The seasonal forcing of pelagic communities invokes a succession of the dominant phytoplankton and zooplankton species. Here, we characterize the seasonal succession of the plankton traits and their interactions using observations and model simulations of the plankton community in the western English Channel. We focus on activity traits that characterize the defensive and feeding abilities of zooplankton and distinguish between low risk, low return ambush feeders and high risk, high return feeding‐current feeders. While the phytoplankton succession depends on traits related to nutrient acquisition and photosynthesis, it also depends on grazing which couples feeding and motility traits across trophic guilds. Despite interannual variations in the species dominating the protist plankton community, the seasonal trait distribution reveals robust and repeatable seasonal patterns, changing between non‐motile cells flourishing in spring and motile community dominating during summer. The zooplankton community is dominated by active feeding‐current feeders with peak biomass in the late spring declining during summer. The model reveals how zooplankton grazing reinforces protist plankton seasonal succession and shows how the physical environment controls the vertical structure of plankton communities, where ambush feeders exhibit a preference for greater depths during summer. We characterize the seasonal succession as trophic trait coupling and conjecture that this coupling leads to a trophic trait cascade where successive trophic levels alternate in their expression of activity traits further up in the food chain.

Modern in situ digital imaging systems collect vast numbers of images of marine organisms and suspended particles. Automated methods to classify objects in these imageslargely supervised machine learning techniquesare now used to deal with this onslaught of biological data. Though such techniques can minimize the human cost of analyzing the data, they also have important limitations. In training automated classifiers, we implicitly program them with an inflexible understanding of the environment they are observing. When the relationship between the classifier and the population changes, the computer's performance degrades, potentially decreasing the accuracy of the estimate of community composition. This limitation of automated classifiers is known as "dataset shift." Here, we describe techniques for addressing dataset shift. We then apply them to the output of a binary deep neural network searching for diatom chains in data generated by the Scripps Plankton Camera System (SPCS) on the Scripps Pier. In particular, we describe a supervised quantification approach to adjust a classifier's output using a small number of human corrected images to estimate the system error in a time frame of interest. This method yielded an 80% improvement in mean absolute error over the raw classifier output on a set of 41 independent samples from the SPCS. The technique can be extended to adjust the output of multi-category classifiers and other in situ observing systems.

Many aquatic microbes form colonies, yet little is known about their abundance and fitness relative to single-celled taxa. The formation of diatom chains, in particular, has implications for diatom growth, survival, and carbon transfer. Here, we utilize an autonomous underwater microscope, combined with traditional microscopy, to develop a novel, multiyear record of the abundance of single-cell and colony-forming diatoms at Scripps Pier, a coastal location in the Southern California Bight. The total abundance of diatoms was lower during the warmer and more stratified conditions from 2015 to early 2016, but increased in cooler and less stratified conditions in mid-2016 to late 2017. Diatom blooms were dominated by chain-forming taxa, whereas solitary diatoms prevailed during low-biomass conditions. The abundance of dinoflagellates, some of which are important diatom predators, is highest when colonies (chains) are most abundant. These observations of the diatom assemblage are consistent with a trade-off between resource acquisition and predator defenses. Solitary diatom cells dominated during conditions with weak nutrient supply because they have a greater diffusive catchment area per cell in comparison to cells living in colonies. In contrast, during bloom conditions when nutrient supply is high and predators are abundant, forming a colony may reduce predation losses to quickly growing microzooplankton predators, and afford chains a higher fitness despite the costs of sharing resources with neighboring cells. These results highlight the contrasting ecology of single-cell and chain-forming diatoms, and the need to differentiate them in monitoring campaigns and ecological models.