Fmsy search warnings

[Nick Fisch]

Hi all, 

We're working on an assessment of Dover Sole using SS3 and we'd just like to make sure we've got our forecast settings correct, as we're seeing some interesting results and warnings. 

Ideally we would like the Fmsy calculation to use the mean relative F between fleets based on the last 5 years of the assessment (we think we've set it up this way). 

Firstly, we're getting a warning about convergence in Fmsy (mostly concerned about #3). 

Warning 1 Adjustment: basis for HCR anchor was not set; setting to 2 to match default before 3.30.24
Warning 2 : depletion_basis is using virgin but HCR anchor is using SSB_unf from benchmark. Are you sure?
Warning 3 : poor convergence in Fmsy, final dy/dy2= -0.138238
Warning 4 : Fmsy/mey is >3x of Fspr are you sure?  check for convergence 
Warning 5 : Fmsy.mey is close to max allowed; check for convergence 
Note 1 Information: A revised protocol for the Fcast_yr specification is available and recommended.
 5 warnings  and 1 note 

We've tried altering the Btarget as was suggested in this thread (Question on MSY estimates) however that did not change the warnings nor Fmsy. In fact, Btarget it appears to be read in correctly in the echoinput.sso, however appears as -999 in r4ss output and in the Forecast-report.sso it does not appear to change from a value of 0.4.

I ran some numerical simulations to ensure the yield curve relative to unfished was correct and it appears so, but would be nice to have the Fmsy search converge. 

Any ideas as to what may be going on? and where might I find the "revised protocol for the Fcast_yr specification" (can't seem to find anything obvious in the manual)?

We're using the most recent version of the executable. 

Any information you can provide would be helpful. 

Thanks, 

Nick

--

Nicholas Fisch

nickc...@gmail.com

[Richard Methot - NOAA Federal]

Hi Nick,

I'm glad that you looked at the forecast-report.sso to look at the iterations searching for Fmsy and probably have compared your profile to the SPR/YPR profile in SS3's report.sso.

I suspect your diagnosis is exactly correct - with selectivity shifted towards larger mature fish, there are some fish that reach spawning size/age before they get exposed to F.  This can allow F to be very high on the selected fish while still preserving enough SSB to produce recruits to support the catch.  This effect is exacerbated at high steepness causing Fmsy to approach F that produces max yield per recruit, as you see in your profile.  When you see this happening it is important to invoke other considerations so that management does not blindly attempt to implement a very high F which can greatly reduce the mean age of fish in the stock and can cause high bycatch considerations.

You also flagged some other warnings.  Some of these are interactions that became recognized in the development of the newest issue of SS3.  Some are described in our recent presentation of the interaction of time-varying growth with reference point calculations.  The warning algorithm is not smart enough to know to be quiet if you are not using time-varying growth.  Please let me know if you have not yet resolved these issues.

Rick