Reply to Errera and Campbell: No, low salinity shock does not increase brevetoxins in Karenia brevis.

Our paper (1) was undertaken to challenge earlier reports that low salinity stress increases brevetoxin production in Karenia brevis (2). Despite independent negative findings by three laboratories (1), Errera and Campbell still assert that low salinity shock increases cellular brevetoxins (3). Their initial report of >14-fold increases (2) lacks experimental controls. Their correction (2) substantially reduces the reported increases to 20–53% but does not address the lack of controls or alter their interpretation. They now (3) refer to results from new experiments that have not yet been published or vetted by peer review. In their rebuttal, theymisrepresent data in Sunda et al. (1). For example, they say that “laboratory A also demonstrated increased brevetoxin cell quota in SP3 by ∼15% after 12 d,” but this increasewaswithin the SDof replicate analyses. Indeed, statistical tests showed no significant changes in brevetoxin per cell following low salinity shock in the experiments conducted by our three laboratories (1). Errera and Campbell (3) challenge our findings (1) because our experiments did not exactly replicate theirs with regard to extraction protocols, analytical methods, K. brevis isolates, and culturing conditions. However, the eight strains we examined (1) included three they used in their experiment (2). Furthermore, we subjected all Karenia strains to the same or similar salinity shift (a decrease from 35 or 36 to 27) as Errera and Campbell (a decrease from 35 to 27) and measured brevetoxins within 3 h to 48 h, the same time range where they said increases were observed (2). Any robust physiological response, such as the reported brevetoxin increase within 3 h or 24 h (2) following low salinity shock, should not depend on variations in culture medium, light conditions, or analytical protocols. Errera and Campbell also cite that our methods for brevetoxin analyses (1) lack “a number of experimental details, including use of internal standards and preparation of toxin standards.” Brevetoxin measurements in all three laboratories in Sunda et al. (1) were based on standard curves for the brevetoxin congeners PbTx-1, -2, and -3. Laboratory A used the same internal standard as Errera and Campbell (2) in each brevetoxin analysis and found extraction efficiencies of 90–95%. Laboratory B used spiked brevetoxin standards before their experiment and found extraction efficiencies of 95%. Extraction efficiencies were not checked in laboratory C, but their brevetoxin quotas for the toxic Wilson strain were within 20% of published values (4). Therefore, accurate brevetoxin measurements are not an issue. Other arguments by Errera and Campbell (3) are misleading. For example, contrary to their assertions, Coulter Counters are routinely used in culture studies to accurately measure algal cell concentrations and volumes (e.g., refs. 4 and 5). Another criticism is that changes in cell volume during the day might have influenced our results. However, the brevetoxin per cell measurements we used to examine toxin differences between paired control and treatment samples are not affected by time-dependent changes in cell volume. Such side issues do not obscure the central flaw in the Errera and Campbell study—the lack of experimental controls. William G. Sunda, Cheska Burleson, D. Ransom Hardison, Jeanine S. Morey, Zhihong Wang, Jennifer Wolny, Alina A. Corcoran, Leanne J. Flewelling, and Frances M. Van Dolah Beaufort Laboratory, National Centers for Coastal Ocean Science, National Ocean Service, National Oceanic and Atmospheric Administration, Beaufort, NC 28516; Fish and Wildlife Research Institute, Florida Fish and Wildlife Conservation Commission, St. Petersburg, FL 33701; and Marine Biotoxins Program, Center for Coastal and Environmental Health and Biomolecular Research, National Oceanic and Atmospheric Administration, Charleston, SC 29412