Functional Stability of a Coastal Mediterranean Plankton Community During an Experimental Marine Heatwave

Abstract

As heatwaves are expected to increase in frequency and intensity in the Mediterranean Sea due to global warming, we conducted an in situ mesocosm experiment for 20 days during the late spring and early summer of 2019 in a coastal Mediterranean lagoon to investigate the effects of heatwaves on the composition and function of coastal plankton communities. A heatwave was simulated by elevating the water temperature of three mesocosms to +3°C while three control mesocosms had natural lagoon water temperature, for 10 days. Further, the heating procedure was halted for 10 days to study the resilience and recovery of the system. Automated high frequency monitoring of dissolved oxygen concentration and saturation, chlorophyll-a fluorescence, photosynthetic active radiation, salinity, and water temperature was completed with manual sampling for nutrient and phytoplankton pigment analyses. High-frequency data were used to estimate different functional processes: gross primary production (GPP), community respiration (R), and phytoplankton growth (μ), and loss (l) rates. Ecosystem stability was assessed by calculating resistance, resilience, recovery, and temporal stability in terms of the key functions (GPP, R, μ, and l). Meanwhile, the composition of phytoplankton functional types (PFT) was assessed through chemotaxonomic pigment composition. During the heatwave, GPP, R, μ, and l increased by 31, 49, 16, and 21%, respectively, compared to the control treatment. These positive effects persisted several days after the offset of the heatwave, resulting in low resilience in these key functions. However, GPP and R recovered almost completely at the end of the experiment, suggesting that the effect of the heatwave on these two rates was reversible. The heatwave also affected the PFT composition, as diatoms, prymnesiophytes, and cyanobacteria were favored, whereas dinoflagellates were negatively affected. By highlighting important effects of a simulated marine heatwave on the metabolism and functioning of a coastal Mediterranean plankton community, this study points out the importance to extend this type of experiments to different sites and conditions to improve our understanding of the impacts of this climate-change related stressor that will grow in frequency and intensity in the future.

Metabolic responses of plankton to warming during different productive seasons in coastal Mediterranean waters revealed by in situ mesocosm experiments

“…Finally, the chl- a sensor data were corrected for non-photochemical quenching (NPQ), as a strong decrease in chl- a occurred in almost all daily chl- a cycles during sunlight. To account for NPQ, chl- a data were linearly interpolated between sunrise and sunset 49 , 50 . This interpolation assumed that spatial homogeneity in the phytoplankton community inside each mesocosm was achieved with the use of a pump to gently mix the water column 51 .…”

“…Phytoplankton µ and l were estimated using the corrected chl- a fluorescence data and a method detailed in Soulié et al 50 and following similar principles as in Neveux et al 53 , 54 . The daily chl- a cycle was separated into two parts: the “increasing period,” which starts from sunrise until the fluorescence maximum is reached (generally a few hours after sunset), and the “decreasing period,” from the time of maximum fluorescence until the following sunrise.…”

“…The instantaneous change in dissolved oxygen is considered to depend on GPP, R, and on F, which represents the physical oxygen exchange between the water and the atmosphere, and A, which encompasses all other phenomena which could affect the dissolved oxygen concentration. A was taken as null in the present work as in most other studies 21 , 50 , 57 . F was calculated as follows (Eq.…”

“…In this equation, k represents the piston velocity coefficient, and the concentration and saturation of dissolved oxygen respectively, and the water column mixing depth, which is the mesocosm water column length in the case of mixed mesocosms 50 , 52 . The k value was taken as k = 0.000156 m min −1 from the literature 58 .…”

“…The k value was taken as k = 0.000156 m min −1 from the literature 58 . Then, k was corrected for temperature and salinity using the high-frequency sensors data and following the procedure described in Soulié et al 50 .…”