Implementation of a Survey Method for Noxious Weeds

Project summary:

In 2004 and 2005, a study was conducted in Mn/DOT District 4 to test the effects on sampling precision of stratifying sampling by ecozones and of the adoption of greater numbers of shorter test segments. Comparison of matching sample statistics from 3-mile and 1/4-mile survey plans in each year indicated that the two plans yielded equivalent estimates of mean acres per roadway mile of each weed; however, precision at the district level was much greater in all cases under the 1/4-mile plan. Weed abundances also varied substantially among ecozones. Furthermore, a combination of computer-based mapping and resampling of the 1/4-mile segments observed in the two years suggested that additional improvements in precision and efficiency are likely to occur if segment lengths are shortened to 125 feet or less. Shorter segments would reduce inspection costs, increase sample sizes, improve precision, and possibly allow conversion from an area-measurement approach to one based on presence or absence of chosen weeds in selected segments.