Reviewer Comments on Priors

[CamFreshwater]

Starting to go through the synchrony manuscript comments. Reviewer 1 notes the following:

It's easy enough to evaluate the sensitivity of beta estimates to this prior, but I was wondering if any of you had strong opinions? @ann-marieH I believe these are the same priors used by FRSSI - did you and Gottfried mull this over at some point?

[ann-marieH]

I'm assuming that "because Catherine checked it" isn't a sufficient response?

let's see what I remember...

• prior dist'n with max obs S as the mean - yes. But that doesn't mean that it's constrained to be no greater than max obs, as the comment seems to imply?
• constraint of 3 times Smax was put in place after some of the stock estimates failed to converge
• we checked visually for odd-looking posteriors ... which I can't remember what we got, exactly - I remember some truncations happening, but 1. not sure if that was for alpha or beta or one of the lag betas and 2. nothing egregious??? (or rather, there were some egregious looking posteriors, which led to some adjusting of priors)
• SD of 1 is what we use for the lognormal ... fairly generic? I don't remember testing the effects of alternate SDs in specific, but this was years ago.

[carrieholt]

Fraser River sockeye were harvested relatively heavily from 1950 until the mid 1990s (~80% exploitation rate). Precautionary management (and the TAM) rule have only been place for the last 1-2 decades, concurrent with a decline in productivity keeping abundances low. It's unlikely that a large portion of the spawner data are above Smax. In addition, the priors on Smax are relatively uninformative, and allow for the possibility of a low Smax estimate.

[CamFreshwater]

Thanks, roughly what I had in mind and I think it will be sufficient. Plus I did do those sensitivity analyses where we played with some of priors on beta and it was relatively insensitive. I'll revisit those and see if they're relevant. Hoping to have a new draft and replies to reviewer comments for all of you by early next week.