To:  Workshop on Analysis of Trawl Survey Experiment Participants

From:  Barbara Stevenson

Date:  12 January 2003

 Re:  Comments

 I am sorry I am not in Woods Hole with you, but even though I cannot attend there are a few things I would like to mention.

 In my talk with Lawrence Dixon I realized I still was unsure as to the purpose of this meeting.  This uncertainty may be what underlies my concern over the process.  Looking at the trawl data, it was very quickly obvious to me that neither the protocol gear nor the misaligned etc. gear was very good as survey gear.  Thus the question of whether the “bad” data can be adjusted or whether the “bad” data is good enough to use is secondary to establishing reliable survey gear and procedure.

  I had gotten the impression that all agreed this was essential and time sensitive.  The spring trawl survey is fast approaching.  Not only has a workshop to begin the process not been scheduled, but I keep hearing from the Inspector General’s office and from other various sources that the Center claims the trawl survey is not essential to stock assessments. 

I am very confused. If the trawl survey is not essential, why is there so much resistance to changing it – the less essential it is, the less the continuity is important.   But for many species – like skates, dogfish, and monkfish – species that it is obvious current survey gear is not designed to catch -  the overfishing definition is solely dependent on a survey index.  It has been stated by Woods Hole personnel that cod, haddock, and yellowtail were the only species that mattered. I always would have disagreed with this statement, but it was the general philosophy until recent years.  It should have been expunged from the center’s thinking long ago. 

• If they have no intention of redesigning the trawl survey so that it is adequate for all/most managed species, then the Center should notify the Councils,  NOAA, and the Congress that:
  • They cannot support any survey index based overfishing definition, and
  • That the law should be changed so that the Councils are not hamstrung by this inability.
• If they do intend to conduct trawl surveys that provide adequate information for species with trawl survey dependent overfishing definitions, then they need to immediately begin development to be implemented by the spring survey.  I would suggest this would be the best use of this workshop, although I suspect the persons attending would have been different if that were the stated purpose.

 Prior to the side-by-side trawls, it was agreed that the purpose was to see how extensive future comparison cruises need to be.  I now hear that the purpose of this meeting is to decide if the “bad” data is usable as is.  Please do not undermine what little faith the industry has in this process.  The side-by-side tows were extremely limited and covered only three areas.  It was obvious that differences in the gear varied greatly from area to area.  I have not attached any charts showing this as they are available on my website bdssr.com. This type of comparison needs to be done in all bottom types, depth strata, and species mix before any conclusion can be made. 

I have understood and my understanding has been reconfirmed several times that there are factions that believe that as long as the variability in the protocol gear is high enough to overlap with the difference between the gears, then the difference is not significant.  This only shows me how bad the protocol gear is.   The variability between gears and within the protocol gear was much higher than the variability between sets of tows on the Seabreeze.  Yes, there is a background level of variability that cannot be avoided.  No, this does not mean one can use that variability as an excuse for any degree of sloppiness.

I have also been assured by others that it is essential to determine the effect of the messed up gear, no matter what the inherent variability within a gear is.  I support the latter.

 My strong recommendation is that since we know the current survey gear and protocol does not work well (or at all) for some of the species for which management is dependent on the survey, step one should be implementation of a new survey gear and protocol.  Step two should be the type of comparative tows needed to compare old gear to new.  And step three should be to attempt to correct the “bad” data.

 I will gladly do what I can towards design and implementation of new survey gear; towards assuring industry vessel participation in the implementation of this gear as well as towards use of industry vessels in the broadreaching comparative surveys that will need to occur.

 Here’s wishing you all a successful workshop.