Further Evaluation of GOSSYM-COMAX in Arkansas and Tennessee

T.C. Keisling, P.B. Hoskinson, D.D. Howard, R.L. Maples, and Clay Owens


 
ABSTRACT

The Gossym/Comax (G/C) computer program was evaluated during the 1990 growing season at both Eastern Arkansas and Western Tennessee as a program aid for cotton production. The G/C recommendations for N and irrigation were compared to methods that were previously used in each state. One experiment and one on farm evaluation was conducted at Marianna, AR. Several experiments on several soil types were conducted at Milan, Jackson, and Ames, TN.

The on farm utility of G/C was assessed by working closely with one East Arkansas cotton producer on a little over 1000 acres of cotton. His assessment was that the program causes you to spend time thinking about the importance of N, irrigation, and the importance of timing.He found the program to be useful as a forecaster of N and irrigation inputs 7 to 10 days in advance of their needs. This advanced notice allowed (1) verification that the need was required and (2) scheduling of maintenance, personnel, and equipment. The farmer found that making multiple runs with different weather scenarios provided him with an idea of plant progression toward maturity under different environments but did not really aid him in making any crop management decisions. He liked the fact that the program was dynamic and attempted to address all agronomic variables as they influenced cotton production. He felt that his involvement with the program was worth a little in excess to $10,000.00 per 1000 acres of cotton. On his farm this year he felt that irrigation, N timing, and plant fruiting projections were excellent. Conversely, he felt that N amounts, plant vegetative development, PIX applications, and harvest projections were less than acceptable.

Tennessee's results showed that after the total season weather and N inputs were entered into the computer, about 80% of the G/C predicted yields were at least 20% less than actual yields while 40% were 35% less than actual yields. Only 1 instance out of 25 was G/C found to be within 10% of actual yield.

A series of treatments were utilized to compare G/C N recommendations to Tennessee N recommendations. The G/C yield ranged from about 10% less than that obtained with recommended N rates. However, in several cases the G/C N rate resulted in the same yields obtained by applying 1/3 as much N. The G/C analysis averaged 36% more N than the normal recommendations and resulted in yield increases between 5 and 10% for 25% of the time. An equivalent relationship was found between yield decreases.

Arkansas direct comparison G/C to normal recommendations experiment showed that irrigation was begun 1 week late. Nitrogen fertilizer was about the same for both systems. The resultant yield was in excess of 1100 lbs/Ac lint with Arkansas showing about a 50 lb/Ac advantage the second year in a row.



Reprinted from 1991 Proceedings Beltwide Cotton Conferences pg. 836
©National Cotton Council, Memphis TN

[Main TOC] | [TOC] | [TOC by Section] | [Search] | [Help]
Previous Page [Previous] [Next] Next Page
 
Document last modified Sunday, Dec 6 1998