CRiSP Harvest models the complete life-cycle of chinook salmon stocks.
CRiSP Harvest is a user-friendly computer model that simulates the harvest of 30 chinook salmon stocks by 25 fisheries over an extended time horizon. The geographic range covered by the model extends from Southeast Alaska to the Oregon coast. Ten stocks and two fisheries from the Columbia River basin are included in the model.
The computational engine of CRiSP Harvest is based upon the forecasting portion of the Pacific Salmon Commission (PSC) Chinook Model. A major goal of the Pacific Salmon Treaty between the United States and Canada is to rebuild depressed chinook salmon stocks by 1998. The PSC Chinook Technical Committee (CTC) developed the Chinook Model to examine alternative management approaches to implement the PSC chinook rebuilding program.
A key feature of the model is the interaction between stocks through annual catch ceilings, or quotas, imposed upon fisheries that harvest multiple stocks. Catch ceilings are the primary PSC management tool. As stocks rebuild or decline at different rates over time, relative harvest rates in fisheries with catch ceilings also change. Single stock models cannot simulate this type of interaction.
CRiSP Harvest enhances the features of the PSC Chinook Model by providing a graphical interface to view and change model parameters, flexible code that permits expansion of the model, vastly increased computational speed and data storage capacity, and a stochastic operational capability providing measures of variability and uncertainty in assessing the impacts of proposed harvest management actions.
CRiSP Harvest produces information to help managers evaluate the effects of changes in environmental factors, such as brood year survival rates, and management actions designed to assist salmon recovery efforts, including:
CRiSP Harvest is written in the C++ language and runs on PC computers and Sun SPARCstations, which are powerful UNIX computers.
CRiSP Harvest is an empirically based stochastic model simulating fish survival over a complete life-cycle.
CRiSP Harvest is an accounting device for tracking the fate of individual chinook salmon cohorts through time. The annual computational cycle consists of the following steps:
Ocean survival rates for ages one through five are assumed constant for all stocks (currently set at 0.5, 0.6, 0.7, 0.8, and 0.9, respectively). Survival rates to age one (also called brood year survival rates, or "EV scalars" for Environmental Variability) are estimated during the calibration process. Brood year survival rates during simulation years can be deterministic or stochastic.
Fishing mortalities during preterminal and terminal fisheries include both legal catches and incidental mortalities, which include "shakers" (sub-legal sized fish caught and released during chinook fisheries) and "CNRs" (legal and sub-legal sized fish caught and released during "chinook non-retention" fisheries directed at other species, such as coho).
Maturation rates are stock and age specific for all stocks. Some stocks have variable maturation rates.
Production functions for both hatchery and natural stocks are estimated from historical data. Changes in hatchery production can be simulated, as well as supplementation (the release of artificially propagated fish in streams to increase or establish natural spawning fish populations).
Model parameters, including cohort sizes, age specific harvest rates, maturity schedules, and incidental fishing mortality rates, are estimated by a technique known as "cohort analysis" or "virtual population analysis." This type of analysis involves the reconstruction of an annual series of abundance estimates using catch and escapement data and making assumptions about natural and incidental mortalities. Coded-wire-tag (CWTs) placed in the snouts of millions of salmon smolts over the past two decades and their subsequent recovery in fisheries, at hatchery racks, and on spawning grounds provide the necessary data for run reconstructions.
Model parameters can be obtained from the PSC Chinook Technical Committee, or advanced users can utilize there own parameters. Parameters are stored in temporary files in ASCII text format which can be read by CRiSP Harvest without modification (with the exception of one file which must be translated from binary format to ASCII format).
CRiSP Harvest has two modes of operation: scenario and Monte Carlo.
CRiSP Harvest allows the operator to change parameters and view results interactively. Parameter values can be changed by using the mouse or keyboard. A 25 year simulation with 30 stocks and 25 fisheries takes less than 10 seconds. Results can be presented in graphical form on the screen immediately after a simulation run, or can be downloaded to data files for archiving or further analysis.
The model can be run in two modes:
A scenario mode tracks cohorts through a single time series starting in 1979 and typically extends for 25 years. The period from 1979 to the present is used as the calibration period. The operator must make assumptions about the future values of key environmental parameters, such as brood year survival rates and maturation schedules.
A Monte Carlo mode tracks cohorts through many time series. Each time series has a different set of values for key environmental parameters, such as brood year survival rates. These values are selected from previously observed values (bootstrapping) or drawn from a probability distribution. Results from Monte Carlo runs are used to study probability distributions of future stock escapements.
A graphical interface allows viewing and modification of variables and model outputs through an interactive map. Mouse clicking on stock and fishery icons permits easy access to parameters.
An analysis tool allows immediate graphical presentation of model results, such as stock abundances, terminal run sizes, escapements, legal fishery catches, and incidental mortalities. For Monte Carlo runs, histograms of stock escapements are prepared.
A report generation tool makes tables of model outputs for inclusion in documents. With SPARC-station software any graphical model output can be captured in a document.
A help tool provides information about stocks and fisheries in the model.
CRiSP Harvest is calibrated against observed terminal run sizes, escapements or catches for individual stocks.
The model must be calibrated to produce useful results. CRiSP Harvest is calibrated by finding a suite of stock and year-specific brood year survival rates that results in model outputs that most closely match user specific terminal run sizes, escapements, or catches for individual stocks during the calibration period. CRiSP Harvest utilizes the same calibration algorithms as the PSC Chinook Model.
The user specifies the brood year survival rates for the simulation period, often taken to be the median, or the average, of the calibration period values. The model results are known to be sensitive to the selection of the EV scalars for the simulation period.
Agencies in the Columbia River Basin have access to CRiSP.
CRiSP Harvest is being used at the University of Washington to analyze the potential impacts of proposed harvest management strategies to help recover Snake River fall chinook stocks listed under the Endangered Species Act. CRiSP Harvest for Sun SPARCstations is available to a network of state, federal and public agencies and tribes in Idaho, Oregon and Washington. The new PC version (currently in beta-testing) will make CRiSP Harvest available to the general public.
Workshops are held to teach model users.
Workshops have been conducted at the University of Washington to train agency personnel in the use of the CRiSP Harvest model. Employees from twenty-five federal, state and regional agencies have undergone training for the CRiSP.1 and CRiSP Harvest models.
The CRiSP model is programmed with a robust and flexible structure that allows for easy future accommodation of new theory with a minimum of reprogramming effort.
The following activities are planned for the near future: