Marine plankton provide essential support to all life in the planet, through the production of oxygen, contribution to nutrient cycling in the ocean, and by being the base of the marine food web. Due to their small size and limited motility, plankton are also highly susceptible to changes in the environment. Monitoring and understanding how the plankton responds to different climatic and anthropogenic effects is imperative in order to predict and prevent damaging outcomes to the ecosystem. Here I focused on four key plankton time-series, from the Continuous Plankton Recorder survey, in the Northeast Atlantic and North Sea.
There is evidence in the literature that grazing plays a major role for the control of phytoplankton populations on a seasonal scale, however there is no consensus on whether bottom-up control from physical variables or top-down control from zooplankton grazing would be dominant from an interannual perspective. Through vector autoregressive modelling, I analysed interactions among the phytoplankton, zooplankton, and environmental indicators. Sea surface temperature emerged as a significant driver of variability, and some evidence for bottom-up control was found from a lack of dependency of diatoms on the other plankton variables. However, the plankton seemed to be mostly regulated by serial correlation and the seasonal cycle.
The serial correlation suggested the presence of nonlinearity in the system, and led to the question of whether linear approximations were suitable for plankton dynamics. I investigated the presence of nonlinearity, stochasticity and deterministic chaos in the plankton. The seasonal cycle was key to stabilise the fluctuations within plankton populations, which appeared to be regulated by a nonlinear stochastic dynamics rather than by deterministic chaos. The lack of significant links between the plankton and the two environmental indicators reinforced the complexity of the plankton system, and implied that it is not likely to exist a single factor as the main driver of plankton variability.