Suivi et modélisation de la dynamique des cyanobactéries dans les lacs urbains au sein de leur bassin versant

Abstract

Urban lakes, often artificial, are fully integrated into the urban environment and perform several roles that are important for the sustainable functioning of cities. These ecosystems are strongly affected by human activities such as urbanization, land use changes and surface imperviousness which raise runoff in volume and speed, causing greater carrying capacity and greater pollutant load to aquatic receptors. Because of these anthropogenic activities, many lakes in the world are eutrophic and frequently affected by cyanobacterial blooms, some species of which are potentially toxic. The objective of this thesis is to contribute to a better understanding of the factors involved in the dynamics of cyanobacteria in urban lakes. Our research was conducted on two study sites: Lake Enghien in France and Lake Pampulha in Brazil. In the French study site, we benefited from new technologies for monitoring phytoplankton to model the dynamics of cyanobacteria in short-term scale. In the Brazilian lake, our main study site, the modelling of the cyanobacteria dynamics was carried out taking into account the inclusion of the lake in its catchment through and integrated modelling: the model used to simule lake cyanobacteria is coupled to a hydrological model for simulating the runoff quantity and the quality, which is inputted into the lake. Both models were calibrated and validated using measurements collected during a 21-month monitoring program. The performance of a spectrofluorometric probe in estimating phytoplankton biomass in this hypertrophic lake was evaluated. On this study site, our results showed that: · The spectrofluorometric probe underestimates the cyanobacterial biomass when (i) Oscillatoria cyanobacteria are present because they have a lower fluorescence per unit of chl-a, (ii) colonial cyanobacteria represent more than 20% of the cyanobacterial biomass and (iii) spectrofluorometric total chlorophyll is above 100 g chl-a.L-1. · The hydrological model showed a good performance for runoff simulation (Nash coefficient is between 0.70 and 0.88 in calibration and 0.72 and 0.78 in validation) and moderately satisfactory results for pollutants simulation. The lake model, developed in steps, showed good predictive ability of the cyanobacteria dynamics in its most complete version, NMAE = 0.26 (calibration) and 0.55(validation). · Runoff is a major source for suspended solids, phosphate and nitrate influxes into Lake Pampulha. Ammonium comes mainly from wastewater discharged into tributaries. Rainfall events are responsible for the mixing of the water column and disrupting algal growth during the wet season. · Cyanobacteria in Lake Pampulha were limited by phosphorus during most of our monitoring. Mixing episodes of the lake water column have different effects on the availability of nutrients in the surface layers. The research carried out in Lake Pampulha allows us to progress on the coupling between the quantitative and qualitative aspects of the water cycle in urban areas and provides a tool that can be used in the simulation of scenarios of the catchment changements and their impact on the lake. The methodology developed in this thesis can be applied elsewhere, taking advantage of existing knowledge on Lake Pampulha.