ESTIMATING TOTAL KILL OF NORTHWEST ATLANTIC HARP SEALS, 1994-1998

Discussions about the magnitude of seal hunts usually revolve around estimates of reported (or landed) catches. It is generally acknowledged, however, that some animals which are killed by hunters are neither recovered (e.g., Malouf 1986) nor accounted for in landed catch statistics. For the purposes of stock assessments and management, it is often assumed that any unreported kills are subsumed, at least implicitly, in estimates of natural mortality (m). In the case of Northwest Atlantic harp seals (Pagophilus groenlandicus = Phoca groenlandica), scientists have long urged that an attempt be made to account for animals that are killed but not landed, in order that total removals can be incorporated directly into stock assessments. This has not yet happened, however. The current population model used in management of the Canadian harp seal hunt does not attempt to account for seals that are killed in various seal hunts but not landed (struck and lost), or those that die as a result of other human activities. It is assumed that all such deaths are implicitly included in estimates of natural mortality. A recent workshop on interactions between harp seals and fisheries recommended that "effort should be made to derive estimates of removals [of Northwest Atlantic harp seals] in addition to annual information on landings" Here, I attempt to do this for the years 1994-1998. My preliminary calculations support the view that unreported kills are sufficiently large and variable that they should be accounted for explicitly in estimates of hunting mortality, both in future stock assessments and in models used to generate scientific advice to management authorities.