Assessing landscape and contaminant point-sources as spatial determinants of water quality in the Vermilion River System, Ontario, Canada.

Abstract

The Vermilion River and major tributaries (VRMT) are located in the Vermilion watershed (4272km2) in north-central Ontario, Canada. This watershed not only is dominated by natural land-cover but also has a legacy of mining and other development activities. The VRMT receive various point (e.g., sewage effluent) and non-point (e.g., mining activity runoff) inputs, in addition to flow regulation features. Further development in the Vermilion watershed has been proposed, raising concerns about cumulative impacts to ecosystem health in the VRMT. Due to the lack of historical assessments on riverine-health in the VRMT, a comprehensive suite of water quality parameters was collected monthly at 28 sites during the ice-free period of 2013 and 2014. Canadian water quality guidelines and objectives were not met by an assortment of water quality parameters, including nutrients and metals. This demonstrates that the VRMT is an impacted system with several pollution hotspots, particularly downstream of wastewater treatment facilities. Water quality throughout the river system appeared to be influenced by three distinct land-cover categories: forest, barren, and agriculture. Three spatial pathway models (geographical, topographical, and river network) were employed to assess the complex interactions between spatial pathways, stressors, and water quality condition. Topographical landscape analyses were performed at five different scales, where the strongest relationships between water quality and land-use occurred at the catchment scale. Sites on the main stem of Junction Creek, a tributary impacted by industrial and urban development, had above average concentrations for the majority of water quality parameters measured, including metals and nitrogen. The river network pathway (i.e., asymmetric eigenvector map (AEM)) and topographical feature (i.e., catchment land-use) models explained most of the variation in water quality (62.2%), indicating that they may be useful tools in assessing the spatial determinants of water quality decline.