.comment-link {margin-left:.6em;}

David @ Tokyo

Perspective from Japan on whaling and whale meat, a spot of gourmet news, and monthly updates of whale meat stockpile statistics

12/02/2006

 

The RMP's Catch Limit Algorithm

Annex D of the IWC Scientific Committee's report for 2006 includes an interesting summary of the process by which the current Catch Limit Algorithm (CLA) was originally evaluated and selected.
THE ROLE OF TUNING IN THE EVALUATION AND SELECTION OF THE CLA

When it came to recommending a single management procedure to the Commission for possible adoption, the Committee noted that comparison of alternative procedures is not entirely trivial, because they have to meet competing objectives. A procedure can always be modified to reduce the risk of depletion of stocks, but at the cost of allowing less catch. Likewise, higher catches can be achieved, but at the cost of a greater risk of depletion of the stock. The Committee recognised that in the presence of this trade-off, it could be difficult to compare the underlying performance of two candidate procedures if they are tuned to achieve different trade-offs between the two main objectives of catch and risk, and if only one tuning is presented.

However, if results from several alternative tunings of each procedure are available, then it may be possible, by interpolating the results if necessary, to draw conclusions about the relative performance of the two procedures.

In 1990 (IWC, 1991) the Committee therefore requested developers to present results a wide range of tunings to ensure that there would be at least some overlap in the range of risk-related performances of the different candidate procedures. In preparation for a final selection between procedures, the fourth and final Comprehensive Assessment Workshop on Management Procedures recommended that selection of a procedure be based on results for three specific tunings of each of the each of the five candidate procedures. These three tunings should be such so as to achieve a median final depletion (ratio of current to unexploited population) of the mature female population after 100 years of 0.60, 0.66 and 0.72, in a specific reference trial. The reference trial chosen was the so-called D1 trial. "The D1 trial was chosen because it reflected the greatest discrimination between the preferred tunings of the various developers; the highest and lowest final population values adopted for tuning purposes reflected the range covered by thesepreferred tunings" (IWC 1992a).

A set of 12 trials was selected for performance comparison, in addition to a number of robustness trials for which it was merely required that a procedure perform "acceptably". Fourteen basic performance statistics were provided for each trial, but comparison between procedures was to be made on thebasis of just 18 composite statistics calculated from the core set of 12 trials (IWC 1992b). The Committee selected one of the five candidate procedures on this basis, for recommendation to the Commission (IWC 1992c). The Commission accepted the recommendation and selected one of the three tunings presented, the 0.72 tuning (IWC 1992d).

DISTINCTION BETWEEN TUNING TARGETS AND PERFORMANCE CRITERIA

The denomination of the selected tuning in terms of a median final depletion to 0.72K in the D1 trial has caused considerable confusion within the secondary literature on the RMP. The tuning level of 0.72 has frequently been erroneously interpreted as a management target, when in fact it was only for comparative purposes for assessing the relative performance of alternative procedures. No special weight was placed by the Scientific Committee on either the D1 trial, nor on the median final depletion as a performance measure.

Of the eight risk-related performance measure used for management procedure selection, only two related to final depletion and six related to the lowest depletion over the 100-year period. The D1 trial was chosen because of the discrimination it offered between procedures, not because it was a typical, central or average trial in any sense. On the contrary, it was a relatively extreme trial in that it was based on the lower extreme of the range of MSY rates (1%, 4% and 7%)considered across the trials. For MSY rates in the middle of the range (i.e. the 4% trials), the median 100-year depletion was considerably higher.

The Committee based its selection of the current CLA on the consideration of a number of performance measures (18 in all). The actual final depletion level achieved by any catch limit algorithm, whether after 100 or 300 years, depends on the details of the specific trial. The current CLA, which has a nominal tuning level of 0.72, will only actually exhibit a median final depletion of 0.72 in the precise reference trial used for tuning purposes.

Very interesting - I admit to being including amongst those who were mistakenly thinking that 0.72K (i.e. 72% of the estimated carrying capacity) was a kind of management target.

As the extract notes, the "D1 trial" is "relatively extreme" in that it assumes an MSY rate of 1%, and that in the case of an MSY rate of 4%, the resulting size of the managed population would be "considerably higher" than that (0.72K). I wonder just how much "considerably higher" is?

Whatever it is, having taken this on board one can understand more keenly than ever why Greg Donovan, Head of Science at the IWC describes as "very limited" the levels of catch that would be permited under the RMP for only abundant whale stocks, and why Doug Butterworth described the RMP as being "so risk averse that the only real scientific basis for questioning its immediate implementation is that it is so conservative that it will waste much of a potential harvest".

(Greg Donovan's article "The International Whaling Commission and the Revised Management Procedure" is a good introduction to the material above)

Labels: ,


Comments: Post a Comment



<< Home

Archives

June 2004   July 2004   August 2004   September 2004   October 2004   November 2004   December 2004   January 2005   March 2005   April 2005   May 2005   June 2005   July 2005   August 2005   September 2005   October 2005   November 2005   December 2005   January 2006   February 2006   March 2006   April 2006   May 2006   June 2006   July 2006   August 2006   September 2006   October 2006   November 2006   December 2006   January 2007   February 2007   March 2007   April 2007   May 2007   June 2007   July 2007   August 2007   September 2007   October 2007   November 2007   December 2007   January 2008   February 2008   April 2008   May 2008   June 2008   July 2008   August 2008   September 2008   October 2008   November 2008   December 2008   January 2009   February 2009   March 2009   April 2009   May 2009   June 2009   July 2009   August 2009   September 2009   October 2009   November 2009   January 2010   February 2010   April 2010   May 2010   June 2010   July 2010   August 2010   September 2010   February 2011   March 2011   May 2013   June 2013  

This page is powered by Blogger. Isn't yours?