Exercises for the Tucannon Model

1. Tucannon Harvest Exercise

The harvest exercise is a game for three. You will assume the role of harvest manager. One classmate assumes control of hydro-electric development. The other assumes control of land use in the watershed. Your job is to find a strategy for setting the harvest fraction as conditions change over time. Meanwhile, your classmates will be changing the simulated conditions. Their job is to test your ability to react to conditions beyond your control.

The "screen capture" shows the monitor of a Macintosh computer midway through a student experiment with the harvesting model. Instructions are given in the scrolling field on the left side of the screen. You are to experiment with the harvest fraction to maximize the total harvest over a 40 year period. In this screen, the classmates have raised the egg loss fraction and lowered the carrying capacity. You can imagine that their goal was to create a challenging situation for the harvest manager. Results for the first 240 months are on display in the main graph, and the "finger" is pointed at the harvest fraction which is currently at 90%. The harvest manager is probably wondering whether to change the fraction before advancing through another 24 months of the simulation.

Experiment with the interactive model until you feel that you have demonstrated your ability to manage the system. Conduct a final simulation that reveals your ability to manage the system under challenging conditions. Document the results of the final simulation with the "Print Map" command.

Use the downloaded model to verify the simulation results in Figure 14.4.

Confirm the 50% harvesting result shown in Figure 14.9. Then select a harvesting fraction between 50% and 90% and run the model to see if the simulated harvest matches what you would expect from Figure 14.10.

The Tucannon model sets the smolt migration loss at 90% and the ocean losses at 35% in the first year and 10% in the second year. Based on discussion with Ted Bjornn, the smolt loss fraction could be reduced and the ocean losses increased. Conduct a test with the smolt migration loss fraction at 50% and the ocean losses set at 75% and 50%. How does the new simulation compare with the simulation in the 2nd exercise?

Use the model to verify the stability check in Figure 14.5. (Caution: your random variations in the smolt migration loss fraction may not match exactly the sequence used in Figure 14.5, so you should not expect to match each and every variation.)

Introduce 90% harvesting midway through the previous simulation to learn if 90% harvesting is sustainable when the system is subjected to random disturbances.

Use the model to verify the simulated impact of development shown in Figure 14.12. Then conduct an additional experiment with 25% harvesting starting in the 240th month of the simulation. You should see annual returns of 3,490 returns and an annual harvest of 870. Can you do better with a higher harvest fraction?

The indicated redds in Figure 14.2 is the product of the adults spawning and the fraction female. The actual redds takes the value of the indicated redds or the maximum number of redds, whichever is smaller. Set the maximum number of redds to 3.27 to represent Bjornn’s estimate that the Tucannon could support 3,270 redds. Run the model to learn if this limit leads to any changes in the results in Figure 14.4.

If you have studied time series analysis, you will have learned about the stationary, autoregressive model to explain the variation in population numbers based on variations one year ago, two years ago, three years ago, etc. Repeat the simulation from the 5th exercise and ask for a table of results. You might imagine that the table of results is similar to time series data that might be collected on the returning adults. But since these results are created by a model, they might be called synthetic data. If you have access to a statistical analysis program, export results to the statistical program. Create a stationary, autoregressive model, and run the program to estimate the coefficients from the synthetic data. Does your analysis reveal a four year lag in the salmon population?

10. Precocious Fish

Draw a new version of the flow diagram in Figure 14.2 to include the possibility that some salmon return to the Columbia after only one year in the ocean. These are precocious fish, so name these fish "precos" in the model. Introduce a new converter called Fraction Precos as an input. Except for their early return, the precos have the same biological properties as the regular fish.

11. Precocious Fish, Part II

Draw a new version of the flow diagram to add the precocious fish to the model. But in this case, you are told that the precocious fish have a higher adult migration loss factor and a smaller number of eggs per redd..

Draw a new version of the flow diagram in Figure 14.2 if you are told that the salmon spend three years in the ocean rather than two. The total life cycle is now five years rather than four. Suppose all other population parameters were the same as the previous model, and suppose the loss fraction during the 3rd in the ocean is negligible. Would the new model reach a dynamic equilibrium? If so, would the equilibrium value of the number of returning adults be higher or lower than the previous model?

Build the model diagrammed in the previous exercise and use it to check your reasoning about the size of the equilibrium population.

14. Combine the Salmon Models?

The model in Chapter 13 provides a detailed simulation of hatchery smolts during their spring migration. Do you think it would be useful to combine the model from Chapter 13 with the life cycle model in this chapter? If the DT is set to 1 day (as in Chapter 13), how many steps would be needed to run the new model to compare to Figure 14.12?