Better Scouting Through Computer Management

Michael R. Williams, J. L. Willers, T. L. Wagner, R. L. Olson, and J. M. McKinion


The `art' of scouting cotton is perplexing at times but always challenging. Sampling remains as the single most difficult factor in making decisions. Time is money to farmers, commercial entomologists as well as researchers. Quality information is essential, but there is a limit on what can or should be factored into decision making. How much is too much, and where should the line be drawn with how little is too little? Traditionally, the sampling methods used for monitoring arthropod populations have varied according to the objectives (Southwood, 1978). Sampling plans built from a research point of view require sample sizes and numbers of observations which are too large for commercial monitoring. They also recommend complete coverage of fields. Sampling by commercial scouts for the purpose of making management decisions varies with available time, pest thresholds, economics, etc.; often resulting in sample sizes which are much too small. Wilson et al. (1989), discusses almost all aspects of sampling in depth.

Both commercial and research workers need rapid, high quality, versatile sampling techniques which give accurate estimations of arthropod populations. Willers et al. (1990), using a Bayesian approach, developed streamlined methods for scouting the major arthropod pests of cotton. Fields were divided into management units (areas where the crop was uniform). Six to 10 quadrats (called stops) of 2 rows by 9 feet were selected at uniform distances beginning at least 75 to 100 feet from the edge. Ten plants were examined for insects and insect damage at each stop. The stops were grouped into transects. Each transect was subdivided into exterior stops (the first and last in each transect) and interior stops (those in the middle of the transects). Analysis of variance tests using several scouting dates indicate that only perimeter areas of a field need to be scouted. The location and stop (obs. x loc.) effects were more often non-significant, indicating that there is no consistent spatial pattern for Heliothis. Therefore, there is no advantage to a scouting pattern which covers the entire field. Similar data on other species indicate similar results. Willers et al. (1990; unpubl. data), has also shown there is no advantage to larger sample sizes.

We still recommend that historical data and known `hotspots' be taken into account, but examination of 3-5 areas of a field with 3 to 5 stops per area can give accurate population estimates. Additionally, a simulation experiment discovered that 5 plants at a stop gave as much information on insect abundance as 10 plants per stop. Thus, examination of as few as 45 plants over a large fairly uniform field can give as accurate and timely information than older methods which utilized more than 100 plants.

Computers have the ability to archive historical data. This capability means that present and past information can be managed for simultaneous use. The most notable advantage of this feature is that "observer" bias and error can be monitored and corrected. Additionally, the ability to recall data from previous years enable managers to anticipate possible trends and thus, prepare.


1. Southwood, T. R. E.. 1978. Ecological Mehtods. John Wiley & Sons, New York.

2. Willers, J.L., D.L. Boykin, J.M. Hardin, T.L. Wagner, R.L. Olson, & M.R. Williams. 1990. A simulation study on the relationship between the abundance and spatial distribution of insects and selected sampling schemes. Proc. Applied Stat. Agric., Kansas State Univ., Manhattan.

3. Wilson, L. T., W. L. Sterling, D. R. Rummel, J. E. DeVay, 1989. Quantitative sampling principles in cotton IPM. In Integrated pest management systems and cotton production. Eds: Raymond E.. Frisbie, Kamal M. El-Zik and L. T. Wilson. John Wiley & Sons, New York. pp 85-119.

Reprinted from Proceedings of the 1994 Beltwide Cotton Conferences pg. 1250
©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