Yield Map Interpretation

Abstract

Yield monitoring and mapping are becoming commonplace with many crops throughout North America. Some fields now have a 3- to 5-yr yield map history. A yield map, however, only documents the spatial distribution of crop yield and does not explain what factor(s) caused the variation. The goal of yield map interpretation is enhanced profitability through better understanding and control of natural and management- induced sources of yield variation. Numerous causes of crop yield variation have been documented, including climate, soil- water relationships, soil physical and chemical properties, site attributes, crop pest infestations, crop inputs and condition, field history, and cultural practices. Proper visual presentation of yield monitor data in yield map form and the accurate identification of characteristic patterns of yield variation are essential for meaningful interpretation of the yield map. Unfortunately, all yield monitor data sets and maps contain inherent error, some of which cannot be easily corrected. Error-induced patterns must be separated from real yield variation in order to make correct interpretations. In general, irregular areas, blotches or speckles, and elliptical patterns are the result of naturally occuring yield-limiting factors. Conversely, rectangles, abrupt boundaries, circles, arcs, and streaks or lines reflect management-induced patterns of yield variation. In addition, divergence of parallel swaths, missing data points, and a repeating sawtooth pattern along field margins usually result from errors associated with Global Positioning System (CPS) signal reception and yield monitor data collection. Yield map interpretation is greatly enhanced by ongoing grower involvement and the insightful use of auxiliary agronomic, spatial, and historical site information. Geographic Information System (GIS) tools are virtually essential to evaluate multiple layers of spatial data. GIS software can be used to systematically and quantitatively evaluate the relationships between yield data and other spatial features. In the future, tools such as data-mining software, and other sophisticated mathematical and spatial models may provide additional power to interpret single-year as well as multi-year yield map information. The Issue Crop yield monitors, Global Positioning System receivers, and yield mapping software are now commonly used in North America to create maps showing the spatial distribution of crop yield. Initially, farmers optimistically hoped to use this information to significantly improve their management efficiency and increase net returns. Unfortunately, yield patterns within a field can vary dramatically from year to year and making management decisions directly from the yield map has proven difficult. Today, growers realize that the full value of yield map information will only be realized when the causes of yield variation are clearly identified. Successful yield map interpretation, which leads to improved farm management, must identify the causes of yield variation in a single year's yield map and long-term yield trends in the same field across multiple crops and growing environments. Background Agriculturists have long known of the existence of spatial variation in crop yields. Early crop modelers from the Cold War era developed sophisticated aerial and satellite techniques to estimate crop yields in regions, and even whole countries. Prior to the early 1990s, growers were well acquainted with perennially high or low yielding portions of individual fields, but transient or inconsistent spatial trends in crop yield may have gone unrecognized. Measuring and mapping crop yields in sub-units within fields were very labor-intensive and yet could produce only the most rudimentary yield map. As such, yield monitoring was typically done on a whole field or even a whole farm basis. The advent of compact, high-speed computers, crop yield monitoring sensors, and the worldwide Global Positioning System has enabled the measurement, capture, and mapping of continuous yield data from every point within a farm field. The popularity of yield monitoring and mapping has extended to numerous grain, forage, vegetable, and specialty crops. Indeed, all major farm implement manufacturers and a number of independent companies now produce yield monitoring systems and related products. Like many innovations, yield monitoring and mapping were eagerly embraced by virtually all segments of the agricultural industry. Growers looked to yield mapping to unlock the secrets of yield variation and lead to a new level of farming efficiency and profitability. Manufacturers and input suppliers viewed yield mapping as the logical complement to variable-rate technologies designed to enable true site-specific crop management. Even environmental advocates and government policymakers looked favorably on yield monitoring and mapping. Perhaps this new approach to farm management would facilitate wiser stewardship of nutrient and pesticide inputs and therefore reduce the risks to air, soil, and water resources. Unfortunately, the translation of yield map information into increased knowledge for making better management decisions has proven more difficult than expected. There are several reasons for this. First, the data recorded by yield monitors contain inherent error and yield maps do not perfectly duplicate the spatial trends in actual crop yield. In addition, using a yield map to improve crop management requires that the underlying causes of crop yield variation be accurately identified; something that a yield map alone cannot do. A further complication is the inconsistency of crop yield patterns across multiple seasons due to differing environmental conditions and their complex interaction with other site features. Even the way yield monitor data are grouped, shaded, and presented on the yield map can affect the level of apparent variability within a field. A realistic understanding of these limitations is the first step toward accurate yield map interpretation. Applied Questions Why has crop yield mapping become so popular in North America ? There are many reasons why yield mapping is so appealing. Growers expect that yield mapping will provide a new level of understanding and therefore, control of yield variation in their fields. In today's agriculture, any technology that improves production efficiency will also probably result in enhanced profitability, an obvious incentive for adoption. Growers have also been quick to realize the potential for using yield monitoring and mapping as a convenient way to conduct on-farm testing to evaluate the performance of various inputs and cultural practices. Will yield maps always improve farm management efficiency ? Not necessarily. Yield maps will only improve farm management efficiency if the patterns of yield variation they display can be correlated with specific causes. This is complicated by the multitude of naturally occurring and management-induced sources of crop yield variation and their potentially complex interactions in time and space. For example, soil type, internal soil drainage and seasonal weather patterns can interact in complex ways that could have many different effects on yield. Finally, the causes identified must be correctable and intervention must be profitable and environmentally benign. What are the keys to successful yield map interpretation ? Accurate interpretation of crop yield maps will lead to improved management decisions. This will require attention to the following, (i) recognition and minimization of inherent error in yield data and maps, (ii) proper selection of yield ranges and color schemes to display yield map data and the accompanying legends (theming), (iii) use of auxiliary site information including thorough historical records, and (iv) ongoing grower involvement in yield mapping, analysis, and interpretation. Summary and Future Outlook In spite of the limitations, yield mapping will continue to grow in importance and utility. The required hardware and software are affordable and subject to continuous improvement. Additional layers of spatial information such as soil type, topography, and remotely sensed images add considerable explanatory value when interpreting a yield map. These companion data are becoming increasingly accessible from commercial providers and via the Internet. In addition, they are now more easily used through powerful analytical tools such as Geographic Information Systems, data-mining software, and spatial-mathematical models.