Résumés
Résumé
Des concentrations en P excédant le seuil d'eutrophisation sont fréquemment mesurées dans l'eau des affluents du fleuve Saint-Laurent au Québec, Canada. Un enrichissement excessif en P des sols agricoles en serait la source. Une norme relative à la saturation en P des sols a été proposée comme critère de risque de contamination en P des eaux de surface. L'objectif de ce travail est d'étudier le lien entre la richesse en P du sol et la teneur en P de l'eau de drainage dans deux agro-écosystèmes du Québec. Le bassin versant de la rivière Boyer (BVB), dominé par des sols en pente, une forte densité animale et des productions fourragères et les Basses Terres de Montréal (BTM), dominées par des sols plats utilisés pour la production intensive de maïs et une faible densité animale sont étudiés. Le degré de saturation en P des sols du BVB est de 8 à 10 % alors que celui des BTM dépasse très souvent 15 %. Le pool du P organique est plus faible dans les sols des BTM que dans le BVB. La concentration moyenne en P de l'eau de drainage est plus élevée dans les sols du BVB (171 µg L-1) que dans ceux des BTM (98 µg L-1). Elle est corrélée à la teneur en P extrait à l'oxalate des sols argileux et à la teneur en P soluble dans l'eau ou à l'index de sorption en P des sols grossiers. Ces relations sont plus étroites pour la couche 0-5 cm de sol que pour les couches plus profondes. Les résultats de cette étude démontrent qu'il est difficile de prédire la concentration en P de l'eau de drainage de sols contrastants. Regrouper les sols selon leur texture améliore la précision de la prédiction de la teneur en P de l'eau de drainage à partir de leurs propriétés.
Mots-clés:
- Phosphore,
- eau de drainage,
- index de P,
- degré de saturation en P,
- gleysols
Abstract
Concentrations of P higher than the recognised threshold for eutrophication are often measured in the tributaries of the St. Lawrence river, Quebec, Canada. An excessive P enrichment of agricultural soils was identified as the potential cause of this phenomenon. This enrichment results in a decrease in the P sorption capacity of soils and an increased risk of P contamination of surface waters. A norm based on soil P saturation degree (DSPS) was proposed as an agro-environmental criterion to reduce this risk. Relationships between DSPS and other soil P attributes and surface runoff P concentration have been reported in the literature but not for tile-drainage water. The objective of this work is to study the relationship between soil P attributes and drainage water P in two agro-ecosystems of Quebec.
The Boyer watershed (BW), which is dominated by soils with significant slopes, a high animal density and forage production, and the Montreal lowlands (ML), dominated by flat soils, low animal density and mostly used for corn (Zea mays L.), were studied. The A, B and C horizons from soils from the BW were sampled at three locations in tile-drained fields from farms in surplus or not of manure N in regards to crop needs. Soils were sampled in the ML according to a gradient in clay content with the same procedures. Soils were characterised for their pH, particle size distribution, Mehlich 3-extractable P (Pm3) and Al, water-soluble P (Pw), organic P (Po) and oxalate-extractable P (Pox), Al (Alox) and Fe (Feox) contents. Grab tile-drainage samples were taken in triplicate and characterised for total (TP), particulate (PP) and dissolved reactive P (RP) and unreactive P (UP).
The soils of the BW are more acidic and have higher P retention capacities than ML soils. The Pw content of BW soils is lower than in ML ones in spite of comparable Pm3 contents. The DSPS (Pox /Alox + Feox) of BW soils is moderate (8 to 10 %) whereas DSPS in ML soils is often >15 %. The soil P organic pool is much smaller in the ML soils than in those from BW, probably because of more frequent tillage and lower manure C inputs.
The average P concentration (TP) in drainage water was higher in soils from the BW (171 µg L-1) than from ML soils (98 µg L-1). The PP was the main P fraction in drainage water from the two ecosystems. The RP was on average 44 % of TP whereas UP was much less. The TP concentration in drainage waters was higher than 0.03 mg L-1 in most cases. The TP concentration was higher in tile-drainage waters from the BW than in the ML even though comparable Pm3 and lower DSPS were found in the BW than in ML soils. These results suggest that other criteria should be included in the prediction of the risk of P contamination of drainage waters. The results of the present study indicate that agricultural practices (crop species, manure inputs, tillage type and frequency) could have a greater influence than soil P status on the TP concentration in tile-drainage waters.
A linear correlation analysis between the logarithm of the TP (TP + 0.5) and that of the different soil P attributes indicated that TP was related to the P extracted by oxalate from clay soils and by water in coarse-textured soils. These relationships were closer in the 0-5 cm soil layer than in deeper strata. Multiple regression analysis between the logarithms of drainage water P concentrations and soil attributes revealed that, when soils were grouped by texture, the prediction of TP was much more accurate than when all soils were considered. The present regulatory approach of assessing the risk of water P contamination by using the soil Pm3 and DSPS only was inadequate in the two considered agroecosystems.
The results of this study indicate that the prediction of the drainage water P concentration with P attributes from contrasting soils is difficult. Grouping soils by texture improves the prediction of Pt from soil P attributes.
Keywords:
- Phosphorus,
- drainage water,
- P index,
- P saturation degree,
- gleysols