Application of a mixed modelling approach to standardize catch-per-unit-effort data for an abalone dive fishery in Western Victoria, Australia. (17th January 2018)

• Record Type: Journal Article
• Authors: Giri, Khageswor; Gorfine, Harry
• Is Part Of: Journal of the Marine Biological Association of the United Kingdom. Volume 99:Number 1(2019)
• Page Start: 187
• Page End: 195
• Publication Date: 2018-01-17
• Subjects: CPUE standardization, -- Linear Mixed Model, -- Generalized Linear Mixed Model, -- abundance index, -- abalone fishery; Biology -- Periodicals; Marine biology -- Periodicals; 578.77
• DOI: 10.1017/S002531541700203X
• Languages: English
• ISSNs: 0025-3154

Abstract

Abstract : Despite the prevalence of catch per unit effort (CPUE) as a key metric in fisheries assessments it can be fraught with inherent problems that often cause its use as an index of abundance to become contentious. This is particularly the case with abalone, a sedentary shellfish targeted by commercial dive fishers around the globe. It is common practice to standardize CPUE to at least partly address issues about how well it reflects the actual abundance of a stock. Differences between standardized and unstandardized trends may lead to controversy between scientists and stakeholders when standardized trends provide a less optimistic picture of stock status. It is within this context that we applied Linear Mixed Model (LMM) and Generalized Linear Mixed Model (GLMM) methods to standardize CPUE for the Western Zone blacklip abalone fishery in Victoria, Australia. This fishery was chosen for our evaluation because it included substantial population losses from a disease shock during the middle of the time series. The effects of diver, reef location, month and their interactions with year were included as random effects in these models and the results compared with nominal geometric means. The two standardization methods provided similar standardized CPUE trends and clearly demonstrated that a large proportion of the variance could be attributed to diver and spatial effects. The GLMM seemed to explain more variability in the data and produced better precision for standardized CPUEs than LMM. The temporal trend in variability attributed to divers and spatial scales reveals the impact of disease as well as any homo/heterogeneity effect. The CPUE trends responded to the impact of disease against a backdrop of declining stock, however when compared with the inter-annual pattern in nominal CPUE, the standardized trends showed that the decline immediately following the onset of disease was less precipitous. In contrast to what appeared to be an increase in the nominal series during the more recent post-disease period, there was only a slight non-significant increase observable in the standardized trends.