Predator-prey interactions between Isochrysis-Galbana and Oxyprhis-Marina. 3. Mathematical Modellling of predation and nutrient regeneration

A mathematical model was derived to simulate ingestion, growth and nitrogen (N) regeneration for the phagotrophic dinoflagellate Oxyrrhis marina. Two types of experimental study were undertaken: prey-deplete O.marina were supplied with Isochrysis galbana in continuous darkness (thus preventing growth of the prey), and predator-prey interactions were also followed in cultures maintained in a light-dark cycle (allowing growth of the prey). During light-dark cycles, Oxyrrhis volume increased more in the light phase than in the dark. Digestion of Isochrysis lasted approximately 0.3 days, with an average maximum ingestion rate of 55 prey predator-1 day-1. During active predation, 30% of Oxyrrhis-carbon (C) was lost from the particulate phase: per day, with this loss falling to 10%: per day at the cessation of herbivory when cannibalism became noticeable. Ingestion was modelled as a function of prey density, C-loss and division as functions of cellular predator C, with cannibalism by Oxyrrhis also included. Two N-regeneration expressions were investigated: one proposed by D.A. Caron and J.C. Goldman (Journal of Protozoology, 35, 247-249, 1988) and an alternative function which related N regeneration to intracellular carbon and N based on the concept of an optimal Oxyrrhis C:N ratio. The latter was more successful in simulating batch culture data and did not require a prior calculation of Oxyrrhis gross growth efficiency. The model of Oxyrrhis numbers, C and N contained only nine parameters whose values were fully obtainable from batch culture experiments. By using this model, we were able to use a single parameter set to simulate the transient dynamics of Oxyrrhis ingesting N-replete and N-stressed prey. Further experiments in which Oxyrrhis grew on Isochrysis in light-dark cycles were simulated by combining the Oxyrrhis model with the nutrient-processing model for Isochrysis of K. Davidson et al. (Journal of Plankton Research, 15, 351-359, 1993). The dynamics of the full predator-prey model were found to be sensitive to the level of sophistication of the prey model; the Quota model was found to be less succesful than the nutrient-processing prey model. Theoretical model runs indicated the importance of being able to simulate changes in both prey numbers and biomass, and also in including realistic equations for nutrient regeneration from predators in microbial predator-prey models.