Voici les éléments 1 - 1 sur 1
- PublicationAccès libreOrigin and behaviour of microorganisms and particles in selected karst aquifer systemsKarst aquifers represent an important groundwater resource world-wide. They are, however, highly vulnerable to contamination due to fast transport and limited contaminant attenuation processes encountered in such systems. Pathogenic microorganisms are among the most frequent and problematic contaminants in groundwater from karst aquifers. It is generally accepted that the presence of faecal indicator bacteria, such as Escherichia coli and enterococci, in groundwater indicates the possible presence of pathogenic bacteria, protozoa and viruses. Monitoring microbial water quality relies, however, on sterile water sampling and subsequent laboratory analyses. Moreover, the sampling intervals and the relatively long time lag between sampling and results are often too long to prevent microbially polluted water entering the distribution network. Therefore, establishing a correlation between the microbial water quality and parameters that can easily be measured online – using it as an “early-warning” parameter – constitutes the first main objective of this work. Most microorganisms in groundwater are harmless. Enumeration of pathogenic and water quality indicator bacteria thus provides an incomplete picture of the microbial assemblages. These microbial communities are of growing interest, as aquifers, which include habitats for a wide range of organisms, are increasingly considered as ecosystems harbouring their proper biocenoses. These ecosystems have to be better understood in order to ensure the sustainable management of groundwater resources. Furthermore, as bacteria are abundant and ubiquitous, thriving even in extreme habitats, the potential use of microbial communities for groundwater biomonitoring has been raised. However, the understanding of microbial diversity, their distribution patterns and their dynamics in karst aquifer ecosystems are actually still in an early stage and rather sparse and constitute therefore the second main objective of this research project. The occurrence, dynamics and transport processes of suspended matter and faecal indicator bacteria were investigated at two karst aquifer systems (the Yverdon-les-Bains and Noiraigue test sites, Switzerland), which are characterised by allochthonous point recharge, and several shallow cave sites located in the unsaturated zone of karst aquifers (the Vers-chez-le-Brandt, Grand-Bochat and Gänsbrunnen test sites, Switzerland). Classical hydrogeological parameters, such as discharge, temperature, electrical conductivity, organic carbon and turbidity, were monitored continuously online, along with event-based, high-frequency analyses of Escherichia coli and enterococci. In order to gain more insight into the origin and behaviour of suspended particles in karst groundwater, particle-size distribution (PSD) was also monitored continuously. Results demonstrated that suspended particles in karst spring water either originated from the remobilisation and scouring of intrakarstic particulate material due to increasing flow velocities (i.e. autochthonous turbidity) and / or the transfer of particles from land surface and sinking streams (i.e. allochthonous turbidity), which entered the system following rainfall events. These allochthonous turbidity events coincided with increased faecal indicator bacteria and organic carbon levels. PSD allowed distinguishing both types of turbidity: autochthonous turbidity consisted of particle concentration increases over a wide range of particle sizes (from colloidal sizes up to 0.1 mm), whereas allochthonous turbidity periods were characterised by a predominance of finer particles (0.9 to 10 µm). PSD is therefore proposed as surrogate indicator for possible microbial contamination of groundwater from karst aquifers. The structure, diversity and temporal variability of microbial communities from a swallow hole draining agricultural land and two connected karst springs (Yverdon-les-Bains karst aquifer system) were investigated using molecular microbiological methods (PCR-DGGE and cloning / sequencing) and related to hydrological and physico-chemical parameters. Storm responses and an annual hydrological cycle were monitored to determine the short- and long-term variability, respectively, of bacterial communities. Statistical analysis of bacterial genetic fingerprints (16S rDNA PCR-DGGE profiles) of spring water samples revealed several clusters that corresponded well with different levels of the allochthonous swallow hole contribution. Microbial communities in spring water samples highly affected by the swallow hole showed low similarities among them, reflecting the high temporal variability of the bacterial communities in the water entering the swallow hole. Conversely, high similarities among spring water samples with low allochthonous contribution provided evidence for a stable autochthonous endokarst microbial community. This autochthonous endokarst microbial community was characterised by a high diversity with only a few dominant species. δ-Proteobacteria, Acidobacteria and Nitrospira species were important members of this community. The high percentage of unknown sequences (i.e. sequences revealing similarities lower than 97 % to already known sequences) suggests that many karst aquifer bacteria are still undiscovered. Moreover, it highlights that karst aquifers are rarely studied, unique habitats and that sequences from shallow subsurface ecosystems are currently underrepresented in databases. In conclusion, the high sensitivity of PSD analysis allows consequently proposing it as an “early-warning” surrogate for real-time monitoring of possible microbial contamination of groundwater from karst aquifers. The method permits optimising water treatment and identifying periods when spring water must be rejected. Furthermore, this study represents a first step towards a better understanding of the microbial ecology in karst aquifer systems and may serve as a starting point for developing biomonitoring tools.