Evaluation of spatiotemporal imputations for fishing catch rate standardization

Research output: Contribution to journalArticle

Abstract

As commercial fishing activity shifts to target different grounds over time, spatial gaps can be created in catch rate data, leading to biases in derived indices of fish abundance. Imputation has been shown to reduce such biases. In this study, the relative performance of several imputation methods was assessed using simulated catch rate data sets. Simulations were carried out for three fish stocks targeted by a commercial hook-and-line fishery off the southwestern coast of Australia: snapper (Chrysophrys auratus), West Australian dhufish (Glaucosoma hebraicum), and baldchin groper (Choerodon rubescens). For high-growth scenarios, the mean squared errors (MSEs) of geometric and linear imputations were lower, indicating higher accuracy and precision than that for base method (constant value) imputations. For low-growth scenarios, the lowest MSEs were achieved for base method imputations. However, for the final standardized and imputed abundance indices, the base method index consistently demonstrated the largest biases. Our results demonstrate the importance of selecting an appropriate imputation method when standardizing catch rates from a commercial fishery that has changed its spatial pattern of fishing over time.

Original languageEnglish
Pages (from-to)1348-1361
Number of pages14
JournalCanadian Journal of Fisheries and Aquatic Sciences
Volume74
Issue number9
DOIs
Publication statusPublished - Sep 2017

Cite this

@article{8819155ed8164febb3792e9601b7a8c6,
title = "Evaluation of spatiotemporal imputations for fishing catch rate standardization",
abstract = "As commercial fishing activity shifts to target different grounds over time, spatial gaps can be created in catch rate data, leading to biases in derived indices of fish abundance. Imputation has been shown to reduce such biases. In this study, the relative performance of several imputation methods was assessed using simulated catch rate data sets. Simulations were carried out for three fish stocks targeted by a commercial hook-and-line fishery off the southwestern coast of Australia: snapper (Chrysophrys auratus), West Australian dhufish (Glaucosoma hebraicum), and baldchin groper (Choerodon rubescens). For high-growth scenarios, the mean squared errors (MSEs) of geometric and linear imputations were lower, indicating higher accuracy and precision than that for base method (constant value) imputations. For low-growth scenarios, the lowest MSEs were achieved for base method imputations. However, for the final standardized and imputed abundance indices, the base method index consistently demonstrated the largest biases. Our results demonstrate the importance of selecting an appropriate imputation method when standardizing catch rates from a commercial fishery that has changed its spatial pattern of fishing over time.",
keywords = "CHOERODON-RUBESCENS, ABUNDANCE INDEXES, STOCK ABUNDANCE, MARINE RESERVES, CONNECTIVITY, FISHERY",
author = "Marriott, {Ross J.} and Turlach, {Berwin A.} and Kevin Murray and Fairclough, {David V.}",
year = "2017",
month = "9",
doi = "10.1139/cjfas-2016-0182",
language = "English",
volume = "74",
pages = "1348--1361",
journal = "Canadian Journal of Fisheries and Aquatic Sciences",
issn = "0706-652X",
publisher = "CANADIAN SCIENCE PUBLISHING, NRC RESEARCH PRESS",
number = "9",

}

Evaluation of spatiotemporal imputations for fishing catch rate standardization. / Marriott, Ross J.; Turlach, Berwin A.; Murray, Kevin; Fairclough, David V.

In: Canadian Journal of Fisheries and Aquatic Sciences, Vol. 74, No. 9, 09.2017, p. 1348-1361.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Evaluation of spatiotemporal imputations for fishing catch rate standardization

AU - Marriott, Ross J.

AU - Turlach, Berwin A.

AU - Murray, Kevin

AU - Fairclough, David V.

PY - 2017/9

Y1 - 2017/9

N2 - As commercial fishing activity shifts to target different grounds over time, spatial gaps can be created in catch rate data, leading to biases in derived indices of fish abundance. Imputation has been shown to reduce such biases. In this study, the relative performance of several imputation methods was assessed using simulated catch rate data sets. Simulations were carried out for three fish stocks targeted by a commercial hook-and-line fishery off the southwestern coast of Australia: snapper (Chrysophrys auratus), West Australian dhufish (Glaucosoma hebraicum), and baldchin groper (Choerodon rubescens). For high-growth scenarios, the mean squared errors (MSEs) of geometric and linear imputations were lower, indicating higher accuracy and precision than that for base method (constant value) imputations. For low-growth scenarios, the lowest MSEs were achieved for base method imputations. However, for the final standardized and imputed abundance indices, the base method index consistently demonstrated the largest biases. Our results demonstrate the importance of selecting an appropriate imputation method when standardizing catch rates from a commercial fishery that has changed its spatial pattern of fishing over time.

AB - As commercial fishing activity shifts to target different grounds over time, spatial gaps can be created in catch rate data, leading to biases in derived indices of fish abundance. Imputation has been shown to reduce such biases. In this study, the relative performance of several imputation methods was assessed using simulated catch rate data sets. Simulations were carried out for three fish stocks targeted by a commercial hook-and-line fishery off the southwestern coast of Australia: snapper (Chrysophrys auratus), West Australian dhufish (Glaucosoma hebraicum), and baldchin groper (Choerodon rubescens). For high-growth scenarios, the mean squared errors (MSEs) of geometric and linear imputations were lower, indicating higher accuracy and precision than that for base method (constant value) imputations. For low-growth scenarios, the lowest MSEs were achieved for base method imputations. However, for the final standardized and imputed abundance indices, the base method index consistently demonstrated the largest biases. Our results demonstrate the importance of selecting an appropriate imputation method when standardizing catch rates from a commercial fishery that has changed its spatial pattern of fishing over time.

KW - CHOERODON-RUBESCENS

KW - ABUNDANCE INDEXES

KW - STOCK ABUNDANCE

KW - MARINE RESERVES

KW - CONNECTIVITY

KW - FISHERY

U2 - 10.1139/cjfas-2016-0182

DO - 10.1139/cjfas-2016-0182

M3 - Article

VL - 74

SP - 1348

EP - 1361

JO - Canadian Journal of Fisheries and Aquatic Sciences

JF - Canadian Journal of Fisheries and Aquatic Sciences

SN - 0706-652X

IS - 9

ER -