Bioluminescence in Dinoflagellates: Evidence that the Adaptive Value of Bioluminescence in Dinoflagellates is Concentration Dependent

Three major hypotheses have been proposed to explain why dinoflagellate bioluminescence deters copepod grazing: startle response, aposematic warning, and burglar alarm. These hypotheses propose dinoflagellate bioluminescence (A) startles predatory copepods, (B) warns potential predators of toxicity, and (C) draws the attention of higher order visual predators to the copepod's location. While the burglar alarm is the most commonly accepted hypothesis, it requires a high concentration of bioluminescent dinoflagellates to be effective, meaning the bioluminescence selective advantage at lower, more commonly observed, dinoflagellate concentrations may result from another function (e.g. startle response or aposematic warning). Therefore, a series of experiments was conducted to evaluate copepod grazing (Acartia tonsa) on bioluminescent dinoflagellates (during bioluminescent and nonbioluminescent phases, corresponding to night and day, respectively) at different concentrations (10, 1000, and 3000 cells mL^?1), on toxic (Pyrodinium bahamense var. bahamense) and nontoxic (Lingulodinium polyedrum) bioluminescent dinoflagellates, and in the presence of nonluminescent diatoms (Thalassiosira eccentrica). Changes in copepod ingestion rates, clearance rates, and feeding preferences as a result of these experimental factors, particularly during the mixed trails with nonluminescent diatoms, indicate there is a concentration threshold at which the burglar alarm becomes effective and below which dinoflagellate bioluminescence functions as an aposematic warning.