Gulf of Maine cod has become the latest species to be potentially threatened amid controversial data. A 22,000 metric ton disparity exists between the 2008 survey, which showed a rebounding stock and was hailed as an example of solid managment, and the 2011 survey which contradicts the prior survey and suggests disaster on the horizon. Few believe that both can be correct. Although officials from Oceana and Pew Environment have arugued that the the 2011 survey is the more credible, experts ranging from council members to top scientists disagree. Saving Seafood interviewed numerous industry leaders, scientists, and individuals with council ties to produce this report detailing eleven of the most-frequently cited problems with the 2011 survey.
by John Cooke and Sarah Hanselman
WASHINGTON (Saving Seafood) 25 January 2012 — The 2011 Gulf of Maine cod stock assessment has many methodological uncertainties, and is based on assumptions that may not be an accurate reflection of the health of the stock. Some of these uncertainties and assumptions were introduced in the 2011 assessment (SAW 53), while others have been persistent (existing in all or some previous assessments). There are numerous problems; we have high highlighted eleven of them:
1. THE DEFINITION OF GULF OF MAINE COD
A persistent problem affecting the accuracy of current and previous assessments is the exact definition of ‘Gulf of Maine cod,’ specifically the stock boundaries, potential sub-populations, and movement between stocks. Under current models, Atlantic cod (in U.S. waters) is assumed to be two completely separate stocks (Gulf of Maine and Georges Bank), with no mixing between them. Recent research indicates that the stock boundary is wrong. Specifically, genetics indicate that cod in southern New England are more related to Gulf of Maine cod and are distinct from cod on eastern Georges Bank. Tagging data also shows frequent movement between Southern New England, the Great South Channel and the Gulf of Maine, and relatively little movement between those areas and eastern Georges Bank. This suggests that the stock boundaries should be revised to assess and manage eastern Georges Bank separately from inshore areas of the Gulf of Maine and southern New England.
2. THE LENGTH-WEIGHT RELATIONSHIP
The length-weight relationship (the equation used to convert fish length into fish weight) and catch-weight-at-age model (the weight of caught fish broken down by age) provided in the survey are unreliable. The survey calculates length-weight relationship based on their own catch; numbers from the commercial and recreational fisheries are not included. However, the fish caught in the survey are different in size from the commercial and recreational landings; where the survey is randomized—aimed at catching fish of all sizes—commercial landings are focused on larger fish, meaning the survey does not accurately depict what was caught commercially and recreationally.
3. CATCH WEIGHT AT AGE
The other model, catch-weight-at-age, is created by averaging the commercial and recreational catch. This creates a different set of problems. By combining the two data sets, the assessment blurs the distinction between what the recreation fishery catches and what the commercial fishery catches, Generally, the recreational fishery will catch smaller fish than the commercial fishery; it is not accurate to assume that the two separate industries are catching the same size and age of cod. By not factoring in this distinction, the model creates further uncertainty in the data. The previous surveys have collected the counts for recreational fishing, but have never included them in the model and the final assessment. Adding new information such as this can introduce a new source of error. Furthermore, no one can account for the accuracy of the recreational discard information with any credibility. It is an educated guess, at best.
4. MEASUREMENT OF AGE
The assessment’s measurements of age are also skewed. Large, older fish, generally older than age 7, are likely under-sampled. The model also groups all fish over the age of 9 into a single category, even though cod can grow past the age of 16. This leads to underestimating the average weight of the cod and potentially underestimating the size of the stock.
5 ASSUMPTIONS ABOUT MORTALITY
The assessment’s assumptions about mortality are also likely flawed. The model used in the assessment estimates a constant natural mortality (M = 0.2) rate among all age groups of cod. However, this model most likely needs to be rethought, as a mortality rate that varies with time and age is probably a more accurate depiction of the population. The assessment also makes the assumption that 100% of discarded fish die. While this may appropriate in some fisheries, tagging data and other scientific studies suggest a range of discard mortalities, less than 100%, in different fisheries (different gear types) and different seasons. Overall the mortality assumptions should be reviewed.
6. CHANGE IN SURVEY VESSELS
In addition to flaws in methodology, there are several troubling uncertainties and flaws in the survey methods and data, some of which are new to the 2011 assessment. For 2011, there was a change in the survey vessel used in the assessment. The Bigelow, the new research vessel, is different from the previous vessel, the Albatross in several important ways. Most importantly, the Bigelow is significantly larger than the Albatross; it’s big enough that it can’t navigate the shallower waters in the region that Gulf of Maine cod inhabit and both commercial and recreational cod fishermen frequent. As a result, it was unable to include in the assessment any of the abundant cod that swim in the shallower waters.
7. CHANGE IN GEAR TYPES
NOAA not only changed vessels for the assessment, but it also changed the gear types used to perform its survey. Because the change in gear and vessels resulted in the vessels being able to catch different amounts of fish, NOAA devised a ratio in order to standardize the catch data between the old and new vessels and the old and new assessments. This method of standardization introduces a new source of uncertainty into the assessment, as it is unclear whether the calibrations are correct.
In addition to a change in research vessels, an anomalously high estimate of recreational catch also provides a new source of potential error in the 2011 assessment. There is a legitimate question in whether the measurement of the size of the recreational fishery is an accurate representation of recreational fishing activity. The 2011 assessment concluded that half of all 2010 recreational fishing activity occurred in just a two-week period in the spring, an unusually high concentration of activity. Further, it concluded that 60% of the recreational catch came from private day boats. Given the size and significance of the charter and party boat industries, few observers believe this to be possible. These anomalous data points more likely reflect errors in the methodology and uncertainty in the results than they do an accurate picture of recreational fishing activity.
9. INSHORE SAMPLES FROM RECENT SURVEYS WERE EXCLUDED
Inshore samples from the recent surveys were excluded from the assessment due to inconsistent sampling even though they provide indications of higher abundance for fish age 0-2. Further, surveys by the Massachusetts Division of Marine Fisheries (MADMF) provide the only reliable measurements of juvenile fish abundance because they survey inshore juvenile habitat. However, they are consistently down-weighted and the MADMF fall survey was completely removed from the final model.
10. BIOMASS ESTIMATES ARE CONTRADICTED BY CATCH PER UNIT OF EFFICIENCY DATA
The biomass estimates of the new assessment, which indicates a dramatically lower stock population than previously thought, is contradicted by catch-per-unit-of-effort (CPUE) data, a measure of efficiency in the fleet. If the assessment was correct, it would imply that the stock of cod has shrunk dramatically, and that efficiency as measured by CPUE would decline, as cod would be more difficult to find. However, the opposite is true in CPUE data, which has been steadily increasing for most of the past decade, indicating a decrease in effort and an increase in landings.
An alternative explanation to the CPUE data is that, as the cod biomass declines, the stock will contract, pooling together in a smaller area and thus making them more efficient to catch. However, preliminary evidence would suggest that this is not happening. If the stock were to start converging, the fleet would converge on them in response. Rather than seeing a convergence of the fleet, fisherman are reporting finding cod in a wide variety of areas, indicating that the CPUE numbers could be a sign of a much larger biomass than has been reported in the assessment.
11. DATA CALLS INTO QUESTION ASSUMPTIONS ABOUT MAXIMUM SUSTAINABLE YIELD
The assessment data also calls into question both the stock estimates and assumptions about maximum sustainable yield (MSY). It has been accepted that a sustainable fishery can harvest around 20% of the cod stock without overfishing or depleting it. But according to the 2011 survey’s estimates, the past decade has seen catches at well over three times this level, with between 60-75% of the spawning stock caught. Catches in numbers that high should show a dramatic impact on the following year’s stock. However, the stock estimates do not show a consistent pattern of overfishing with little to no correlation between the population estimate and catch levels from the previous year, even when those levels would suggest a depletion of the stock. This suggests that either the stock estimates are too low or that the appropriate maximum sustainable yield should be reconsidered.
edited by Bob Vanasse