Salmon Tipping Points: Modeling Study

Using the last 60 years of data on pink salmon from Prince William Sound, we evaluated the role of spawner abundance, hatchery releases of the same broodline, total run size of the previous year, abundance of competitors, sea surface temperature, northern Gulf of Alaska Oscillation, and proxies of ocean acidification on wild pink populations. The pink salmon populations were analyzed using metrics of harvest value, market prices, and salmon population resilience.

To do this we simulated broodlines and subjected pink salmon populations to alternative combinations of (1) recruitment variability, representing stochastic population change, (2) climate-driven regime shifts between low- and highly productive ecosystem states, (3) changes in pink salmon prices, (4) time lags between when decision makers are able to detect regime shifts and when appropriate harvest rules are adjusted, and (5) variability in harvest implementation.

Sidebar
We focused on Prince William Sound because the region accounts for ~50% of the global production of hatchery pink salmon with local hatcheries releasing 600–700 million pink salmon fry annually. Because of the unique life cycle of pink salmon, fish that spawn in even years or odd years form two distinct populations called broodlines. Fish in the even year broodline will never reproduce with fish in the odd year broodline!

Sidebar on definition of recruitment and what recruitment variation means.  Also broodlines.

1. Delays in adjusting policies in the face of regime change can lower the financial performance of the fishery. It can also lead to higher chances of fishing moratoriums and lower cumulative recruitment (i.e. lower ecological health).
2. More responsive policy reduces the uncertainty in harvest value and variability in the fishery, but by how much depends on specific ecological relationships.
3. Policies that anticipate climate-driven regime change may not significantly improve the overall financial performance of the fishery, but they can reduce uncertainty and improve ecological outcomes by reducing fishing effort. They can also lower the chance of having to close the fishery.
4. High levels of recruitment variation, as historically experienced through boom/bust cycles in these fisheries, can lead to years of very high harvests and very profitable fisheries, but also increased chances of fishery closures.
5. Recruitment variation is the main driver of the boom-bust cycle with PWS pink salmon.
6. Price variation narrows and lowers distributions of economic impact, while widening distribution of recruitment (which also declines). If harvests are positive they are lower and vary more, with increased probability of zero harvests. More boom bust cycles but booms are diminished. Better information on recruitment allows managers to manage in a way that reduces the magnitude of boom bust cycles.
7. → Price declines lead to less harvest because fewer fishermen want to fish. → The model show there may be price declines, and the survey shows the price declines will affect behavior.
8. Slow management responses could cost the industry money (makes it less valuable), lead to lower harvests and lower recruitment, lead to lower annual harvests if positive and higher chance probability of harvest being zero.

• A number of fisheries in the U.S. have experienced sudden changes that existing management procedures are not adapted for (see also https://pubmed.ncbi.nlm.nih.gov/33981495/). Examples include:
  • Marine heat waves off the US west coast (2014 – present)
  • Collapse of snow crab, and other crab stocks in Alaska (fall 2022)
  • Pacific cod closure in Alaska (2020)
• Fisheries management systems are inherently variable (environmental variability, economic variability, variability imposed by harvest implementation), however decisions still need to be made in the face of uncertainty
• Our scenario analysis found that the biggest influence on the value of the fishery and harvests was how quickly management decisions could be adapted to new environmental conditions. Detecting these sudden changes relies on consistent data collection, collaboration and communication across science, industry, community and management networks.
• Developing harvest control rules that incorporate non-stationary (either with time – varying control rules, incorporating environmental drivers into stock assessments, or including regime changes) offer promising avenues for continuing to manage fish stocks sustainably into the future, and preserving the livelihoods of fishermen for generations.