Because of their large population sizes and rapid cell division rates, marine microbes have, or can generate, ample variation to fuel evolution over a few weeks or months, and subsequently have the potential to evolve in response to global change. Here we measure evolution in the marine diatom Skeletonema marinoi evolved in a natural plankton community in CO2-enriched mesocosms deployed in situ. Mesocosm enclosures are typically used to study how the species composition and biogeochemistry of marine communities respond to environmental shifts, but have not been used for experimental evolution to date. Using this approach, we detect a large evolutionary response to CO2 enrichment in a focal marine diatom, where population growth rate increased by 1.3-fold in high CO2-evolved lineages. This study opens an exciting new possibility of carrying out in situ evolution experiments to understand how marine microbial communities evolve in response to environmental change.

Densities of submerged vegetation and those of associated animals tend to co‐vary. This relationship is often attributed to the positive correlation between the density of vegetation and its protective value against predators. However, two counteracting basic elements underlying this paradigm limit its generality. That is, increasing vegetation density should result in decreased predator–prey encounters, whilst at the same time predator–prey encounters should increase as animal densities increase. These two mechanisms should thus counteract each other when the densities of vegetation and associated animals, including both prey and predators, co‐vary. Experimental designs that expose fixed densities of prey and predators to varying densities of vegetation assess only the former mechanism and may thus not properly evaluate the protective value of vegetation in such conditions. By contrast, designs that mimic the naturally co‐varying organism densities test both mechanisms and thus their counteractive impacts on predator–prey encounters. We compared the outcomes of the two alternative designs and carried out additional experiments to explain the putative discrepancy. Increasing vegetation density (mimics of Potamogeton pectinatus) enhanced prey (Daphnia magna) survival only when fixed densities of prey and predators (Perca fluviatilis or Rutilus rutilus) were used. When the animal densities were allowed to co‐vary with vegetation density, vegetation had no impact on prey survival. Instead, prey survival was determined by the aggregate density of prey and predators, shaped by the species‐specific traits of the latter. Thus, the impact of the increased animal densities overrode the impact of the increased vegetation density on predator–prey encounters. It may be insufficient to attribute the co‐variation of vegetation, prey and predator densities simply to the association between vegetation density and its protective value. Increased food resources and reduced competition within vegetation may promote prey and thereby also predator abundance to a greater extent than previously thought.

Productivity and trophic status of aquatic systems is traditionally quantified by chlorophyll a measurements. Environmental conditions and ecological interactions cause variability in chlorophyll a abundance. In coastal ecosystems, shallow and complex bathymetry reduces vertical heterogeneity, but promotes horizontal heterogeneity. However, coastal monitoring programs and scientific surveys are primarily focused on the vertical dimension. Here we demonstrate the spatial patchiness of chlorophyll a in coastal waters. We collected horizontally detailed and extensive in situ chlorophyll a data from the coastal Baltic Sea (SW Finland), covering the ice-free season of an annual cycle. Altogether, more than 200,000 observations were logged by an automated underway measurement system equipped with an optical sensor connected to a flow-through system. We analyzed the spatial heterogeneity of calibrated chlorophyll a data by using multiple statistical approaches, and quantified the chlorophyll a patches using a rolling average filter. We were able to identify patches and quantify their abundance and size for each of the 11 sampling campaigns. On average, 285 patches, ranging from 0.6 to 3142 m in size, were observed on the 830 km sampling transect. The average size of the patches was 237 (95% CI 226-248) m, most patches being between 10 and 1000 m. Our results show that patches of chlorophyll a can be effectively identified and quantified by modern in situ optical instrumentation. Such information is both theoretically and practically relevant. First, these results increase our understanding of the overall heterogeneity of the coastal environment. Further, they demonstrate the value of knowing the magnitude and occurrence of chlorophyll a patchiness in accurate detection of changes in coastal ecosystems caused by increased inputs of nutrients.

Multiple biogeochemical processes in estuaries modulate the flux of nutrients from land to sea, thus contributing to the coastal filter. The role of particle dynamics in regulating the fate of terrestrial nutrients in estuaries is poorly constrained. To address this issue, we resolved the particle size distribution of suspended material, and quantified size-fractionated particulate nitrogen (PN) and phosphorus (PP), in a stratified mesotrophic estuary (Pojoviken, Finland). We also carried out a mixing experiment where the effects of salt-induced flocculation on particle size distribution and concentrations of PN and PP were examined. The experimental results showed that saltinduced flocculation at already very low salinities increases the total particle concentration and mean particle size, indicating transfer of dissolved material into particulates. Correspondingly, a significant increase in PP and particulate iron (Fe) was observed in the experiment results, suggesting coupled flocculation of P-containing organic matter (OM) and ferrihydrite. Particle dynamics in the field data were dominated by processes occurring downstream of the flocculation zone. Primary production created a downward flux of autochthonous OM particles, promoting passive aggregation by random collisions with terrestrial material in the water column. Maximum particle concentrations were observed at and below the halocline. The highest PN and PP concentrations were observed in the subhalocline layer, 3.5 and 0.14 μmol L À1 , respectively. Molar ratios of N:P in this material were >40, consistent with typical marine snow in the early stages of microbial processing. Our study provides a mechanistic overview of the biogeochemical drivers of particulate nutrient dynamics in stratified estuarine environments.