Figure 2.

Feedback and feedforward processes allow living systems to persist in fluctuating environments by allowing them to minimize fluctuations in fitness-defining variables (e.g. predation risk). Copepods and other zooplankton combine feedback and feedforward processes to avoid predation in sunlit surface waters. Copepods feed in the surface waters (epipelagic zone) where phytoplankton is abundant. However, feeding in sunlit, illuminated surface waters exposes copepods to visual predators. (a) Copepods can detect predators via their setae, which are mechanoreceptors. When setae bend, this may elicit a neurophysiological response in the brain (the controller), triggering the copepod to swim away (effector). This escape behaviour is a type of feedback process—detecting predators causes copepods to move away from predators until they are no longer detectable. Feedback processes are reactive in that they occur after changes in their internal state, z(t), such as bending of setae owing to predator presence, have occurred. (b) Feedforward processes, such as diel vertical migrations, occur when organisms respond to some external environmental cue, e(t), here indicated by a light blue circle, to control an internal variable such as predator exposure. An internal model allows organisms to ‘pull the future into the present’ [25] by acting, in the present, on some cue that is correlated to a likely future environmental state. In this case, the change in light (dI/dT), which precedes periods of high predation risk during day time, is used as a predictive cue to adjust depth (i.e. light-cued vertical migration) in order to escape predation. This feedforward mechanism allows zooplankton to move to deeper depths (the mesopelagic zone) proactively at sunrise, before surface waters (epipelagic zone) become sunlit and predation risk by visual predators increases (c). Feedforward mechanisms may be combined with feedback mechanisms that allow organisms to respond to predators after they are detected. In (a,b), light blue arrows correspond to the feedforward process while dark blue arrows correspond to the feedback process. The grey arrow back from ‘effector’ to ‘internal model’ in (b) indicates that internal models can change as the environment changes, a feature of general adaptive systems (GAS). These changes to internal models may occur via learning or other mechanisms by which organisms update their internal models or of those of their offspring.