Fishermen focus on how ADFG sets, achieves escapement goals

Editor’s note: This is the eighth in the Morris Communications series, “The case for conserving the Kenai king salmon.”

Each spring, as the early-run king salmon start returning to the Kenai River, the Alaska Department of Fish and Game begins a four-month effort to manage fishing in a way that ensures enough salmon swim past fishermen of all types to meet escapement goals. 

“‘Escapement’ is actually what escapes fisheries and lives to spawn,” said ADFG biologist Tim McKinley, who helped draft the current king salmon escapement goals during the fall and winter of 2012 and 2013.

ADFG sets the escapement goals, which are the number of fish that need to return to produce healthy returns in the future.

On the Kenai, managers try to meet escapement goals for several runs: late- and early-run kings, and sockeyes. 

For each run, ADFG wants a target range of fish to return and spawn — too few fish and too many fish are both problematic for future returns.

At high spawning levels, density dependent factors can reduce survivability and therefore the number of salmon that return in the future from that spawning year. 

That means that too many fish on the spawning grounds can produce smaller future runs, although it can be very hard to determine how many fish are “too many.” 

That leaves ADFG to try to manage precisely within a set range of fish.

In 2012 and 2013, ADFG had to find ways to get enough kings in the Kenai River unharmed without exceeding the top end of the sustainable escapement goal range for sockeyes, which is 1.2 million.

The late-run king salmon sustainable escapement goal, or SEG, is 15,000 to 30,000 fish.

According to ADFG’s final estimate, 15,395 late run kings escaped in 2013. That number is based on what were counted by the sonar, and subtracting the in-river harvest and additional fish to account for catch and release mortality.

It’s also a number for much debate, and managing multiple goals at one time in one river — especially when sockeye are abundant and kings are not — is not the most difficult part of the equation.

In recent years, the goals themselves have become the focus dispute for fishermen, biologists and stakeholder advocates concerned about catches and the health of the river.

The arguments stem, in part, from disagreements about counting methods. 

ADFG has used different tools over the past decade to estimate how many fish are swimming into the river, and how many of those are not caught by sport anglers, and instead allowed to spawn. 

Different tools have different “currencies,” or ways to count the fish also known as “enumeration.” ADFG must update escapement goals periodically to reflect new enumeration methods, the most recent returns and new knowledge about what number of spawners result in the best returns.

In March 2013, the Board of Fisheries voted to approve the new late-run Kenai king salmon goal produced by ADFG. The board vote was a formality, as under the law the board must accept sustainable escapement goals without alteration.

The board does have the ability to choose optimum escapement goals, or OEG, which can be higher than the SEG for allocative purposes. 

In 2011, the board raised the late-run sockeye salmon optimal escapement goal to enhance the in-river sport fishery. The change raised the upper end of the goal from 800,000 to 1.4 million, allocating 200,000 for the in-river sport fishery. That was proposed by the Kenai River Sportfishing Association, or KRSA.

The sustainable escapement goal for that run is 700,000 to 1.2 million salmon, also set in 2011. 

The goals for Kenai River kings are based on a run reconstruction that McKinley and other ADFG scientists modeled using 30 years worth of data from both sonars, and other enumeration methods and relative abundance metrics such as mark-recapture, netting projects and catch per unit effort.

The late-run reconstruction, “is trying to take all of that into account,” McKinley said.

The reconstruction produced official estimates of total run, in-river run, escapement and recruitment for late-run kings, he said.


Based on that, the department determined the ideal escapement, that would produce the best runs in the future, using the counts from the new DIDSON sonar. DIDSON stands for dual-frequency identification sonar, and it produces more distinct results between kings and sockeyes. The lack of distinguishing between species was the main drawback to the previous “split-beam” counters used for in-season management up until 2011.

The new late-run Kenai River goal of 15,000 to 30,000 is lower numerically than the previous goal of 17,800 to 35,500 used with the less precise split-beam counter. But that doesn’t mean that the department thinks fewer fish need to spawn, it just means the fish are measured in a different currency, McKinley said.


Questions linger

Fishermen, however, have questions about the methodology  — from concerns about how the enumeration methods are incorporated into the goals, to issues with the Bayesian statistical method that underlies the models.

Bayesian statistical methods use the language of probability to quantify uncertainty in the model parameters, including the data that drives the model.

According to ADFG’s late-run escapement goal report, the use of Bayesian methods incorporates a more realistic assessment of uncertainty than classical statistical methods, and allows the effects of measurement error and missing data to be incorporated into the analysis. 

A peer review of the escapement goal reports released Dec. 16 on the ADFG website was largely favorable. ADFG asked three fisheries professionals to review the reports.

University of Washington fisheries professor Ray Hilborn, University of Rhode Island professor Jeremy Collie and National Marine Fisheries Service research biologist Robert Kope each reviewed, and commented, on the reports.

Each reviewer’s comments were published anonymously by ADFG.

One wrote: “The analyses are very thorough, and carefully explore and characterize the uncertainty in both the data and the resulting estimates of parameters and reference points. The use of a state‐space model in a Bayesian framework allows for incorporation of nearly all available data as well as evaluation of the uncertainty in those data. In my view, this a far superior approach to conventional spawner‐recruit analyses where all these data are condensed into time series of spawner and recruit abundance, and most of the uncertainty is ignored.”

The others largely agreed with that assessment, although they offered certain critiques of other components.

Prior to the peer review release, fishermen voiced their concerns about ADFG’s escapement goals. 

That might be one of the only points that the Upper Cook Inlet sport and commercial fishermen agree on right now, although they’ve taken opposing positions on the goals in the past.

Given the questions surrounding counting methods, and the ways in which those have changed, both organizations and fishermen have asked if it’s fair to develop a model and goal based upon them.

One of the reviewers said the methodology accounted for that properly, although he noted that he didn’t “know enough about the individual data sources to critically review all of the assumptions.”

The United Cook Inlet Drift Association, or UCIDA, has questioned other aspects of the methodology. After ADFG did its run reconstruction and determined the range of returns that would produce the most fish in the future, a “safety factor” was added.

According to the late-run escapement goal report, produced this past January, a range of 13,000 to 28,000 kings would be expected to provide yields of at least 90 percent of maximum sustained yield even at the minimum end of the range. 

That was adjusted upward by 2,000 fish at each end of the goal in part because of reduced productivity in recent years, according to the report, which means that the historical productivity may not be as reliable at precisely predicting future performance.

Jeff Fox, a former ADFG Kenai Peninsula area management biologist who now works as a consultant for UCIDA, said the upward adjustment for the “safety factor” also can be seen as an allocation for in-river users. 

More fish in the river makes it easier to catch one, added UCIDA Executive Director Roland Maw. According to Fox, UCIDA wouldn’t have questioned adding fish at the
lower bound of the goal. 

But at the upper end of the goal, extra spawning fish could equate to smaller returns in the future.

When the escapement changes from 12,000 to 13,000 fish, the return increases by about 1,800 fish, Fox said. But at the upper end of the goal you start losing fish in subsequent years when you add more spawners, he said.

That goes back to the density-dependent issues, which appear in the numbers generated by the run reconstruction.

Smaller escapements in the late 1990s produced large returns in the early to mid-2000s. Then, exceptionally high escapements in 2003, 2004 and 2005 far greater than the top end of the goal of 35,500 produced the smallest number of recruits into the fishery seen in the reconstruction, according to the escapement goal report.

In 2004, 63,770 kings escaped out of a total run of 99,690, according to the run reconstruction. Those spawners contributed just 21,280 fish to future runs, according to the model.

According to the report: “The relative role of density-dependent and density-independent factors for late-run Kenai River chinook salmon also remains uncertain.
Most of the cohorts represented in recent small runs have originated from large escapements, which is consistent with density dependence playing a large role. But these runs also coincide with a statewide decline in chinook salmon productivity thought to be due to factors, like marine survival, not related to stock-specific chinook salmon density.”

KRSA Executive Director Ricky Gease questioned the data in that relationship, however.

Gease pointed to the different counts going into the model in that time frame as a possible source of some issues in the resulting numbers.

According to Gease, ADFG’s numbers indicate less than a 1:1 to return from the 2003 to 2006 escapements. But Gease and other fishermen don’t recall seeing an unusually large number of fish on the river those summers, he said.

A total run of 99,690 fish, as the reconstruction said was seen in 2004, would have meant daily runs of 2,500 to 3,000 kings at the peak of the run. Anecdotally, Gease said sport fishermen didn’t see harvests that would have resulted from those runs.

“The numbers don’t make sense. And yet, those are the numbers used to reduce the escapement goal,” he said.

KRSA fishery management consultant Kevin Delaney, a former director of ADFG’s Division of Sport Fish, said there was a significant amount of uncertainty associated with those numbers. Although he said the Bayesian state-space model used for the run reconstruction represented ADFG’s best effort to make sense of several data sources, it still came with uncertainty.

KRSA has used some of the same numbers, however, in its own work. The nonprofit has a “Save Our Kenai Kings” campaign with daily social media updates, and has referred to a decline from 100,000 fish to 20,000 fish.

From 1986 to present, the only time nearly 100,000 fish show up in the run reconstruction was 2003, 2004 and 2005. 

Gease said that while the organization has questions about the numbers, those are the official data, and KRSA is using the best available information. He added that the organization is referring to runs of those sizes in the mid-1980s when the salmon management plans were created.

The largest run in the reconstruction in the 1980s was 82,190 kings in 1987.

Delaney was working for ADFG at the time. He said that he remembers a sonar estimate of that same run putting the number of kings in the river at 90,000, closer to the number KRSA is using in
its campaigns.

Delaney said that the more technical information provided to the public is, the harder it is to understand. KRSA tries to provide information “in a way that attracts attention and is not erroneous, not false,” he said.

He also noted that certain estimates of the large runs from the 2000s provided numbers above 100,000 kings.


Changing the threshold, again

The 2013 change was just the most recent in ADFG’s ongoing effort to review goals and ensure they reflect the most current information.

When the department switches to a sonar counter five miles upriver from the current DIDSON being used at mile 8.6, it’s possible that they will have to revise the goal once again, although that’s not yet known, McKinley said. The upriver sonar is counting more kings than the lower river sonar because it covers the river from bank-to-bank while the lower river sonar is missing some fish that swim behind it.

Even without the sonar plans, ADFG generally reviews goals every three years, on the same cycle as the Board of Fisheries discusses the rivers.

When the goal came up last March, sportfishermen and setnetters both commented on the best range for late-run Kenai kings.

Sportfishermen, including KRSA, generally wanted a higher goal. Setnetters asked for a lower goal. Both groups said biology was behind their concern; but both also stood to benefit from their preferred escapement range.

If ADFG was tasked with meeting a higher minimum goal for late-run kings, more would have to make it into the river, and sportfishermen might get additional harvest opportunity. 

At times of low abundance, that could come at the expense of commercial fishermen, particularly setnetters, targeting sockeyes in Cook Inlet if ADFG restricts fishing time to allow more kings to make it to the Kenai.

On the other hand, if the minimum goal is lower, the department might be able to offer commercial fishermen in the ocean and along Cook Inlet beaches a little extra fishing time, without worrying about conserving as many kings.

This time around at the Upper Cook Inlet meeting set to begin Jan. 31, 2014, members of the public have again asked for changes.

The Kenai River Sportfishing Association is asking the board to revise the goal upward to an optimal escapement goal of 20,000 to 40,000 late-run kings.

In March 2013, Gease said the range that was adopted by the BOF seemed to reflect the department’s desire to allow harvest, and didn’t best reflect the range of escapements the fishery has seen in the past. A higher goal would have reflected the peak levels of production that could be possible, Gease said.

Now, his organization is asking the board for that change.

Gease said escapement goals can be based around maximum sustained yield or maximum sustained returns. KRSA’s proposed goal is based on maximum sustained returns, Gease said.

Basing the goal on that metric can help ensure that in-river users have ample fish to harvest, he said. That will preserve a sport fish priority on Kenai kings, Gease said.

The new goal also reflects another KRSA concern.

“We don’t have return data from escapements below 23,000 fish,” Gease said, referring to the recent escapements in 2012 and 2013 that have been at or less than that level. “It’s all based on the modeling.”

When the board adjusted the Kuskokwim River goal in 2013, members noted that in the absence of data about what happens at a lower run, you don’t use a lower escapement goal, Gease said.

ADFG has not yet provided comment on the proposed change to the Kenai River goal, but will do so before the meeting begins in January. 

Gease said there’s another concern that KRSA’s new goal would address: declining salmon sizes.

Fishermen are seeing changes in the fish returning to the Kenai each summer. It appears, based on weir data and anecdotal evidence, that fish are smaller and a larger proportion of them are male, Fox said.

Trophy-sized kings are no longer a common catch on the Kenai. The last was reported in 2009. Setnetters, likewise, have reported that the kings they do catch are generally small ones.

But it’s the large fish that spawn best, Gease, Maw and Fox all agreed.

Larger females, greater than 30 inches, have more eggs than their smaller counterparts, making them more likely to spawn, Fox said.

There’s also a larger portion of males in the fish seen in the river. A skewed sex ratio doesn’t help with spawning either, Fox said.

Gease said KRSA’s goal would help account for those issues. More total fish swimming upriver could mean that more of those fish are healthy spawners.

But Fox and Maw don’t think a larger goal like KRSA is proposing is the best solution. Another option would be to look at how many fish the river can support, and set a goal based on habitat and other factors they said.

They’re not the only one with ideas for how to change the escapement goals. 

Mark Ducker, a setnetter, has asked the BOF for a biological escapement goal of 12,000 to 28,000 late run kings, which is the number of fish referenced in the 2013 escapement goal report.

Trustworthy Hardware’s Scott Miller cited the need to protect fish for sportfishermen in the future, and proposed a biological escapement goal of 17,800 to 35,700 fish.

Escapement goals defined

Biological Escapement Goals (BEGs) and Sustainable Escapement Goals (SEGs) are the most important goals used for management. These two goals are established based on the number of salmon, by stock and river system, that need to escape to spawn to provide for sustained yields in the future. BEGs and SEGs are determined through ADFG research programs.

While BEGs and SEGs are set by ADFG, Optimum Escapement Goals (OEGs) and Inriver Goals are established by the Board of Fisheries. Inriver Goals require ADFG to leave enough returning salmon unharvested to meet the BEG or SEG and to make a certain number of salmon available for inriver harvests. The OEG also can add fish to the BEG or SEG for inriver fishermen, but also may add fish for escapement when there are uncertainties in the data used to establish a BEG or SEG. Not all of the salmon stocks that ADFG manages have inriver goals or OEGs.

Source: Alaska Department of Fish and Game


Next week: A look at the forces and factors impacting king salmon in their ocean habitat.

Molly Dischner can be reached at Comments on this series can also be sent to

Dipnetters work a bank near the mouth of the Kenai River alongside a commerical fishing boat looking to net sockeye salmon. Salmon swim a daunting course past setnets, drifters, dipnetters and anglers in order to reach their natal spawning grounds in the lakes and tributaries of the Kenai River. -AP file photo

Dipnetters work a bank near the mouth of the Kenai River alongside a commerical fishing boat looking to net sockeye salmon. Salmon swim a daunting course past setnets, drifters, dipnetters and anglers in order to reach their natal spawning grounds in the lakes and tributaries of the Kenai River. -AP file photo