136 



5 

 spring simulation (p.ES-8). While the possibility of modeling simulation of 

 summer stratified conditions is indicated (p 4-1), the authors state 'the present 

 conclusions are not conditional upon future modeling results." Yet harmful algal 

 bloom events, their nutrient regulation and trophic consequence are seasonal 

 issues, with the summer stratified period being the critical period for such 

 blooms and their nutrient regulation and potential disruption of normal food web 

 structure and energy flow. Thus, the Assessment's focus on the winter-spring 

 diatom bloom is irrelevant to the Assessment issues. The extrapolations (p.4- 

 17-20) of the distributional patterns, outfall dilution contours, variability, and 

 expected nutrient enrichment above ambient condition from the winter-spring to 

 the stratified conditions, phytoplankton community composition, blooms and 

 trophic relationship during the summer are not appropriate. Summer conditions 

 require separate analysis. 



EPA's dubious understanding of the nutrient-phytoplankton issues to be 

 resolved is also evident in other ways. Significance is attributed (p. 4-27) to 

 Kelly et al.'s failure to find correlations among various variables measured in 

 the 1992 synoptic monitoring program, i.e. there was "a general lack of 

 correlation between nutrient concentrations and phytoplankton or zooplankton 

 species composition". There is a wealth of data indicating that (at least among 

 marine phytoplankton) indicator species of various nutrient levels do not occur. 

 The absence of such correlations, therefore, is not indicative that variations in 

 nutrient concentrations do not regulate plankton community composition. This 

 is contrary to the Assessment's interpretation that the negative evidence from 

 Massachusetts Bay that species composition there is not influenced by nutrients 

 is, therefore, indicative that the MWRA outfall will have no effect on species' 

 composition. The Assessment makes another serious error of extrapolation 

 (p. 4-24) in its confusion between phytoplankton responses as biomass vs. 

 changes in community structure. Based on the log-log yield-dose (i.e. primary 

 production vs. DIN concentration) relationship presented (Figure 4-5), the report 

 concludes that "additional inputs of nutrients into the bay would not result in 

 ....changes in the phytoplankton community". This interpretation cannot be 

 applied. Figure 4-5 allows one to conclude only that annual primary production 

 increases with annual DIN inputs. It provides absolutely no information as to 

 whether changes in species composition or other alterations in community 

 structure also accompanied the changing DIN inputs plotted in Figure 4-5. It 

 therefore cannot be used by EPA as supporting evidence for the latter. 



The EPA report appears to be procrustean in being biased in seeking 

 support for its conclusions regarding nutrient-phytoplankton relationships. Thus 

 (p. 4-37), after acknowledging that phytoplankton species selection may be 

 influenced by nutrient ratios, the report concludes "there is no a priori reason 

 that such shifts will cause blooms of toxic species". In reaching this conclusion, 

 a significant literature is overlooked, with justification for this conclusion 

 seemingly based primarily on a tenuous summer event in Chesepeake Bay. 

 The report wrongly asserts (p. 4-37) "to date, there appears to be little empirical 

 evidence demonstrating that anthropogenic nutrient additions cause outbreaks 

 of toxic phytoplankton species". The authors, in fact, cite the work of Smayda 

 (1989, 1990) in other contexts, but ignore the strong global evidence presented 

 by him for an apparent linkage between nutrient buildup and harmful bloom 

 events. The report dismisses the Geraci et al. report (p. 4-42) as only 

 "circumstantial" (p.4-43) that 14 humpback whales died of STX poisoning after 

 consumption of mackerel. The report (p.3-43) indicates 81 dead fin whales 



