[STAR Panel Rec.] Non-random or non-representative sampling

[chantelwetzel-noaa]

The 2021 STAR panel recommended further evaluation of sampling processes for fishery length and age data to ensure random sampling. This was identified as a medium priority for the next stock assessment.

"Continue to investigate the potential sources of bias to the nature of fishery dependent agedata sampling, particularly with respect to the implications of non-random sampling of aged data even when using a conditional age-at-length approach. Consider improvements to both sampling approaches and the preparation and treatment of the age data in assessments."

Conflict between ages and lengths and the appearance of non-random sampling in the collection of otoliths or the selection of otoliths for ages has been an ongoing concern in fishery data for lingcod. This was also discussed extensively at the 2017 STAR Panel. In the 2017 STAR panel, the fishery marginal age compositions were removed from both model areas.

The 2021 assessment opted to treat fishery ages as conditional-age-at-length data to account for potential non-random sampling of ages; however, there was still conflict in the model between lengths and ages that may or may not be due to non-random sampling.

This assessment should examine the fishery data by year to identify if there appears to be non-random sampling of ages based on the length data to understand these data. Depending on the outcome of this analysis, we should consider how these fishery data should be treated for the 2027 assessment.

[chantelwetzel-noaa]

Closing and nesting under #198