Delineation and Analysis of Site-specific Management Zones

Site-specific management requires the development of agronomic strategies for sub-field management zones that are subject to a unique combination of potential yield-limiting factors. Creation of unique management areas has focused mainly on soil type surveys, yield mapping, and soil fertility management based on grid soil sampling. This paper discusses the use of the fuzzy k-means unsupervised continuous clustering algorithm to delineate withinfield management zones using soil and field characteristics (i.e., landscape features). Soil electrical conductivity and topographic attributes such as elevation and slope measured in two claypan soil fields were used for the clustering process. Measures of cluster performance indicated no advantage of dividing these fields into more than four or five management zones. Classification using this procedure explained 10 to 35% of the variation in grain yield. Year to year differences in the appropriate number of management zones was attributed to weather and crop type. 1.0 INTRODUCTION The key concept of site-specific management is to identify and manage spatially coherent regions within the geographic area defined by field boundaries. In order to attain maximum efficiency of crop inputs, these regions or management zones should represent a homogenous combination of potential yield-limiting factors. Determination of sub-field areas is difficult due to the complex combination of factors that may affect crop yield. In some cases the factors affecting yield may interact with each other (e.g., nitrogen and phosphorus fertility). Additionally, improperly defining management zones may be no better than uniform management of the field. A number of procedures have been used to delineate within-field management zones for site-specific management. One approach uses relatively stable soil properties such as soil electrical conductivity (EC) and/or landscape features in conjunction with soil-landscape models to estimate patterns of soil variability. Topographic attributes and landscape position data have been widely used to map within-field areas of high and low productivity based on water availability (Jones et al., 1989; Jaynes et al., 1994; Sudduth et al., 1997). In these studies, footslope positions generally out-yielded side-slope positions unless ponding resulted from poor drainage. Soil EC has also been used to investigate yield variability caused by soil water differences (Jaynes et al., 1994; Sudduth et al., 1994). Kitchen et al. (1998) compared the use of traditional soil surveys and a map overlay approach based on topsoil depth and elevation to delineate within-field management zones. They concluded that the map overlay method has the advantage of being based on georeferenced measurements that are repeatable, unlike traditional soil surveys. However, the map overlay approach is dependent upon arbitrary classification criteria defined by the user that establishes the number of classes and class breaks for each variable. Because of the continuous nature of soils data, classification systems that allow any one observation to belong to exactly one class are often inappropriate. Fuzzy or continuous classification procedures were developed for use in situations where the class boundaries are not, nor can they be, sharply defined. In contrast to crisp classification, continuous classification procedures allow individuals to have partial class membership (i.e. an individual can belong to more than one class) (Burrough et al., 1992). Continuous classification has been widely * Presented at the Second International Conference on Geospatial Information in Agriculture and Forestry, Lake Buena Vista, Florida, 10-12 January 2000. used for soil classification and delineation of management units (McBratney and deGruijter, 1992; Odeh et al., 1992; Boydell and McBratney, 1999; Lark, 1998). Previous work (Fraisse et al., 1999) showed that soil EC, elevation, and slope were the most useful attributes for the delineation of within-field management zones. The objective of this study was to develop a systematic procedure to delineate within-field management zones using soil EC, elevation, and slope. The use of unsupervised fuzzy classification for delineation of within-field management zones was evaluated. Further analysis of the identified zones was conducted using a commercial GIS software package. Finally, measures of cluster performance and yield data were used to evaluate the “goodness” of the resulting potential management zones and to determine the optimal number of zones for a given field. 2.0 MATERIALS AND METHODS 2.1 DESCRIPTION OF RESEARCH FIELDS Soil EC and elevation data were collected on two adjacent fields, 13 ha (Field 1) and 21 ha (Field 2) in size, located near Centralia, Missouri. The soils of the area are characterized as claypan soils, primarily of the Mexico-Putnam association (fine, montmorillonitic, mesic, Udollic Ochraqualfs). These soils are poorly drained and have a restrictive, high-clay layer (claypan) occurring below the topsoil. Elevation data were obtained using a total station surveying instrument and standard mapping procedures (Figure 1). A drainage channel transverses Field 1, dividing it in two. For the area east of the drainage channel elevation ranges from 257.52 m in the northwest corner along the drainage outlet to 264.71 m on the eastern boundary. The area west of the drainage channel exhibits a similar range in elevation with its highest point being near the southern boundary. Field 2 elevation ranges from 257.44 m on the west edge to 264.67 m at the southeast corner. Figure 1. Elevation maps of the research fields: Field 1 (top) and Field 2 (bottom). 0 100 200 300 257.44 258.89 260.33 261.78 263.22 264.67 Field 2 Elevation (m) 257.52 258.96 260.40 261.84 263.27 264.71 Field 1 Elevation (m)