From the Jersey Coast Angler Associations's June 2008 Newsletter:

Summer Flounder Stock Assessment Technical Update
By Bruce Freeman

This year is especially important for assessing the stock of summer flounder, known locally as fluke. Stock assessment scientists from coastal states, federal government, academic institutions, Mid Atlantic Fishery Management Council, and the Atlantic States Marine Fisheries Commission will present and review the most up-to-date biological information and catch data for this species. This process is termed a benchmark assessment. The results of this assessment will be used over the next several years in setting catch limits for fluke, both recreationally and commercially.

Three workshops are being held this winter and spring leading up to the final fluke benchmark stock assessment this June. The previous benchmark assessment completed in 2006 utilized a Virtual Population Analysis (VPA) model for determining the allowable coast-wide harvest. While this model has been used for many years, not only for fluke but many other fishes, it has several shortcomings. Its basic design relies upon a retrospective view, i.e., it uses data we know to have already occurred. For example, by monitoring the catch of both the recreational and commercial fishing we know fairly accurately the least number of fish that existed. By determining the size of these fish, we can approximate the age. If we go back in time and add what we estimate to be those fish which die from natural causes and those which have died incidental to fishing, we can determine the number of fish that make up a particular year class. We can then add the various year classes together to gain an insight into the total stock of fluke.

The VPA model is useful in looking backwards in time to determine what has occurred, but it is poor in predicting what will occur in the future. The most recent information shows that the VPA model under estimated fishing mortality by about 30%, over estimated the number of young fish entering the fishery by about 5%, and over estimated the total number of fluke by about 10%.

So far as the February and April workshops are concerned, it has been agreed that the VPA model is not adequate for making future stock projections and needs to be replaced. Three other models were considered with one of them being rejected outright. The working group is considering two possible models, the SS2, now used for Pacific Coast fisheries and the ASAP model. Each has its strong points and weaknesses. It remains to be decided as to which will be used for the fluke stock assessment.

More important than selecting which model to use are the variables used in the model. Either model appears to give the same outcome given that the same variables are used in the calculations. One of the variables used is the number of fish removed by recreational and commercial fishermen, both those landed and those killed incidentally. While the NMFS stock assessment scientists feel that the landings from both the recreational and commercial fisheries are reasonably accurate, recreational fishermen remain very skeptical that the MRFS fishery landings data used in the assessment are correct.

Since the New Jersey catch of summer flounder accounts for about 40% of the coast-wide total, any inaccuracy in the NJ numbers will have a large impact on the overall stock assessment. In addition, there remains concern that the 10% incidental mortality for the recreational fishery and 80% for the commercial fishery may not be correct.

However, even more important is the rate of natural mortality (M). Prior fluke stock assessments have used M= 0.20 or an 18% annual mortality rate. Some biologists argue that M is set unreasonably low and it should be set at 0.3 or 0.4. Some argue that we need to use sex specific mortalities since males mature earlier in life than females and do not reach as old an age. Just four years ago no fluke over seven years old were found, now males as old as 12 years and females to 14 years have been aged. Any assigned mortality rate by sex can lead to very different model results. Nevertheless, the scaling effect of natural mortality has the most impact on the results. For example, if we use as M 0.2 vs. 0.4 and run the models, we see that in the first instance the fluke stock would be about halfway to our goal of maximum sustained yield. In the second instance the fluke stock would be near the maximum sustained yield level and we would not be over fishing the stock.

The working group will meet one last time in late May in an effort to resolve the mortality issue as well as several other technical issues prior to the June Stock Assessment Workshop meeting. I will be at the May meeting and will report the progress.

Back to Top