Indicateurs agroenvironnementaux adaptés à la gestion de projets ciblés sur la prévention de la pollution diffuse par le phosphore
|Authors:||Michaud, Aubert Raymond|
|Advisor:||Laverdière, Marc R.|
|Abstract:||Reduction of non-point source phosphorus (P) related to agricultural land use has been identified as a priority for Missisquoi Bay, a water body impaired by cyanobacteria blooms and shared by Quebec province, Canada, and Vermont state, USA. The objectives of a research initiated in 1997 within the Beaver experimental watershed (11 km2) were 1) to describe the non-point source P transfer to the aquatic ecosystem through landscape and agricultural production systems descriptors and 2) to evaluate the effectiveness of surface runoff management on P transfer. A simulated rainfall study conducted on an array of 36 runoff plots deployed on three benchmark soils of the watershed demonstrated the high level of interaction among cropping (tilled vs hay) and manure treatment and the site-specific influence of soil physical and chemical properties on P export and bioavaibility. Variance analysis indicated significant contributions of treatments in explaining total P concentration in runoff as follows: Soil type effect > Cropping effect > Soil*Cropping interaction > Manure effect. Annual exports of dissolved reactive P and total P averaged 0.65 and 1.53 kg-P ha-1, respectively, over the six years watershed monitoring period. These export loads were temporally episodic in nature and occurred dominantly during peak stream flow events and under late-winter/early-spring conditions when overland runoff was generated through snowmelt or precipitation on thawing or saturated soils. Under base flow conditions, the flux-stream discharge relationship indicated that P point sources contributed between 3 to 11% of annual exports. During elevated stream flow conditions, high bioavailability and particulate ratio of P exports indicated in-stream P storage and transformation mechanisms, as well as a significant contribution of subsurface P transport to the main stream channel. Covariance analysis of water quality data using stream flow as a covariate highlighted a landscape-driven hydrologic control on the spatial pattern in P transfer, as well as an influence of manure P sources management. Temporal trend in water quality from the downstream station also indicated a 25% reduction in total P flow-weighted concentration during high flows in response to the establishment of riparian buffers and catch basins along the stream main stem.|
|Document Type:||Thèse de doctorat|
|Open Access Date:||11 April 2018|
|Collection:||Thèses et mémoires|
All documents in CorpusUL are protected by Copyright Act of Canada.