@article{GENERAL,
      recid = {7541},
      author = {Cullen, John J. and MacIntyre, Hugh L},
      title = {The case for using the Most Probable Number (MPN) method  in ballast water management system type approval testing},
      journal = {Conference Proceedings of BWTC},
      address = {2017-01-13},
      number = {GENERAL},
      abstract = {Recently, the U.S. Coast Guard (USCG) rejected the Serial  Dilution Culture-Most Probable Number (SDC-MPN) method for  enumerating viable phytoplankton cells in ballast water  discharge as an alternate to their prescribed method  — the Environmental Technology Verification (ETV)  Protocol. This method distinguishes living from dead  organisms using vital stains and motility. Succinctly, the  USCG position has been that the ETV Protocol is a reliable  and repeatable efficacy test and the SDC-MPN method is not.  New evidence and an expanded consideration of published  research supports a fundamentally different assessment.

A  peer-reviewed quantitative evaluation of ETV vital stains  for 24 species of phytoplankton has conclusively  established that the ETV Protocol, even with observations  of motility, is not reliable for all species. In contrast,  published results suggest that errors in the method were  small for the limited number of locations studied to date.  It is possible that the communities tested in these were  dominated by species that can be classified accurately  using vital stains. Even so, it must be acknowledged that  the reliability and accuracy of vital stains is untested  for thousands of species of phytoplankton.

Introduced in  1951, the SDC-MPN method for phytoplankton is an  established approach for use with multi-species  communities. As applied to ballast water testing, SDC-MPN  is much less vulnerable to methodological uncertainties  than has been assumed. Notably, all species of  phytoplankton need not be cultured in the conventional  sense. Rather, a single viable cell in a dilution tube need  grow only enough to be detected — a requirement known  to have been met by otherwise uncultured species. Further,  delayed restoration of viability after treatment with  ultraviolet radiation (UV) is not a problem: organisms  repair UV damage quickly or not at all, consistent with the  assumptions of the test.

Two critical methodological  failures could compromise protection of the environment in  ballast water testing: living organisms that do not stain  or move, and viable organisms that do not grow to detection  in the MPN cultures. These can be assessed with  complementary measurements, but importantly, the relative  protection of each method can be evaluated by comparing  counts of living cells from the ETV Protocol with counts of  viable cell from SDC-MPN in untreated samples. Available  evidence provides no basis for concluding that either  method is consistently less protective. However, as applied  in ballast water testing, the statistical estimate of MPN  is less precise. On this basis, SDC-MPN is worse for a  single test. But, counter-intuitively, it is more  protective of the environment when five consecutive tests  must be passed for type approval, because the likelihood of  one false rejection out of five tests is higher and five  false passes would be exceedingly rare. Addressing only the  science, we conclude that both the ETV Protocol and the  SDC-MPN method, though imperfect, are currently appropriate  for assessing the efficacy of ballast water management  systems in a type-approval testing regime. In closing, we  show proof of concept for a rapid assay of viability,  benchmarked against SDC-MPN, that could be well suited for  routine assessment of treatment system performance.},
      url = {http://library.imarest.org/record/7541},
}