This application forecasts future in-season harvest rates of one
exploited species in a competitive multispecies harvest using past
and contemporary data on species catches and harvesting effort.
Publication:
Smith, B.D., T.F.
Shardlow, and A.W. Argue. 1999. An error identification approach
to forecasting the harvest rate of one species in a multi species
fishery illustrated using two salmon fisheries. North Am. J. Fish.
Manage. 19:421-435
Abstract:
For a multispecies fishery managed to achieve a harvest rate target
for a single species, an early and accurate forecast of an in-season
or the end-of-season harvest rate index can assist a manager with
a decision to control effort.
We present a model for rapidly forecasting an in-season or end-of-season
harvest rate index for one species, the species of interest, in
a multispecies fishery using catch and effort data collected for
that fishery.
These harvest rate predictions can be statistically evaluated
against harvest rate indices calculated independently from demographic
and biological data obtained for the fishery.
The model structure was defined such that an estimate of both
measurement error and process error could be obtained. The data
required are the historical time series of catches for the species
caught in the fishery, fishing effort, and harvest rate indices
for the species of interest calculated independently of the catch
and effort data.
We illustrate the model with chinook salmon Oncorhynchus tshawytscha
as the species of interest in the west coast of Vancouver Island
troll fishery and the Strait of Georgia sport fishery.
Other important species caught in these multispecies fisheries
are coho salmon O. kisutch, sockeye salmon O. nerka,
and pink salmon O. gorbusha.
Addendum:
The publication accompanying this model did not consider the
use of AIC as a criterion for model selection. However the model
application offered on the webpage does incorporate AIC.
|