| 论文摘要： |
Cyanobacterial surface scum (here defined as visible Cyanobacteria colonies accumulating at the lake surface) is a harmful phenomenon that negatively affects water quality, human and animal health. Colony-forming Microcystis is one of the most important and ubiquitous genera that can suddenly accumulate at water surfaces. Turbulent water motion, e.g., generated by wind, can vertically disperse this scum layer, which later can re-establish by upward migration of Microcystis colonies. However, the role of wind-generated turbulence in scum formation and development is still poorly understood. Here we present results from a laboratory mesocosm study where we analysed the processes of scum formation and its response to wind-generated turbulence at low wind speed (<= 3.6 m s(-1)). Microcystis colony size and flow velocity at the water surface and in the bulk water were measured using a microscope camera and particle tracking velocimetry. The surface scum formed by aggregation of colonies at the water surface, where they formed loose clusters of increasing size. The presence of large colony aggregations or of a surface film determined the stability of the scum layer. For the largest applied wind speed, most of the aggregations were broken down to sizes <2 mm, which were dispersed to the bulk water. The surface scum recovered quickly from such disturbances after the wind speed decreased. We further observed reduced momentum transfer from wind to water with the growing scum layer. The presence of the scum increased the threshold wind speed for the onset of flow and reduced the flow velocities that were generated above that threshold. This effect was likely caused by the presence of a film of surface-active material at the water surface (surface microlayer), which is related to the presence of Microcystis. Both the small-scale turbulence and surface microlayer might play an important, yet largely unexplored role in Microcystis surface scum development in aquatic ecosystems. Improved understanding of the interplay of both processes will be instrumental for improving current mechanistic models for predicting surface bloom dynamics. (C) 2019 Elsevier Ltd. All rights reserved. |
