Associations between fish health and Pfiesteria spp. in Chesapeake Bay and mid-Atlantic estuaries

Abstract

In response to concerns that there may be an association between harmful algal bloom (HAB) species and fish health, including the widespread use of fish health as one indicator of a possible HAB warranting further investigation, evidence for such an association was evaluated in Chesapeake Bay and other mid-Atlantic estuaries (1999–2001). A statistical approach was used, without invoking causality, to test whether there is an association between the prevalence of externally-visible lesions in fish populations above background levels and the presence of Pfiesteria spp. in co-located water and fish samples. Externally visible anomalies (e.g. ulcers, necrosis, parasites, etc.) were recorded for Atlantic menhaden ( Brevoortia tyrannus) and all other fish collected. Polymerase chain reaction (PCR) techniques were used to test for the presence of Pfiesteria spp. in water samples collected at routine and rapid response sampling events. No actively toxic Pfiesteria was found during this study. Fine-scale (within a given sample site) and broad-scale (estuary-wide sampling) comparisons showed positive associations between externally-visible fish lesions in menhaden populations and the presence of Pfiesteria spp. in co-located samples. Logistic regression modeling of Pfiesteria detection probabilities as a function of prevalence of menhaden with lesions was significant ( P = 0.0096). Reductions in the false positive (tests indicating Pfiesteria presence when its absent) and false negative (tests indicating Pfiesteria is absent when it is actually present) rates occurred when the minimum sample size threshold increased from 1 to 30 fish ( P = 0.003–0.001). This association served as a useful field indicator of potential HAB activity that could warrant further field investigation and testing.