Introduction

Precision nutrient management requires consideration of the unique combinations of crop and soil responses within a field. Additionally, those fields receiving animal wastes require consideration of soil-landscape interactions that may move waste constituents from place-to-place or otherwise change nutrient availability from those wastes. The motivations behind precision nutrient management include (i) more prudent use of production resources (ex. fertilizers), (ii) potentially decreased environmental degradation (proper placement of nutrient sources to minimize loss with runoff and percolating water, and (iii) increased agricultural sustainability through optimum use of natural resources (soil, water) and decreased production risk. In managing nutrients like nitrogen (N) and phosphorus (P) across agricultural landscapes, producers need to assess in-field variation of soil properties (biological, chemical and physical) that influence nutrient transformations within the soil as well as crop growth and physiology responses affecting nutrient removal from the soil.

Typically speaking, soil sampling protocols for precision agriculture have proven expensive, and a barrier to adoption of the new paradigm. Further, while many fields have shown large variations in soil test values, much of that variation was observed among high and very high soil test values, with limited interpretational significance. The advent of yield monitoring, and with it the prospect of using yield to estimate nutrient removal, has given new life to variable rate nutrient management. Nutrient removal is related to both grain yield and grain nutrient concentration, and landscape factors causing variation in both parameters will be of interest to nutrient managers using this technique. Another alternative, directed sampling, is emerging as a consequence of quantitative descriptions of the relationships between terrain attributes and static soil properties that affect soil nutrient supply, crop growth and nutrient demand. Directed sampling is a type of stratified sampling that subdivides fields into management areas based on prior knowledge of variability, commonly derived from digital elevation maps. Research is needed to evaluate these new approaches and to illuminate the landscape factors of interest, should either directed sampling or nutrient removal be promising approaches to variable rate fertilization of soils.

Objectives

The objective of this research is to determine whether nutrient removal has promise as an approach to guide variable rate fertilization of soils. This assessment will be done on both manured and unmanured fields because patterns of amendment application, and the interaction of these patterns with terrain attributes, can vary. Selected soil properties will be monitored, either on a grid or using a soil-landscape modeling approach that will identify portions of the field having similar environments.

Specific objectives of this portion of the overall project are:

1. to determine the variation in the static soil properties related to soil nutrient supply, buffer capacity and crop yield (e.g., organic matter content, clay content, drainage) for eight fields (manured and unmanured) in the Crider-Pembroke soil association in the Pennyroyal soil physiographic region;

2. to examine the similarity of these soil property-landscape relationships within landscape segments, across the eight fields; and

3. to determine the spatial variation of grain yield and grain nutrient concentrations within the four fields and relate these to soil properties and/or pedo-transfer functions that translate terrain attributes into static soil properties with minimal additional field sampling.

4. To use these results to evaluate the nutrient removal approach, as opposed to the grid and directed (smart) sampling approaches, as a basis for variable rate fertilization of these fields.

Background

Grid Sampling - The most common method of assessing soil variability for precision agriculture is through grid sampling. This type of sampling can be particularly useful when no other information is known about soil variability within a field prior to sampling. Critical to the ability of the sampling grid to represent the variability of soils within a field is the sampling intensity of the grid. To accurately represent the mesoscale (<10 m) variability in soils using uniform grid sampling, Pocknee et al. (1996) suggest that sampling must occur at least at the same scale as, if not smaller than, to ensure that multiple samples are taken from each soil unit. An additional concern over grid sampling is that the composite samples used do not adequately represent each grid cell because samples are not taken from throughout the grid, but are clustered at the center of each grid cell and represent only a small area in the center of each cell (Pocknee et al., 1996).

Once analyzed, data from the grid sampling are then used to generate application maps that are used to guide fertilizer application rates. Commonly, geostatistical techniques, such as kriging, are used to interpolate the sample data to create the application map. However, these methods normally require large numbers of samples and are not transferable to other similar soil landscapes. Furthermore, these methods are data dependent and do not account for the primary causes of the soil variability—processes of soil formation (Moore et al., 1993; Pocknee et al., 1996). Instead of just characterizing the observed patterns, a more valuable outcome would be to link the observed patterns to the processes of soil development.

Directed Sampling - An alternative to uniform grid sampling is directed sampling. It is a type of stratified sampling design that subdivides fields into management areas based on prior knowledge of variability in soils, landscape, and prior land use (Pocknee et al., 1996). This type of sampling has advantages over grid sampling in that it recognizes systematic soil variability within a field. However, the resultant maps still divide fields into discrete units that are presumed homogeneous for important agronomic variables.

Soil Variability and Crop Response

Common soil properties used to create management zones and determine application rates include soil test P, soil test K, and soil pH. These soil properties are not always directly related to crop response in the upcoming growing season. Further, they are dynamic soil properties that can vary with time. So, while soil nutrient status is a dynamic soil property that may change from year to year, there are several more static soil properties that are more stable over time and often have a profound influence on crop response, e.g., organic matter content, clay content, and soil drainage.

Soil organic matter (SOM) is a reservoir of many essential plant nutrients (N, P, S). The physical and chemical properties of SOM also (i) improve soil aggregate stability, which increases soil stability, increases infiltration, and improves aeration, and (ii) increase moisture retention and available water capacity. The distribution of SOM in the landscape is largely determined by properties and processes related to topography, such as the redistribution of high-SOM material through erosion and deposition, and soil wetness.

Similarly, the processes of erosion and deposition bring subsoil materials closer to the surface and incorporate them into the plow layer. The subsoil materials tend to have a higher clay content, lower organic matter content, lower available water holding capacity, lower available nutrients, and higher bulk density (Frye et al., 1982). Frye et al. (1982) studied the effects of past erosion on the productivity of two Kentucky soils, and found that corn grain yields were up to 24% lower on eroded soils as compared to uneroded soils.

The effects of SOM variability and past erosion are further exacerbated by within-field and seasonal variations in available moisture. For example, Langdale et al. (1979) found that the reduction in corn yield attributed to past erosion was greatest in drier years. This can be attributed to the reduction in available water holding capacity caused by the past erosion. Within a field, we also know that lower landscape positions have greater moisture because of greater collection of surface and subsurface runoff. During the growing season, particularly in drier years, these lower landscape positions should provide more available moisture, and is less limiting to crop growth.

Variations in grain nutrient concentrations with spatial soil attributes have not been well documented, but variations in grain N and P with changing soil nutrient availability have long been reported in small-plot research. Climatic (seasonal) variation in grain N and P is also well known, such that lower yields due to water stress are often associated with higher grain N and P levels.

Effects of Precision Agriculture Management on Biological Properties Related to Nutrient Management

Although there is substantial evidence that topographic features and long-term management practices affect soil biological processes, there is little evidence that precision agriculture management can influence soil biological processes if they are known to differ between locations. For example, Sutherland et al., (1993) showed that nitrogen-15 was enriched in field depressions because denitrification enhanced nitrogen-14 depletion. Furthermore, soil variables and nitrogen-15 abundance of plants and soils followed similar patterns. The effects of long-term organic N addition on acidification of soil by nitrification are well-demonstrated (Doran and Smith, 1987).

Spatial variability of soil biological processes makes it difficult to attribute significant effects to management. Robertson et al. (1988), for example, examined spatial relationships between basic soil processes in a successional community, which suggested that nitrogen transformations were autocorrelated at scales < 40 m and corresponded less with the topographic

scale than with the change in plant community composition at the same scale.

Hypothesis

Our hypotheses are (i) that the spatial patterns of crop yield and grain nutrient concentration can be predicted from spatial patterns of biological, chemical and physical soil attributes, with due consideration of seasonal effects on both plant parameters, and (ii) that nutrient removal and/or directed sampling are likely to prove superior to grid sampling in agricultural and environmental nutrient management.

Research Plan

The proposed research will be conducted over a three-year period with the objective of evaluating all crop components grown in the rotations commonly grown in the target fields. The first year of the study will begin the grid soil attribute and crop component sampling as well as the landscape modeling component (to “direct” the directed sampling). The second year of the study will continue the grid crop sampling and begin the directed soil and crop sampling as well as early results analysis. The final year will finish the directed soil and crop sampling and complete the results analysis, as well as dissemination of project results.

Site Selection - The field study will be done in eight fields located in the Pennyroyal physiographic region of central and southwest Kentucky. Initial sites are located in Caldwell and Marion Counties. Both Crider and Pembroke are agriculturally significant soils in Kentucky, supporting a vast acreage of tobacco, corn, and other row crops. All sites will be in corn, soybean and wheat-doublecrop soybean rotations not exceeding 3 years in total duration. At least two of the fields will have been regularly amended with animal waste. It is believed that animal waste application may change the relationship between certain soil properties (including soil test values) and the crop’s removal of nutrients. Such fields may also have different spatial patterns in nutrient availability and different relationships between nutrient availability and terrain attributes. Rates of manure are not the issue, but the “usual” pattern of manure delivery to these fields is an issue of consequence.

DEM Generation and Terrain Analysis - A field survey DEM with approximate 10-m horizontal resolution and 0.1-m vertical precision will be created from a ground survey using a kinematic global positioning system. The raw data will be processed to a regular 10-m grid using ANUDEM (Hutchinson, 1995). Primary and secondary terrain attributes (slope gradient, slope aspect, slope curvature, upslope contributing area, etc.) will be calculated from each of the DEM using the Terrain Analysis Programs for the Environmental Sciences (TAPES) program (Moore, 1992), or other similar algorithms.

Soil and Crop Sampling - At each site, we will grid sample soil and crop yield at about 100 m (1 acre) intervals and also use the topographic variability as quantified using the DEM to direct soil sampling using a stratified random sampling design (Petersen and Calvin, 1986). Grain samples will also be taken for nutrient analyses. We will define three to five intervals of one of these terrain attributes calculated from the DEM, and delineate areas of each terrain attribute class. From each class, or stratum, an equal number of randomly selected soil sampling points will be chosen (Gessler et al., 1995).

Field & Laboratory Analysis - Cores will be returned to the laboratory for morphological description and both physical, and chemical analysis. The thickness, color, and presence of redoximorphic features of individual horizons will be described (Soil Survey Division Staff, 1993). For each diagnostic soil horizon we will determine organic carbon content by dry combustion (Nelson and Sommers, 1996), clay content by the pipette method (Gee and Bauder, 1986), water holding capacity, soil pH, and extractions for bioavailable nutrients (Mehlich III). Soil physical properties to be determined include penetration resistance, surface shear, bulk density and surface hydraulic conductivity. Grain will be analyzed for N, P, K, Ca, Mg, and S (also Cu and Zn in animal waste amended fields.

The question is raised whether easily measured soil biological properties can be related to mappable features in a field and bare relationship to manageable crop productivity differences. The rate of the processes that occur will be a proxy measurement for the size and composition of the underlying microbial community. Since N and P are the important amendments in most crop production systems, we will examine those biological processes most directly associated with the delivery of N and P: to wit, the induction period and potential for nitrification and denitrification, the short term mineralization rates of applied and endogenous organic N as a function of management, and the potential mineralization of organic P applied in manure as assessed by phosphatase activity.

Expected Benefits

Improved knowledge of crop nutrition, soil and landscape relationships in the Crider-Pembroke soil association—a common and agriculturally significant soil association. A clearer understanding whether yield based nutrient removal or soil testing can guide fertilizer rate recommendations. This has important implications to the question of what will be the basis for agricultural and environmental nutrient management in the next century. A determination of whether directed sampling, based on landscape segmentation after terrain analysis, can improve our understanding of the soil factors driving crop growth, nutrient demand and the nutrient supply potential of field soils.

Deliverables

Deliverables will include scientific journal articles, extension publications, conference presentations, and workshops.