By Jenny Neyman
Redoubt Reporter
For sport anglers fishing from the banks of the Kenai River or dip-netters wading into the muddy river mouth, sockeye salmon are easy to count — it’s one at a time, as each fish is reeled in or hauled by a net to shore.
For Alaska Department of Fish and Game biologists, tracking numbers of sockeye in the Kenai River is no simple matter. Several factors are involved — where the fish are in the river, the user group for which they are allocated and what technology is used to track them.
Since 2008, a sockeye at the Kenai River Mile 19 sonar station counts as 1.4 to Fish and Game, while, this year only, one sockeye caught in the upper Kenai counts as one, and only one, fish.
The difference is the result of an upgrade in sonar technology used to track sockeye escapement numbers in the Kenai, and a correction to the way those numbers are related to the Alaska Board of Fisheries’ intention to allocate a certain number of sockeye salmon in the upper river for sport fish harvest.
It’s mostly apples-to-apples, oranges-to-oranges scientific shuffling that doesn’t have a net effect on the fish or those fishing for them, but a few oranges of upriver sport fish-designated sockeye have gotten mixed up with the apples of the in-river escapement goal in past years, and a change to separate the two means 60,000-plus extra fish this year over the past two for fishermen to sink their hooks, nets and teeth into.
Up for the count
For the past eight years, Fish and Game has been working toward upgrading Bendix sonar technology, used in Alaska since the 1960s, to newer, more accurate dual-frequency identification sonar DIDSON technology. Bendix sonar had become so antiquated that it’s difficult to maintain, and the main individual who developed the system in the state has died, leaving no source of technical support, said Jeff Fox, area management biologist with Fish and Game.
Redoubt Reporter file photo. A computer monitor displays images of fish swimming in the Kenai River as recorded by a DIDSON sonar.
The switch to DIDSON is a statewide effort. On the Kenai Peninsula, DIDSON and Bendix sonar systems were used side by side in the Kenai and Kasilof rivers from 2004 to 2007 in order to compare accuracy. DIDSON sonar has a better ability to detect fish with better coverage of the water column. It has a longer range, higher power level, better interaction with boundary layers without losing fish images and a 29-degree field of view divided into several beams. DIDSON produces detailed fish images on a screen, can show direction of fish travel and the data can be stored and archived.
Bendix had between a 4-inch and 2-inch field of view, which missed many fish, especially close to shore. Its visual display was just a spike on an oscilloscope, which can be difficult to interpret and couldn’t be saved or stored. When a large glut of fish swam by, one spike on Bendix may actually have meant several fish.
It’s not that Bendix was junk, Fox said. It was cutting edge at the time it was developed and was a huge step up in the Kenai over the aerial surveys used to count fish before the 1960s.
“The average return in those days was 1 million fish. It’s now 4 million, so you can’t really say sonar didn’t help. It’s actually a very good tool,” Fox said. “The problem is it was developed when the escapement goal was 150,000 to 300,000. And then it got moved up to 250,000 to 400,000, then 400,000 to 700,000. So it’s not any stretch for it to have problems. It wasn’t developed for what we’re doing now.”
Fish and Game knew Bendix sonar wasn’t completely accurate and estimated an error range of 20 percent. But comparison with the DIDSON showed Bendix had actually been underestimating sockeye by more like 40 percent in the Kenai. That means one sonar count by Bendix is 1.4 as counted by DIDSON.
But even with DIDSON, the sonar count produced to estimate escapement isn’t each individual fish, because DIDSON isn’t a perfect system, either. It can’t detect every fish swimming by, especially if they’re in a large group, and can’t tell the difference between a small king or an early pink salmon and a sockeye. Fish and Game uses a fish wheel to check apportionment of the run — to estimate what percentage is sockeye versus other species. Species apportionment error isn’t much of a concern, Fox said, since small kings are relatively few and pinks don’t show up en masse until August, right around when the sockeye sonar counter is removed anyway.
When is a fish more than a fish?
Still, DIDSON isn’t a perfect system. It’s better than Bendix by far, Fox said, but DIDSON doesn’t count each and every fish that swims by the sensor at River Mile 19, either. That’s where the apples and oranges come in. Sonar counts produce an index of fish, not an actual, one-sonar-count-equals-one-fish tally.
“Sonar counts aren’t fish,” Fox said. “It’s an index. We know it’s an index. It’s reasonably close, especially with the DIDSON counts, but it’s still not accurate. We’re not counting all the fish.”
The fact that sonar counts aren’t one-to-one fish counts isn’t the problem, Fox said, since the escapement goal is measured as a sonar index. On the Kenai, Fish and Game’s in-river sustainable escapement goal for sockeye in runs less than 2 million fish is 650,000 to 850,000. That’s 650,000 to 850,000 sonar index counts of fish. That escapement range was established thinking in terms of Bendix counts, so one Bendix count equals one step toward the lower-bound, 650,000 threshold.
At average fish-passage rates, DIDSON sonar counts 1.4 times more fish than Bendix sonar. Since 2008, Bendix has been offline and Fish and Game has used only DIDSON sonar to monitor sockeye escapement in the Kenai. So to manage toward the escapement goal, Fish and Game figures that the 650,000 lower-end threshold, set in terms of Bendix counts, is equivalent to 910,000 in DIDSON counts. That doesn’t mean the number of fish has changed or that the escapement goal has been increased since 2008 — it’s just a different measure of tracking when that goal has been met.
“It’s very simple to convert the escapement goal from Bendix to DIDSON. It’s Fahrenheit to Celsius. Water still freezes when it freezes. It’s not a big deal,” Fox said.
The problem comes when factoring actual, individual sockeyes into the sonar index count. The Kenai’s 650,000 lower-end sustainable escapement goal for sockeyes in runs less than 2 million is made up of two factors — a 500,000 in-river escapement, measured by sonar index, and 150,000 sockeyes allocated by the Board of Fish for an upriver sport fish harvest. The 150,000 sport allocation comes from harvest surveys of sport anglers. And to a fisherman, one fish reeled in is one fish, period — not an index of fish, and certainly not 1.4 fish.
But in 2008 when Fish and Game switched exclusively to DIDSON and started converting escapement numbers, it converted the entire escapement estimate by multiplying by the 1.4 factor, including the 150,000 sport fish allocation. In 2008 and 2009, then, Fish and Game figured that the 650,000 lower end of the escapement goal, set in terms of Bendix counts, converted to 910,000 using DIDSON as the counter. The department managed Kenai fisheries accordingly, being conservative with commercial fishing openings and harvest limits so the 910,000 threshold could be met.
But the 150,000 upriver sport fish allocation of sockeyes should be measured in actual fish, not as a sonar index of fish, and shouldn’t be subject to the 1.4 conversion factor. That’s a realization that dawned on Fox just this winter, he said.
“It wasn’t until this last winter that I was arguing with somebody that I realized — ‘Wait a minute, we are screwing this up,’” Fox said. “And I explained it to him and he carried the ball — ‘Oh, yeah. You are screwing this up.’”
The conversion error meant the Kenai sockeye escapement was underestimated by about 60,000 fish in 2008 and 2009. Fish and Game should have converted the base, 500,000 in-river escapement by 1.4 to 700,000 and added in the 150,000 up-river sport allocation without the conversion factor, for an overall, lower-end escapement of 850,000. Instead, they managed for 910,000, which put 60,000 extra fish up the river. Had the run been stronger, the error would have meant even more missed fish and an even greater overescapement.
Running a correction
In biological terms, 60,000 fish isn’t a big deal, Fox said. Overescapement can be harmful to the health of future runs, if too many fish spawn and over-compete for available resources. But 60,000 isn’t going to impact future sockeye runs, he said.
“Typically if we bust the (escapement) goal by two or three times, it’s an issue. Sixty thousand is no big deal,” Fox said. “We did an analysis a few years ago. If our (escapement estimate) number was 100 percent off— so we thought it was 500,000 and it was really 1 million, it would cost us 10 percent average in production (of future runs). So it’s not a sky-is-falling kind of thing.”
Where 60,000 sockeyes do have an impact is in allocation. Fish and Game uses a more-restrictive fisheries management plan for Kenai sockeye runs less than 2 million, as was the case with the runs in 2008 and 2009. Opportunities for commercial fishermen were limited in order to hit the lower end of the in-river escapement goal, but that may not have been necessary if managers had realized they were undercounting the escapement.
“It’s more an allocation issue, and problems being caused in other fisheries,” Fox said. “(Sixty thousand extra fish) in the river to a sport fisherman it’s like, ‘Oh, hey, wow. That’s great.’ To a commercial fisherman, it’s like, ‘Whoa, wait a minute.’
“We’ve had closures the last two years over 30,000 and 40,000 fish. A higher escapement goal is harder to hit, and the way we hit it is by closing people.”
Making waves
Fish and Game announced that it would change how it calculated escapement for the 2010 Kenai sockeye run at a public meeting July 2, to correct for the 60,000 fish error. Presenters took heat from many in the audience for making a hasty decision, instituting the change outside the Board of Fish cycle and announcing all this during fishing season, when people are too busy fishing to consider fish politics or escapement estimation strategies.
“I was under the impression that the Legislature put money out for this DIDSON to get better management so you’d understand exact numbers,” said Gary Hollier, a commercial fisherman, of Kenai. “When you look at page 94 of the reg(ulation) book, it says ‘fish past the sonar counter at Mile 19.’ It doesn’t say anything about Bendix or DIDSON or units of fish. I’ve never heard of a fish unit until I got your message here a little bit ago. So it’s a little bit confusing.”
“Did the department also take this on their own account to adjust these numbers? I think the Board (of Fish) wanted accurate fish past there, not an adjustment. And when you have a 4-to-1 spawner return ratio average statewide and Cook Inlet is going to be down next year to 1-to-1, something’s wrong. Somebody should be really embarrassed. This Kenai River is driven to its knees,” Hollier said at the meeting.
Fox said the technology change is not new, nor is the escapement estimate conversion from Bendix to DIDSON. The only thing that’s new this year is the change in accounting for the 150,000 upriver sport fish allocation, and that’s being done this summer, as opposed to next year, to correct an error that resulted in less fish for commercial nets the past two years.
“A lot of them just wanted to wait one more year, but it’s like, ‘Well, wait. It’s closures.’ I mean the last two years we closed (commercial fishing) for less than that,” Fox said.
The Kenai sockeye escapement goal is set to be revised this fall as it is. The new goal will be based on DIDSON counts and factor in changes in catch allocations, so there will be no need for conversions. The department will address the new escapement goal with the Board of Fish at its February meeting this winter, and the board will also consider several in-river goal allocative proposals. In the meantime, Fox said that the escapement conversion was needed to remain in line with the board’s wishes for the upriver sport allocation, and that dictating scientific methodology is the department’s responsibility, not something in which the board is involved.
“The board passed a 150,000 in-river allocation based on a statewide harvest survey, which is their best interpretation of real fish. They then turned around to the department and said, ‘Apply that to the in-river goal.’ And the only tool we had was Bendix. We now know that that tool didn’t accurately or correctly index all the fish. We have a better tool now based on the past four years, and that’s what we’re using. That’s about as simple as I can put it,” he said.
Paul Shadura, a commercial fisherman and president of the Kenai Peninsula Fishermen’s Association, thanked the department for catching and correcting the error.
“I appreciate you guys coming down and making the change. I think it is very difficult for the department to come forward and make these implementations prior to the season, but it’s necessary. It’s the right, responsible thing to do,” he said. “What were talking about is 60,000 real fish here, and it does have some management implications.”
As of Friday, Fish and Game released its in-season Upper Cook Inlet sockeye run estimate. The Kenai River sockeye run as of Friday was estimated at 2.2 million fish, meaning run management now falls under the less-restrictive guidelines for run sizes between 2 million and 4 million fish.
The 2 million-plus run means the Kenai in-river escapement goal changes from a minimum of 650,000 (in Bendix units, or 850,000 in DIDSON corrected for the 150,000 sport harvest allocation) to a minimum of 750,000 (Bendix units, or 990,000 in corrected DIDSON units). Additional commercial fishing openings were announced over the weekend, dip-netting in the Kenai opened to 24 hours a day through July 31, and the sport fish bag limit for sockeye over 16 inches increased to six fish beginning Saturday, not including in the Russian River and upper Kenai fly-fishing-only waters near the Russian River.
With the sockeye run returning stronger than was originally forecasted and the conversion of sonar complete, Fox said the issue is over. It’s time to go fish.
“It’s just correcting a problem,” he said. “It’s a done deal already.”
