Data files of figures 1, 2 and 3

We investigated optimal conditions for characterization of bioactivity of lytic compound(s) excreted by Alexandrium tamarense based on a cell-bioassay system. Allelochemical response of the cryptophyte Rhodomonas salina indicated the presence oflytic compound(s) in a reliable and reproducible way and allows for quantification of this lytic effect. The parameters tested were the incubation time of putatively lytic extracts or fractions with the target organism R. salina, different techniques for cell harvest from A. tamarense cultures and the optimal harvest time. A three hour incubation time was found to be optimal to yield a rapid response while accurately estimating effective concentration (ECso) values. Harvest of A. tamarense cultures by filtration resulted in loss of lytic activity in most cases and centrifugation was most efficient in terms of recovery of lytic activity. Maximum yield of extracellular lytic activity of A. tamarense cultures was achieved in the stationary phase. Such optimized bioassay guided fractionation techniques are a valuable asset in the isolation and eventual stmctural elucidation of the unknown lytic substances.

Model results of sensitivity experiments for marine nitrogen cycle in four NetCDF files

The marine nitrogen (N) inventory is thought to be stabilized by negative feedback mechanisms that reduce N inventory excursions relative to the more slowly overturning phosphorus inventory. Using a global biogeochemical ocean circulation model we show that negative feedbacks stabilizing the N inventory cannot persist if a close spatial association of N2 fixation and denitrification occurs. In our idealized model experiments, nitrogen deficient waters, generated by denitrification, stimulate local N2 fixation activity. But, because of stoichiometric constraints, the denitrification of newly fixed nitrogen leads to a net loss of N. This can enhance the N deficit, thereby triggering additional fixation in a vicious cycle, ultimately leading to a runaway N loss. To break this vicious cycle, and allow for stabilizing negative feedbacks to occur, inputs of new N need to be spatially decoupled from denitrification. Our idealized model experiments suggest that factors such as iron limitation or dissolved organic matter cycling can promote such decoupling and allow for negative feedbacks that stabilize the N inventory. Conversely, close spatial co-location of N2 fixation and denitrification could lead to net N loss.