In iantaylor-NOAA/Lingcod_2021: Estimate the Status of Lingcod off the US West Coast in 2021

Likelihood profiles were conducted by fixing a given parameter that was estimated in the base model at a range of values in turn while estimating the remaining parameters. These profiles were conducted for female $M$, $h$, and $R_0$. We choose to not profile over male $M$ because male and female $M$ are highly correlated, and therefore, it is instructive to see the estimates of male $M$ given a fixed value for females.

```{asis, opts.label = "north"} Female $M$ was most informed by length-composition data and secondarily by the Washington recreational index and the \gls{s-tri} (Figure \@ref(fig:profile-piner-femaleM)). This information conflicted with age-compositions from the Washington recreational fleet (left panels of Figure \@ref(fig:profile-piner-femaleM)). Data from the commercial FG fleet suggested a much smaller $M$, closer to that of the previous assessment.

```{asis, opts.label = "south"}
Female $M$ was most informed by mainly length-composition data and
secondarily age-composition data from the \gls{s-wcgbt} because it was the only source
for which ages were included in the base model
(Figure \@ref(fig:profile-piner-femaleM)).
Information in the surveys provided little information with respect to female $M$ and
was slightly conflicting for the Deb WV index, which suggested a lower $M$.

From the stock trajectories, it is apparent that the upper range of investigated female $M$ values are implausible given that they lead to a highly unlikely increase in stock size relative to all other values (Figure \@ref(fig:profile-ssb-femaleM)). Male $M$ changed non-linearly with fixed values of female $M$, where the most similar estimates of $M$ between the sexes occurred near values of 0.3 (Table \ref{tab:profile-M}). The current status of the stock was highly dependent upon the assumed value of female $M$ (Figure \@ref(fig:profile-depl-femaleM)).

There was less information in the data available to estimate $h$ than female $M$, which can be seen with the smaller differences in likelihoods across fixed values compared to female $M$ (Figure \@ref(fig:profile-piner-steepness)). This was not surprising, as estimating $h$ is notoriously difficult. ```{asis, opts.label = "north"} Almost all values of $h$ that were explored are equally likely, but the age-composition data appear to have pushed the estimate higher than the prior (lower left panel of Figure \@ref(fig:profile-piner-steepness)).

```{asis, opts.label = "south"}
All data types contributed to the estimate of $h$, but the information in the ages
conflicted with the lengths for all sources except the commercial FG fleet and
the California recreational ages in the Deb WV index
(Figure \@ref(fig:profile-piner-steepness)).
This index includes information on discarded fish, and thus may be sampling slightly
smaller ages than the other recreational and commercial sources.
Information in the surveys provided little information with respect to female $M$ and
was slightly conflicting for the Deb WV index, which suggested a lower $M$.

The largest changes in the time series of spawning stock biomass with different values of $h$ are in the historical period before the main recruitment deviations are estimated or before there are reliable composition data (Figure \@ref(fig:profile-ssb-steepness)) The current status of the stock was highly dependent upon the assumed value of $h$ (Figure \@ref(fig:profile-depl-steepness)).

```{asis, opts.label = "north"} The length data were highly informative with respect to $R_0$, in particular the commercial TW lengths and recreational lengths from Oregon and Washington (Figure \@ref(fig:profile-piner-logR0)). All investigated values of $R_0$ led to the different estimates of the time series of spawning stock biomass with the population increasing in scale incrementally with $R_0$, which is not surprising (Figures \@ref(fig:profile-ssb-logR0) and \@ref(fig:profile-depl-logR0)). Though, the current fraction unfished never reached 1.0 for any value investigated and remained above the minimum stock size threshold of 0.25 for all values except 8.0.

```{asis, opts.label = "south"}
Many data sources contributed to the estimate of $R_0$,
including prior information, which happened to have a minimum at the cross roads of
where length and age information intersected
(Figure \@ref(fig:profile-piner-logR0)).
Values greater than nine led to large increases in the current estimate of
spawning stock biomass and a stock size that was larger than that from the unfished state
(Figures \@ref(fig:profile-ssb-logR0) and \@ref(fig:profile-depl-logR0)).

iantaylor-NOAA/Lingcod_2021 documentation built on Oct. 30, 2024, 6:42 p.m.