 |
||||||||||
 |
|
|
|
|
||||||
|
|
|
|
|
||||||
|
|
|
|
|||||||
|
|
|
|
|||||||
|
|
||||||||||
Canadian Science Advisory Secretariat
Research Document - 2003/068
Harvest Simulations for 2003-2006 Harp Seal Management Plan
By Hammill, M. and Stenson, G.
Abstract
The northwest Atlantic harp seal population is currently estimated to number around 5 .2 million animals. The current large size of the herd coincides with a failure of northwest Atlantic cod stocks to recover from a period of intensive overfishing that resulted in closure of the Atlantic cod fishery at the beginning of the 1990s. Although the potential impact of harp seals on the recovery of Atlantic cod are inconclusive, the large size of the herd combined with improvements in market conditions have lead to requests for an increase in the Canadian allowable harvest level. Current management objectives to increase the economic return to the industry consider that a smaller population size is acceptable as long as it remains above the Precautionary Reference level (N70), which at 70% of the estimated maximum population size is about 3.85 million animals. Owing to uncertainty associated with current estimates of population size, it was suggested that the lower 60% confidence limit serve as a metric to determine when N 70 had been attained. The width of the 60% confidence intervals increases as time since the last survey increases, reflecting an increase in uncertainty concerning the estimate of overall population size. Regular and frequent surveys are necessary to reduce the uncertainty surrounding these estimates. To determine the impact of various harvest levels on northwest Atlantic harp seals, a simplified Excel model incorporating uncertainty was constructed. The model results were similar to the model used previously to estimate abundance of this population to 2000. Harvests ranging from 75,000 - 500,000 over the next three years were examined for their potential impact on the northwest Atlantic population. Assuming that the age structure of the Canadian and Greenland harvests does not change, and that no changes occur in reproductive and natural mortality rates, the replacement yield for the current population of about 5.2 million animals is approximately 255,000 animals. With the exception of one run assuming a harvest of 75,000 animals, the scenarios assumed harvest levels that exceeded current estimates of replacement yield and consequently resulted in an overall population decline. Harvest levels of 275,000 animals result in only a slight change in abundance until 2009, after which the population begins to decline at a more rapid rate. Harvests as high as 500,000 animals for three years, followed by harvests of 275,000 animals per annum resulted in the population dropping rapidly to N 70 by 2009. Harvest scenarios that examined a variable take within a three year period had a similar impact on the population as fixed harvest levels that removed the same total number of animals during the three year period. Including additional uncertainty such as the variability surrounding the actual fraction of the established quota that is harvested, and reporting rates, increased the overall uncertainty around the modeled population estimates. Further simulation testing is needed to examine the performance of the model and the usefulness of using the 60% C.I. as a metric for population size in response to failures in model assumptions and additional uncertainty.
