Page Banner

United States Department of Agriculture

Agricultural Research Service

Research Project: Sustainable Pest Management Strategies for Arid-land Crops

Location: Pest Management and Biocontrol Research

Title: The Problem of False Positives in Protein Marking Techniques

Authors
item Sheller, Frances -
item Rosenheim, Jay -
item Hagler, James

Submitted to: Integrated Pest Management Symposium Workshop Proceedings
Publication Type: Proceedings
Publication Acceptance Date: February 1, 2009
Publication Date: March 24, 2009
Citation: Sheller, F.J., Rosenheim, J.A., Hagler, J.R. 2009. The Problem of False Positives in Protein Marking Techniques. Integrated Pest Management Symposium Workshop Proceedings. 19.6 pg 35.

Interpretive Summary: Protein marking is a valuable technique in the study of insect movement in agriculture. It can be implemented on a large scale and is relatively inexpensive to use. Unlike other marking techniques, protein marking is a quantitative method. Whether an individual is considered marked or not is dependant on a threshold that is chosen by the experimenter. The traditional method of choosing a threshold for considering a sampled individual 'marked' results in some risk of false positives, where an unmarked individual is misclassified as marked. In dispersal studies where the recapture rate of marked individuals is low, false positives can significantly affect estimates of dispersal rates. Using simulations, we demonstrate the problems produced by false positives. We introduce two possible approaches that can minimize this problem. First, populations can be doubly marked as a means of reducing the incidence of false positives. Second, we introduce new algorithms for choosing a threshold that will decrease the incidence of false positives and allow data to be corrected for anticipated rates of false positives. Together, these methodologies should enhance researcher confidence in the data generated from dispersal studies using protein marking techniques.

Technical Abstract: Protein marking is a valuable technique in the study of insect movement in agriculture. It can be implemented on a large scale and is relatively inexpensive to use. Unlike other marking techniques, protein marking is a quantitative method. Whether an individual is considered marked or not is dependant on a threshold that is chosen by the experimenter. The traditional method of choosing a threshold for considering a sampled individual 'marked' results in some risk of false positives, where an unmarked individual is misclassified as marked. In dispersal studies where the recapture rate of marked individuals is low, false positives can significantly affect estimates of dispersal rates. Using simulations, we demonstrate the problems produced by false positives. We introduce two possible approaches that can minimize this problem. First, populations can be doubly marked as a means of reducing the incidence of false positives. Second, we introduce new algorithms for choosing a threshold that will decrease the incidence of false positives and allow data to be corrected for anticipated rates of false positives. Together, these methodologies should enhance researcher confidence in the data generated from dispersal studies using protein marking techniques.

Last Modified: 7/30/2014
Footer Content Back to Top of Page