Characterizing bias, precision, and sampling efficiency of commonly used electrofishing sampling designs in midwestern lentic systems

Balto, George C

Permalink

https://hdl.handle.net/2142/117844 Copy

Description

Title

Characterizing bias, precision, and sampling efficiency of commonly used electrofishing sampling designs in midwestern lentic systems

Author(s)

Balto, George C

Issue Date

2022-12-09

Director of Research (if dissertation) or Advisor (if thesis)

Stein , Jeffrey A

Department of Study

Natural Res & Env Sci

Discipline

Natural Res & Env Sciences

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

M.S.

Degree Level

Thesis

Keyword(s)

• sampling design
• assessment
• population
• sport fish
• bias
• precision
• efficiency

Abstract

As the change in aquatic environments continues in response to anthropogenic pressures, so does the need for natural resource managers to be able to accurately assess populations under ecological monitoring programs. Population assessments used in the management of recreational fisheries can employ various types of sampling designs, each of which has the potential to alter the accuracy and precision of the resulting data. In this study, I used a four-lake, multi-season sport fish assessment dataset to answer two questions. First, what effects do fixed, random, stratified random, and hybrid sampling designs have on the accuracy and precision of common population parameter estimates? Second, what is the effect of the sampling effort used within those sampling designs on the accuracy and precision of parameter estimates of the sport fish community? Sampling design had significant effects on accurately characterizing the catch per unit effort (CPUE) of abundant species such as Bluegill (Lepomis macrochirus) and Largemouth Bass (Micropterus Salmoides), as well as proportional stock density (PSD) when adequate numbers of fish were collected, while all designs tested performed equally well in estimating relative weight (Wr). At fixed levels of effort, the simple random sampling design produced the least biased estimates of CPUE and PSD across impoundments and seasons while no single design consistently produced the most precise parameter estimates. In optimizing sampling effort, I found that up to four 15-minute electrofishing runs under a simple random design was an optimal compromise between accuracy and precision while not requiring an excessive amount of sampling effort. This research served to evaluate the feasibility of implementing a standardized sampling design on small to medium-sized midwestern lakes and impoundments where fixed site designs had been historically used. Poor performance of a stratified random sampling design was attributable to habitat homogeneity in the study systems, and future work should focus on the role of system size and habitat complexity as factors influencing the effectiveness of sampling design in standardized assessment programs.

Graduation Semester

2022-12

Type of Resource

Thesis

Copyright and License Information

Copyright 2022 George Balto

Owning Collections