Clearance rate regulation in mussels: Adding the effect of seston level to a model of internal state-based regulation

• We tested a model where clearance rate varies with seston level and mussel energy density.
• The experiment was inconclusive with respect to the effect of energy density.
• Mean clearance rate varied non-monotonously with seston level and we present a new model.
• We were able to estimate a subset of parameter values.
• Short-term variability was highest at intermediate seston level and was much larger than noise.

We tested a model of clearance rate regulation that specifies that clearance rate is driven by the density of energy reserves and is modulated by factors such as biochemical signalling and seston concentration (Fréchette, M. 2012. A model of clearance rate regulation in mussels. J. Sea Res. 73:32–40). Mussels were fed natural seston at high and low concentrations for 71 days to manipulate the density of energy reserves (“position effect”). Clearance rate of each group was assayed in a two-level repeated-measures design experiment, within and among days. The effect of position and its interactions was not significant. It is likely that the low seston level treatment was too mild to generate phenotypic flexibility in clearance rate. The effect of time, however, was significant at all scales. Seston variability allowed examining the functional response of mean clearance rate to seston level. Short-term clearance rate was bounded by seston concentration at very low and high seston concentration. Between concentration extremes, however, short-term clearance rate was not regulated by actual seston concentration. On some occasions hysteresis dominated the functional response, implying delayed regulation, presumably in response to gut fullness. Hourly-averaged mean clearance rate exhibited a clearly-defined non monotonous response to seston level. Therefore we modified the initial model formulation in order to include the effect of low seston level. We found an excellent fit between the modified model and the data. Therefore, in spite of appearing erratic at intermediate seston level, short-term clearance rate patterns are embedded within a well-defined functional response. Understanding these patterns apparently requires that seston level, its temporal variability and the dynamics of gut fullness be taken into account.