Thomas Gast & Associates. 2021. Analysis and model evaluation of long-term data collected at the Willow Creek outmigrant trap. Report 20190910YTFP for the Trinity River Restoration Program (TRRP). Thomas Gast & Associates Environmental Consultants, Arcata, California. Available: www.trrp.net/library/document?id=2492.
The Willow Creek outmigrant trap on the Trinity River has been operated annually since 1989. It was installed and continues to be operated primarily to assess outmigration timing and duration of salmonids, Chinook in particular. In 2014 Hayden and Heacock (Hayden and Heacock, 2014) developed the HDAT Model that models RT80 (when 80% of the Chinook juveniles have passed the trap) based on accumulated daily averaged water temperatures in Hoopa. We evaluated that model with a longer time series and investigated additional temperature threshold models using both the Hoopa and Pear Tree water temperature data to explain RT80 and RT50 (the date at which 50% of the Chinook juvenile run pass the trap). RT50 has more yearly variation than RT80 and was investigated as an alternative metric to RT80. We also investigated additional explanatory variables such as emergence date, redd counts, discharge, length frequency, and fish condition. Bioenergetics analysis was used to compare the growth of Chinook juveniles between emergence and outmigration as well as between years. We also evaluated the mark-recapture and flow-based abundance estimates for juvenile Chinook and compared the pre-ROD flow abundance estimates to the post-ROD flow abundance estimates.
The HDAT model using the full time series has lower correlation, but clearly has value. The best single variable explanatory model that we tested for RT50 and RT80 used a threshold accumulated daily averaged thermal unit (ATU) and the Pear Tree temperature time series. Several interesting multivariate models were identified that include emergence date and redd counts as well ATU, but do not have sufficient time series data for practical use. These should be evaluated further as more data becomes available.
The bioenergetics analysis shows that the potential consumption rate for juvenile Chinook is variable in February to the end of April and deceases to a lower constant level during the ROD flows. The decrease in water temperature associated with the decrease in potential consumption is apparent in the temperature record at the Pear Tree site. Exploratory models indicate that juvenile Chinook consumption is positively influenced by higher discharge in February, March, and April, but negatively influenced by redd counts the previous fall and discharge in May.
The flow-based abundance estimates average 54% of the mark-recapture abundance estimates pre-ROD (2004) and 37% of the mark-recapture estimates post-ROD. The best run timing models use the flow-based estimates, but the mark-recapture method should produce better estimates. This discrepancy warrants further investigation.
Juvenile Chinook abundance in post ROD years has significantly increased versus pre-ROD estimates despite the low abundance estimates for 2015 and 2016. During pre-ROD flows the juvenile Chinook abundance trend was either flat or decreasing depending on estimation method. Redd count data is not correlated to outmigrant abundance and has not increased, indicating that there has not been a concomitant increase in spawning.
These analyses indicate that warmer water temperatures during the initial time of the ROD-flows would increase the potential for growth and suggest that the warmer water will encourage earlier outmigration. Larger smolt size at time of ocean entry may enhance ocean survival. Water quality issues and risk of parasitic infection increase and may reduce survival later in the summer in the Klamath River. Selective withdrawal from Trinity Lake could provide warmer water temperatures during late April and May; however, the operational feasibility, effects on other fishes and aquatic life, and economic tradeoffs will need full evaluation prior to implementation.