Date of Award:

8-2020

Document Type:

Thesis

Degree Name:

Master of Science (MS)

Department:

Watershed Sciences

Committee

Phaedra Budy

Committee

Mary M. Conner

Committee

Mark C. McKinstry

Abstract

Estimating demographic parameters, such as survival and abundance, with accuracy and precision is vital for detecting trends in populations and assessing the effectiveness of management actions. In most cases, a lack of capture data make estimating parameters very challenging. The use of new technologies to increase the amount of remotely collected data is increasing, but brings new limitations and analytical issues to be resolved. One of those new technologies is the use of a mobile floating PIT-tag antenna to detect PIT-tagged fish. The issue that arises with this technology is determination of the status of detected tags (i.e., live fish or ghost tag; a tag left in the environment when a fish dies or sheds its tag). The objective of this study was develop a method to determine the status of tags detected with a mobile floating antenna. I determined the movement dynamics of known ghost PIT-tags and contrasted them with known live fish. I used a random forest analysis to develop a classification model for the two different states. I found ghost tags exhibited movement that is to be expected based on sediment transport analysis and that the distance and direction moved, and response to changes in flow were the most important variables for correctly determining tag status. This study provides a useful framework for developing models for other systems, even though the relative importance of specific predictor variables will most likely vary with location and species of interest.

Checksum

813a8bd22a86ca736e00603ed91cf2d0

Share

COinS