Ageing issue with the 2013/2014 year classes

[jysullivan]

I identified a difference in the peak year class in our NSEI age compositions compared to the Feds. The Feds show the peak in the 2014 year class, and while the State is showing 2014 to be large, the peak is in the 2013 year class. Below is a comparison between the Feds and State in the longline survey age composition data, where “year class” is the year of sampling minus the observed age. You can see in both data sets that the 2014 year class came in strong as age-2s in 2016, but that the State is showing the peak year class as 2013 (they are age-5’s in 2018).

After talking with Kevin McNeel with the Age Determination Unit (ADU), it is possible that an edge effect may have caused bias in the State’s ageing. An edge effect is fast growth that occurs on the otolith during the growing season, and it can be difficult for age readers to interpret this as a full year of growth or as a partial year of growth. If our readers had a consistent bias of calling this a full year of growth instead of a partial year of growth, it may explain why our peak in the age composition occurs with the 2013 instead of the 2014 year class. I think the Federal survey and fishery occurs slightly earlier on average than the State survey and fishery, which means that our samples may be subject to ambiguous edge effects more frequently than the Federal samples. It is also possible that because oceanographic conditions are different in Chatham Strait and the continental slope, this growth could look different or be more exaggerated in Chatham. Kevin shared this cartoon from the Committee of Age Reading Experts (CARE) manual, which shows how an otolith edge could be interpreted as a full year of growth or a partial year of growth:

If a portion of what we are calling 2013 year-class fish are actually from 2014, this would have two big impacts on our stock assessment: 1) the average weight for this large year class would be smaller, and 2) a smaller proportion of them would be assumed to be mature. This means we could be over-estimating total biomass and spawning stock biomass.

The ADU is seeking travel approval to attend the Committee of Age Reading Experts (CARE) summit on April 9. This summit allows ADFG to validate and conduct quality control and assurance for our ageing methods. If travel is approved, ADU will be able to resolve this issue quickly by subsampling the affected age range and comparing observed ages between the ADU and Alaska Fisheries Science Center (AFSC) labs. It is critical we address this issue as soon as possible because it will impact quota setting in 2019.

[jysullivan]

Clarification from Kevin:

Mid-summer capture dates are the most difficult to interpret, because we don’t fully understand how quickly sablefish otoliths can grow after winter and there appears to be variation between specimens within a given year. So, the NOAA longline survey otoliths that are collected in June/July would also be a subjective interpretation. For all agencies, we are more confident about edge interpretation for fish caught in early Spring or late Fall.

I just asked Delsa (AFSC) to get access to the known-age 2018 sablefish otoliths and Jodi and I will target smaller otoliths to get an idea where the error might be coming from. Also, I will send you the results of the current sablefish age structure exchange when all of the data is compiled so you can see the differences between the four labs that routinely age the species.

[jysullivan]

Information from Delsa Anderl (NOAA age lab) on 2020-02-03:

At our last CARE meeting, we identified one source of differences in some (not all) of the otoliths. In some cases, a weak mark was identified as a first annulus while some considered it a check. This 1st year problem may be easier to resolve among the ageing labs. In my lab at this time, our bigger question is what to do with marginal growth. The amount of seasonal marginal growth in recent years seems to be different from what we observed from past patterns. Specifically, now we sometimes see what appears to be an annulus forming along the otolith edge in the fall months. This is not something we were seeing in the past. Does that mean the annulus formation happens later in the calendar year now so we may be under-ageing young spring and summer caught fish? Could this change be due to warmer waters? The decision to add a year or not to the number of annuli observed on the otolith could be the main cause of why we do not agree on the dominant year classes.

[jysullivan]

Update 2020-02-19: Known age sablefish otoliths have been shipped to ADFG.