Effects of the toxic dinoflagellate Heterocapsa circularisquama on the valve movement behaviour of the Manila clam Ruditapes philippinarum

The effects of the toxic dinoflagellate Heterocapsa circularisquama on the valve movements of adult Ruditapes philippinarum (Bivalvia, Veneridae) were investigated. The valve movement behaviour was continuously measured using a non invasive Hall element sensor-automated data acquisition system. Individual clams acclimated to the experimental conditions were subjected to different food regimes over a 96 h period: (1) unfed in filtered seawater for the first 24 h, (2) batch-fed1, in which clams were given one ration of non-toxic Isochrysis galbana at the beginning of the second 24 h, (3) batch-fed2, in which clams were given a mixture of I. galbana and H. circularisquama (5 x 10(4) cells ml(-1)) for the third 24 h, and (4) starved for 24 h in filtered seawater after exposure to the toxic alga. For each food regime, the valve activity was analyzed by quantifying: (1) the duration of valve opening (DVO, %), (2) the amplitude of valve opening (AVO, mm), and (3) the frequency of valve adductions (FVA, adcluctions h(-1)). The real-time monitoring of the clams' valve movements proved that R. philippinarum detects H. circularisquama in the mixture of food at as low as 0.5 cells ml(-1). Typically, the addition of the toxic alga induced a short and incomplete valve closure reaction. A strong correlation between the closure reaction time and H. circularisquama concentration was found (R-2 = 0.91), the closure reaction occurring faster with increasing concentrations. Reopening of the valves was characterized by a significant increase in the FVA starting from 5 cells ml(-1) (ANOVA, P<0.001) and a significant decrease in the AVO at 500 cells ml(-1) (Wilcoxon, P<0.05). After exposure, clams transferred to clean seawater resumed normal valve adduction activity, but showed a significant increase in the DVO after exposure to 0.5 and 5 cells ml(-1) (ANOVA, LSD, P<0.05), and a significant increase in the AVO (ANOVA, P<0.01). Such increase could have been associated with recovery behaviour. Our results indicate that the Manila clam is highly sensitive to the toxic H. circularisquama and detects efficiently its presence at extremely low concentrations, in mixture of food, and in a time- and concentration-dependent manner. The Manila clam and the valve movement monitor could thus be used as an early biological warning system to monitor the outbreaks of this toxic alga along marine coastal areas and in bivalve farms. However, field experiments are required to provide further evidence for the use of this monitor in natural environment where several factors may interfere with the valve movement behaviour. (c) 2009 Elsevier B.V. All rights reserved.