Although it seems obvious that with more data, the predictive capacity of ecological models should improve, a way to demonstrate this fundamental result has not been so obvious. In particular, when the standard models themselves are inadequate (von Bertalanffy, extended Ricker etc.) no additional data will improve performance. By using time series from the Sir Alister Hardy Foundation for Ocean Science Continuous Plankton Recorder, we demonstrate that longterm observations reveal both the prevalence of nonlinear processes in species abundances and an improvement in out-of-sample predictability as the number of observations increase. The empirical results presented here quantitatively demonstrate the importance of long-term temporal data collection programs for improving ecosystem models and forecasts, and to better support environmental management actions.

High-throughput sequencing of ocean biomes has revealed vast eukaryotic microbial diversity, a significant proportion of which remains uncharacterized. Here we use a temporal approach to understanding eukaryotic diversity at the Scripps Pier, La Jolla, California, USA, via high-throughput amplicon sequencing of the 18S rRNA gene, the abundances of both Synechococcus and Synechococcus grazers, and traditional oceanographic parameters. We also exploit our ability to track operational taxonomic units (OTUs) temporally to evaluate the ability of 18S sequence-based OTU assignments to meaningfully reflect ecological dynamics. The eukaryotic community is highly dynamic in terms of both species richness and composition, although proportional representation of higher-order taxa remains fairly consistent over time. Synechococcus abundance fluctuates throughout the year. OTUs unique to dates of Synechococcus blooms and crashes or enriched in Synechococcus addition incubation experiments suggest that the prasinophyte Tetraselmis sp. and Gymnodinium-like dinoflagellates are likely Synechococcus grazers under certain conditions, and may play an important role in their population fluctuations.

Summary The Synechococcus cyanobacterial population at the Scripps Institution of Oceanography pier in La Jolla, CA, shows large increases in abundance, typically in the spring and summer followed, by rapid declines within weeks. Here we used amplicon sequencing of the ribosomal RNA internal transcribed spacer region to examine the microdiversity within this cyanobacterial genus during these blooms as well as further offshore in the Southern California coastal ecosystem (CCE). These analyses revealed numerous Synechococcus amplicon sequence variants (ASVs) and that clade and ASV composition can change over the course of blooms. We also found that a large bloom in August 2016 was highly anomalous both in its overall Synechococcus abundance and in terms of the presence of normally oligotrophic Synechococcus clade II. The dominant ASVs at the pier were found further offshore and in the California Current, but we did observe more oligotrophic ASVs and clades along with depth variation in Synechococcus diversity. We also observed that the dominant sequence variant switched during the peak of multiple Synechococcus blooms, with this switch occurring in multiple clades, but we present initial evidence that this apparent ASV switch is a physiological response rather than a change in the dominant population.

The dinoflagellate lineage Syndiniales currently consists entirely of parasitic species that fall into five well-supported clades. Environmental sequencing studies worldwide have found an abundance of Syndiniales in a variety of marine ecosystems, but very little is known about the majority of Syndiniales species including two entire clades which have only been observed in sequence data. Syndiniales are known to have a wide range of hosts, but only a few dozen interactions have been confirmed through observation of actual infections. Here, we describe the diversity of Syndiniales found at the Scripps Institution of Oceanography pier over the course of a year based on 18S sequencing. We find Syndiniales to be the most species (amplicon sequence variant)-rich taxonomic group and for its members to be present and abundant throughout the year. We used several analytical techniques to identify potential parasite-host interactions which we were then able to visualize over time. Using mock communities and size fractionation of seawater, we suggest that the majority of Syndiniales sequences that are found in environmental studies belong to the free-living dinospore stage rather than representing active infections.