Evaluation des risques environnementaux dans le parc marin du Saguenay-Saint-Laurent (Quebec, Canada

Since its foundation in 1998, the Saguenay-St. Lawrence Marine Park (SSLMP), located in the Province of Quebec, is protecting a significant portion of the St. Lawrence Estuary and the Saguenay Fjord. In this large coastal area, the important biodiversity and complexity of the ecosystem are facing human activities and their multiple impacts. In support to conservation goals, we adapted the relative risk model (RRM) as a management tool to highlight the zones the most at risk, identify the main sources of stress and develop environmental strategies to reduce global risk level in the Park.;To build such a model we focussed on variables governing environmental risk in this marine area and proceeded to an update of the necessary dataset to develop our tool. The relative significance of main land-based inputs on the global environmental contamination were first developed and then we examined the chemical and microbial contamination of several compartments of the ecosystem in the main semi-enclosed bays and streams of the Saguenay St. Lawrence Marine Park (SSLMP) . Our study provided a first estimation of each source of contamination, particularly local ones, as only a few unpublished data were previously available. Results showed a ubiquitous contamination but a heterogeneous distribution of the contamination in the ecosystem of the marine park in relation with the location of human activities. Particularly, urban effluents were highlighted as an important source of stress at a local scale but a low contributor to global contamination compared to diffuse external inputs from the St. Lawrence and Saguenay rivers, which relative apportionment was estimated from 88 to 100% depending on the analysed pollutant. This huge discrepancy between sources is mainly driven by their daily discharge but also by sorne local hydrodynamic factors discussed in this paper. Moreover, a closer look at the polycyclic aromatic hydrocarbons (PAH) distribution in intertidal sediments revealed the influence of maritime traffic. Finally, unexpected data were subject to a special focus. In the Bay of Sainte-Marguerite, an important site for belugas (endangered species of the Park), in absence of any local source of contamination, the highest level of contamination for PAH and sorne metals was recorded. Sorne hypotheses are discussed to explain observations and provide insight for further works on contamination fate in this bay.;The information derived from the bulk analysis and multi-contaminant data provided good insights to enforce our understanding of anthropogenic stress in the SSLMP. Then, dataset was integrated to develop our tool. This first deployment of the methodology over a marine conservation area al/owed to depict the regional environmental pressure from multiple anthropogenic sources of stress in the SSLMP. The RRM highlighted the mouth of XVI the Saguenay Fjord and the coastal area of the Lower St. Lawrence Estuary as the subregions to be the most at risk and incriminated vessel traffic, sewage discharges and urban runoff as the main sources of stress in those areas. Moreover, the flexibility of this risk model allowed us to proactively explore future management options (to deal with environmental risks) and assess their probable benefits. The RRM predicts that the implementation of wastewater treatment plans (to solve contamination problems) and a reduction of the navigation pressure would decrease the global risk from 25 to 32 % in the sub-regions identified as the most at risk. Finally, uncertainty and sensitivity analysis was used to provide preliminary insights on the RRM behaviour and performance, and to assist in guiding possible fine tuning in the data integration and focusing future studies to optimize the RRM and its uses.