SSC Review of the Northeast Region Discard Estimation Methods for Groundfish Quota Monitoring and Annual Catch Limits 22 February 2010 Report of R . O ’ Boyle ( Review Panelist

General Comments A response to each of the review's terms of reference is provided below. Provided here are general comments pertinent to all terms of reference. The overall approach to discard estimation was designed-based as opposed to model-based. The former makes no attempt to explain discarding as a function of influential factors whereas the latter does. Discarding is a function of processes related to the stock (e.g. local concentration, population age/size structure), fishing operations (e.g. gear mesh) and management (e.g. small fish prohibitions). Ideally, it would be useful to develop a discard estimation model which describes the key processes and in doing so provides the capability to project discard rates in sampling strata where data are sparse. During the review meeting, it was noted that the initiation of fleet sector management in 2010 will introduce considerable uncertainty into how discard practices may change. Thus, at least in the short term, a design-based approach is warranted. It may be possible to explore modeling of the discard process as the database under sector management accumulates. Certainly, the results of the current analyses could be used as Bayesian priors in these modeling efforts. As a number of the papers noted, catch is equal to landings plus discards, the latter composed of observed and unobserved portions. The review primarily addressed how to estimate unobserved discards. Palmer (2010a: table 5) estimates that during 2004 – 2008, total discards for a number of fleets exploiting New England groundfish was generally less than 15% by weight of the catch. Wigley (2010a: Appendix Table1) provides information that indicates that unobserved longline, otter trawl and gillnet trips during 2005 – 2008 were about 16% of the total catch. On the other hand, it was assumed in these analyses that landings are known with certainty. Palmer (2010b) outlines landings allocation method which takes a precautionary approach to overestimate landings when data are missing to ensure that quotas are not exceeded. As updated information on landings becomes available during a year, the kept estimates used in the discard calculations will change, further changing the overall estimates of catch. It will be important to evaluate the relative contribution of variation in landings and discards for each fishery. It is important to keep in perspective the contribution of unobserved discards to the total catch. It would be useful for NMFS to provide monitoring reports that identify how much catch is based upon unobserved …