Conserving biological diversity of native fish and protecting their habitats

 

 

November 15, 2004

 

MEMORANDUM

 

TO:ÊÊÊÊÊ Dr. Peter Dygert, NOAA Fisheries

FR:ÊÊÊÊÊÊ Bill Bakke, Director

RE:ÊÊÊÊÊÊÊÊÊÊÊ Comments on the statesâ proposed biological assessment for expanding the kill of wild steelhead in the lower Columbia River in 2005

_______________________________________________________________________

 

The Native Fish Society has reviewed the biological opinion that proposes to increase the kill of wild winter steelhead in the Lower Columbia River commercial spring chinook fishery.Ê These steelhead are federal protected species under the Endangered Species Act and listed as a threatened species.Ê The existing limit on the kill of ESA-listed wild steelhead in the commercial chinook fishery is 2%. The biological opinion submitted to NOAA Fisheries on October 29, 2004 by the states of Oregon and Washington argues for an increase in steelhead kill to 6% in order to harvest more hatchery produced spring chinook in the gill net commercial fishery.Ê The following are our comments on this biological opinion.

 

The states of Oregon and Washington present an argument for increasing the kill of ESA-listed steelhead in the commercial fishery based on increasing trends in wild steelhead abundance in the last few years.Ê They present information to support this contention, however, the information is inadequate becauseÊ the states use base lines that are not parallel and the abundance estimates are not as accurate as they claim.Ê

 

The data base lines employed by the states are one problem. On page 6, table 2 presents complete information from 1996 to 2004 showing the Willamette Falls wild steelhead count and the count from the tributaries above the falls.Ê Figure 3 shows the recruits per spawner from 1993 to 1998. Why did the data bases not cover the same years and why did these time series not include all the years of data available? The counts at Willamette Falls are less than accurate.Ê In most years (6 out of 8 years) the count of wild steelhead at the falls is lower than the counts in the tributaries (ranging from 156 to 3,468 fish) an in the other years the count at the falls is higher than the count in the tributaries (ranging from 4049 to 8,044 fish).Ê In only one year did the count at the falls come close to the count of spawners in the tributaries (2003 with a difference of 156 fish).Ê This suggests that the data used to confirm the abundance of steelhead in the Willamette ESU is weak and at best inaccurate. Yet this information is used to support the proposal to increase the kill of wild steelhead in the lower Columbia River.

 

On page 8 Figure 3 shows that all tributaries are out of compliance in at least one year with theÊ standard for recruits per spawner (1.2) in the Oregon Native Fish Conservation Policy adopted by the ODFW commission in 2003. This same figure also shows that these tributary wild steelhead populations are below replacement form one to three years out of the six years evaluated. It is not explained why only the years 1993 to 1998 were used for this evaluation.Ê A longer time series would provide more information about the abundance and apparent health of these populations.

The information in Figure 3 emphasizes those few years when some populations were in excess of 3 recruits per spawner, but this situation is not typical for these populations.Ê

 

On page 10 the BA notes that two of the three wild steelhead populations in Oregon monitored by dam counts have declined since 2002.Ê The Clackamas River declined in 2003 and increased in 2004.Ê Again this information is used to argue that the wild steelhead are increasing.ÊÊ

 

The Oregon data base supporting the increasing abundance of wild steelhead is not impressive.Ê The data series are artificially truncated, the abundance estimates are inaccurate for the Willamette and the dam counts show two of three streams have decreasing numbers.

 

The streams evaluated in the BA by Oregon and Washington represent a few streams where data exists to create a limited time series data base on wild steelhead.Ê The majority of streams with wild steelhead, many of them with small numbers of spawners, are not included in this analysis of steelhead abundance.Ê If the states are able to convince NOAA Fisheries that steelhead are abundant enough to warrant additional mortality in the commercial fishery, NOAA Fisheries will be ignoring streams that have not been inventoried. NOAA Fisheries is being asked to accept the proposal to increase the kill of wild steelhead on a small and incomplete data base for only a few large streams with relative large populations.

 

The BA states: ãThe estimate of intrinsic productivity (Ricker alpha (a) parameter with a 95% confidence interval) for the Clackamas was 1.57 (CI 0.7 ö 3.53) (Table 6).Ê The fact that the lower confidence interval was below 1.0 suggests that this population may not have been self-sustaining at recent levels of productivity.äÊ Until the last few years when ocean conditions have improved, most wild steelhead populations have not been doing well.Ê To suggest that increasing the kill of wild steelhead just when they get favorable ocean survival rates, allowing some populations, but not all, to increase in abundance, is not supporting recovery of these populations.Ê The states are less concerned about wild steelhead recovery in this BA than they are in trying to maximize the harvest of hatchery spring chinook, which means they must find a way to justify the additional kill of wild steelhead. That statesâ goal should be to convert improved smolt to adult survival to spawners rather than increase the kill of adult spawners.ÊÊ

 

The statesâ proposal uses estimated intrinsic productivity for populations of wild steelhead.Ê First of all these are estimates and they have not been verified.Ê Secondly, intrinsic productivity is free of density dependent mortality, in other words, the populations do not have enough spawners to seed the habitat so smolt production is high due to less competition.Ê The states have used this rationale for years in harvest management.Ê By over harvesting salmon populations and reducing the number of spawners they create the condition for high survival rates and smolt yield.Ê This ignores the function of nutrient enrichment for streams from spawners, gene conservation and effective population size for natural populations.Ê Intrinsic productivity estimates are used in the BA to show that that some wild steelhead populations can increase in abundance when there is a higher smolt to adult survival and therefore it is a justification to increase the harvest rate because their inherent productivity has not been harmed by years of poor escapement.Ê In contrast, a healthy wild steelhead population is a function of abundance where density dependent mortality becomes a selective factor, nutrient enrichment of streams is increased by abundant spawners, effective breeding populations are increased and genetic variation is increased for natural selection to work upon in a constantly changing environment.Ê The BA ignores these factors in its assessment because these factors do not support their primary interest of increasing the kill of steelhead to allow a higher harvest of hatchery chinook.Ê

 

On page 18 of the BA the WDFW provides Figure 7 showing recruits per spawner for the years 1993 to 1999.Ê This figure tells the same story that Oregon told, that is, the five populations inventoried had poor recruitment in the years 1993 and 1994; in years 1995, 1996, and 1997 only the Coweeman population had relatively high recruitment and the others poor recruitment; and in the year 1998 the Coweeman did not make replacement while the others had relatively higher recruitment, and in 1999 all five populations had relatively higher recruitment rates.Ê This data indicates that the higher recruitment for all populations happened only in one of the seven years.Ê The problem this data points out is that these five populations are not healthy and they could stand a longer period of higher recruitment to improve chances for recovery. This data would be improved if there were more years included in the time series, but it is clear that increasing the kill of wild steelhead populations just when their survival rate has improved is not a sound management decision.

 

On page 19 of the BA the statesâ declare their premise: ã The improved recruit per spawner ratios and the generally improved abundance in recent years indicate that most populations are responding favorably to improved marine survival conditions; therefore, at least in the short term, a modest increase in mortality due to the proposed fishery actions should be countered by the much stronger effect of favorable ocean conditions.äÊ The statesâ are basing their proposal on the ãresilienceä of a few populations they have data on, saying that because these few populations have responded in a positive way to increased smolt to adult survival due to improved ocean conditions, additional fishing mortality would be offset by improved productivity.Ê The statesâ are willing to expand the risk for wild steelhead during the initial period of rebuilding for all populations in order to gain harvest access to hatchery chinook.Ê Their stance is one of promoting additional risk to ESA-listed populations rather than trying to capture the benefits of improved survival to rebuild these populations.Ê NOAA Fisheries should request a time series data base that includes more years and more current years to assess the health of wild steelhead in the lower Columbia River.Ê Also, we recommend that NOAA Fisheries require the states to hold the line on rebuilding these wild populations rather than expand the risk to them through increased harvest impacts.Ê

 

Wild salmonids do not instantly respond to favorable environmental conditions.Ê Improved environments may increase survival, but it takes populations that have had low survival rates time to generate enough smolts to improve adult spawner abundance.Ê This lag time is critical when framing a recovery program for ESA-listed species that have had an ecological break. It takes several generations to increase adult abundance and smolt numbers.Ê To move in at the beginning of a cycle of improved survival, using that to justify increasing mortality as the statesâ are in their harvest BA, is an action that is not focused on conservation management for an ESA-listed species. It is an attempt to justify increasing the risk to ESA-listed populations in order to accomplish another management priority.Ê NOAA Fisheries should not be temped by this kind of reasoning since it is your responsibility to recover the species, delist them, and when the states get side tracked by other priorities, to remind them of their duty to recovery.

 

The BA states that travel time of wild steelhead in the Columbia River is limited.Ê The BA relies on two studies, one from the 1950s.Ê The states admit that the data base on travel time is poor and the information they have are estimates.Ê No real data exists to support the conclusions the states make about travel time and harvest impacts on any particular population or ESU.Ê However, the states do recommend important research that needs to be completed to improve the data on run timing.Ê The Native Fish Society recommends that this research be funded so that harvest accounting by ESU for steelhead can be accomplished.Ê On April 9^th of this year I requested that NOAA Fisheries develop a harvest accounting methodology by ESU and NOAA Fisheries concurred, the problem was time and people to do the work.Ê I understand the problems, but this BA would be on surer footing if harvest impacts by ESU for the lower Columbia River fishery were available.Ê

 

The problem the states have in this BA is that the timing of the chinook commercial fishery is at the same time that ESA-listed steelhead are abundant in the river.Ê There is some variation in run timing by population, but all populations are affected and the bulk of them may sustain major impacts from this fishery. The states created a model to estimate travel time in the Columbia but admit that ãThese results suggest that the modeled assumptions about travel times do not encompass actual fish behavior and that episodes of slower and faster travel time must occur.ä So the model and its assumptions are not factual, they are speculative and based on assumptions and estimates.Ê

 

The BA discusses various run prediction methods, but none of them worked very well, either under predicting or over predicting the run strength.Ê The states say that ãthe more accurate methods require smolt abundance and age distribution data, which are not available for most winter steelhead populations.äÊ The conclusion is that run size prediction needs more work before it is a useful tool in managing harvest of steelhead and providing the protection needed for rebuilding.Ê

 

The BA discusses using in-season updates to manage the fishery impact on steelhead, but found that these were not timely enough to provide the precision needed for a short fishery.Ê They concluded, ãIt is not clear whether this information can be used in-season to accurately manage the fishery.ä

 

The Native Fish SocietyÊ agrees with the statesâ recommendation that, ãManagement of the fishery could benefit from a study that accurately estimates multiple encounter rates for steelhead and the impact of multiple handles on post release survival rates.äÊ This recommendation is clear indication that the states are not comfortable with their estimates on impacts to steelhead in the fishery being proposed in the BA.Ê While we agree with the need to improve impact rate analysis we do not support this fishery in order to have that analysis perfected.Ê

 

The states admit that evaluation of harvest impacts by ESUs is beyond their reach at this time.Ê They therefore recommend that ãAt this time, until a better understanding of ESU-specific impacts can be determined, the states propose to manage the spring fisheries on the aggregate wild winter steelhead run in the Columbia Basin, including all four ESUs.ä The problem with the ãaggregateä approach to harvest evaluation is that it does not work effectively as a conservation strategy.Ê This is illustrated by example when one takes into account the failure of this approach to management of Oregon coastal coho.Ê The over harvest of coho and over-estimation of spawner abundance through extrapolation from high use spawner index areas all combined to deplete wild coho salmon and contributed to their eventual listing under the ESA.Ê It is unreasonable to expect a different result by applying this same failed methodology to the harvest of ESA-listed steelhead.Ê The states are promising to commit the same errors they proved were a problem for coho salmon.Ê The index streams are few and tend to be larger systems with larger runs and they are also systems where data can be collected.Ê What is the risk of designing a harvest management program for these streams and ignoring impacts on the bulk of the streams supporting small runs with no accurate counting capability?Ê It is the same set up that depleted coho on the coast.Ê The only populations that may survive this kind of fishery configuration are those few streams that have been included in the evaluation process.

 

The states say that the Tribal fishery in zone 6 above Bonneville Dam is not a problem even though streams supporting steelhead in two ESUs are involved.Ê It is remarkably cavalier for the states to say ãAlthough information is not available for recent years, the impact of the treaty Indian fishery on wild winter steelhead appears to be minor, and should not affect the analysis of impact rates in the non-Indian fisheries.ä Why are these fisheries not monitored for impacts on wild steelhead and how do the states and NOAA Fisheries justify the lack of monitoring by saying the impact ãappears to be minorä without supporting evidence? Recovery of ESA-listed steelheadrequires evaluation of impacts, including harvest.Ê This approach to evaluation of Tribal fisheries is similar in all respects to the approach taken by the states in the non-Indian fisheries below Bonneville dam prior to the listing of steelhead under the Endangered Species Act.Ê At that time it was assumed by the states that there was no impact.Ê The annual status reports documented zero steelhead taken in the chinook commercial fishery.Ê Then the steelhead were listed and now the states have to provide a data rich justification for expanding risk to steelhead.Ê This approach did not work in the river below Bonneville, so why do the states and NOAA Fisheries believe it will work now above Bonneville?Ê The Native Fish Society recommends that NOAA Fisheries, the states and the tribes develop a harvest monitoring plan to evaluate the impact on all ESA-listed steelhead in each ESU in Zone

6.

 

On page 38, the BA discusses the impact of the spring chinook commercial fishery on early returning summer steelhead.Ê These fish are listed under the Endangered Species Act and come under its protection.Ê The analysis provided does not evaluate the impact that this proposed fishery will have on these early migrants.Ê It assumes that because the summer steelhead numbers are small at the time of the fishery, that the harvest impact will be small.Ê But what is the impact on summer steelhead with early run timing.Ê Since run timing is a heritable trait, the proposed increase in harvest impact could have a large effect on early returning fish in one or more tributaries.Ê There is no discussion about how the early returning summer steelhead migrate through the lower Columbia River.Ê It could be that this migration timing exposes some populations to higher rates of harvest impact than other populations.Ê The Native Fish Society recommends that a more thorough analysis be developed on summer steelhead impacts, including migration timing, disproportionate impacts on some populations, and how likely the expanded kill rate would reconfigure the biological structure of summer steelhead by exposing the early run fish to higher harvest impacts.

 

On page 40, the BA discusses the virtues of mesh size in gill nets and their effect on steelhead catch.Ê It is clear from this analysis that the 9-inch mesh size is more likely to reduce the catch and handling of steelhead.Ê The BA states, ãCatch rates of steelhead were consistently and significantly less in 9-inch nets than in 8-inch nets during the 2000 studiesä (Beamesderfer and Parker 2001).Ê The BA also states that the harvesters prefer the 8-inch mesh nets.Ê Given that the evaluation of mesh size and steelhead catch supports the use of the 9-inch mesh net, the Native Fish Society recommends that the 9-inch mesh nets be required in this fishery whether this BA is adopted or not.

 

The BA also discusses the use of ãexcluder netsä to help steelhead to avoid commercial salmon nets.Ê The BA says, ãPreliminary information suggests that nets with the excluder may handle fewer steelhead than the regular tangle nets; however, a study designed specifically to test the effect of the excluder panel on steelhead and chinook catch rates has not been completed at this time.äÊ The Native Fish Society recommends that the excluder nets be fully evaluated, but based on the information gathered so far, we request that the excluder nets be required in the fishery whether this BA is adopted or not.Ê

 

The states conclude the following: ãThe states are proposing that NOAA Fisheries make a no-jeopardy determination and consider granting approval under the ESA for a wild winter steelhead harvest impact rate up to 6% for 2005.äÊ Based on the information provided in this BA the Native Fish Society does not support the statesâ request.Ê The reason this request is not viable is due to the lack of information and the preliminary nature of the data presented on steelhead migration timing, impact rates by population and ESU, missing but important data such as genetic monitoring of the catch, and the narrow subset of the impact analysis on only a few wild steelhead populations in rivers where accurate dam counts are available.Ê The real impact on wild winter steelhead and those that are listed under the ESA is missing from this analysis.Ê It is our concern, and it is supported by the admitted problems identified in this BA, that the expanded kill of wild winter steelhead is likely to have a large but unknown impact on small populations that are not included in this analysis.Ê In addition, this proposal to increase the kill rate on wild steelhead comes at a time when favorable ocean survival conditions are contributing to the increased abundance of wild steelhead in the lower Columbia River.Ê This BA proposes to increase the kill rate just when these fish are beginning to recover from years of low returns where many populations were below replacement.Ê At some point the states have to recognize that the goal of the ESA is to rebuild wild steelhead rather than use an increasing trend to support increased harvest.Ê

 

The states are even conflicted in their claim when they say: ãThe overall risk to recovery or rebuilding of wild winter steelhead populations as a result of increasing the impact rate from 2% to 6% for 2005 is negligible.ä and then admit that ãIn making this proposal for an increased impact rate up to 6% for 2005, the states recognize that there is some level of uncertainty in the management of commercial fishery impacts on wild winter steelhead.ä

 

NOAA Fisheries is responsible for recovery of protected species under the ESA.Ê The states would not have conducted this analysis if the wild steelhead were not listed.Ê But the BA analysis is not convincing. ESA-listed steelhead will not be adequately protected in a fishery with expanded kill rates. It is for this reason that NOAA Fisheries ought to reject the BA and the proposal to expand the kill of ESA-listed steelhead in 2005 commercial fisheries in the lower Columbia River. The standard that should be foremost in NOAA Fisheries decision is whether this proposed fishery will impede recovery of the ESA-listed steelhead.Ê The Native Fish Society believes it will given the uncertainty acknowledge in this BA by the states

 

On the other hand the states have proposed important studies to be conducted to improve conservation management of ESA-listed steelhead in a mixed stock fishery.Ê It is the recommendation of the Native Fish Society that these studies be funded and their results incorporated into management.Ê In addition, the information that exists on the benefits of a 9-inch mesh net and the use of excluder nets should be a requirement in this fishery. The Native Fish Society is always interested in seeing the findings of research and evaluation applied to management and it is clear that progress has been made and that these improvements in harvest management must be applied.ÊÊ

 

 

 

 

 

ÊÊ

Ê

 

 

 

 

 

 

 

 

 

Ê