An economic analysis of binomial sampling for weed scouting

Abstract

Full-count random sampling has been the traditional method of obtaining weed densities. Currently it is the recommended scouting procedure when using HERB, a herbicide selection decision aid. However, alternative methods of scouting that are quicker and more economical need to be investigated. One possibility that has been considered is binomial sampling. Binomial sampling is the procedure by which density is estimated from the number of random quadrats in which the count of individuals is equal to or less than a specified cutoff value. This sampling method has been widely used for insect scouting. There has also been interest in using binomial sampling for weed scouting. However, an economic analysis of this sampling method for weeds has not been performed. In this paper, the results of an economic analysis using simulations with binomial sampling and the HERB model are presented. Full-count sampling was included in the simulations to provide a benchmark for comparison. The comparison was made in terms of economic losses incurred when the estimated weed density obtained from sampling was inaccurate and a herbicide treatment was selected that did not maximize profits. These types of losses are referred to as opportunity losses. The opportunity losses obtained from the simulations indicate that in some situations binomial sampling may be a viable economic alternative to full-count sampling for fields with weed populations that follow a negative binomial distribution, assuming no prior knowledge of weed densities or negative binomial k values.

Evaluating the Potential for Site-Specific Herbicide Application in Soybean

“…Random Sampling to Determine Suitability of Field for Site-Specific Management. For the four fields in which 10 random 9.3-m 2 samples were taken, it appears that this quicker, less-costly scouting technique, suitable for making whole-field decisions (Krueger et al 2000), may help identify those fields for which site-specific management may be most appropriate. For GR soybean, there was a strong correlation between expected gain in TNR calculated using grid cell weed densities and gain in TNR calculated using the 10 random samples ( Figure 3).…”

Evaluation of Scouting Methods in Peanut (Arachis hypogaea) Using Theoretical Net Returns from HADSS™1

“…Accurately estimating weed populations and determining what herbicide to apply to control weeds is challenging. Several attempts have been made to develop scouting procedures for row crops (Gold et al 1996;Krueger et al 2000;Wiles et al 1992). These approaches often require intensive scouting with numerous stops within a field.…”

Economics and Effectiveness of Alternative Weed Scouting Methods in Peanut

“…that the purpose of scouting is primarily to assist in making sound weed management decisions rather than solely to estimate weed population densities. Previous studies have shown that the cost of additional sampling rapidly exceeds the value when the objective is to make a herbicide decision (Krueger et al 2000, Wiles et al 1993).…”

“…Attempts are being made to develop better scouting procedures in row crops that are less time intensive, but will still be cost effective (Gold et al 1996;Jordan et al 2003;Krueger et al 2000;Wiles et al 1992b). The more time spent scouting and gathering information in the field, the better the recommendation (Krueger et al 2000;Wiles et al 1992a); however, grid sampling is not practical in production agriculture because of the cost associated with this level of scouting (Clay et al 1999). Murdock and Murray (2002) found that estimating weed population densities in cotton and peanut was just as effective as counting weeds when using the data to generate herbicide recommendations in combination with a decision support system.…”