Land use and riparian effects on prairie wetland sediment properties and herbicide sorption coefficients.

Sorption of commonly used herbicides by wetland sediment can provide important information for herbicide fate modeling. The influence of sediment properties on herbicide sorption as a result of different land uses in the wetland catchment is unclear. We examined the effects of land use on the physiochemical properties of wetland sediments and the associations between these sediment properties and herbicide sorption characteristics. Bottom sediments were sampled in 0- to 5- and 5- to 10-cm sections from 17 wetlands under five different land use classes: (i) ephemeral wetlands with no riparian vegetation in a cultivated catchment (ECNR), (ii) ephemeral wetlands with riparian vegetation in a cultivated catchment (ECR), (iii) ephemeral wetlands in a grassland catchment established 4 yr ago (E4G), (iv) ephemeral wetlands in a brome grass catchment established 20 yr ago (E20G), and (v) semi-permanent (SP) wetlands in a multiple-land-use catchment. Sediments were analyzed for total organic carbon (TOC), total inorganic carbon (TIC), pH, electrical conductivity, exchangeable cations (EXCAT), total cation exchangeable capacity (CEC), and percent clay (%clay). Sediment herbicide sorption partition coefficient (Kd) was measured for trifluralin, atrazine, 2,4-D, and glyphosate. The sorption of the herbicides in the sediment increased in the order of 2,4-D < atrazine < glyphosate < trifluralin. The sorption of 2,4-D, atrazine, and trifluralin was positively correlated to TOC, EXCAT, and CEC but negatively correlated to %clay. Glyphosate sorption was negatively correlated to pH, TIC, EXCAT, and %clay. Overall, wetland sediments that were recently cultivated (ECNR and E4G) had lower TOC, TIC, EC, EXCAT, CEC, and Kd values (2,4-D, trifluralin, and atrazine) than sediments that had not been recently cultivated (ECR, E20G, and SP). The ECR wetland sediments had the largest Kd for all four herbicides, suggesting that land use and riparian vegetation have a significant impact on herbicide sorption.

Regional assessment of herbicide sorption and degradation in two sampling years.

Sorption and degradation of the herbicide 2,4-D [2,4-dichlorophenoxyacetic acid] were determined for 123 surface soils (0 to 15 cm) collected in 2002 and in 2004 between 49 degrees to 60 degrees north longitude and 110 degrees to 120 degrees west latitude in Alberta, Canada. The soils were characterized by soil organic carbon content (SOC), pH, electrical conductivity, soil texture, cation exchange capacity, carbonate content, and total soil microbial activity. The 2,4-D sorption coefficients, Kd and Koc, were highly variable with coefficients of variation of 89 and 59%, respectively, at the provincial scale. Both Kd and Koc were well described by regression models with SOC and soil pH as variables, regardless of scale. Surprisingly, variations in 2,4-D mineralization were much smaller than variations in sorption. Variability in total 2,4-D mineralization was particularly low, with a coefficient of variation of only 7% at the provincial scale. Average 2,4-D half-lives in ecoregions ranged from 1.7 to 3.5 d, much lower than the field dissipation half-life of 10 d reported for 2,4-D in general pesticide property databases. Regression models describing degradation parameters were generally poor or not significant because 2,4-D mineralization was only weakly associated with measured 2,4-D sorption parameters and soil properties. As such, regional variations in herbicide sorption coefficients should be measured or calculated based on soil properties, to assign distinct pesticide fate model input parameters when estimating 2,4-D off-site transport at the provincial scale. Spatial variations in herbicide degradation appear less important for Alberta as 2,4-D half-lives were similar in soils across the province. The rapid mineralization of 2,4-D is noteworthy because 2,4-D is widely used in Alberta and perhaps adaptation of soil microbial communities allowed for accelerated degradation regardless of soil properties or the extent of 2,4-D sorption by soil.

2,4-D Mineralization in soil profiles of a cultivated hummocky landscape in Manitoba, Canada.

The objective of this study was to quantify 2,4-D (2,4-dichlorophenoxyacetic acid) mineralization in soil profiles characteristic of hummocky, calcareous-soil landscapes in western Canada. Twenty-five soil cores (8 cm inner diameter, 50 to 125 cm length) were collected along a 360 m transect running west to east in an agricultural field and then segmented by soil-landscape position (upper slopes, mid slopes, lower slopes and depressions) and soil horizon (A, B, and C horizons). In the A horizon, 2,4-D mineralization commenced instantaneously and the mineralization rate followed first-order kinetics. In both the B and C horizons, 2,4-D mineralization only commenced after a lag period of typically 5 to 7 days and the mineralization rate was biphasic. In the A horizon, 2,4-D mineralization parameters including the first-order mineralization rate constant (k(1)), the growth-linked mineralization rate constant (k(2)) and total 2,4-D mineralization at the end of the experiment at 56 days, were most strongly correlated to parameters describing 2,4-D sorption by soil, but were also adequately correlated to soil organic carbon content, soil pH, and carbonate content. In both B and C horizons, there was no significant correlation between 2,4-D mineralization and 2,4-D sorption parameters, and the correlation between soil properties and 2,4-D mineralization parameters was very poor. The k(1) significantly decreased in sequence of A horizon (0.113% day(-1)) > B horizon (0.024% day(-1)) = C horizon (0.026% day(-1)) and in each soil horizon was greater than k(2). Total 2,4-D mineralization at 56 days also significantly decreased in sequence of A horizon (42%) > B horizon (31%) = C horizon (27%). In the A horizon, slope position had little influence on k(1) or k(2), except that k(1) was significantly greater in upper slopes (0.170% day(-1)) than in lower slopes (0.080% day(-1)). Neither k(1) nor k(2) was significantly influenced by slope position in the B or C horizons. Total 2,4-D mineralization at 56 days was not influenced by slope positions in any horizon. Our results suggest that, when predicting 2,4-D transport at the field scale, pesticide fate models should consider the strong differences in 2,4-D mineralization between surface and subsurface horizons. This suggests that 2,4-D mineralization is best predicted using a model that has the ability to describe a range of non-linear mineralization curves. We also conclude that the horizontal variations in 2,4-D mineralization at the field scale will be difficult to consider in predictions of 2,4-D transport at the field scale because, within each horizon, 2,4-D mineralization was highly variable across the twenty-five soil cores, and this variability was poorly correlated to soil properties or soil-landscape position.