Skip to main content
ARS Home » Research » Publications at this Location » Publication #161984
Title: FREQUENCY ANALYSIS OF YIELD FOR DELINEATING MANAGEMENT ZONES

Author
item Diker, Kenan
item Buchleiter, Gerald
item Farahani, Hamid
item Heermann, Dale
item Brodahl, Mary

Submitted to: International Conference on Precision Agriculture Abstracts & Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 3/4/2002
Publication Date: 7/17/2002
Citation: DIKER, K., BUCHLEITER, G.W., FARAHANI, H., HEERMANN, D.F., BRODAHL, M.K. FREQUENCY ANALYSIS OF YIELD FOR DELINEATING MANAGEMENT ZONES. INTERNATIONAL CONFERENCE ON PRECISION AGRICULTURE ABSTRACTS & PROCEEDINGS. 2002.

Interpretive Summary: Analysis of the multi-year yield data was a problem. The common method to analyze the multi-year yield data results in smoothing over time. This paper reports on a methodology that we developed using a "frequency analysis" procedure to create yield response zones that can be translated into potential management zones. The procedure seemed to be effectively separating the stable low and high yielding zones from the fluctuating ones over years. Yield response zones developed with this methodology agreed about 50% with the zones delineated by two other techniques. We expect producers and crop consultants will find this methodology useful in creating site-specific management plans.

Technical Abstract: Use efficiency of agricultural inputs and farm profitability could be improved by identifying management zones for variable rate applications. Management zones may be delineated by using soil properties such as soil texture, series, electrical conductivity and soil color. Using one of these soil properties may not be sufficient for delineation of management zones, because, the yield in any given field or management zone is a product of interaction between many soil properties and production inputs. Therefore, multi year yield maps may give better insight to determine potential management zone areas. This research was conducted to develop a methodology to delineate yield response zones by using two-state frequency analysis conducted on yield maps for three years on two commercial corn fields near Wiggins, Colorado. The zones delineated by frequency analysis were also compared with those created by electrical conductivity (EC) and bare soil color driven management zone areas. For each year, yield data below the mean of whole field were assigned to state 0 and rest of the field was assigned to state 1 for reclassification/standardization to account for the requirements of two-state frequency analysis, temporal variability. The frequency analysis of state 1 resulted in 4 classes over the fields reflecting how many years State 1 occurred. Classes producing statistically similar yield were combined resulting in three potential management zones. A zone was identified by number of years that yield was greater than the average yield in a given year. Results showed that the frequency analysis of multi year yield data is an effective way to assess yield variability in time and space. Comparison of yield zones developed by frequency analysis to soil color and EC driven zones indicated that field zoning for site-specific agriculture could successfully be achieved by frequency analysis of multi year yield data. Management zones determined from EC and bare soil color driven management zone areas were comparable to zones from frequency analysis of yield.