Increased flux of carbon and nutrients from human activities in river basins were linked to acidification and deepwater hypoxia in estuaries and coastal areas worldwide. Annual loads (1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011) of suspended particulate matter (SPM), dissolved organic carbon (DOC), total nitrogen (TN) and total phosphorus (TP) were assessed at the Lake Ontario inlet of the St. Lawrence River (SLR) (7110 m 3 s -1 ) and its estuarine outlet at Québec City (12,090 m 3 s -1 ). Internal loads from the Ottawa River (1950 m 3 s -1 ), seventeen other tributaries, urban wastewaters, atmospheric deposition and erosion were also estimated. Erosion (65% of SPM, 29% of TP), inflow from Lake Ontario (42% of DOC, 47% of TN) and Ottawa River (28% of DOC) contributed important flux to the estuary. Loads from other tributaries (20 and 27% of TN and TP at Quebec City) largely exceeded municipal sources (6% of exported TN and TP) and require future remediation. Aquatic plants fixed 277,000 t of C, 49,000 t of N and 7000 t of P (May-Sept.), delaying the nutrient flux to the estuary and turning the SLR into a nutrient sink over summers of lowest discharge. Degradation of exported organic C could consume 5.4-7.1 million t O 2 year -1 in the estuary whereas SLR flux of N and P represent 31-47% and 7-14% of total annual estuarine flux, respectively. Carbon and Nitrogen flux from freshwaters partly explain the decline in pH and oxygen concentrations in deep estuarine waters thus highlighting the need to reduce diffuse sources of nutrients in the entire watershed.

The aim of this work is to evaluate and discuss river water quality trends over the last decade in ten watersheds where diffuse pollution represents more than half of the annual load of phosphorus (P) and nitrogen (N). Trend analyses taking into account flow data indicate a significant reduction of total P in eight rivers, of ammonia N in five rivers, of nitrate + nitrite in four rivers, of total filtered N in three rivers and of suspended solids in two rivers. An increase of turbidity was observed in four rivers and, for fecal coliforms, no trends. P decrease can be explained by reduced mineral P inputs on cropped lands related to means such as agro-environmental fertilization plans and addition of phytase in pig and poultry feed. However, for seven of them, median P concentrations remain at least two times greater than the Québec water quality guideline for protection of rivers against eutrophication. Concentrations of other parameters remain problematic in some rivers too. These results indicate the need to continue the efforts for further diffuse pollution reduction. Future work should better quantify actions taken at the watershed scale to reduce diffuse pollution.

La contamination bactériologique des eaux de surface en milieu agricole préoccupe les intervenants des secteurs agricole et de la santé publique au Canada. Afin de connaître l’influence de l’agriculture sur cette contamination, un suivi des coliformes fécaux (CF) a été réalisé dans 16 cours d’eau au Québec méridional. Ces cours d’eau drainent des sous-bassins versants de 15 à 100 km2qui comportent une large gamme de densité animale (10 à 520 unités animales•km-2) et supportent une agriculture surtout intensive. Le suivi des CF aux deux semaines pendant trois ans a montré que leur concentration dans les cours d’eau augmente avec la densité animale dans les sous-bassins versants. Des relations développées par régression permettent d’évaluer l’influence des élevages d’animaux et des humains sur les CF mesurés dans les cours d’eau de sous-bassins versants comparables. L’étude a aussi montré l’influence des sources de CF en amont immédiat des stations d’échantillonnage. Ces sources de contamination contribuent à augmenter les valeurs médianes de CF de 3 à 20 fois en saison estivale (mai à octobre) et jusqu’à plus de 100 fois en saison hivernale (novembre à avril). Les relations présentées constituent des outils d’aide à la gestion du risque de contamination bactériologique. Ces outils comportent toutefois plusieurs limites à l’heure actuelle. Ils pourraient être améliorés et complétés en prenant en compte d’autres paramètres.