Materials And Methods

Measurements were made in a 32 ha field within the Walnut Creek watershed in central Iowa (41°58 ' N, 93°43' W). A detailed description of the soils, geology, and farming practices within the watershed can be found in Hatfield et al. (1999). The landscape within the watershed is characterized by gentle swell-swale relief of several meters (Daniels and Handy, 1966). Surface drainage is poorly developed, resulting in numerous closed depressions or potholes that have been extensively tile drained in the past 100 years.

Soils within the field were formed from till from the most recent substage of the Wisconsin glaciation. The toposequence of soils within the field range from well-drained Clarion loam (fine-loamy, mixed, superactive, mesic Typic Hapludolls), to the somewhat poorly drained Nicollet loam (fine-loamy, mixed, superactive, mesic, Aquic Hapludolls), to the poorly drained Canisteo silty clay loam (fine-loamy, mixed, superactive, calcareous, mesic Typic Endoaquolls) and Harps loam (fine-loamy, mixed, superactive, mesic Typic Calciaquolls), and ending with the very poorly drained Okoboji mucky silty loam (fine, smectitic, mesic Cumulic Vertic Endoaquolls) (Figure 14.1).

Prior to planting in 1992, a grid was laid out across a 250 m by 250 m area in the southern half of the field (Figure 14.1). Detailed information about the grid and soil properties not reported here can be found in Cambardella et al. (1994) and Novak et al. (1997). The grid spacing was 25 m in both the easterly and northerly directions. Additional grid points were established at closer spacings but were not included in this study. Within 1 m of each grid point, three 6-cm diameter soil cores were taken to a depth of 15 cm and composited for analysis. The mass of organic carbon was measured using dry combustion methods with a Carlo-Erba NA1500 NCS elemental analyzer (Haake Buchler Instruments, Paterson, NJ) after carbonates had been removed with 2M H2SO4.

+ Sampling grid • Transects

Soils

55 Nicollet 90 Okoboji 95 Harps 138 Clarion 507 Canisteo

FIGURE 14.1 Grid locations for soil sampling and apparent electrical conductivity, ECa, measurements, and transects where additional ECa measurements were made superimposed on the soils map for a 32 ha field.

+ Sampling grid • Transects

Soils

55 Nicollet 90 Okoboji 95 Harps 138 Clarion 507 Canisteo

FIGURE 14.1 Grid locations for soil sampling and apparent electrical conductivity, ECa, measurements, and transects where additional ECa measurements were made superimposed on the soils map for a 32 ha field.

Atrazine sorption was determined on duplicate 4 g samples of air-dried soil that had passed a 2 mm sieve. Soil was equilibrated with 15 ml of solution containing 1.5 mg L-1 atrazine dissolved in 0.01M CaCl2 for 72 h. The solution was then centrifuged, and atrazine in the supernatant analyzed with high-performance liquid chromatography (HPLC) as outlined by Novak et al. (1994).

At the same time soil samples were collected, apparent electrical conductivity measurements (ECa) were made with an EM38 electromagnetic induction meter (Geonics Limited, Ontario, Canada). A single reading was made at each grid point with the meter in the vertical dipole orientation and suspended 20 cm above the soil surface. Readings were taken with the meter suspended above the soil surface rather than on the surface so that the results were directly applicable to a companion transect survey. A vertical rather than horizontal orientation was used so that slight variations in the spacing between the meter and soil surface would have minimal effect on the readings (Rao and Davidson, 1980) during the transect survey. A horizontal orientation may have resulted in values better correlated to the surface soil properties measured for this study. Two days after making ECa measurements at the grid locations, a transect survey of the entire 32 ha field was made. The EM38 was attached to a wooden boom suspended 20 cm above the ground and pulled by a five-wheel, all-terrain vehicle equipped with a GPS (Ambuel et al., 1991). Electrical conductivity was recorded along with position at 2 s intervals while driving transects across the field (Figure 14.1).

Contour maps of the values collected within the intensive grid area were produced using Surfer software (Golden Software, Inc., Golden, CO) based on a 41 x 41 grid produced by a linear kriging interpolation procedure embedded within the software. A contour map of ECa from the transect survey of the entire 32 ha field was produced in a similar manner for a 50 x 100 grid using the relationship found between ECa and Kd.

Descriptive statistics and regressions were computed using standard methods (Draper and Smith, 1966; Snedecor and Cochran, 1967). Correlograms were calculated using the method described in Davis (1973). The spatial structure of the mapped properties was quantified using Moran's I statistic (Moran, 1950; Upton and Fingleton, 1985). Moran's I is similar in concept to correlation and ranges from -1 to 1 where values near -1 indicate that like members of a group are evenly interspersed across the field like the colored squares of a checkerboard, and values near 1 indicate that members are grouped closely together in space. Moran's I was calculated using the Excel 97/2000 Visual Basic routine written by Sawada (1999).

0 0

Post a comment