. Agropecuarias (FONAIAP) Cumaná, Edo. Sucre, Venezuela.

Source: PAHO. Veterinary Public Health Program, Division of Disease Prevention and Control.

Acknowledgments to the national agricultural and health authorities of all affected countries, for shring their information and to Dr. Amelia la Barbera, of the National Fund for Research on Farming and Livestock (FONAIAP) of the State of Sucre, Venezuela, for her excellent technical support and expertise.

Return to Index
Epidemiological Bulletin, Vol. 21 No. 2, June 2000