Attack is not the best form of defense: Lessons from harmful algal bloom dynamics

The potential roles that allelopathic, grazing deterrence and predator-killing chemicals have in shaping planktonic food webs are considered making use of models in which variable prey selection and stoichiometry are described. Allelopathy does not aid the growth of a poor competitor for resources unless some other factors initially elevate the biomass of the producer and hence increase the levels of allelopathic chemicals to exceed a critical level. Killing a predator is only of use for a strong resource competitor; otherwise removal of a predator that also controls the growth of competitors is disadvantageous. Killing a predator that also controls the biomass and activity of another predator (e.g. the toxic action of diatom-originating polyunsaturated aldehydes on copepods, thus releasing grazing pressure by copepods upon microzooplankton) also appears disadvantageous. The best recourse for a moderate-to-poor resource competitor appears to be predator deterrence; low-level growths of such organisms probably provide little natural selective stimulation for the evolution of counter-measures by predators. Factors that disturb the natural balance of the system, such as disproportionate N:P(:Si) eutrophication, may provide deterrence-producing algae with a window of opportunity from which they may then form HABs. However, the failings of non-HAB species to bloom are just as important in dictating the success of a HAB species under any given set of conditions. The simulation of such events thus requires inclusion of non-HAB as well as HAB descriptions, and also of their predators, in a modeling framework capable of capturing the dynamics of prey selection and rejection.