Dynamic factor analysis to reconcile conflicting survey indices of abundance. (2nd May 2021)

• Record Type: Journal Article
• Authors: Peterson, Cassidy D; Wilberg, Michael J; Cortés, Enric; Latour, Robert J
• Editors: Anderson, Emory
• Is Part Of: ICES journal of marine science. Volume 78:Number 5(2021)
• Page Start: 1711
• Page End: 1729
• Publication Date: 2021-05-02
• Subjects: catchability -- coastal shark -- data conflict -- fishery-independent sampling -- relative abundance; Ocean -- Periodicals; Fisheries -- Periodicals; Fishes -- Periodicals; Marine biology -- Bibliography -- Periodicals; 551.4605
• DOI: 10.1093/icesjms/fsab051
• Languages: English
• ISSNs: 1054-3139

Abstract

Abstract: Stock-wide trends in fish relative abundance are challenging to obtain when a single, comprehensive survey is unavailable, and multiple, spatially, and/or temporally fragmented surveys must be relied upon instead. Indices of abundance from multiple surveys frequently have differing trends, resulting in obscured true abundance patterns of the resource. We use an age-structured simulation model of two coastal shark species in the southeast United States to explore the performance of dynamic factor analysis (DFA) for reconciling multiple indices of abundance that are in conflict. Survey-specific time-variation in catchability was induced to generate conflicting indices of abundance. Key simulation sensitivities included survey variability, abundance pattern in the resource, and missing years of survey data. We caution against using DFA when there is no contrast in the underlying stock abundance or when trends in catchability in all surveys result in no survey that is representative of stock abundance. When multiple representative surveys were available, DFA proved useful across species in estimating stock-wide trends from conflicting survey indices with different selectivities, catchabilities, variances, and, to a lesser extent, with missing data. Our results suggest that resolving contrasting patterns among multiple time-series of relative abundance can improve understanding of the temporal trend in stock abundance.