C05__Attachment_8_AFWO_NOAA_(2013)_MEMO_to_USBR_about_Fall_Flow.pdf

MEMORANDUM

TO: IN, REC1~AMAT1ON NORTHERN CALIFORNIA AREA MANAGER

FROM: “137ic~’ AND NICHOLAS HETRICK (94FWS) — ~

SUBJECT: 2013 FALL FLOW RELEASE RECOMMENDATION /~4~J/—d. /A9’~’

CC: ROBIN SCHROCK (TRRP) /

DATE: AUGUST 12, 2013

Background

A significant fish kill occurred in the lower Klamath River in September 2002. Though estimates vary, the US Fish and Wildlife Service (Service) reported that a minimum of 34,000 adult fish, primarily fall-run Chinook salmon, died during the event, (Guillen 2003a). Carcasses were observed between September 18 and October 1,2002 within the lower 36 miles of the Klamath River, extending from the estuary upstream to Coon Creek Falls. The Service (Guillen 2003b) reported that:

“Low river discharges apparently did not provide suitable attraction flowsfor migrating adult salmon, resulting in large numbers offish congregating in the warm waters ofthe lower River. The high density offish, low discharges, warm water temperatures, and possible extended residence time ofsalmon created optimal conditions for parasite proflferation andprecipitated an epizootic ofIch and columnaris.”

The Yurok Tribe (Belchik et al. 2004) concluded that:

“the clinical cause ofmortality was massive infections of ich and columnaris.

This fact was confirmed by direct observations, as well as pathology reports by USFWS and CDFG.”

California Department of Fish and Game (Turek et al. 2004) concurred with the findings of the Service’s and Yurok Tribe’s causative factors reports, adding that:

‘flow is the only controllablefactor and tool available in the Klamath Basin... to manage risks against future epizootics and major adultfish kills.”

Several flow-related evaluations and management actions have been implemented in the past to reduce the likelihood of occurrence of an adult fish kill, including the development of criteria for triggering the release of supplemental flows during the fall-run Chinook salmon migration season (Clarke 2010; Hayden 2012; TRRP 2012a) as well as supplemental flow releases from Lewiston Dam on the Trinity River in 2003, 2004, and 2012.

Snowpack and precipitation were below average during the fall/winter 2012-20 13 throughout southern Oregon and northern California, resulting in below average river flows in the region.

• Hydrologic forecasts released by the California Nevada River Forecast Center (CNRFC) predict below average discharge in the lower Klamath River during the 2013 adult fall-run Chinook salmon migration season (Appendix A). Mean monthly discharge for the lower Klamath River is predicted to be 2,168 cfs in August and 2,076 cfs in September, based on inflow predictions and current operation plans that guide managed flow releases from Iron Gate and Lewiston dams.

These predicted mean monthly flows are similar to mean monthly flows experienced in the lower Klamath River during the 2002 fish kill (Table 1; Figure 1; Appendix A). Anticipated flow accretions to the lower Klamath River are about 50% of those observed in 2012. Discharges in the lower Klamath River during the 2013 adult fall-run Chinook salmon migration season are predicted to be equivalent to about 90-95% exceedances (Appendix B).

Escapement of fall-run Chinook salmon to the Klamath Basin in 2013 is projected to be the second largest on record. The Pacific Fishery Management Council’s Salmon Technical Team estimated that 272,400 adult fall-run Chinook salmon will return to the Klamath River (PFMC 2013); which is about 110,000 fish greater than the adult run size associated with the 2002 fish kill (CDFW 2013). This is important as a large run size combined with low river discharge were reported as the primary contributing factors in the 2002 fish kill (Guillen 2003b; Belchik et al.

2004; Turek et al. 2004). Similarly, below average stream discharge has been associated with Ich outbreaks in fish populations in other rivers (Maceda-Veiga et al. 2009).

Given the concerns described above, the Bureau of Reclamation (Reclamation) requested that the US Fish and Wildlife Service and the National Marine Fisheries Service provide technical assistance in assessing the current and predicted hydrologic conditions for the time period overlapping with the 2013 adult fall-run Chinook salmon migration season in the lower Klamath River, and in developing preventative and emergency measures that would reduce the risk of the occurrence of an adult fish kill, while being conservative of limited water resources given the dry hydrologic conditions. This memorandum contains only technical analyses and recommendations regarding adult fall-run Chinook salmon in the Klamath-Trinity Basin. It does not contain any analyses regarding the potential effects of the fall flow releases on any species listed under the Endangered Species Act (ESA) and does not address, nor is it intended to address, compliance with the ESA or any biological opinions issued under the ESA.

Review of 2012 preventative fall flow releases

During spring 2012, Trinity River Restoration Program (TRRP) staff, TRRP partners and Reclamation’s Klamath Basin Area Office jointly developed 1) preventative flow release criteria designed to minimize the risk of a fish disease outbreak and subsequent fish kill (TRRP 2012a), and 2) emergency flow release criteria designed to reduce the severity of a fish kill (TRRP 2012b). The preventative flow release measures identified by the TRRP Fall Flow Subgroup in 2012 were implemented by Reclamation, with supplemental flows originating primarily from Lewiston Dam with a lesser amount of water released from Iron Gate Dam for ceremonial purposes at the request of the Yurok Tribe (Figure 2). Following the recommendations of the Subgroup, BOR targeted a discharge of 3,200 cfs in the lower Klamath River from August 15- September21 (Figure 2). A fish kill did not occur during the 2012 adult salmon migration season in the lower Klamath River, despite dry hydrologic conditions and an unprecedented return of 302,100 fall-run Chinook salmon to the Klamath Basin (CDFW 2013). While it is not known to what extent the preventative flow releases contributed to averting a fish kill, measures taken in 2012 did contribute to reducing water temperatures by up to 1 .4°C in the lower Klamath River (Magneson 2013; Figure 2) and a fish kill did not occur. Similar decreases in water temperatures of about 2.1 °C and 1.6°C were observed in the lower Klamath River during the 2003 and 2004 fall flow releases (Zedonis 2004, 2005).

Table I. Discharge (cfs) in the Klamath River near Klamath gage (U.S. Geological Survey Site #11530500) in August and September 2002 and predicted discharge in 2013.

Year August September 2002 2,327 1,993 2013 (predicted) 2,168 2,076 Long term average 3,170 3,170

A I’

3500 —

• S I’ I “

— I

3000 : 2002 -

~. •I . 2012 projected

— 2012

2500 2013 projected

00 00 03 00 00 00 tO 00 00 tO 00 a a a a a a a a a a D D S S S S S S S S S 03 03 03 03 03 03 0) @3 0) 0) 444444<444 <tflLL L~ U~ tfl U~U~

~ r. a r~tho r~i J~ ~~mto a, r-j Lfl~ — q No

— — . — r4 r.j r-~ en — — ,-l CN (N (N En

Date

Figure 1. Observed flows in the lower Klamath River (RKM 13) in 2002 and 2012 (includes preventative fall flow augmentation) and preseason flow forecasts for 2012 (includes ceremonial pulse event for the Yurok Tribe released from Iron Gate Dam) and 2013 (without preventative or emergency fall flow augmentation or ceremonial release flow recently requested by the Hoopa Valley Tribe). Vertical green lines depict the primary period of the fall-run Chinook salmon migration season in the lower Klamath, August 15 through September 21.

USGS 11530500 KLAMATH R MR KLAMATH CA USGS 11630600 KLAMATH R MR KLAMATH CA

5000 25.0

24.0

~1 74.0

22.0 n.e

22.0

I

IL. 04.0 1

57.0 —

~290 0912 0010 01z25 Sep01 Sep20 SeplS Sep22 Sep29 ~q 11 RIse 10 RIle 25 Sep01 SepOO Sep25 Sep22 sep29 5912 2912 2912 2912 2912 2012 2012 2012

2012 2012 2012 2912 2012 2912 2002 2012 09119 naieIm., t..oraL,st 0.11w nodS., tole.orot’we

Doug o0~ dI.cbwg. PeeIod of .ppro4od dote c.llg ainIao, t.eratas P.rIed of appro,ed dot.

Figure 2. Discharge and water temperature in the Lower Klamath River during August and September of 2012.

Review of past recommendations for preventative fall flow releases

Several minimum flow recommendations for the lower Kiamath River have been reported in the literature for the fall time period, ranging from 2,500 cfs to 3,200 cfs (Table 2).

Recommendations of Turek et al. (2004) and Strange (2010a), however, were made without consideration of recent run sizes that exceeded previous maximum adult returns observed since comprehensive fall-run Chinook salmon monitoring was initiated in 1978.

While not independent of flow, water temperature can also be a critical parameter in affecting adult salmon behavior (Goniea et at. 2006). Strange (20 lOb) identified an adult Chinook salmon migration threshold of 23 C in the Klamath River, which is important because thermal migration barriers can lead to crowding of adult migrant fish and therefore, conditions conducive to fish-to-fish disease transmission and fish kills.

Recommendation for 2013 preventative fall flow releases

Given the large forecasted run size for 2013 and that preventative flow measures were taken in 2012 and a fish kill did not occur, implementing the 2012 fall flow plan (TRRP 20l2a, 2012b) is likely to pose a lower risk for the occurrence of a fish kill in 2013 than the risk associated with other flow recommendations presented in Table 2. It is not possible, however, to assess if a fish kill would have occurred had discharge in the lower Klamath River been lower than 3,200 cfs experienced in fall 2012. While we do know that a fish kill did not occur in 2012, there is considerable uncertainty with regard to the specific discharge that flow augmentation should target in the lower Klamath River to prevent a fish kill.

We acknowledge that Reclamation has multiple obligations to consider in managing water resources in the Klamath-Trinity Basin. In addition, hydrologic conditions in 2013 are drier than those measured in 2012. For example, the 90% forecasted end of September storage for Trinity Reservoir in 2013 is 1.3 million acre feet (MAF), which is about 28% lower than the 1.8 MAF experienced in 2012. In addition, the 2013 projected end of September carryover storage is similar to that observed in 2009, which contributed to water temperature concerns in the Trinity River and resulted in the use of the auxiliary outlet of Trinity Reservoir. Similarly, the hydrologic conditions in the Upper Klamath Basin are also drier in 2013 than 2012.

Table 2. Review of previous for minimum discharge recommendations for the lower Klamath River during the fall-run Chinook salmon migration season.

Projected Adult Fall Author Minimum Flow Recommendation Chinook Salmon Run Size Turek et al. (2004) 2,200 cfs (Klamath near Orleans None specified.

~ +Trinity at Hoopa) -~ 2,500 cfs in Lower Klamath

Strange (2010a) 2,500 cfs in Lower Klamath Less than 170,000 Strange (201 Oa) 2,800 cfs in Lower Klamath Greater than 170,000 TRRP (2012a) 3,200 cfs in Lower Klamath 380,000

Given the large fall-run Chinook salmon run size predicted for 2013 and the dry hydrologic conditions being experienced throughout the Klamath Basin, we recognize the need to provide supplemental flows in the Lower Klamath River to prevent a fish kill using a strategy that minimizes risk while conserving limited water resources. We also recommend that an adaptive management approach be taken that incorporates real-time environmental and biological conditions. In general, ourjoint recommendations are as follows, with more detail following and in the emergency fall flow recommendation section.

• Initiate preventative flow augmentation in the lower Klamath River (RKM 13) to a minimum of 2,800 cfs when the cumulative harvest of Chinook salmon in the Yurok Tribal fishery in the Estuary area meets or exceeds a cumulative total of 7,000 fish (Appendix C). The accounting of harvest should commence starting July 4 and we recommend all Chinook salmon, regardless of race, count toward the cumulative total.

• Fall flow augmentation should be initiated by August22 if the fish metric is not triggered.

• Fall flow augmentation should continue until September21 unless mean daily water temperature at rkm 13 is projected to be >23°C, in which case flow augmentation to maintain a minimum of 2,800 cfs should continue until daily water temperature at rkm 13 is projected to be <23°C or until the end of September when seasonal air temperatures typically cool and contribute to water temperatures suitable for upstream migration (Figure 4).

• Implement real-time flow-temperature management using the RBM1O and SN Temp water temperature models developed for the Klamath and Trinity rivers and NOSA Weather Service weather projections to manage flows in assessing the 23°C water temperaturemigration threshold emergency flow release.

• Implement fish pathology monitoring to determine the need for a fish pathology/mortality emergency release, and

• Monitoring should occur during the fall-run Chinook salmon migration period in the lower Klamath River to inform the need and timing of preventative and emergency flow releases based on real-time environmental conditions (Figure 3; Appendix D).

Preventative Fall Flow Management Action (flow criteria is for the lower Klamath River @RKM13

Cumulative Harvest in the Yurok Estuary NoFlow

N0Tribal Fishery >7,000 Augmentatiofl Fish or August22

See Emergency Fall Flow

YES Manageme t A

No FlowLower Klamath (rkm N0 Emergency Fal Flow Management Actions**

13) flow ‘c 2.800 cfs Augmentation

Based on Water Temperature (@RKM13) and/or Fish Pathology!

Mona fly Criteria

Yes See Emergency Water

Fa I F ow Temperature No Additional Fish Health Management Mean Daily Water NO Flaw NO Fish Pathology/

Increase flow to 2,800 cfs Act Ons Temperature Augmentation Mortality Criteria until September 21. If mean ?23’C for 3 Exceeded daily water tempei’ature consecutive 4~ys alter Septernoor 2110 —-to ciflergenc ‘~a(0r tern rattiit crtlcroii YES

YES

See Emergency Before Alter

Fall Flow September September22 21 Managemert

Actions Double flow in the lower Increase flow to 3,200

If the fish metric or minimum flow criteria are not met the emergency cfs, then decrease bac Maintain flow Klamath River for fall (low criteria and management actions are still in effect, to 2,800 cfs when water at 2,800 cfs 7 days.

temperature decreases until Fish Pathology/Mortality Criteria following the recommendations in below 23’C fort roe September ac

TRRP 2012a: (1) the confirmed diagnosis of severe lch infection of the gills consecutive days in 5% or greater of the sampled adult salnonids or (2) Observed mortality of> 50 dead adult salmonids in a 20 km indet reach in 24 hr coupled with confirmed presence of Ich by USFWS Fish Health Center.

Figure 3. Flow chart depicting proposed 2013 preventative and emergency fall flow release criteria and management actions.

2 203420 2 2035 is —2036

~16 2008

14 2010

4~ C.

E 4’ ~ 16

4~ 4D 4~ 4~ a a a C. C. S S

‘ “ “ 31 31 31 oc i ‘~

— c. r ‘n n. ~o n~ 6 r~ 4 — th— na n

Date

Figure 4. Mean daily water temperature (C) from July through October in the lower Kiamath River (rkm 13), 2003-2012 (upper) and mean for all years (lower).

— 3 40 00 00 o- a a a a — Z~ZW4, 3131319 (~; ~ .. th n r.~ th ~b n in 6 r~ — 6

— — •4 — r r,~ — r.i t4

Date

Fish Metric

We recommend that fall flow augmentation should be initiated once the cumulative harvest of Chinook salmon in the Estuary Area by the Yurok Tribal fishery exceeds 7,000 adults. The tallying of the cumulative harvest of Chinook salmon in the Estuary Area to commence on July 4 and the 2013 Yurok Tribal commercial fishery in the estuary will open on August 10. The use of a fish metric as a real-time indicator of the initiation of upriver migration of fall-run Chinook salmon will entail significant coordination among the Yurok Tribe who collect the fishery data and federal managers that will be implementing the fall flow augmentation.

Initiation ofFall Flow Augmentation ~fFish Metric Is Not Met

Fall flow augmentation should be initiated by August22 if the fish metric is not triggered. The reasoning behind this date is as follows:

• The short time period provided to develop the fish-based metric precluded an in-depth evaluation of temperature and flow data, which may have a significant influence on harvest and effort data in the Estuary and Middle Klamath areas. As a result, the metric may not be conservative enough to ensure that flow augmentation will occur.

• It is anticipated that there will be large numbers of fall Chinook salmon in the Estuary Area due to the large projected inriver run. In addition, the 2013 forecast is expected to be more accurate than the 2012 projection and is comprised of a higher proportion of Klamath fall Chinook salmon stocks (Shasta River, Iron Gate Hatchery, Bogus Creek, and mainstem Klamath River) which tend to enter the estuary and river earlier than Trinity River stocks.

• In four of the five years (80%) included in the break point analyses we conducted to define the fish metric, large numbers of fish were harvested in the Estuary Area by August 22 and harvest in the Middle Klamath Area starts to increase in the following weeks, suggesting that the upriver migration of the run had commenced by this date.

Ending ofFall Flow Augmentation

Fall flow augmentation should be continued through September21 and can end after this date if mean daily water temperature in the lower Klamath River at rkm 13 remains below 23°C (TRRP 2012). See “Recommendation for emergency fall flow releases for 2013 — Water Temperature Criterion” if the mean water temperature in the lower Klamath River exceeds 23°C after September21.

Recommendation for emergency fall flow releases for 2013

The recommended triggers for emergency flow releases in 2013 are two-tiered, both of which are intended to minimize the potential for the occurrence of an epizootic disease outbreak and resulting fish-kill. The first phase recommends that flow in the lower Klamath River be increased to 3,200 cfs at rkm 13 when the fish metric criterion is met or exceeded and mean daily water temperature (actual and/or predicted) at rkm 13 is? 23° C for three consecutive days. The second phase of the emergency release is based on the fish pathology/mortality criteria adopted by the Trinity River Restoration Program - Fall Flow Subgroup’s recommendation for 2012 (TRRP 2012a; TRRP 2012b).

Water Temperature Criterion

Water temperature is widely known as a critical factor for influencing the upstream migration of adult salmonids (Goniea et al. 2006, Strange 2010b), with Strange (2010b) identifying mean daily water temperature threshold of 23°C for the migration of adult Chinook salmon in the Kiamath River.

We recommend the use of a water temperature criterion when the fish metric is met or exceeded and mean daily water temperature (actual and/or predicted) is ? 23° C for three consecutive days to trigger the increase of flow in the lower Klamath River (rkm 13) to 3,200 cfs. While it can be expected that water temperatures will occasionally exceed this temperature threshold, prolonged periods of water temperatures above this threshold can lead to large densities of fish in the lower river as they migrate from the estuary. Therefore, we recommend a three consecutive day period for this water temperature trigger to avoid reacting to short (one or two day) temperature increases above the water temperature criterion. Maintaining mean water temperature below this temperature threshold for the following three days will allow adult to migrate upstream and reduce fish density in the lower river. This emergency action is intended to eliminate thermal migration barriers and reinitiate upstream migration of adult fish, thereby reducing the extended residence time of adult fish in thermal refugia and as a result, conditions conducive to fish-to-fish disease transmission and associated fish kills.

We also expect that adult Chinook salmon will resume upstream migration on the onset of periods of declining river temperatures that would result from increasing flows from 2,800 cfs to 3,200 cfs in the lower Klamath River. This real-time management concept is supported by the findings of Strange (2010b) who reported that adult Chinook salmon key into periods of declining river temperature during their upriver migration to take advantage of brief windows of thermal opportunity.

Fall flow augmentation should continue until September21 unless mean daily water temperature at rkm 13 is projected to be >23°C, in which case flow augmentation to maintain a minimum of 2,800 cfs should continue until daily water temperature at rkm 13 is projected to be <23°C or until the end of September. In early October, mean water temperatures in the lower Klamath River are generally decreasing (Figure 4) due to seasonal decreases in air temperatures and most of the fall-run Chinook salmon have commenced their upstream migration based on harvest data from the Middle Klamath Area (Appendix C, Figure 12)

Fish Pathology/Mortality Criterion

Two primary fish health monitoring efforts will be relied upon to determine the need for a diagnostic Ich survey which could trigger an emergency fall flow release.

1. Adult fish health monitoring will be conducted by the Yurok Tribal Fisheries Program (YTFP) in the lower reach of the Kiamath River to determine the presence and severity of Ich and columnaris infection throughout the fall Chinook salmon run. Additionally, Tribal. Fisheries crews will count and examine all pre-spawn mortalities to determine possible cause of death. Pre-spawn mortality due to columnaris, wounds (hook, net, or seal bites), and other causes will be documented but will not be used as diagnostic criteria for identifying an imminent Ich epizootic. Results of these sampling efforts will be used to determine if a more intensive diagnostic Ich survey is needed.

2. In addition to the directed fish health monitoring conducted by the YTFP, the Klamath Fish Health Assessment Team (KFHAT) will implement its response plan if moribund or dead fish are observed in any areas of the Klamath or Trinity rivers. KFHAT response plan documents can be found at the following link:

http:/fvovw.kbmpnetiimages/storiesfpdf/KFHAT/FinalResPlan_AppendicesUpdatedMarch2Ol 1 .pcff

These efforts will provide information on the disease incidence observed in adult salmonids in -the lower Klamath River or the numbers/condition of dead or moribund fish throughout the Klamath-Trinity Basin. We recommend that the criteria used to institute a diagnostic Ich survey should be:

I. Prevalence of severe Ich infection in 5% or greater of the weekly adult fish health monitoring samples collected by resource agencies, with Ich infection to be confirmed by the Service’s California-Nevada Fish Health Center (CNFHC) from fixed samples, or

2. Observed mortality of> 50 adult salmonids (Chinook and/or coho salmon and steelhead), regardless of cause, in a 20-km reach within a 24-h time period. Recently deceased fish (<24 hours post death) will be differentiated from older mortalities (>24 hours post death) by the presence of at least one clear eye. Data on the presence of hook scars, gill net marks, predator wounds (seal/sea lion, lamprey, otter), and condition of gills (i.e.

columnaris) will also be collected to determine other possible causes of mortality.

Diagnostic Ich Survey

If either of the criteria established for the adult fish health monitoring effort are met, an intensive sampling of adult salmonids will be initiated to collect live or recently deceased fish. Ich diagnostic surveys will be performed by the CNFHC that will provide a pathology report documenting the findings of these surveys. These efforts will focus on determining the level and severity of Ich infection or the possible cause of death in the event of large numbers of dead fish are observed. We recommend the level and severity of an Ich infection that would trigger an emergency release be defined as a confirmed observation of a minimum of 5% of the sampled fish having 30 or more parasites on one gill arch. The recommended action is to augment flows in the lower Klamath River to double the preexisting flow for 7 consecutive days.

If possible, a minimum of 60 adult salmonids should be sampled within a consecutive 2-day period. While a sample of 60 fish is desired, it may not always be achievable and the minimum acceptable sample size is set at 30 fish. These fish should be live or recent mortalities (<3 hours).

Criteria for triggering an emergency flow release and recommended management action based on the level and severity of an Ich infection are as follows.

The Service’s CNFHC will provide a pathology report documenting the findings of the diagnostic survey to Brian Person of Reclamation, to other federal, state and tribal co-managers, and to the KFHAT group. It is recommended that an emergency release be implemented immediately upon BOR’s receipt of a positive pathology report to limit fish mortalities associated with a potential Ich epizootic.

Additional Considerations

While modest increases of river discharge in the Klamath Basin from summer rainstorms are not uncommon during the adult fall-run Chinook salmon migration period, these rain events are typically short in duration and occur with limited frequency. The sustained release of a substantial volume of water from one or both Klamath Basin dams, as recommended above, would mark a departure from the natural flow regime (Poffet al.

1997, Lytle and Poff 2004) of the Klamath and Trinity rivers because the duration of the elevated flows is unnatural. Modification of flow can be expected to have cascading effects on the ecological integrity of rivers and the organisms that depend on them (Poff et

al. 1997). However, both the Klamath and Trinity rivers are extensively managed systems and given existing water withdrawals and conveyances, their hydrology already deviate significantly from the natural conditions.

Fish Pathology/Mortality Emergency Criteria Management action

1. The confirmed diagnosis of severe Ich infection of the gills in 5% or greater of a desired sample of 60 adult salmonids (3 infected out of a 60-fish sample). Following the 5% threshold criteria, a confirmed diagnosis of 2 or more individuals having a severe Ich infection would meet the criteria for a sample size of less than 60 but greater than the minimum of 30 fish.

Recommend immediate Emergency Fall Flow release with a 7 day duration pulsed spike to double pre-existing flows in the Lower Klamath River.

Or

2. Observed mortality of> 50 dead adult salmonids in a 20 km index reach in 24 hr coupled with confirmed presence of Ich by USFWS Fish Health Center.

1~

References

Belchik, M., D. Hillemeier, and R.M. Pierce, R.M. 2004. The Klamath River fish kill of 2002;

Analysis of Contributing Factors; Yurok Tribal Fisheries Program. 42pp.

California Department of Fish and Wildlife (CDFW). 2013. Klamath River Basin fall Chinook spawner escapement, in-river harvest and run-size estimates, 1978-2012. l3pp.

Clarke, E. 2010. Memorandum Trinity Management Council (TMC), Jennifer Faler, Interim Executive Director, Trinity River Restoration Program (TRRP). RE: Klamath River special flow releases. Available from the Trinity River Restoration Program:

http://odp.trrp.netfDatafDocuments/Details.aspx?document=472

Foott, J.S. 2002. Pathology report. FHC Case No. 2002-139. USFWS. Anderson, California.

Goniea, T. M., M. L. Keefer, T. C. Bjornn, C. A. Peery, D. H. Bennett, and L. C. Stuehrenberg.

2006. Behavioral thermoregulation and slowed migration by adult fall Chinook salmon in response to high Columbia River water temperatures. Transactions of the American Fisheries Society 135:408—419.

Guillen, G. 2003a. Klamath River fish die-off, September 2002: report on estimate of mortality.

US Fish and Wildlife Service. Report Number AFWOF-01-03. 28pp.

Guillen, G. 2003b. Klamath River Fish Die-off, September 2002: Causative factors of mortality.

US Fish and Wildlife Service. Report Number AFWOF-02-03. l28pp.

Klamath River Technical Advisory Team (KRTAT). 1997. Inseason adjustment of inriver fishery harvest controls based on catch per unit effort. Report to the Klamath Fishery Management Council. 34 pp.

Hayden, T. 2012. Memorandum to the fall flows subgroup. RE: 2010 and 2011 Fall flow release criteria and evaluation process. Available from the Trinity River Restoration Program:

www.trrp.net.

Lytle, D. A. and N. L. Poff. 2004. Adaptation to natural flow regimes. Trends in Ecology and Evolution 19(2): 94-100.

Maeeda-Veiga A., H. Salvado, D. Vinyoles & A. De Sostoa. 2009. Outbreaks of Ichthyophthirius multifiliis in redtail barbs Barbus haasi in a Mediterranean stream during drought. Journal of Aquatic Animal Health 21(3): 189-194.

Magneson, M.D. 2013. The Influence of Lewiston Dam Releases on Water Temperatures of the Trinity River and Lower Klamath River, CA, April to October 2012. U. S. Fish and Wildlife Service, Arcata Fish and Wildlife Office, Arcata Fisheries Data Series Report Number DS 2013-30, Arcata, California.

Pacific Fishery Management Council (PFMC). 2013. Preseason Report III. Council adopted management measures and environmental assessment part 3 for 2013 ocean salmon fishery regulations. 5 ‘pp.

Poff, N. L., J. D. Allan, M. B. Bain, J. R. Karr, K. L. Prestegaard, B. D. Richter, R. E. Sparks, J.

C. Stromberg. 1997. The natural flow regime; a paradigm for river conservation and restoration. BioScience 47: 769-784.

Strange, J.S., 2010a. Summary of scientific evidence to guide special flow releases to reduce the risk of adult fall Chinook salmon mass disease mortality in the lower Klamath River Available from the Trinity River Restoration Program: www.trrp.net.

Strange, J.S. 2010b. Upper thermal limits to migration in adult Chinook salmon: evidence from the Kiamath River Basin. Transactions of the American Fisheries Society 139:1091—1108.

Trinity River Restoration Program (TRRP). 2012a. Memorandum to Brian Person, Reclamation Northern California Area Manager. 2012 fall flow release recommendation. May 31, 2012.

Trinity River Restoration Program (TRRP). 2012b. Memorandum to Brian Person, Reclamation Northern California Area Manager. 2012 fall flow release recommendation addendum.

August 16, 2012.

Turek, S., M. Rode, B. Cox, G. Heise, W. Sinnen, C. Reese, S. Borok, M. Hampton, and C.

Chun. 2004. September 2002 Klamath River fish-kill: final analysis of contributing factors and impacts. California Department of Fish and Game. l83pp.

Zedonis, P. 2004. Lewiston Dam releases and their influence on water temperatures of the Trinity and Klamath Rivers, CA; April to October, 2003. Report AFWO-F01-04. U.S. Fish and Wildlife Service, Arcata Fish and Wildlife Office, Arcata, CA 95521. 34 pp.

Zedonis, P.2005. The Influence of Lewiston Dam Releases on water temperatures of the Trinity and Klamath Rivers, CA., April to October, 2004. U. S. Fish and Wildlife Service, Arcata Fish and Wildlife Office, Arcata Fisheries Technical Report Number TR2005-03, Arcata, California. 3lpp.

Appendices

Appendix A. Predicted discharge for the lower Kiamath River (U.S. Geological Survey Site #11530500) with no preventative flow release. (Data from CNFRC downloaded on June 30, 2013).

Date 50% flow IGD Flows Lewiston forecast river Flows flow without dam releases (from CNRFC)

7/30/2013 7/31/2013 8/1/2013 8/2/20 13 8/3/20 13 8/4/20 13 8/5/2013 8/6/2013 8/7/2013 8/8/2013 8/9/2013

8/10/20 13 8/11/2013 8/12/20 13 8/13/20 13 8/14/20 13 8/15/20 13 8/16/20 13 8/17/20 13 8/18/20 13 8/19/2013 8/20/2013 8/21/2013 8/22/20 13 8/23/20 13 8/24/2013 8/25/20 13 8/26/20 13 8/27/20 13 8/28/20 13 8/29/2013 8/3 0/20 13 8/31/2013

Flow estimate at KNK with no supplemental flows 2,320 2,302 2,287 2,282 2,261 2,243 2,226 2,212 2,201 2,189 2,180 2,164 2,151 2,136 2,126 2,118 2,110 2,096 2,084 2,193 2,184 2,178 2,174 2,166 2,163 2,152 2,144 2,141 2,136 2,132 2,126 2,121 2,121

1,000 1,000 1,000 1,000 1,000 1,000 1,000 1,000 1,000 1,000 1,000 1,000 1,000 1,000 1,000 1,000

Date 50% flow IGD Flows Lewiston Flow forecast river Flows flow without dam releases (from CNRFC) estimate at KNK with no supplemental . flows

9/1/2013 667 1,000 450 2,117 9/2/2013 660 1,000 450 2,110 9/3/2013 651 1,000 450 2,101 9/4/2013 644 1,000 450 2,094 9/5/2013 639 1,000 450 2,089 9/6/2013 634 1,000 450 2,084 9/7/2013 630 1,000 450 2,080 9/8/2013 628 1,000 450 2,078 9/9/2013 626 1,000 450 2,076

9/10/2013 627 1,000 450 2,077 9/11/2013 623 1,000 450 2,073 9/12/2013 621 1,000 450 2,071 9/13/2013 619 1,000 450 2,069 9/14/2013 619 1,000 450 2,069 9/15/2013 622 1,000 450 2,072 9/16/2013 628 1,000 450 2,078 9/17/2013 626 1,000 450 2,076 9/18/2013 624 1,000 450 2,074 9/19/2013 623 1,000 450 2,073 9/20/2013 624 1,000 450 2,074 9/21/2013 621 1,000 450 2,071 9/22/2013 616 1,000 450 2,066 9/23/2013 618 1,000 450 2,068 9/24/2013 614 1,000 450 2,064 9/25/2013 611 1,000 450 2,061 9/26/2013 611 1,000 450 2,061 9/27/2013 613 1,000 450 2,063 9/28/2013 615 1,000 450 2,065 9/29/2013 616 1,000 450 2,066 9/30/2013 611 1,000 450 2,061 10/1/2013 608 1,000 450 2,058 10/2/2013 614 1,000 450 2,064 10/3/2013 620 1,000 450 2,070 10/4/2013 609 1,000 450 2,059 10/5/2013 618 1,000 450 2,068

Appendix B. Exceedance table based on monthly average flows, not daily flows, for the Lower Kiamath River (U.S. Geological Survey Site #11530500) using years 1911-2012.

Exceedance July August September

0.05 9,646 4,869 4,875

0.10 8,894 4,397 4,247

0.15 7,770 4,326 4,113

0.20 7,352 4,131 3,943

0.25 6,650 3,683 3,773

0.30 6,397 3,532 3,605

0.35 5,455 3,447 3,415

0.40 5,177 3,279 3,346

0.45 4,793 3,170 3,139

0.50 4,477 2,982 3,032

0.55 4,265 2,956 2,968

0.60 4,083 2,901 2,857

0.65 3,924 2,861 2,758

0.70 3,789 2,787 2,691

0.75 3,574 2,672 2,598

0.80 3,313 2,574 2,538

0.85 3,230 2,372 2,501

0.90 2,960 2,200 2,447

0.95 2,518 1,876 2,003

Appendix C. Development of a Fish Metric to Inform the Timing of Fall Flow Augmentation for the Lower Klamath River in 2013.

Introduction

In 2012 the Trinity River Restoration Program Fall-Flow subgroup developed flow recommendations to protect the forecasted large inriver run of fall-run Chinook salmon expected to enter the Klamath/Trinity Basin and prevent a fish-kill (TRRP 2012). Recommendations made by the group included a temporal component (August 15-September 21) based on Yurok Tribal net harvest data collected in the Klamath River estuary, which was used as a proxy for inriver run timing.. With a large fall-run Chinook salmon inriver run forecast for 2013 (PFMC 2013a, Figure 1) and the expected low flows in the lower Klamath River in August and September, the Bureau of Reclamation sought technical assistance from the US Fish and Wildlife Service and National Marine Fisheries Service to develop recommendations for augmenting fall flows, while being conservative of the limited water resources given the dry hydrologic conditions.

One component of the recommendation developed by the Fall-Flow subgroup in 2012 that needed refinement was to better define and support the period when flow augmentation would be implemented. A fish abundance-based metric and associated real-time monitoring was deemed desirable to inform the timing of flow augmentation rather than relying on fixed dates as specified in the 2012 plan.

The projected 2013 inriver run of adult fall-run Chinook salmon is 282,400 fish, approximately 110,000 greater than the 2002 inriver run (Figure 1), of which the majority are predicted to be age-4 fish (PFMC 2013a). The projected age composition of the run is pertinent in that the age-4 predictions are generally more accurate than the age-3 predictions (PFMC 2013b). The 2012 inriver run was 79% of the preseason projection with this error partially attributed to the relatively low precision and accuracy in the preseason forecast for the age-3 component of the run at large stock sizes which comprised 82% of the 2012 run (KRTT 2012). In addition to the age composition of the 2013 run being skewed towards age-4 fish (69%; O’Farrell, pers. com.), the run is expected to be dominated by Klamath stocks (Iron Gate Hatchery, Klamath River mainstem, Bogus Creek, Shasta River, Scoff River and Salmon River) based on the distribution of spawners observed in 2012 (CDFW 2013, Figure 2) and these stocks tend to enter the river earlier than Trinity stocks (Polos and Craig 1994, Strange 2007). Therefore, we expect the number of fall-run Chinook salmon that return to the Klamath Basin will be closer to the forecast as compared to 2012 and that many of the fish will return earlier than average.

The goal of this analysis was to develop a fish metric that can be used as an indicator of the first substantial increase of fall-run Chinook salmon in the lower Klamath River, indicating that the inriver run and subsequent upstream migration has commenced. This fish metric is intended to be used as a trigger to initiate fall flow augmentation. The benefit of this real-time management approach is its potential to more efficiently use limited water resources as needed to protect the large predicted return of Klamath Basin fall-run Chinook salmon, rather than relying on fixed dates to start and end flow augmentation.. However, it is critical that an abundance-based metric be conservative so that augmented flows are released in time to protect the run, with the need amplified by the low flows projected for August and September (similar to those in 2002). A metric that is not conservative enough may result in large numbers of adult fall-run Chinook salmon entering the river and commencing their upstream migration under flow conditions that are similar to those that occurred during the 2002 fish-kill.

An ideal metric for guiding the management decision of when additional water should be released would be based on the density of fish holding above the estuary in the mainstem Klamath River in the reach where the 2002 fish-kill occurred. This is an area where adult and juvenile salmonids often congregate in high densities in thermal refugia when warm mainstem Klamath River water temperatures inhibit migration. Another potential metric would be the abundance of Ich theronts in this reach of the mainstem Klamath River. This information, in combination with the fish density data, could be used to determine the potential for an Ich epizootic. While the development of this fish metric has focused on the abundance of adult fall-run Chinook salmon, the abundance ofjuvenile salmonids as well as other fish species that may be holding in thermal refugia should be considered because they can also be infected by Ich and can possibly be the source of the initiation of an Ich epizootic. At this time, however, fish and Ich theront density information are not available. As an alternative we chose to use the harvest data from the Yurok Tribal fishery as a proxy for fish density in the mower mainstem river because the historic information were readily available and implementation and tracking of the metric in real-time is feasible.

Methods.

We evaluated the Yurok Tribal net harvest and effort data from the Estuary and Middle Klamath monitoring areas provided by the Yurok Tribal Fisheries Program (Figure 3). These data were selected for exploratory analyses because they: (1) are assumed to provide an indirect mea~ure of fish abundance and run timing, (2) were collected in the area (Middle Klamath) where the 2002 fish-kill occurred, (3) were quickly accessible given the limited time provided to us to develop a fish metric, and (4) the data are updated every 24 hours during the fishery to facilitate a real-time use of the metric to inform flow augmentation decisions. Utilizing harvest data for this task requires two assumptions:

• The number of fall-run Chinook salmon that have escaped estuary harvest is positively associated with the number of estuary harvested fish.

• Fish that escape estuary harvest will soon-after arrive at, and potentially hold in, the section of the Klamath River considered most susceptible to fostering a disease outbreak in returning adult salmon.

The data consisted of weekly estimates of Chinook salmon harvest and fishing effort from July 4 through November30 for years 2001 through 2012 (Williams, pers. com.). The Estuary Area is an area of intense fishing effort and harvest and rigorous monitoring, especially during years when a commercial fishery is conducted, and can indicate when large numbers of fall—run Chinook salmon have migrated into the estuary. The Middle Klamath Area is the area where the 2002 fish-kill occurred (Guillen 2003), where fall-run Chinook salmon initiate their upstream migration and where they are susceptible to temperature induced migration delays (Strange 2010), potentially leading to high fish densities. These conditions, in combination, can contribute to the initiation of an Ich epizootic (Guillen 2003, Turek et al. 2004, Belchik et al.

2004).

Following discussions with the Yurok Tribal biologists concerning the harvest data, it was decided that the data from five years (2001, 2002, 2003, 2007 and 2009) were most appropriate for analyses for the following reasons:

• the Kiamath Basin fall-run Chinook salmon inriver runs during these years were large (Figure 1),

• commercial fisheries occurred in the Estuary Area during these years, and

• the commercial fisheries began between July 29 and August 1.

Data from 2011 and 2012 were not considered because thecommercial fisheries started on August21 and August 19, respectively, which significantly shifts the timing of effort and harvest in the Estuary Area (Figure 4). While weekly data were provided for each year from July 4 through December 4, our analyses focused on the time period from July 4 through November 6 as it encompasses the initiation and the end of the fall-run Chinook salmon migration through the Estuary and Middle Klamath areas. In evaluating catch-effort data, the period was limited from August 1 through October 2 when significant fishery effort and harvest occur and comparable effort data (net-hours) were available. Graphic display of the data used the last day of the week rather than the first day of the week so cumulative data were representative of the sum of weekly data up to that date. While the data are not continuous, line graphs were used for display purposes to facilitate comparative display of the data.

Weekly harvest, effort and catch-effort (CE) data for the Estuary Area and Middle Klamath Area were plotted to evaluate any obvious patterns that could be further evaluated as fish abundance metrics. Weekly values of harvest, effort or CE data were graphed as well as cumulative values of harvest and effort throughout the period. Additionally, proportions and cumulative proportions of each year’s harvest and effort were plotted.

Differences in the timing of harvest between the Estuary Area and the Middle Klamath Area were examined to determine potential patterns that could be used to infer run timing into the Middle Klamath Area by the harvest in the Estuary Area. This would allow for flow augmentation to be linked to the fall-run Chinook salmon abundance in both the Estuary and Middle Klamath areas. Yurok Tribal biologists expressed concerns with using the Middle Klamath data for a fish metric because the fishing effort is typically lower during the early part of the run, especially when a commercial fishery is occurring, so the data may not be adequate as an indicator for fish abundance.

The relationship between effort and CE were investigated to see if these variables could be used as indicators of abundance. Ideally, CE data could be used as an indicator of the abundance of fish in the Estuary Area; but due to the intensity of the fishery and the variable removal (via harvest and upstream migration) and addition (via fish moving into the Estuary from the ocean) of fish this in not likely the case. It was speculated that increases in effort could indicate when the large numbers of fish were in the estuary due to fishers reacting the presence of fish or that effort could influence CE, possibly decreasing it with the increase in effort. Harvest was not evaluated for these relationships because of the lack of independence between harvest and effort and harvest and CE since these variables are used to calculate harvest (Equation 1):

Harvests = Z1=1(Effort1 * CE1) Equation 1

• .where harvest in week t is estimated by summing the daily harvest estimates generated by multiplying the effort on day i by the catch-effort (CE1) on day i.

The object of the fish metric is to initiate fall flow augmentation so increased flows in the lower Klamath River coincide with large numbers of fish exiting the Estuary and migrating into the lower Klamath River. Following the graphical display of data, it appeared that a fish abundance metric could be developed by looking at the large and abrupt increases in harvest or inflection points of cumulative harvest that occur in the Estuary Area fishery.

The cumulative harvest estimates from all years display a very similar pattern: (1) a period of relatively little harvest before the population of returning salmon arrive en masse, (2) a sharp increase in the amount of harvest that continues for several weeks, and (3) a plateau in harvest during the latter part of the harvest season. Being able to estimate when the cumulative harvest curves begin their quick acceleration would allow us to also estimate when the bulk of the population of returning adult salmon were about to be entering the river. Given the common shape-characteristics of the cumulative harvest curves under consideration, we opted to estimate the beginning of the accelerated arrivals of adults using break-point analysis, applying the model and estimation techniques of Muggeo (2003). Rather than considering a year’s cumulative harvest as a single curve, we instead considered each as a set of continuous piece-wise linear segments. Each segment potentially has a unique slope, and changes in slopes occur at break points. For example, consider the segmented relationship between a response variable (Y) and a single, continuous, explanatory variable (X) with a single break-point (w). A model for the mean of Y is:

E(Y)= fl+fl~X+f?~(X-i~)+ where the “+“ is a logical expression indicating a 1 if X — i~ >0, and and 0 if not, According to this parameterization, the slope of the relationships between Y and X is f~i if X ≤ 4’ and (/31 + /32) if X > i4’. While this model can describe the relationship between the response and explanatory variable, the likelihood is not differentiable at the break points. To combat this issue, likelihood estimation is carried out under an iterative procedure based on a first-order Taylor’s expansion (Muggeo, 2003).

Results

Estuary Area — Harvest, Effort and Catch-Effort

Harvest, effort and CE data in the Estuary Area show some general trends among the years evaluated but also substantial variability (Figure 5). Harvest data exhibit some distinct peaks in mid-August and early September, possibly coinciding with the peaks of Klamath origin and Trinity origin fish entry into the estuary. The 2007 data are unique in that the run appeared to enter the river later than in the other years. Fishing effort typically has one peak but it occurs over a five week period from early August through early September. The large CE values that occur in July and October can be attributed to low effort inflating the CE estimates. CE was variable with no consistent trend when the analysis t was limited to August and September, with peaks occurring in late August to mid-September (Figure 6). Data on the proportion of harvest and proportion of effort also showed the similar trends in the timing of peak harvest and effort and the variability throughout the season (Figure 7).

Cumulative and cumulative proportion of harvest and effort showed the same general trends in peaks and timing of harvest and effort data although the relative trends from week to week can be distinguished by changes in the slope of the line segments (Figure 8 and 9). The cumulative proportion of harvest in the Estuary Area show that three years (2001,2003, and 2009) exhibited similar trends in cumulative harvest through late August while the cumulative harvest line is shifted earlier for 2002 and later for 2007. The later run timing observed in 2007 may have been due to the run being composed of primarily Trinity origin fish (61% based on the distribution spawners) which tend to enter the river later and also the development of a berm at the mouth of the Klamath River which is believed to hinder the migration of salmonids into the estuary (Hilliemier pers. corn). The later run timing of fall Chinook salmon was also observed by Strange (2008) and a protracted spawning duration in the upper mainstem Klamath River (Gough, pers. com.).

Middle Klamath Area — Harvest. Effort and Catch-Effort

Harvest and CE in the Middle Klamath Area were highly variable in magnitude and timing of peaks (Figure 10). Effort was relatively stable throughout the period evaluated except for July and the large peaks in August and September in 2009. The proportion of harvest and proportion of effort data exhibited high variability throughout the season (Figure 11).

Cumulative harvest indicates that the pattern of harvest was similar up to mid-September in three years (2001, 2002 and 2003) but increased earlier in 2009 and later in 2007 (Figure 12).

Cumulative effort showed similar trends except in 2002 when effort was substantially greater than in other years. Cumulative proportion of total harvest shows that increase in harvest in the Middle Klamath Area was variable, occurring from mid-August to mid-September (Figure 13).

Timing of Harvest in the Estuary and Middle Klamath Areas

The Estuary and Middle Klamath areas show similar trends in cumulative proportion of harvest within years, with harvest occurring in the Estuary Area earlier than in the Middle Klamath Area as is expected…