Imperfect detection and the determination of environmental flows for fish

Challenges, implications and solutions

D.C. Gwinn, L.S. Beesley, P. Close, B. Gawne, P.M. Davies

Research output: Contribution to journalArticle

26 Citations (Scopus)

Abstract

Relationships between river flow characteristics and fish community/population dynamics (i.e. flow-ecology relationships) underpin methods to determine and monitor environmental water allocations. Quantifying these relationships can be difficult, and consequently, most environmental flow strategies for fish conservation in Australian rivers are based on general flow-ecology relationships as opposed to statistical predictions. Of those studies that have investigated relationships between flow and fish, most have not accounted for incomplete and variable detection of fish by the sampling methods, thus making the implicit assumption that sampling efficiency is invariant. This important assumption is rarely met, leading to inconsistent research findings and spurious results, and a reliance on generic flow-ecology principles for defining flow management strategies. We illustrate how and when detection probability varies when sampling freshwater fish and the consequences to scientific inference about fish-flow relationships. Methods for accounting for imperfect detection of fish are identified and tools to increase the efficiency of experimental designs while reducing sampling cost are discussed. These tools include methods for borrowing information among experimental components and simulation techniques to optimise sampling designs. We argue that, due to the very nature of sampling designs to quantify flow-ecology relationships (e.g. sampling at different flow magnitudes/regimes), the challenge of imperfect detectability is particularly relevant to environmental flow science. We encourage the broader adoption of methods that account for imperfect detection to improve inference about fish-flow relationships and increase the successful application of environmental flows for managing fish communities.
Original languageEnglish
Pages (from-to)172-180
Number of pages9
JournalFreshwater Biology
Volume61
Issue number1
DOIs
Publication statusPublished - Jan 2016
Externally publishedYes

Fingerprint

fish
sampling
ecology
detection
methodology
water allocation
rivers
river flow
experimental design
freshwater fish
population dynamics
method
prediction
monitoring
river
cost
simulation
water

Cite this

Gwinn, D.C. ; Beesley, L.S. ; Close, P. ; Gawne, B. ; Davies, P.M. / Imperfect detection and the determination of environmental flows for fish : Challenges, implications and solutions. In: Freshwater Biology. 2016 ; Vol. 61, No. 1. pp. 172-180.
@article{e15629c2389641ae8cc4a4adddb2cb90,
title = "Imperfect detection and the determination of environmental flows for fish: Challenges, implications and solutions",
abstract = "Relationships between river flow characteristics and fish community/population dynamics (i.e. flow-ecology relationships) underpin methods to determine and monitor environmental water allocations. Quantifying these relationships can be difficult, and consequently, most environmental flow strategies for fish conservation in Australian rivers are based on general flow-ecology relationships as opposed to statistical predictions. Of those studies that have investigated relationships between flow and fish, most have not accounted for incomplete and variable detection of fish by the sampling methods, thus making the implicit assumption that sampling efficiency is invariant. This important assumption is rarely met, leading to inconsistent research findings and spurious results, and a reliance on generic flow-ecology principles for defining flow management strategies. We illustrate how and when detection probability varies when sampling freshwater fish and the consequences to scientific inference about fish-flow relationships. Methods for accounting for imperfect detection of fish are identified and tools to increase the efficiency of experimental designs while reducing sampling cost are discussed. These tools include methods for borrowing information among experimental components and simulation techniques to optimise sampling designs. We argue that, due to the very nature of sampling designs to quantify flow-ecology relationships (e.g. sampling at different flow magnitudes/regimes), the challenge of imperfect detectability is particularly relevant to environmental flow science. We encourage the broader adoption of methods that account for imperfect detection to improve inference about fish-flow relationships and increase the successful application of environmental flows for managing fish communities.",
keywords = "Detection probability, Environmental flow, Murray-Darling Basin, River regulation, Water management",
author = "D.C. Gwinn and L.S. Beesley and P. Close and B. Gawne and P.M. Davies",
note = "Cited By :5 Export Date: 17 January 2018",
year = "2016",
month = "1",
doi = "10.1111/fwb.12684",
language = "English",
volume = "61",
pages = "172--180",
journal = "Freshwater Biology",
issn = "0046-5070",
publisher = "Wiley-Blackwell",
number = "1",

}

Imperfect detection and the determination of environmental flows for fish : Challenges, implications and solutions. / Gwinn, D.C.; Beesley, L.S.; Close, P.; Gawne, B.; Davies, P.M.

In: Freshwater Biology, Vol. 61, No. 1, 01.2016, p. 172-180.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Imperfect detection and the determination of environmental flows for fish

T2 - Challenges, implications and solutions

AU - Gwinn, D.C.

AU - Beesley, L.S.

AU - Close, P.

AU - Gawne, B.

AU - Davies, P.M.

N1 - Cited By :5 Export Date: 17 January 2018

PY - 2016/1

Y1 - 2016/1

N2 - Relationships between river flow characteristics and fish community/population dynamics (i.e. flow-ecology relationships) underpin methods to determine and monitor environmental water allocations. Quantifying these relationships can be difficult, and consequently, most environmental flow strategies for fish conservation in Australian rivers are based on general flow-ecology relationships as opposed to statistical predictions. Of those studies that have investigated relationships between flow and fish, most have not accounted for incomplete and variable detection of fish by the sampling methods, thus making the implicit assumption that sampling efficiency is invariant. This important assumption is rarely met, leading to inconsistent research findings and spurious results, and a reliance on generic flow-ecology principles for defining flow management strategies. We illustrate how and when detection probability varies when sampling freshwater fish and the consequences to scientific inference about fish-flow relationships. Methods for accounting for imperfect detection of fish are identified and tools to increase the efficiency of experimental designs while reducing sampling cost are discussed. These tools include methods for borrowing information among experimental components and simulation techniques to optimise sampling designs. We argue that, due to the very nature of sampling designs to quantify flow-ecology relationships (e.g. sampling at different flow magnitudes/regimes), the challenge of imperfect detectability is particularly relevant to environmental flow science. We encourage the broader adoption of methods that account for imperfect detection to improve inference about fish-flow relationships and increase the successful application of environmental flows for managing fish communities.

AB - Relationships between river flow characteristics and fish community/population dynamics (i.e. flow-ecology relationships) underpin methods to determine and monitor environmental water allocations. Quantifying these relationships can be difficult, and consequently, most environmental flow strategies for fish conservation in Australian rivers are based on general flow-ecology relationships as opposed to statistical predictions. Of those studies that have investigated relationships between flow and fish, most have not accounted for incomplete and variable detection of fish by the sampling methods, thus making the implicit assumption that sampling efficiency is invariant. This important assumption is rarely met, leading to inconsistent research findings and spurious results, and a reliance on generic flow-ecology principles for defining flow management strategies. We illustrate how and when detection probability varies when sampling freshwater fish and the consequences to scientific inference about fish-flow relationships. Methods for accounting for imperfect detection of fish are identified and tools to increase the efficiency of experimental designs while reducing sampling cost are discussed. These tools include methods for borrowing information among experimental components and simulation techniques to optimise sampling designs. We argue that, due to the very nature of sampling designs to quantify flow-ecology relationships (e.g. sampling at different flow magnitudes/regimes), the challenge of imperfect detectability is particularly relevant to environmental flow science. We encourage the broader adoption of methods that account for imperfect detection to improve inference about fish-flow relationships and increase the successful application of environmental flows for managing fish communities.

KW - Detection probability

KW - Environmental flow

KW - Murray-Darling Basin

KW - River regulation

KW - Water management

UR - http://www.scopus.com/inward/record.url?scp=84954378156&partnerID=8YFLogxK

UR - http://www.mendeley.com/research/imperfect-detection-determination-environmental-flows-fish-challenges-implications-solutions

U2 - 10.1111/fwb.12684

DO - 10.1111/fwb.12684

M3 - Article

VL - 61

SP - 172

EP - 180

JO - Freshwater Biology

JF - Freshwater Biology

SN - 0046-5070

IS - 1

ER -