New Lobster Model Provides Improved Basis for Management
Contact: Yong Chen, School of Marine Sciences, 207-581-4303
ORONO- Scientists at the University of Maine and Maine Department of Marine Resources (DMR) have developed a new stock assessment model for the American lobster. The product of more than four years of work, the stock assessment model has been accepted by the Atlantic States Marine Fisheries Commission (ASMFC) as the basis for future population assessment of New England’s most valuable fishery.
Yong Chen of the UMaine School of Marine Sciences (SMS), UMaine post-doctoral researcher Minoru Kanaiwa and Carl Wilson of the DMR developed the model with support from DMR and Maine Sea Grant. Larry Jacobson of National Marine Fisheries Service (NMFS), an adjunct professor in SMS, also helped develop the new model which consists of thousands of lines of computer code reflecting the most up-to-date scientific methodology, information and data from government agencies and the fishing industry.
Over the last two decades, fishermen and federal scientists have been at loggerheads over the status of the Gulf of Maine lobster population. Fishermen have steadily increased the annual harvest, but the NMFS continues to list the lobster as “overfished.” Fishermen also report that their traps catch an abundance of egg-bearing female lobsters which, by law, they must return to the sea. In 2000, the apparent contradiction between lobster abundance and “overfished” status led to a concerted effort to develop a new scientific model.
The ASMFC requires states to address overfishing in the lobster industry by 2008.
When Chen came to UMaine in 2000, one of his first tasks was to develop that new model. He received his Ph.D. at the University of Toronto in 1995 and before coming to Maine, had done fisheries stock assessment modeling in Newfoundland, New Zealand and Australia. He brought experience in modeling rock lobsters, sea urchins, groundfish, crab and even the lowly sea cucumber. For lobster, his goal was to develop a model that “reflects biological realism”, he says.
In an oversimplified view, assessing lobster stock status takes three steps, each of which can become complex. First, scientists develop a population dynamics model that accounts for fish life history processes such as growth and maturation and other biological factors as well as mortality from natural and human causes.
Second, they calculate a “biological reference point,” a number that has been at the heart of the debate between fishermen and regulators. In lobster fisheries management, that point is known as “F10%.” It gives an indication of just how much fish harvesting affects reproduction and thus the ability of a stock to sustain itself. Scientists estimate the F10% number by calculating the fishing mortality that would make the egg production drop to ten percent of the egg production that would be in an unharvested population.
In the third step, scientists compare fishing mortality, estimated in their population dynamics model, to the F10% number. If the mortality of a harvested population is above the F10% number, the lobster is declared to be overfished.
Problems with the previous lobster stock assessment model and approach are well known. For example, it assumes that natural mortality is constant over time and known. Abundance data used in the model only include data from the NMFS bottom trawl survey which is not efficient in sampling lobsters and does not have a good coverage of inshore areas. Over 75% of the lobster landings in the Gulf of Maine come from inshore areas. In addition, the previous assessment model used an out-of-date statistical method. It takes estimates of F10% and fishing mortality from different models with different data, making them incompatible.
The complex life history of American lobster makes it difficult to develop population dynamics models with biological realism, notes Chen. As lobsters molt, pass from one life stage to another and move between deep and shallow waters, the factors at work are poorly understood, as are mortality rates in the wild.
Finally, the composition of the lobster population by age and sex is not well known, nor is the efficiency with which federal agency trawling runs sample the lobster population. There are no abundance-based biological reference points, which are required to decide if a fish population is “overfished,” as opposed to using F10% to decide if the fishery is “overfishing” the lobster.
Since 2000, the DMR has been sampling the lobster population by trawling in state waters. Although controversial within parts of the lobster industry, that activity has generated data that can be compared directly with the NMFS trawls. According to a recent paper by Chen and Maine DMR scientists, state and federal trawl data underscore the relationship between lobster size and location. As with other marine species, small lobsters tend to stay closer to shore. Large animals tend to travel farther and to spend more time in deep water.
One explanation for the abundant harvests of the past decade, suggests Chen, is a high level of recruitment, an estimate of new individuals reaching harvestable size in a given year. “They’re (fishermen) fishing hard. There’re no doubts about it. But recruitment is very strong every year,” says Chen. “So, although we are ‘overfishing’ the lobster population every year (with a high fishing mortality), the lobster stock abundance remains high and the stock is not ‘overfished’ because of strong recruitment.”
Chen, Wilson, and Kanaiwa have tested their model, as well as four other models, against a simulation backed by actual harvest data. Last fall, Chen presented the Maine model to a team of scientific reviewers that was evaluating its performance. The reviewers suggested modifications, which Chen incorporated, and then recommended that ASMFC adopt the Maine model.
The focus of Chen’s work this year will be to continue testing the model against the computer simulation. He will propose and test new biological reference points that can better reflect the status of the lobster fishery. He has received funding from ASMFC to adapt the model for use in managing lobster populations on Georges Bank and in southern New England waters.
Michael Errigo, a UMaine Ph.D. student in Chen’s lab, has received a three-year fellowship to work on the model testing. The National Marine Fisheries Service, the National Sea Grant College Program and UMaine contributed a total of $114,390 in stipend and tuition support. Errigo had the top ranked application for the annual fellowship program which focuses on fishery population dynamics.