Pfiesteria Species Identified in Ships ’ Ballast Water and Residuals : A Possible Vector for Introductions to Coastal Areas

Phytoplankton species most likely to survive in ballast water and unpumpable ballast residuals are those that form resistant resting stages, have alternative modes of nutrition, or both. In our field studies of ballasted ships arriving to Chesapeake Bay and NOBOB (No-Ballast-On-Board) vessels arriving to the Great Lakes, we monitored for the HAB species Pfiesteria piscicida and P. shumwayae using polymerase chain reaction (PCR)-based methods. During 2001, we boarded 10 vessels arriving to Chesapeake Bay, and P. piscicida was detected in 1 ship (10%), while P. shumwayae was detected in 3 ships (30%). Of the residual NOBOB water samples tested, P. piscicida was detected in 2 of 34 samples (6%), and P. shumwayae was detected in 1 of 34 samples (3%). P. piscicida and P. shumwayae each were detected in only 1 of 33 (3%) residual NOBOB sediment samples. Thus, Pfiesteria is found in low frequency in ships’ ballast tanks. Further investigation is required to determine whether these strains are toxic. Steidinger, K. A., J. H. Landsberg, C. R. Tomas, and G. A. Vargo (Eds.). 2004. Harmful Algae 2002. Florida Fish and Wildlife Conservation Commission, Florida Institute of Oceanography, and Intergovernmental Oceanographic Commission of UNESCO. 318 controls. Detection of Pfiesteria indicates presence of vegetative or cyst forms, since the probes make no distinction between Pfiesteria life stages. Results and Discussion During 2001, we found P. piscicida in ballast water of 1 of 10 vessels (10%) arriving to Chesapeake Bay and P. shumwayae in ballast water of 3 vessels (30%; Table 1). The ships containing Pfiesteria arrived from Immingham (U.K.), Antwerp (Belgium), and Amsterdam (The Netherlands). Pfiesteria was detected at lower frequency (3–6% of tanks) in residual ballast water, sediment, or both, in 34 NOBOB ballast tanks arriving to the Great Lakes (Table 1). The NOBOB vessels that contained Pfiesteria in ballast water were from Western Europe, having taken on ballast water most recently in Antwerp (Belgium), Amsterdam (The Netherlands), or Hull (U.K.), and in 1 of 3 cases, had undergone open-ocean exchange. Pfiesteria was detected at a similar frequency in residual NOBOB sediments (Table 1). The NOBOB vessels that contained Pfiesteria in ballast sediments originated in China or Western Europe, but had taken on ballast in numerous ports including Inchon (South Korea), Hong Kong, Ghent or Antwerp (Belgium), Sept. Iles or Montreal (Canada) as well as the Mediterranean (Augusta, Sicily and Ravena, Italy). There was no consistent pattern with respect to Pfiesteria presence (water or sediment) and ballast origin or whether ballast had been exchanged. In 10 of 11 cases where NOBOB water samples were analyzed in replicate, results were consistent between subsamples. In the one instance in which Pfiesteria was not found in both subsamples, it was not detected in water withdrawn from the top of a sampling container that had been left sitting for approximately 1 hour, but was detected in water withdrawn from the bottom. Further, there were 3 instances where Pfiesteria (P. piscicida or P. shumwayae) was detected in residual water but not in sediments. Likewise, there were 2 instances where Pfiesteria was detected in sediments but not in water samples from the same tank. These differences may be the result of encystment and subsequent sedimentation—likely during rapid changes in salinity during ballasting operations. However it could also be related to sampling—puddles can form where there is no accumulated sediment, and water is often collected in different locations than sediment within a tank, to ensure that samples are relatively mud-free (for better DNA extraction efficiency). The disagreement of sediment and water results could also be due to patchy Pfiesteria distribution within the tank, as well as molecular probe sensitivity (where relatively small differences in cell abundance can determine whether the sample exceeds the detection limit). We also monitored for Pfiesteria at several lower Chesapeake Bay sites: at the bay mouth adjacent to a sewage outfall (4 sampling events), in the Pagan River, Virginia (3 sampling events), and at nearby coal terminals (2 sampling events). There were 3 samplings in which multiple samples were collected that yielded at least one water sample positive for Pfiesteria (33%; Table 1). However, P. shumwayae was not detected in sediments collected from two locations (Pagan River upstream and downstream, stations separated by approximately 1.3 km). There were some differences in results among independent, replicate water samples assayed for Pfiesteria using the two different PCR-based methods, with only 20 of 28 samples giving the same result (either positive or undetected). Only 1 sample tested positive using both methods; 8 other samples tested positive with one method only. Both methods are very sensitive, and differences between the assays are likely due to small differences in cell abundance within the samples, sample storage protocols, DNA extraction or amplification efficiencies between replicates, or a combination of these factors. Given such sensitivity, a more rigorous way to compare PCR-based methods would be to assay subsamples of the DNA after extraction, rather than using DNA extracted from independent filters. It should be noted that methods for toxicity testing have not been developed for Pfiesteria present in ballast tanks. Despite the methodological issues associated with molecular detection, however, this study shows that Pfiesteria is transported in ships’ ballast water and residual sediments. While its distribution within and between ballast tanks is variable, it is clear that commercial shipping could contribute to Pfiesteria’s global dispersal and potential for toxic blooms. This finding is particularly relevant in view of Pfiesteria’s disjunct global distribution, and the similarity of European P. piscicida strains to those in the U.S.A. (Jakobsen et al., 2002; Rublee et al., this Proceedings). Acknowledgements We thank members of the Great Lakes NOBOB sampling team for their efforts (particularly Phil Jenkins), Eric Schaefer and Cory Williams for running the PCR assays, and Leslie Kampschmidt for laboratory assistance. We also thank Rob Brumbaugh of the Chesapeake Bay Foundation for boat time and his assistance in collecting samples from the Pagan River. This work was conducted under the multiinstitutional Great Lakes NOBOB Project funded by the Great Lakes Protection Fund, the National Oceanic and Atmospheric Administration (NOAA), the U.S. EnvironTable 1 Presence of Pfiesteria in ships’ ballast water arriving to Chesapeake Bay and in residual water and sediments arriving to the Great Lakes; n = number of tanks sampled. Data are compared with presence of Pfiesteria at Chesapeake Bay monitoring sites (bay mouth, adjacent to 2 coal piers and Pagan River). Location P. piscicida P. shumwayae Chesapeake Bay ballast water 1 (n = 10) 3 (n = 10) NOBOB residual water 2 (n = 34) 1 (n = 34) NOBOB residual sediments 1 (n = 33) 1 (n = 33) Chesapeake Bay water 0 (n = 9) 3 (n = 9) Chesapeake Bay sediments 0 (n = 3) 0 (n = 3)