2015 South Delta Chinook Salmon Survival Study

By Rebecca Buchanan, University of Washington

Denise Barnard, Pat Brandes, Kristen Towne, Jack Ingram, Ken Nichols

U.S. Fish and Wildlife Service

Josh Israel, U.S. Bureau of Reclamation

Collated and Edited by Pat Brandes, U.S. Fish and Wildlife Service

16 April 2018

1

Table of Contents Introduction ...... 5 Project Objectives ...... 6 Study Design and Methods ...... 6 Sample size analyses ...... 6 Study Fish ...... 7 Tags (transmitters) ...... 8 Activation ...... 8 Surgeon Training ...... 8 Tagging ...... 9 Tag Validation ...... 9 Transport to Release Sites ...... 10 Transfer to Holding Containers ...... 11 Fish Releases ...... 12 Dummy-Tagged Fish...... 13 Fish Health Assessment ...... 14 Tag Life Test ...... 14 Tag Retention ...... 15 Mobile Monitoring ...... 15 Statistical Methods ...... 16 Data Processing for Survival Analysis ...... 16 Distinguishing between Detections of Chinook Salmon and Predators ...... 18 Constructing Detection Histories ...... 21 Survival Model ...... 22 Durham Ferry Survival Model ...... 23 Survival Model ...... 27 Parameter Estimation ...... 29 Analysis of Tag Failure ...... 32 Analysis of Surgeon Effects ...... 34 Analysis of Travel Time ...... 34 Route Selection Analysis ...... 35 Survival through Facilities ...... 35

2

Results ...... 35 Transport to Release Sites ...... 35 Fish Releases ...... 36 Dummy Tagged Fish ...... 36 Fish Health Assessment ...... 37 Tag Retention Study ...... 37 Detections of Acoustic-Tagged Fish ...... 38 Durham Ferry Releases ...... 38 Medford Island Release ...... 40 Tag-Survival Model and Tag-Life Adjustments ...... 42 Surgeon Effects ...... 43 Survival and Route Entrainment Probabilities ...... 43 Release Group 1 ...... 43 Release Group 2 ...... 45 Pooled Release Groups ...... 46 Comparison between Release Groups ...... 46 Comparison between years ...... 46 Travel Time ...... 47 Mobile Monitoring ...... 49 Discussion...... 49 Survival in 2015 ...... 49 Mobile Monitoring ...... 51 Summary ...... 52 Project Objectives ...... 52 References ...... 56 Acknowledgements ...... 58 Figures ...... 59 Tables ...... 76 Appendices ...... 133 Appendix A...... 133 Appendix A1 ...... 161 Appendix B ...... 178

3

Appendix C...... 183 Appendix D...... 190 Appendix E...... 198

4

Introduction The Chinook Salmon Survival Study in 2015 was conducted to estimate juvenile Chinook salmon survival through the and Delta (and routes contained within) and relate it to temperature, flow, exports with the physical barrier at the head of . These data were to be combined with that gathered in the past, to better understand factors influencing survival, in and through the Delta, and route entrainment. The spring of 2015 was very dry and the water year type was classified as critical. It was the fourth year of a drought. Two estimates of survival were made: one in the second week of April and one during the last week of April/first week of May, in 2015. Both releases were to be made at Durham Ferry. However, water temperatures at the Durham Ferry release site in the fourth week of April were nearing lethal levels for Chinook salmon, necessitating the movement of the release site to Medford Island, where water temperatures were lower, for the release of the remaining fish from the fourth week of April (second week of releases). Unfortunately, there were no provisions for holding study fish and dummy-tagged fish at the Medford Island release site, so they were held at the hatchery for 24 hours after tagging, prior to being transported to the Medford Island site for release. The release at Medford Island allowed additional information to be generated on survival in the lower reaches of the Delta to be gathered – which has been lacking in previous years, due to the poor survival to downstream reaches.

A total of 1,287 juvenile Chinook salmon tagged with VEMCO V4 acoustic tags were released into the San Joaquin River in mid-April through early May of 2015: 645 at Durham Ferry on April 15–19, 151 at Durham Ferry on April 29-30, and 491 at Medford Island on April 30–May 2. Acoustic tags were detectable on hydrophones located at 43 stations throughout the lower San Joaquin River and Delta to (i.e., Mallard Slough) and Benicia Bridge. Detection data were also available from 150 acoustic tags implanted into several species of predatory fish released in the Delta in April – May 2014, and another 150 tags implanted into predator fish released in the Delta in April 2015: 78 striped bass, 128 largemouth bass, 34 channel catfish, and 60 white catfish. A rock barrier was in place at the head of Old River during the Chinook salmon study in 2015. An additional rock barrier was installed at False River blocking direct access from the San Joaquin River to False River in 2015; this emergency-drought- rock barrier installed on False River changed the tidal flows in and out of Frank’s Tract. However, because the False River barrier was not near completion until late May, which was after the final tag detection in the lower part of the study area, it was not expected to affect the outcome of the survival study.

5

Project Objectives There were five specific objectives of the project:

a) Determine survival of emigrating salmon smolts from Durham Ferry and Mossdale through the Delta to Jersey Point and Chipps Island. b) Determine the survival for tagged juvenile salmon taking different pathways to Chipps Island (e.g. Old River route versus the San Joaquin River route). c) Identify reach-specific mortality of tagged fish. d) Identify the proportions of fish entering Turner Cut and at the head of Old River and factors influencing them. e) In conjunction with past results, assess the role and influence of flow and exports, a barrier at the head of Old River, mortality by reach, and the proportion entering Turner Cut, on survival through the Delta.

With the change in release site from Durham Ferry to Medford Island for most of the second week of releases, the first objective was modified to add estimating survival from Medford Island to Jersey Point and Chipps Island. We also added an objective of measuring survival from Jersey Point to Chipps Island, with good confidence. In the past, we have had too few detections at Jersey Point to either make estimates or feel confident in the survival estimates from Jersey Point to Chipps Island when they could be made (SJRGA 2013). The release at Medford Island also allowed three hypotheses about survival in the lower reaches of the Delta to be evaluated. a) Survival to Chipps Island is low for fish entering Turner Cut or Columbia Cuts. b) Tagged fish entering Turner Cut or Columbia Cuts do not reach the Fish Facilities or make it back to the mainstem San Joaquin River. c) Tagged fish enter Threemile Slough, and may move into the and thus would not be detected at Jersey Point, but could be detected at Chipps Island if they survive to that point.

Study Design and Methods Sample size analyses Analyses were conducted to determine the needed sample sizes to estimate survival from Mossdale to Chipps Island, given various assumptions about survival and detection probabilities

6

(Appendix A. Appendix A1). Four conclusions resulted in an initial study design of releasing two groups of 648 fish at Durham Ferry for the study.

• With a single release of 648 fish at Durham Ferry, we can expect to estimate key regional survival parameters under all scenarios except those with low or very low survival downstream of Mossdale. • High detection probabilities at Chipps Island facilitate estimation of model parameters but do not fully compensate for very low survival probabilities at a release size of 648. • With a single release of 1,296 fish (pooled across 2 groups of 648) at Durham Ferry, parameters were estimable in most cases unless survival was very low. • Reasonable parameter estimates may be attainable under conditions of very low survival if detection probabilities at Chipps Island are very high (e.g., 0.99) and the two release groups of 648 are pooled.

As mentioned above conditions in the field during the second week of releases (high water temperatures) resulted in a change from this design; only part of the second release was made at Durham Ferry while the balance of the second release was made downstream at Medford Island.

Study Fish Fall-run Chinook salmon were originally scheduled to be tagged at the Merced River Hatchery in 2015, but due to a low number of returning fish and lack of juvenile fish for tagging studies our request for fish from Merced River Hatchery was denied and fish from Hatchery (MKRH) were substituted instead. We utilized the same tagging trailer on the hatchery grounds in which we tagged Steelhead for the Six-Year Survival Study.

For each of the two tagging weeks, approximately 800 salmon were sorted by size to ensure a maximum tag weight to body weight ratio of less than 5%. For the sorting process, we weighed and measured fish and used a length criterion of 100 mm as acceptable to add to our stock tank in the hatchery building. Prior to each day’s tagging, 200 fish were transported to four 166.56 L (44 gallon) perforated cans that were placed in a raceway near the tagging trailer. Each can contained a minimum of 151.4 L (40 gallons) of water, which easily met space recommendations for fish holding (52 grams/ l.8 L [1 gallon])(Peven et al 2005). Fish were carefully netted into 19 L (5 gallon) anesthesia buckets filled to a 10 L line as needed for tag implantation. The average size of Chinook salmon tagged during the study was 98 ± 6 mm fork length (FL) (mean ± standard deviation), with a range of between 83–119 mm and weighed 11.5 ± 2.0 g (with range of 8.7–22.1 g).

7

Tags (transmitters) Tags were weighed and soaked in saline water for at least 24 hours prior to tag activation. Chinook salmon were tagged with VEMCO V4 180 kHz transmitters that weighed on average 0.416 grams in air (SD = 0.002), range (0.411 g - 0.427 g). Tags were 11 mm long, 3.6 mm in height, and 5.7 mm in width (http://vemco.com/products/v4-v5-180khz/; accessed 6/15/15). The average tag weight to body weight ratio was 3.71% (SD=0.56%) for the 1,298 fish tagged, which was well below the recommended 5% ratio (Peven et al 2005). None of the fish had a tag weight to body weight ratio of greater than 4.8%. Tags were custom programmed with two separate codes in 2015: a traditional Pulse Position Modulation (PPM) style coding along with a High Residence (HR) coding. The HR component of the coding allowed for detection at high residence receivers. High residence receivers were placed where signal collisions were anticipated from many tags emitting signals at the same time to the same receiver (CVP, CCF). The PPM identification code is transmitted at 180 kHz, on average every 30 seconds. The HR code is transmitted at 170 kHz and is randomly spaced in between the PPM transmission. The advantages of this programming are that 1) HR pings are transmitted on a different frequency from PPM, which allows different receivers to co-exist and support different research goals in the same study (e.g., survival and positioning studies can use the same fish tags); 2) PPM transmissions provide greater detection range than the Hybrid (PPM with HR) transmission used in 2014; 3) HR transmissions are spaced and transmitted between PPM transmissions rather than transmitted over a 1.5 second period once every 60 seconds. These changes in the tag programming were expected to increase tag detection probabilities at HR sites. For example, tags transmitted 1 HR ping approximately every 5 seconds rather than 8 pings in a 1.5 second window once every 60 seconds. The VEMCO V4 180 kHz tag specifications listed the 95% and 50% estimated battery life of the transmitters as 31 and 38 days, respectively. Activation Tags were activated approximately 24 hours prior to tag implantation using a VEMCO tag activator (Figure 1). Each tag was shipped in a uniquely coded vial. Because the tags had no external identifiers, the codes on the vials were used to track the identity of the individual tags (e.g., manufacturing serial number and tag code). The vials and individual tags were placed into labeled pillboxes after activation. For all releases, time of activation was estimated to the nearest minute. Surgeon Training Chinook salmon surgeon training took place between April 7 and April 10 at the MKRH. Because surgeons received formal surgery training for steelhead six weeks prior to the Chinook salmon training,

8

surgeons practiced their tagging technique on the smaller Chinook during training. U.S. Geological Survey’s Columbia River Research Lab (USGS-CRRL) standard operating procedures (SOP) were used (Liedtke et al. 2012). Salmon were necropsied after tagging to assess surgical proficiency. Two of the surgeons were tagging Chinook salmon for the first time, and one surgeon had eight years of experience.

Tagging Study fish were taken off feed 24 hours prior to transmitter implantation. Study fish (n=1,301) were tagged over two tagging weeks: 4/14/2015 - 4/17/2015 (648) and 4/28/2015 - 5/1/2015 (653). During each tagging session Chinook salmon were surgically implanted with VEMCO V4 tags following procedures based on a standard operating procedure (SOP; Appendix B) developed by the USGS-CRRL, but modified as necessary during the training week. The SOP directed all aspects of the tagging operation, and at least one quality assurance check was made during each tagging day to ensure compliance with the SOP guidance.

Prior to transmitter implantation, fish were anesthetized in a 39.5 mg/L AQUI-S 20E solution until they lost equilibrium. Fish were removed from anesthesia, checked to determine if the fish was suitable for tagging, measured (fork length (FL) to nearest mm), and weighed to nearest 0.1 g. As each fish was removed from the anesthetic solution, fish were examined for attributes that sometimes led to a fish being rejected for tagging. Rejection criteria included fin, eye, and operculum damage, as well as disease, descaling, and size. Fish were replaced if they obtained any observable injury during surgery. Average surgery times were 2:30 minutes (range: 1:23 min to 4:27 min). Once tags were inserted into the fish’s body cavity, fish were placed and held for 10 minutes in 19 L (5 gal) buckets filled with 10 liters of water with high dissolved oxygen concentrations (130-150%) to recover from anesthesia effects. Each tagger generally tagged fish in groups of 3 fish that were placed into two 19L buckets, one containing 1 fish and the other containing two fish. Three surgeons were utilized to tag the fish and each surgeon had an assistant. Three individuals (runners) helped to move fish into and out of the tagging trailer.

Tag Validation Tags were monitored during fish recovery with 180 kHz hydrophones placed in recovery buckets to validate the operational status of each tag prior to transportation to the release site (Figure 2). In 2015, VEMCO provided the study with two VRHR prototype receivers that were capable of detecting

9

both the PPM and HR code of each tag (Figure 2). After validation and a 10 minute recovery period, the three fish held across two buckets were combined into one 19 L (5 gal) perforated bucket. Buckets were perforated, starting 15 cm from the bottom of the bucket, to allow water exchange within the transport truck holding tank. Each bucket was covered with a snap-on lid and labeled for use in placing fish in the transport truck, and to ease transfer to the holding cans at the release site fish.

Transport to Release Sites After tagging, the 19L (5 gallon) perforated buckets usually containing three tagged Chinook salmon were loaded directly into a transport tank on a flat-bed truck (Figure 3). There were five buckets, over the course of the two week tagging period, that only had two tagged fish in the bucket, due to the tags not coding out, and having no tag replacements. Immediately prior to loading, all fish were visually inspected for mortalities or signs of poor recovery from tagging (e.g. erratic swimming behavior). Fish that died or were not recovering from surgery were replaced with a new tagged fish.

In order to minimize the stress associated with moving fish and for tracking smaller groups of individual tagged fish, three specially designed transport tanks (one large and two small) were used to move Chinook salmon from the MKRH, where the tagging occurred, to the release sites at Durham Ferry or Medford Island (Figure 3). The transport tanks for Chinook salmon were designed to securely hold a series of 19 L perforated buckets with fish. Tanks had an internal frame that held 21 or 30 buckets in individual compartments to minimize contact between containers and to prevent tipping. Buckets were also covered in the transport tanks with stretched cargo nets to assure buckets did not tip over or lids did not come off. Two small transport tanks (each holding up to 21 buckets) were mounted on the bed of an 8 m (26 ‘) flatbed truck that was equipped with an oxygen tank and hosing to deliver oxygen to each of the tanks for the morning transport (Figure 3). One larger transport tank (containing 30 buckets) was mounted on a second 8 m flatbed truck with an oxygen tank and hosing and used for the afternoon transport. Two trips to Durham Ferry were made each tagging day, with the morning and afternoon sessions of tagged fish being transported separately for the first week of releases and for the first day of the second week of releases (April 14 – April 17 and April 28) (Table 1). Because river temperatures rose to nearly lethal levels at Durham Ferry, only 162 study fish (fish from the first day) were transported to the Durham Ferry release site during the second week of tagging (Table 1). The remaining 491 study fish tagged during the second week of tagging were held at MKRH for 24 hours after tagging and then transported to a ferry dock near Medford Island (Table 2). Medford Island is further downstream than Durham Ferry on the San Joaquin River (Figure 4) and was used as a release site because water

10

temperatures were more suitable for release. No 24 hour holding period was possible at the Medford Island location, so fish were held for 24 hours at the MKRH before being transported there.

Water temperature and dissolved oxygen (DO) in the transport tanks were recorded after loading buckets into transport tanks but before leaving the MKRH, and at the release site after transport, but prior to unloading buckets (Table 3). The temperature and DO were also measured in the river at the holding site at Durham Ferry and at the ferry dock at Medford Island, prior to moving fish into holding containers or onto the boat for release, respectively (Table 3). Transfer to Holding Containers When water temperatures were relatively high in the river and higher than that in the transport tank, an attempt was made to allow temperatures to increase in the transport tank by opening the lids of the transport tank and allowing the buckets to sit in the tank for a maximum of 1 hour (at Durham Ferry) or it until it was time to load them onto the boat for release on the ebb tide (Medford Island). When water temperatures in the river were 50 C or higher than the water temperatures in the transport tanks, as was the case for the first truck delivery to Durham Ferry on 4/28 (transport tank 1 and 2; Table 3), we increased the water temperature in the transport tanks by opening the lids of the transport tanks and allowing the water to warm from solar radiation before unloading the buckets (Table 3). While this generally worked to increase water temperature in the transport tanks slightly it did not increase the water temperatures substantially over a reasonable amount of time (1 hour maximum).

Opening the transport tank lids to increase water temperature in the transport tanks was also done for most of the Medford Island releases, even if the water temperature between the transport tank(s) and the river was within the 50 C difference, to allow the water temperature in the perforated buckets to better match that in the river while waiting for the beginning of the ebb tide when fish were moved onto the boat for release downstream.

In two cases (4/28 tank 3, and 4/30 tank 3; Table 3), opening the transport tank lids for increasing the water in the transport tanks was not going to be sufficient and an alternative approach was used. For these two groups, tagged fish were tempered in buckets by adding ½ bucket of river water to “sleeves” and placing perforated buckets into “sleeves”. (A sleeve is a similar sized, non- perforated bucket that allows more water to stay in the perforated bucket than would be the case without placing it in a sleeve.) Perforated buckets retained some remaining transport tank water in them. We also added a bubbler for oxygenation and waited up to one hour for the fish to acclimate to

11

an intermediate water temperature between the transport tank and river, before putting them into the holding cans at Durham Ferry or loading them onto the boat for release at Medford Island.

Once acclimated if necessary, or immediately after arrival at Durham Ferry or after being held until the beginning of the ebb tide at Medford Island, tagged fish were either moved into holding cans at Durham Ferry or onto the boat for movement downstream to the release site at Medford Island, respectively.

At the Durham Ferry release site, 19 L “sleeves” were filled half-way with river water, placed in the back of a pick-up truck then driven a short distance to alongside the transport truck on the levee road. The perforated buckets were then unloaded from the transport tank and placed into the water- filled sleeves in the back of the pick-up truck and driven down to the river’s edge. Once at the river’s edge, perforated buckets within sleeves were unloaded from the pick-up truck and carried to the river. Perforated buckets were separated from the sleeves at the shore and carried to the holding containers in the river about a meter from the shore.

Once at the river’s edge at Durham Ferry, the tagged Chinook salmon were transferred from 19L buckets to 120 L (32 gal), perforated, plastic garbage cans (“holding cans”) held in the river. The perforated holding cans had hole sizes of 0.79 cm. Five buckets containing three fish each were emptied into each perforated holding can. Each bucket and holding can was labeled to track each specific tag and assure fish were transferred to the correct holding can for later release at the correct time. Tagged salmon were held in the perforated holding cans at Durham Ferry for approximately 24 hours prior to release.

For tagged salmon released near Medford Island, a group of perforated buckets was taken out of the transport tank and placed into sleeves ½ full of river water, then loaded onto a boat, for movement to the release site downstream of the ferry dock. Between 5 and 8 perforated buckets in sleeves were unloaded from the transport tanks at a time, and loaded onto the boat, for movement downstream to the release site (Table 2).

Fish Releases The Chinook salmon, held in perforated garbage cans at Durham Ferry, were placed into “sleeves” (Figure 5) and transported downstream by boat to the release location (Figure 6) in the middle of the channel, downstream of the holding location (Figure 7). We placed perforated holding cans into

12

sleeves prior to moving the fish downstream and moved them downstream prior to release to potentially reduce initial predation of tagged fish immediately after release, because predators may congregate near the holding location or follow the scent from within the perforated cans as they move downstream. Releases were made every six hours after the 24 hour holding period, at approximately 1500 and 2100 (the day after tagging), and 0300 and 0900 (two days after tagging) at Durham Ferry (Table 1).

Immediately prior to release, each holding container was checked for any dead or impaired fish. At the release time the lid was removed and the holding container was rotated to look for mortalities. The container was then inverted to allow the fish to be released into the river. After the holding container was inverted, the time was recorded. As the holding containers were flipped back over, they were inspected to make sure that none of the released fish swam back into the container. Once the release was completed, the information on any dead fish was recorded and the tags removed. The tags were bagged and labeled and returned to the office to have the individual tag identified.

At the Medford Island site, tagged fish were kept in perforated buckets within “sleeves” and moved onto the boat. After loading the buckets, the boat moved downstream to the release site. Once at the release site, each bucket lid was removed and fish were checked for mortality or impairment just prior to release. After checking each bucket for impaired or dead fish, the bucket was emptied over the side of the boat. Fish were released one bucket at a time, until all buckets on the boat were emptied. Five to eight buckets, with 3 fish per bucket, were released per trip (Table 2). The release time was identified to the nearest minute. Fish were released on the beginning of the ebb tide at Medford Island. Multiple trips to the release location at Medford Island were made for each truckload.

Dummy-Tagged Fish In order to evaluate the effects of tagging and transport on the survival of the tagged fish, several groups of Chinook salmon were implanted with inactive or “dummy” tags. Dummy tags in 2015 were systematically interspersed into the tagging order for each release group. For each day of tagging and transport, 15 fish were implanted with dummy tags and included in the tagging process (Table 1; Table 2). Procedures for tagging these fish, transporting them to the release sites, and holding them at the release site or at the hatchery were the same as for fish with active tags. Dummy-tagged fish were evaluated for condition and mortality after being held at the Durham Ferry release site for approximately 48 hours or just after the release of study fish from one transport truck at Medford Island (Table 2). Fish released at Medford Island were held at the hatchery for approximately 24 hours after

13

tagging to recover from the tagging experience prior to being transported to the ferry dock near Medford Island. Dummy-tagged fish were assessed qualitatively for percent scale loss, body color, fin hemorrhaging, eye quality, and gill coloration (Table 4).

A fish necropsy was performed on each dummy-tagged fish to assess surgical implantation of tags. Each fish was examined and given a score based on the scoring rubric adopted by the USGS-CRRL. Variables examined included suture quality, incision apposition, presence of fungus, organ damage, peritoneal apposition, and signs of expulsion. A total composite score was given for each fish (Table 5). Fish Health Assessment As a part of the 2015 South Delta Chinook salmon survival studies, the U.S. Fish and Wildlife Service’s CA-NV Fish Health Center (CNFHC) conducted a general pathogen screening and smolt physiological assessment. The health and physiological condition of the study fish can help explain their performance and survival during the studies. Pathogen screenings during past studies using Merced River Hatchery Chinook have regularly found infection with the myxozoan parasite Tetracapsuloides bryosalmonae, the causative agent of Proliferative Kidney Disease (PKD). However, fish from Merced River fish hatchery were not used in 2015, and PKD is not a problem at the MKRH where the study fish came from in and where the tagging occurred in 2015. The objectives of this element of the project were to evaluate the juvenile Chinook salmon used for the studies for specific fish pathogens including Tetracapsuloides bryosalmonae and assess smolt development from gill Na+ - K+- ATPase activity to determine potential differences in health between groups. For a complete description of methods see Appendix C. Two additional groups of 15 dummy-tagged fish were held for approximately 48 hours and assessed for parasites and other diseases on April 19 and May 1 by CNFHC (Table 1). Another group of 30 dummy tagged fish were held at the hatchery for 72 hours (instead of 48 hours, due to a scheduling conflict), and sampled for fish health by CNFHC on May 4.

Tag Life Test Two in-tank tag-life studies of Vemco V4 tags were implemented in 2015. The first study used 50 tags, began May 4, and ended when the final tag transmitted its last detection on June 18; one tag did not activate and was removed from the study. The second study began May 14, and ended June 28; this study also used 50 tags, but also had one tag that failed to activate and was removed from the study. In total, 98 tags were used in the tag-life study.

14

Tag Retention The objective of the tag retention component of the study was to evaluate survival, tag retention, and the condition of incision and sutures over time in Chinook salmon that were surgically implanted with acoustic V4 dummy tags. On April 10, 2015, 75 study fish (25 per tagger) were implanted with inactive VEMCO acoustic tags that were marked with a colored dot representing a particular surgeon. However, it was not possible to track each individual fish that was used. A total of 150 fish (75 control and 75 tagged) were held in a 2,271L (600 gallon) fiberglass tank in the hatchery building at the MKRH. All fish were sorted by size to ensure tag burden did not exceed 5%, similar to the released fish. The average weight of these tagged fish at the time of tagging was 11.5 g (SD=2.06 g) and ranged between 8.7 g and 17.6 g. The average FL was 98.8 mm (SD=6.01 mm) and ranged between 83 mm and 109 mm. Control fish were visually assessed to be similar in size but were not weighed or measured.

The tagged and un-tagged retention fish were held for 31 days. On May 11, 2015, all fish were assessed for mortality and external characteristics (parameters defined in Table 4). At this time, the holding tank was discovered to be inadequately covered, resulting in five unaccounted Chinook salmon. Presumably these fish jumped from the holding tank. One shed tag was also recovered from an adjacent tank. The remaining 69 tagged fish were euthanized with a lethal dose of Aqui-S 20E and necropsied for evaluation of parameters as defined in Table 5. Mobile Monitoring During and after completion of both weeks of fish releases, we conducted pilot mobile telemetry monitoring to locate some of the acoustic tags that were still active in the survey area. We conducted two mobile monitoring surveys around the Durham Ferry release site, two surveys downstream from the Durham Ferry release site, and one survey around the Medford Island release site (Figure 4). The first survey was conducted on April 17, 2015, between the Durham Ferry release site and approximately 1 km downstream of the release site. The second survey was conducted on April 21, between approximately 1 km upstream and 1 km downstream from the Durham Ferry release site. The third survey was conducted on May 8, began at the mouth of the Stanislaus River (at the Two Rivers Marina), and followed the San Joaquin River approximately 17.5 km downstream. The fourth survey was conducted on May 15 between the end of the third survey and just past the head of Old River, for about 13.5 km downstream. The fifth survey was conducted on May 21, started approximately 1 km upstream from the Medford Island release site, went 10.6 km downstream, and ended midway into the

15

eastern side of Frank’s Tract. We did not conduct any mobile monitoring in Old River, near the trashracks of the CVP, or outside or inside of .

The mobile monitoring surveys were conducted using a VR100 or VEMCO VRHR, connected to a 180 kHz hydrophone hanging over the side of a boat about 10–30 mm under the surface of the water to detect active tags in the survey area. The boat remained mostly in neutral and was allowed to drift downstream with the current in the middle of the channel. Whenever a code or cluster of codes was detected during surveys 1 and 2, the time of detection was recorded. For surveys 4 and 5, GPS coordinates were also linked to each code and time of detection. During survey 3, GPS coordinates were only linked to detections observed during the first half of the survey due to equipment failure (the remaining detections are only linked to times). The equipment failure occurred at about the time the boat passed the Durham Ferry release site (Figure 4). After completion of the surveys, the data from the VRHR were downloaded and checked against the recorded data to ensure that all codes detected by the receiver were documented.

Statistical Methods Data Processing for Survival Analysis The University of Washington received the database of tagging and release data from the US Fish and Wildlife Service. The tagging database included the date and time of tag activation and tagging surgery for each tagged Chinook salmon released in 2015, as well as the name of the surgeon (i.e., tagger), and the date and time of release of the tagged fish to the river. Fish size (length and weight), tag size, and any notes about fish condition were included, as well as the survival status of the fish at the time of release. Tag serial number and two unique tagging codes were provided for each tag, representing codes for different types of signal coding. Tagging data were summarized according to release group and surgeon, and were cross-checked with Denise Barnard (USFWS) and Pat Brandes (USFWS) for quality control.

Acoustic tag detection data collected at individual monitoring sites (Table 6) were transferred to the US Geological Survey (USGS) in Sacramento, . A multiple-step process was used to identify and verify detections of fish in the data files and produce summaries of detection data suitable for converting to tag detection histories. Detections were classified as valid if two or more transmissions were recorded within a 30 minute time frame on the hydrophones comprising a detection site from any of the tag codes associated with the tag. The University of Washington received the primary database

16

of autoprocessed detection data from the USGS. These data included the date, time, location, and tag codes and serial number of each valid detection of the acoustic Chinook salmon tags on the fixed site receivers. The tag serial number indicated the acoustic tag ID, and was used to identify tag activation time, tag release time, and release group from the tagging database.

The autoprocessed database was cleaned to remove obviously invalid detections. The University of Washington identified potentially invalid detections based on unexpected travel times or unexpected transitions between detections, and queried the USGS processor about any discrepancies. In 2015, the cleaning process found that the usual procedure of accepting as valid all detections that included two or more transmissions within 30 minutes resulted in three detections at the Central Valley Project (CVP) that appeared invalid, based on the release date and pattern of other tag detections. In particular, two tags had detections at the CVP that began at least 21 days before the release date, and one tag had a single detection at the CVP that began 7 weeks after the previous detection, which was at Benicia Bridge. In all cases, the CVP detections were defined by only two transmissions per 30 minute time period. After consultation with Dale Webber at VEMCO and examination of all tag detections defined by only two transmissions, a modified procedure was implemented that required four or more transmissions within 30 minutes to define a valid detection at the CVP, and two or more transitions within 30 minutes for valid detections at all other sites. All changes were noted and made to the database by the University of Washington. All subsequent analysis was based on this cleaned database.

The information for each tag in the database included the date and time of the beginning and end of each detection event when a tag was detected. Unique detection events were distinguished by detection on a separate hydrophone or by a time delay of 30 minutes between repeated hits on the same receiver. Separate events were also distinguished by unique signal coding schemes (e.g., PPM vs. HR). The cleaned detection event data were converted to detections denoting the beginning and end of receiver “visits,” with consecutive visits to a receiver separated either by a gap of 12 hours or more between detections on the receiver, or by detection on a different receiver. Detections from receivers in dual or redundant arrays were pooled for this purpose, as were detections using different tag coding schemes.

The same data structure and data processing procedure was used to summarize detections of the acoustic-tagged predatory fish. Detections of the predatory fish were compared to detections of the Chinook salmon tags to assist in distinguishing between detections of salmon and detections of predators.

17

Distinguishing between Detections of Chinook Salmon and Predators The possibility of predatory fish eating tagged study fish and then moving past one or more fixed site receivers complicated analysis of the detection data. The Chinook survival model depended on the assumption that all detections of the acoustic tags represented live juvenile Chinook salmon, rather than a mix of live salmon and predators that temporarily had a salmon tag in their gut. Without removing the detections that came from predators, the survival model would produce potentially biased estimates of survival of actively migrating juvenile Chinook salmon through the Delta. The size and direction of the bias would depend on the amount of predation by predatory fish and the spatial distribution of the predatory fish after eating the tagged salmon. In order to minimize bias, the detection data were filtered for predator detections, and detections assumed to come from predators were identified.

The predator filter used for analysis of the 2015 data was based on the predator filter designed and used in the analyses of the 2011, 2012, 2013, and 2014 data (SJRGA 2013, Buchanan et al. 2015, 2016, Buchanan et al. 2018). Those predator filters in turn were based on predator analyses presented by Vogel (2010, 2011), as well as conversations with fisheries biologists familiar with the San Joaquin River and Delta regions and the predator decision processes used in previous years (SJRGA 2010, 2011). The filter was applied to all detections of all tags. Two data sets were then constructed: the full data set including all detections, including those classified as coming from predators (i.e., “predator-type”), and the reduced data set, restricted to those detections classified as coming from live Chinook salmon smolts (i.e., “smolt-type”). The survival model was fit to both data sets separately. The results from the analysis of the reduced “smolt-type” data set are presented as the final results of the 2015 Chinook salmon tagging study. Results from analysis of the full data set including “predator-type” detections were used to indicate the degree of uncertainty in survival estimates arising from the predator decision process.

The predator filter was based on assumed behavioral differences between salmon smolts and predators such as striped bass and white catfish. All detections were considered when implementing the filter, including detections from acoustic receivers that were not otherwise used in the survival model. As part of the decision process, environmental data including river flow, river stage, and water velocity were examined from several points throughout the Delta (Table 7), as available. Hydrologic data were downloaded from the California Data Exchange Center website (http://cdec.water.ca.gov/selectQuery.html) on 14 September 2016, and from the California Water Data

18

Library (www.water.ca.gov/waterdatalibrary/ ) on 18 July 2016. Environmental data were reviewed for quality, and obvious errors were omitted.

For each tag detection, several steps were performed to determine if it should be classified as predator or salmon. Initially, all detections were assumed to be of live smolts. A tag was classified as a predator upon the first exhibition of predator-type behavior, with the acknowledged uncertainty that the salmon smolt may actually have been eaten sometime before the first obvious predator-type detection. Once a detection was classified as coming from a predator, all subsequent detections of that tag were likewise classified as predator detections. The assignment of predator status to a detection was made conservatively, with doubtful detections classified as coming from live salmon. In general, the decision process was based on the assumptions that (1) salmon smolts were unlikely to move against the flow, and (2) salmon smolts were actively migrating and thus wanted to move downriver, although they may have temporarily moved upstream with reverse flow.

A tag could be given a predator classification at a detection site on either arrival or departure from the site. A tag classified as being in a predator because of long travel time or movement against the flow was typically given a predator classification upon arrival at the detection site. On the other hand, a tag classified as being in a predator because of long residence time was given a predator classification upon departure from the detection site. Because the survival analysis estimated survival within reaches between sites, rather than survival during detection at a site, the predator classifications on departure from a site did not result in removal of the detection at that site from the reduced data set. However, all subsequent detections were removed from the reduced data set.

The predator filter used various criteria on several spatial and temporal scales, as described in detail in previous reports (e.g., SJRGA 2013, Buchanan et al. 2015, 2016, Buchanan et al. 2018). Criteria fit under various categories, described in more detail in SJRGA (2013): fish speed, residence time, upstream transitions, other unexpected transitions, travel time since release, and movements against flow. The criteria used in the 2011, 2012, 2013, and 2014 studies were updated to reflect river conditions and observed tag detection patterns in 2015 (Table 8). There were several new receiver sites installed in 2015 that were added to the predator filter: BDF1 = Below Durham Ferry 1 (A3), BDF2 = Below Durham Ferry 2 (A4), COL = Columbia Cut (F2), and SJD = San Joaquin River at Disappointment Slough (A13) (Table 6 and Figure 4). In addition, there was a receiver placed at the San Joaquin Shipping Channel (SJS = A10) again in 2015, after not being used in 2014.

19

No new criteria were developed for the 2015 study, and the general approach used in the 2013 and 2014 studies was used for the 2015 study. The overall maximum far field residence time, maximum total visit length, and maximum time since release allowed remained set at 360 hours (15 days) for all detection sites. More information on criteria used can be found in reports on the 2013 and 2014 studies in Buchanan et al. (2016) and Buchanan (2016), respectively. Predator filter components are summarized below for the new detection sites in 2015, as are a subset of criteria for select other sites; other criteria are defined in Table 8 for all sites.

DFU, DFD = Durham Ferry Upstream (A0) and Durham Ferry Downstream (A2): maximum total visit length = 15 hours for DFU, and 94 hours for DFD.

BDF1, BDF2 = Below Durham Ferry 1 (A3) and Below Durham Ferry 2 (A4): ignore flow and velocity measures, allow long travel time to accommodate initial disorientation after release; maximum total visit length = 114 hours.

BCA, MOS, HOR = Banta Carbona (A5), Mossdale (A6), and Head of Old River (B0): maximum total visit length = 120 hours for BCA, 130 hours for MOS, and 140 hours for Head of Old River.

SJL, ORE = San Joaquin at Lathrop (A7) and Old River East (ORE): maximum total visit length = 176 hours for SJL, and 249 hours for ORE.

SJS, MAC, TCE/TCW, MFE/MFW, OSJ = San Joaquin River Shipping Channel (A10), MacDonald Island (A11), Turner Cut (F1), Medford Island (A12) and Old River at San Joaquin River (near Old River Mouth; B5): maximum total visit length = 10 hours for SJS and TCE/TCW, 20 hours for MAC, and 60 hours for MFE/MFW and OSJ.

COL = Columbia Cut (F2): ignore flow and velocity measures; maximum total visit length = 60 hours; a single transition is allowed from Middle River at Middle River (C3).

SJD = San Joaquin at Disappointment Slough (A13): allow longer residence time if transition water velocity was low. Repeated visits require arriving with opposite flow and velocity conditions to departure conditions. Maximum total visit length = 60 hours and total time actually detected per visit must be less than 3 hours for downstream transitions. Allow limited transitions from Jersey Point (G1), Middle River at Middle River (C3), and Old River at the San Joaquin (B5); arrival and departure should not be against flow direction (departure flow ignored at Jersey Point).

20

WCL, OR4, RGU/RGD = West Canal (B3), Old River at Highway 4 (B4), and radial gates at the Clifton Court Forebay (D1, D2): maximum total time in SW interior Delta (WCL, OR4, MR4, MID, RGU/RGD, CVP, or CVP tank) or near facilities after leaving lower San Joaquin River = 120 hours.

MR4, MID = Middle River at Highway 4 (C2) and Middle River at Middle River (C3): maximum total visit length = 60 hours at MR4, and 174 hours at MID. Maximum total time in SW interior Delta or near facilities after leaving lower San Joaquin River = 120 hours.

JPT/JPE/JPW, FRE/FRW, TMN/TMS, SBS = Jersey Point (G1), False River (H1), Threemile Slough (T1), and Spoonbill Slough (T3): maximum total visit length = 145 hours for JPT/JPE/JPW, and 10 hours for FRE/FRW, TMN/TMS, and SBS.

MAT/MAE/MAW, BBR = Chipps Island (G2) and Benicia Bridge (G3): maximum total visit length = 50 hours.

The predator scoring and classification method used for the 2011 – 2014 studies was used again for the 2015 study, resulting in tags being classified as in either a predator or a smolt upon arrival at and departure from a given receiver site and visit; for more details, see SJRGA (2013). All detections of a tag subsequent to its first predator designation were classified as coming from a predator, as well.

Constructing Detection Histories For each tag, the detection data summarized on the “visit” scale were converted to a detection history (i.e., capture history) that indicated the chronological sequence of detections on the fixed site receivers throughout the study area. In cases in which a tag was observed passing a particular receiver or river junction multiple times, the detection history represented the final route of the tagged fish past the receiver or river junction. In particular, if a fish was observed even far downstream in one route but then returned to the river junction and finally selected the other route, then survival and detection in the later route were modeled. Detections from the receivers comprising certain dual arrays (2 lines of receivers per array) or triple arrays (3 lines of receivers per array) were pooled, thereby converting the dual and triple arrays to redundant arrays: the San Joaquin River just downstream of the release site at Durham Ferry (DFD, site A2), at Banta Carbona (BCA, site A5), near Mossdale Bridge (MOS, site A6), near the Head of Old River (HOR, site B0), Jersey Point (JPT/JPE/JPW, site G1), and Chipps Island (MAT/MAE/MAW, site G2). There were too few detections at the radial gates at the entrance to Clifton Court Forebay in 2015 to model transitions to that site, with or without the effect of gate status (open

21

or closed). Unlike in previous years, detections on the receivers located in the San Joaquin River just upstream of the head of Old River (B0) were used in the survival model for Durham Ferry releases. For Medford Island releases, detections were used in the predator filter but not in the survival model from Turner Cut (site F1), Middle River at Highway 4 (C2), Old River at Highway 4 (B4), West Canal (B3), the radial gates at the Clifton Court Forebay (D1, D2), Threemile Slough (T1), and False River (H1).

Survival Model For each release site, a multi-state statistical release-recapture model was developed and used to estimate Chinook salmon smolt survival and migration route parameters throughout the study area. The models were based on the multi-state release-recapture models used in previous Chinook salmon tagging studies (SJRGA 2013, Buchanan et al. 2015, 2016, Buchanan et al. 2018), but simplified to the reduced data structure observed in 2015. Although the 2015 study placed receivers at most of the sites used in previous years, there were few or no detections at many of the interior Delta and downstream sites; this required simplification of the model.

Sites new to the study in 2015 included the two sites below Durham Ferry (BDF1 = A3 and BDF2 = A4) and Columbia Cut (COL = F2). Sites BDF1 and BDF2 were incorporated into the statistical model used with the Durham Ferry release groups, and COL was included in the model used with the Medford Island release group.

For both release sites, the release-recapture model used parameters that denote the probability of detection ( Phi ), route entrainment (ψ hl ), Chinook salmon survival ( Shi ), and transition probabilities equivalent to the joint probability of movement and survival (φkj, hi ) (Figure 8 and Figure 9, Appendix E.:

Table E1). For each dual array, unique detection probabilities were estimated for the individual receiver lines comprising the array: Phia represented the detection probability of the upstream line at station i in route h, and Phib represented the detection probability of the downstream line. A comparable parameterization was used for triple arrays, using parameters Phia , Phib , and Phic to represent the detection probabilities of the upstream, middle, and downstream lines, respectively, that comprised the triple array. The arrays at both Jersey Point (G1) and Chipps Island (G2) were triple arrays (Figure 4). Some sites that were composed of two receivers lines were modeled as a single (“redundant”) array, in which detections from the multiple lines were pooled and only a single overall detection probability was estimated; this was performed to improve model fit.

22

The model parameters are:

Phi = detection probability: probability of detection at telemetry station i within route h, conditional on surviving to station i, where i = ia, ib for the upstream, downstream lines in a dual array, respectively, and i = ia, ib, ic for the upstream, middle, and downstream lines in a triple array, respectively.

Shi = perceived survival probability: joint probability of migration and survival from telemetry station i to i+1 within route h, conditional on surviving to station i.

ψ hl = route entrainment probability: probability of a fish entering route h at junction l (l =1, 2, 3), conditional on fish surviving to junction l.

φkj, hi = transition probability: joint probability of route entrainment and survival; the

probability of migrating, surviving, and moving from station j in route k to station i in route h, conditional on survival to station j in route k.

λ = joint transition and detection probability: joint probability of moving downstream from Chipps Island, surviving to Benicia Bridge, and detection at Benicia Bridge, conditional on survival to Chipps Island.

Durham Ferry Survival Model Fish moving from Durham Ferry to Chipps Island had the opportunity to use the same migration routes as in previous years; see Buchanan et al. (2015, 2016) for descriptions of the possible migration routes. In 2015, a barrier blocked most access to Old River from its distributary point (“head”) with the San Joaquin River. Nevertheless, culverts allowed some passage of both water and fish. The routes and study area exit points modeled for the Durham Ferry release groups in 2015 are summarized as follows (Figure 8):

A = San Joaquin River: survival B = Old River: survival F = Turner Cut: survival

23

G = Jersey Point, Chipps Island, Benicia Bridge: survival, exit point H = False River: exit point R = : survival

The survival model used for the Durham Ferry releases in 2015 was primarily a simplification of the full, two-part model used in 2013 and 2014. However, unlike in previous years, the receivers located in the San Joaquin River just upstream of the head of Old River (HOR = B0) were included in the Durham Ferry model to help identify where mortality occurred between Mossdale and the receivers just downstream of the head of Old River (SJL = A7, and ORE = B1). Because so few tags were observed entering the Old River route or past the city of Stockton in the San Joaquin River route, the two part- model used in 2013 and 2014 was reduced to a single model, and only those sites with detections (or complementary to sites with detections) were included, along with Jersey Point, False River, Chipps Island, and Benicia Bridge (Figure 8). In particular, no spatial detail was incorporated into the Old River route portion of the model, because there were too few tags detected in that route to model detailed transitions. Only the overall probability of getting from the Old River East (ORE = B1) receiver to Chipps

Island (MAT/MAE/MAW = G2) was modeled in the Old River route: SB1 .

In the San Joaquin River route, route selection probabilities were modeled at Rough and Ready

Island (as in 2014; Buchanan et al. 2018) using parameters ψ A2 and ψψRA22=1 − , and at the Turner

Cut junction using parameters ψ A3 and ψψFA33=1 − . Route selection was also modeled at the Jersey

Point/False River junction, using parameters ψ G1 and ψψHG11=1 − . The transition probabilities to the Jersey Point/False River junction from the MacDonald Island receivers (MAC = A11) and the Turner Cut

receivers (TCE/TCW = F1) were modeled using parameters φA11, GH and φF1, GH , respectively (Figure 8).

The joint probability of survival from Chipps Island to Benicia Bridge and detection at Benicia Bridge was modeled as the “last reach” parameter λφ= GG2, 3P G 3 .

In addition to the model parameters, derived performance metrics measuring migration route probabilities and survival were estimated as functions of the model parameters. Both route selection and route-specific survival were estimated for the two primary routes determined by routing at the head of Old River (routes A and B):

ψψAA= 1 : probability of remaining in the San Joaquin River at the head of Old River,

24

ψψBB= 1 : probability of entering Old River at the head of Old River.

The probability of surviving from the entrance of the Delta near Mossdale Bridge (site A6, MOS) through an entire migration pathway to Chipps Island was estimated as the product of survival probabilities that trace that pathway:

SA= SSSS A6078,2 B A AG : Delta survival for fish that remained in the San Joaquin River past the

head of Old River,

SB= SSS A6 B 0 BG 1, 2 : Delta survival for fish that entered Old River at its head, where

SSBG1, 2= B 1 .

The parameter SAG8, 2 represents the probability of getting to Chipps Island (i.e., , site MAT/MAE/MAW) from site A8 (SJG). This parameter represents multiple pathways around or through the Delta to Chipps Island, as well as both the San Joaquin River route and the Burns Cutoff route past Rough and Ready Island (Figure 4):

SSSAG8,2= A 8(ψψ AAG 2 9,2+ RRG 2 S 1,2) ,

where SAG9, 2 and SRG1, 2 represent the probability of survival to Chipps Island from the Navy Drive

Bridge receivers (SJNB = site A9) and the Rough and Ready Island receivers (RRI = site R1), respectively:

SSAG9, 2= A 9SA10(ψφ A 3 A 11, GH + ψφF3 F 1, GH ) ψφG 1 GG 1, 2 and

SSRG1, 2 = R1S A 10(ψφ A 3 A 11, GH+ ψφ F3 F 1, GH) ψφ GG 1 1,G 2 .

In the event that site SJS was removed from the model, the parameters SA9, AF = SSAA910 and

SR1, AF = SSRA110 were estimated directly in the model.

The routes from the lower San Joaquin River through the water export facilities were not modeled for Durham Ferry releases in 2015 because no Durham Ferry tags were observed taking those

25

routes. Had there been such tags, additional transitions would have been modeled that connected sites A11 and F1 to Chipps Island but bypassed Jersey Point, because Jersey Point is neither required nor expected to be passed by fish that were salvaged at the water export facilities and released upstream of Chipps Island.

The parameter SSBG1, 2= B 1 includes all possible routes for transiting between the beginning of the Old River route and Chipps Island, including salvage and transport from the water export facilities.

Using the estimated migration route probabilities and route-specific survival for the two primary routes (A and B), survival of the population from Mossdale (site A6) to Chipps Island was estimated as:

STotal=ψψ A SS A + B B .

In previous years, it was possible to estimate survival from Mossdale through the Southern Delta (SD) region, which is bounded downstream by the MAC and TCE/TCW receivers in the San Joaquin River route, and by the water export facility entrance receivers (CVP, RGU) and Highway 4 receivers (OR4, MR4) in the Old River route (Figure 4). In 2015, no Durham Ferry tags were detected in the Old River route past the ORE receivers (site B1), and it was not possible to estimate Southern Delta survival in the Old River route without assuming 100% detection probability at the water export facilities and Highway 4. Southern Delta survival in the San Joaquin River route was estimated as:

SA() SD= SSSS A 6 B 0 A 7 A 8() SD ,

where

SASDA8( )= S 8(ψψ AA 2 S 9+ RR 2 SS 1) A 10 .

As in past years (e.g., Buchanan et al. 2015), route selection probabilities for the route A subroutes were estimated as

ψAA= ψψ A13 A : probability of remaining in the San Joaquin River at the head of Old River and at Turner Cut,

ψAF= ψψ A13 F : probability of remaining in the San Joaquin River at the head of Old River, and

entering Turner Cut at its junction with the San Joaquin River, where ψψFA33=1 − .

26

The probability of reaching Mossdale from the release point at Durham Ferry, φAA1, 6 , was defined as the product of the intervening reach survival probabilities:

φφAA1,6= AAAAAA 1,22345SSSS .

This measure reflects a combination of mortality and possible residualization upstream of Old River, although the Chinook salmon in this study were assumed to be migrating (i.e., no residualization).

Individual detection histories (i.e., capture histories) were constructed for each tag as described above. Each detection history consisted of one or more fields representing initial release (field 1) and the sites where the tag was detected, in chronological order. Detection on both lines of receivers in a dual array was denoted by the code “ab”, detection on only the upstream receiver was denoted “a0”, and detection on only the downstream receiver was denoted “0b”. For triple arrays, there were seven possible detection combinations on the three receiver lines comprising the array, conditional on detection at the array: “abc”, “ab0”, “a0c”, “0bc”, “a00”, “0b0”, and “00c.” Detection histories were parameterized as in previous analyses (e.g., Buchanan et al. 2015, 2018), modified for triple arrays as necessary. Under the assumptions of common survival, route entrainment, and detection probabilities, and independent detections among the tagged fish in each release group, the likelihood function for the survival model for each release group is a multinomial likelihood with individual cells denoting each possible capture history.

Medford Island Survival Model Fish released at Medford Island were expected to move downstream toward Chipps Island, either via the San Joaquin River, Old River and Frank’s Tract, or an interior Delta route via the water export facilities and salvage (Figure 4). However, because of the tidal nature of the release site and the resulting strong reverse flows experienced there several times each day, and also allowing for momentary confusion by newly released fish, it was also possible that fish released at Medford Island would move upstream temporarily before turning back downstream, either via the San Joaquin River again or Columbia Cut. Fish that eventually turned back downstream past the Medford Island receivers (site A12) after initial upstream movement were modeled as moving downstream, using the “last route” protocol described above. Fish that did not eventually pass Medford Island after moving upstream were modeled as moving upstream either to MacDonald Island (A11) or Columbia Cut (F2). Fish that moved to MacDonald Island and did not then return downstream could have maneuvered to Chipps Island via

27

Turner Cut and the water export facilities or Frank’s Tract. However, too few of the tags assigned to the MacDonald Island route were subsequently detected at or near the water export facilities or Frank’s Tract to model subroutes from MacDonald Island, and only the overall probability of reaching Chipps Island from MacDonald Island was modeled (Figure 9).

Fish that moved upstream to Columbia Cut, either directly after release or after an initial movement to MacDonald Island, could have moved either north toward the mouth of Old River (site B5) or south toward the Middle River receivers at MID (site C3). Fish that moved from Columbia Cut to site A12, A13, or B5 were assigned to a northward route from the Medford Island release site, using the “last route” protocol of detection history construction. Thus, only the southward movement from F2 to C3 was modeled (Figure 9). Fish detected at site C3 (MID) could alternatively have come from the release site via the Medford Island receivers (site A12); regardless of how they arrived at site C3, fish at that site were then expected to move toward Chipps Island, either via Frank’s tract or via the water export facilities (Figure 4). Data were too sparse to model those migration route separately, so only the overall probability of getting from site C3 to Chipps Island (G2) was modeled: φCG3, 2 (Figure 9).

Fish that moved downstream from release at Medford Island were expected to be detected at Medford Island and then either at the San Joaquin receiver near Disappointment Slough (SJD = site A13) or in Old River near its mouth (OSJ = site B5); from either of those sites, fish would then move toward the Jersey Point/False River junction, and then to Chipps Island (Figure 9). Thus, the routes and study area exit points modeled for the Medford Island release group in 2015 are summarized as follows (Figure 9):

A = San Joaquin River: survival B = Old River: survival C = Middle River: survival F = Columbia Cut: survival G = Jersey Point, Chipps Island, Benicia Bridge: survival, exit point H = False River: exit point

Because of the complex and tidal nature of the river and Delta system in the region of Medford Island, no attempt was made to estimate route selection separately from survival for the Medford Island

28

release, aside from route selection at the Jersey Point/False River junction. All other transitions used the joint survival and route selection parameter, φ (Figure 9).

Performance metrics estimated for the Medford Island release group included the overall probability of surviving to Chipps Island from various sites, including MacDonald Island (A11), Disappointment Slough (A13), Old River Mouth (B5), Medford Island (A12), Columbia Cut, and the release site:

SAG11, 2= φ AG 11, 2 ,

SAG13, 2= φ AGHGGG 13, ψφ 1 1, 2 ,

SBG5, 2= φ BGHG 5, ψφ 1 GG 1, 2 ,

SAG12, 2=φ AA 12, 13 SS AG 13, 2 ++ φ AB 12, 5 BG 5, 2 φφ AC 12, 3 CG 3, 2 ,

SFG2, 2= φφ FC 2, 3 CG 3, 2 , and

SMF, G 2=φ MF , A 12 SSS A 12, G 2 ++ φφ MF , F 2 F 2, G 2 MF , A 11 A 11, G 2 .

As done for the Durham Ferry release groups, detection histories were constructed for each tag using the “last route” protocol described above. Dual and triple line detections were represented using codes “a”, “b”, and/or “c” as appropriate; see the description of detection histories for the Durham Ferry model for more information. Under the assumptions of common survival, route entrainment, and detection probabilities, and independent detections among the tagged fish in each release group, the likelihood function for the survival model for each release group is a multinomial likelihood with individual cells denoting each possible capture history.

Parameter Estimation The survival models were fit first for each of the two release groups individually, where “release group” is defined by date: group 1 = April 15-19, and group 2 = April 29-May 2. Next, the models were fit to the data pooled across release groups to provide population-level estimates. For cases where tags were released at both Durham Ferry and Medford Island (i.e., release group 2, and the pooled analysis), a joint model was fit that equated detection probabilities at sites common to both release-specific

29

models (e.g., Jersey Point, Chipps Island, Benicia Bridge, and MacDonald Island). Transition probabilities were not equated across release sites. Unlike in previous years, the sparse detection data observed from the Durham Ferry releases prevented estimating unique detection probabilities by release group for the population-level analysis. In each case, the multinomial likelihood models described above were fit numerically to the observed set of detection histories according to the principle of maximum likelihood using Program USER software, developed at the University of Washington (Lady et al. 2009). Point estimates and standard errors were computed for each parameter. Standard errors of derived performance measures were estimated using the delta method (Seber 2002: 7-9). For each model fit, goodness-of-fit was assessed visually using Anscombe residuals (McCullagh and Nelder 1989).

Sparse data prevented some parameters from being freely estimated for some release groups. Transition, survival, and detection probabilities were fixed to 1.0 or 0.0 in the USER model as appropriate, based on the observed detections. Several sites in the Durham Ferry model had no detections of Durham Ferry tags: R1, F1, G1, H1, G2, and G3. It was necessary either to remove these sites from the model, or else assume 100% detection and estimate transition probabilities to those sites (or route selection probabilities) at 0. Sites that had no detections and whose transitions to them also came only from sites that themselves had no detections (i.e., G3 for the first release group) had to be removed from the model completely, without an attempt to estimate transition probabilities to those sites. Also of concern were sites with only a single tag detection. This occurred for several sites from the Durham Ferry release groups: A11 and B1 from the first release group, and A3 and A4 from the second release group. Detection probabilities cannot be estimated when only a single tag is detected, even with a dual or triple array. Thus, it was necessary to assume 100% detection at these sites to make inferences on transition probabilities to these sites; estimation of transition probabilities from these sites and sites with only two detections was not attempted because of the low sample size. Additionally, for the second Durham Ferry release group, no attempt was made to estimate survival past the Durham Ferry Downstream (DFD) receivers at site A2. Detection data were pooled from the two lines comprising the dual array at DFD for the first release group to improve model fit.

For the analyses of data where tags were released at both release sites (i.e., release group 2 and the pooled release groups), it was still necessary to assume 100% detection at Old River East (site B1). However, it was now possible to estimate detection probabilities at sites A11, G1, and G2 for Durham Ferry tags by assuming common detection probabilities for Durham Ferry and Medford Island tags (i.e., use the Medford Island tags to estimate detection probabilities at A11, G1, and G2). Data from the

30

Medford Island release improved the detection probability estimates at shared sites, which in turn provided more confidence in transition probabilities to those sites.

Although several tags were detected at False River from the Medford Island release group, only one such detection was included in the survival model detection histories because the others were all subsequently detected at Jersey Point. Because it was not possible to estimate the detection probability at False River from a single tag detected, False River was omitted from the Medford Island model (and from the Durham Ferry model when the two models equated detection probabilities). This had the effect of preventing estimation of route selection at the Jersey Point/False River junction, and meant that the parameter φxG,1= φψ xGH , G 1 was estimated directly, where site x represents site A13 or B5

from the Medford Island model, and A11 from the Durham Ferry model (transitions from site F1 were removed because no tags were detected at F1). The single array of receivers at Middle River at Middle River (MID = site C3) and the fact that none of the tags detected at that site was subsequently detected at Chipps Island meant that it was not possible to estimate the detection probability at that site. Thus, it was necessary to remove C3 from the Medford Island model. Because the migration route to Chipps Island via C3 was not observed to be successful in 2015, removing this site from the model should not bias estimates of other model parameters. The Medford Island receivers (MFE/MFW = site A12) were also removed from the Medford Island model, because they were too close to the release site to allow

for independent detections at that site. Thus, parameters φFG2, 2 , φMF, A 13 , and φMF,5 B were estimated directly from the simplified Medford Island model. In order to improve model fit, detection data were pooled across the lines within the dual arrays at SJD (site A13) and OSJ (B5), and at the triple arrays at Jersey Point (G1) and Chipps Island (G2).

The very sparse detection data observed for Durham Ferry releases in 2015 meant that it was not possible to compare route-specific survival estimates for the two primary routes through the Delta: the San Joaquin River route and the Old River route from the head of Old River. It was possible, however, to use detections of tags released at Medford Island to compare survival to Chipps Island from the Disappointment Slough receivers (A13) to those from the Old River Mouth receivers (B5), using a two-sided Z-test on the log scale:

ˆˆ ln(SSAG13,2) − ln ( BG5,2) Z = , Vˆ

31

ˆ ˆ ˆˆ Var( SAG13,2) Var( SBG5,2) 2, Cov( SA13,2 G S BG 5,2) where V = +− . ˆ22 ˆ ˆˆ SAG13,2 SBG5,2 SSA13,25,2 G BG

The parameter V was estimated using Program USER. Also tested was whether tagged Chinook salmon smolts showed a preference for route at the head of Old River, using a two-sided Z-test with the test statistic:

ψˆ − 0.5 Z = A . SE (ψˆ A )

Statistical significance was tested at the 5% level (α=0.05).

The effect of release group on the values of the model survival and transition probability parameters was examined by testing for a change in parameter estimates for the second Durham Ferry release group compared to the first release group. For each model survival and transition probability parameter θ , where θφ= kj, hi or θ = Shi , the difference in parameter values between the first and second release groups was defined as

∆=θ θθ12 − ,

for model parameter θR for Durham Ferry release group R ( R =1, 2 ). The difference ∆ was estimated ˆ   by ∆=θ θθ12 − . The null hypothesis of no difference was tested with a two-sided Z-test. Only those parameters that were estimated separately for both release groups and were based on at least four detections at the upstream boundary of the reach were considered; these restrictions resulted in tests for parameters φAA1, 0 , φAA1, 2 , andφAA1, 6 . A family-wise significance level of α=0.10 was selected, and the Bonferroni multiple comparison correction was used, resulting in a test-wise significance level of 0.0333 for 3 tests (Sokal and Rohlf 1995).

Analysis of Tag Failure Six acoustic hydrophones and receivers were used in the 2015 tag-life study. Each receiver experienced times when no tags were detected, which suggested receiver outage and lack of tag monitoring during those periods. Tag detections were compared across all tags detected on each receiver to identify potential unmonitored periods. Tags whose final detections occurred just prior to the unmonitored periods were interval-censored. This affected two tags detected on receiver 300949,

32

both of whose final detections came immediately before an unmonitored period that ran from June 13 to June 15. Interval censoring essentially imputes missing data, in this case either at the beginning of the unmonitored period (i.e., the final tag detection tag) or at the end of the unmonitored period. Imputing missing failure times at the time of final tag detection is conservative for estimating tag survival, but non-conservative for estimating fish survival adjusted by tag survival estimates (i.e., the shorter the estimated tag survival, the higher the adjusted fish survival estimate). Imputing missing failure times at the end of the unmonitored period, on the other hand, is non-conservative for tag survival estimation but conservative for fish survival estimation. Because the overall focus of the study is to estimate fish survival, missing tag failure times were imputed at the end of the unmonitored period (June 15, 2015 16:31:19) for these two tags. Additionally, tag failure times were right-censored to improve model fit; the censoring time was selected based on comparison of model fit.

All tags used in the first tag-life study were activated on May 3, the day before the start of the study. For the second tag-life study, 25 tags were activated on May 3, and 24 tags were activated on May 13, although detections were first recorded on May 14. The tag survival model (the 4-parmaeter vitality curve from Li and Anderson (2009), adjusted for interval censoring) was fit separately first by tag activation date and then by tag-life study. In each case, the model fit was compared to the model fit attained by pooling across all tags using the Akaike Information Criterion (AIC; Burnham and Anderson 2002).

The fitted tag survival curve was used to adjust estimated fish survival and transition probabilities for premature tag failure using methods adapted from Townsend et al. (2006). In Townsend et al. (2006), the probability of tag survival through a reach is estimated based on the average observed travel time of tagged fish through that reach. For this study, travel time and the probability of tag survival to downstream sites were estimated separately for the different routes (e.g., San Joaquin route and Interior Delta route). Standard errors of the tag-adjusted fish survival and transition probabilities were estimated using the inverse Hessian matrix of the fitted joint fish-tag survival model. The additional uncertainty introduced by variability in tag survival parameters was not estimated, with the result that standard error estimates may have been slightly low. In previous studies, however, variability in tag-survival parameters was observed to contribute little to the uncertainty in the fish survival estimates when compared with other, modeled sources of variability (Townsend et al. 2006); thus, the resulting bias in the standard errors was expected to be small.

33

Analysis of Surgeon Effects Surgeon (“tagger”) effects were analyzed in several ways. The simplest method used contingency tests of independence on the number of tag detections at detection sites throughout the study area and upstream of the head of Old River. Specifically, a lack of independence (i.e., heterogeneity) between surgeon and the distribution of detections at key sites was tested using a chi- squared test (α=0.05; Sokal and Rohlf, 1995). Detections from sites downstream of Medford Island were pooled within surgeon categories in order to achieve adequate cell counts for this test. Detections were omitted from other sites that had fewer than 5 detections for any surgeon. This meant that assessment of potential surgeon effects was limited to Banta Carbona, Mossdale, and the sites downstream of Medford Island.

Lack of independence may be caused by differences in survival, route entrainment, or detection probabilities. A second method visually compared estimates across surgeons of cumulative survival throughout the region of the study area that had sufficient detections to estimate survival for all three surgeons, i.e., from Durham Ferry to Garwood Bridge and from the Medford Island release site to Chipps Island and Benicia Bridge (confounded with BBR detection probability). A third method used Analysis of Variance to test for a surgeon effect on individual reach survival estimates, and an F-test to test for a surgeon effect on cumulative survival from both of the release sites (Durham Ferry and Medford Island); individual routes were not considered for this analysis because of sparse detection data. Finally, the nonparametric Kruskal-Wallis rank sum test (Sokal and Rohlf 1995, ch. 13) was used to test for whether one or more surgeons performed consistently poorer than others, based on individual reach survival or transition probabilities through key reaches. In the event that survival was different for a particular surgeon, the model was refit to the pooled release groups without tags from the surgeon in question, and the differences in survival estimates due to the surgeon were examined. The reduced data set (without predator detections), pooled over release groups, was used for these analyses.

Analysis of Travel Time Travel time was measured from release at Durham Ferry or Medford Island to each detection site. Travel time was also measured through each reach for tags detected at the beginning and end of the reach, and summarized across all tags with observations. Travel time between two sites was defined as the time delay between the last detection at the first site and the first detection at the second site. In cases where the tagged fish was observed to make multiple visits to a site, the final visit was used for

34

travel time calculations. When possible, travel times were measured separately for different routes through the study area. The harmonic mean was used to summarize travel times.

Route Selection Analysis There was a physical barrier installed at the head of Old River in 2015 that effectively blocked most entry to Old River. Thus, no analysis of the factors affecting route selection (entrainment) at that river junction was performed. Furthermore, there were too few tags detected on any of the receivers near the Turner Cut junction to implement a route selection analysis for that junction in 2015. Thus, no route selection analysis was performed for the 2015 study.

Survival through Facilities In similar studies of acoustic-tagged steelhead (Buchanan et al. 2015, 2016), a supplemental analysis has been performed to estimate the probability of survival of tagged fish from the interior receivers at the water export facilities through salvage to release on the San Joaquin or Sacramento rivers. This analysis combined detections at Chipps Island with detections at Jersey Point and False River, and compared detection counts to counts of detections at the CVP holding tank and the interior receivers in the Clifton Court Forebay (site RGU). In 2015, there was only 1 tag detected inside Clifton Court Forebay, and no tags detected in the CVP holding tank. Thus, there were too few tags detected at either of the facilities to complete an analysis of salvage through the facilities for Chinook salmon in 2015.

Results Transport to Release Sites There were 2 mortalities observed after transport to the release site (Table 3). Both were from the first transport to Durham Ferry on 4/17. Water temperature in the river at the release sites ranged from 15.4° C to 24.7° C, during the study, with the average during the first week being lower (16.5° C) than for the second week (21.6° C) (Table 3). Dissolved oxygen levels ranged between 6.58 and 13.42 mg/l for all measurements in the transport tanks or in the river (Table 3). Fish in transport tank 3 at Durham Ferry on 4/28, had temperature differences between the transport tank and the river that were substantially greater than 5° C. Water temperatures at the release site were 6.7° C greater than that in the transport tank, as the river water temperature was 24.7° C (Table 3). Despite efforts to acclimate the fish to the water river water before putting them in the river, it is likely the river water temperatures were lethal to the fish.

35

Moving the release site downstream to Medford Island was somewhat helpful, but water temperatures at that release site reached as high as 22.3° C on May 2. Water temperatures were initially 5.0° C higher in the river than in the transport truck after transport on May 1 for transport 1, but holding the fish in the transport tank after arrival at the Medford Island ferry dock allowed them to sufficiently warm so they were within the 4 to 5° difference before unloading (Table 3). The second group of tagged fish transported to Medford Island on 4/30 (transport 3) was acclimated by taking the perforated buckets out of the transport tank and inserting them into sleeves filled ½ full of river water and allowing them to warm for 15 to 30 minutes before loading them onto the boat for transport to the release location. The water temperature in the river was only 4.3° C higher than in the transport tank for this group.

Water temperatures during transport were generally similar between transport tanks within about 0.5° C, with the exception of 4/28 and 4/30 where water temperatures between the warmest and coolest tanks was 1.9o C and 0.9o C, respectively (Table 3 and Appendix D.).

Fish Releases There were 12 mortalities that occurred after holding and prior to release in the 2015 Chinook salmon study (Table 3). Eleven of the 12 mortalities were from the April 28 transport of fish to Durham Ferry (Table 3). Due to the high mortalities and temperatures approaching 25o C, the decision was made to release the remaining fish at the Medford Island site. For the final three releases (April 30, May 1, and May 2), tagged fish were unloaded from the truck into sleeves at the ferry dock and transported to the Medford release site by boat. There were no mortalities just prior to release for the final three days of releases at Medford Island (Table 2 and Table 3).

Dummy Tagged Fish One dummy-tagged fish was found dead after transport, while another one was found dead approximately 5 hours after it was put into its holding container at Durham Ferry, both on 4/28 (Table 1). An additional group of dummy-tagged fish was being put into the same holding container when the dead fish was found approximately 5 hours after it was put into the same holding container. Two others from the 4/28 third transport tank were found dead, and one was missing after being held for 48 hours (Table 1). Of the 60 dummy-tagged Chinook salmon transported to the Durham Ferry holding site, 4 died either prior to, or after being held for 48 hours, and one was missing after being held for 48 hours (Table 1 and Table 9).

36

No mortality was observed for any of the 36 dummy-tagged fish held at the hatchery for 24 hours prior to transport to the release location at Medford Island (Table 2 and Table 9). Dummy-tagged fish were assessed after each truckload of releases was made at Medford Island. No fish from either the Durham Ferry or Medford Island groups were observed to have fin hemorrhaging, light body or gill color, while three had bulging eyes (Table 9). All remaining fish were found swimming vigorously. Mean scale loss for the various groups assessed ranged from 5.0 to 23.3% (Table 9). Mean FL of the ten groups of dummy tagged fish ranged from 95.7 to 106.0 mm. These data indicate that the fish used for the Chinook study in 2015 appeared to be in generally good condition, after being held in the river for 48 hours (Durham Ferry releases) or after being held at the hatchery for 24 hours and transported to Medford Island (Table 9) with the exception of the 4/28 transport groups held at Durham Ferry. Seventeen of the 55 dummy tagged fish examined fish at Durham Ferry were found to have stitched organs. Nine of the 36 dummy tagged fish taken to Medford Island were found to have stitched organs. The mean composite score of all groups ranged between 1.0 and 2.1 (Table 10) indicating generally good surgical techniques, although several had stitched organs. It is not clear how stitched organs could affect survival and if these dummy tagged fish are representative of test fish that were released. Fish Health Assessment In Chinook salmon health assessment groups, mortality over the holding period was low (0–4%) and no significant parasitic infections were detected. Differences in gill Na+/K+-ATPase activity were observed in Chinook salmon sample groups; however the differences were small and likely would not affect migration or survival (Appendix C.).

Tag Retention Study Six dummy tagged fish jumped from the tank during the tank retention study. These were not included in the assessment of mortality during holding. No mortalities occurred among the 75 un- tagged control fish. After the 31 day holding period, all the tagged fish appeared to have normal vigor, body color, fin hemorrhaging (i.e., none), gills, and eyes. The average scale loss of the tagged fish after 31 days was 7.51% (SD=5.24%), which corresponds to 38 fish with normal scale condition (0–5% scale loss), 30 fish with partial scale loss (6–19% scale loss) and two fish that were descaled (≥20% scale loss) (Table 11). The scale condition after holding generally improved since the scale assessments during tagging on Day 0 (26 fish with normal scale condition and 49 fish with partial scale loss), apart from the two descaled fish. In 54 retention fish the acoustic tag resided directly over the incision, in one fish the

37

tag resided anterior to the incision, and in 12 fish the tag resided posterior to the incision (Table 11). In one retention fish the location of the tag was not recorded (Table 11). In the remaining tagged fish without a score for tag position, the tag was completely expelled (Table 11). Three additional fish showed preliminary signs of tag expulsion (e.g., signs of rubbing of the tag against the peritoneum). There was no method used to track individual fish, so it is uncertain whether the tag burden of this fish contributed to its tag loss. However, this fish was one of only two fish to lose its anterior suture (no fish lost their posterior sutures after 31 days) and also showed a broken anterior suture pattern. The inadequate suturing of the incision may have been a contributing factor to tag loss. One tag retention fish displayed a broken anterior suture pattern, and three fish showed a broken posterior suture pattern. None of these fish showed signs of expelling their tags. The anterior ventral suture insertion point (Table 5) showed the least amount of irritation (average=0.78, out of min=0 and max=4; Table 12). The remaining suture insertions points all displayed relatively consistent irritation, ranging between an average of 0.93 and 0.97. Two tagged fish displayed partially open incisions, and no fish displayed gaping incisions. The remaining fish showed good incision apposition, including the fish that expelled its tag. Five fish showed partially unhealed and one fish displayed completely unhealed incisions, with the remaining fish showing good incision healing. Four fish displayed organ inclusion, mainly the encapsulation of the pyloric caeca. Ten percent of the tagged fish displayed some amount of fungus growth, which mainly resided on the suture material. This likely indicates that fungus was not an issue during holding at the MKRH during the 2015 South Delta Salmon Survival Study.

Detections of Acoustic-Tagged Fish Durham Ferry Releases A total of 810 Chinook Salmon were tagged in 2015 for the survival study for release at Durham Ferry. Of those 810 fish, a total of 14 fish died either during transport (2) or during the holding period prior to release (12); the 14 pre-release mortalities were all removed from the release group, and the remaining 796 tagged salmon were released (Table 1). Of the 12 fish that died during the holding period, there were 5 fish whose tags could not be recovered; although those 5 fish were not released, it was not possible to remove their tags from the release data used in estimation of the survival model. This meant that estimates of survival from the release site included survival during the holding period, prior to release. Additionally, two fish that were released had HR codes that were not verified, so they were not used in the survival modeling. Thus, data from a total of 796+5-2 = 799 fish were used

38 in the survival analysis for Durham Ferry releases in 2015. Only 151 tags were released at Durham Ferry from the second release group, but 156 are used in the survival analyses as the five tags that were unable to be recovered are from this group.

Of these 799 tagged fish used in the survival modeling from the Durham Ferry releases, 401 (50%) were detected on one or more receivers either upstream or downstream of the release site (Table 13), including any predator-type detections. The large majority (361) of the detected tags came from the first (mid-April) release group; only 40 tags from the second (late April) release group (Table 13). A total of 346 (43%) were detected at least once downstream of the release site. All 401 Durham Ferry tags that were detected were observed upstream of Mossdale. Only 21 Durham Ferry tags (3%) were detected in the study area at or downstream of Mossdale; no tags from the second release group were detected in the study area (Table 13). Only 1 tag was detected in the Old River route. Eight (8) of the 21 tags detected in the study area were later observed back upstream of the study area, and were not assigned to a route within the study area. This left 12 tags assigned to the San Joaquin River route, and 1 tag assigned to the Old River route for survival analysis (Table 13).

The large majority of Durham Ferry tags detected downstream of the head of Old River were detected in the San Joaquin River (Table 14). Twelve (12) were detected at Lathrop (A7), and successively fewer were detected at sites farther downriver, with only 3 tags detected at Garwood Bridge (SJG = site A8), 2 detected at Navy Drive Bridge (SJNB = site A9), 1 detected at both the San Joaquin Shipping Channel (SJS = site A10) and MacDonald Island (MAC = site A11), and none detected farther downstream (Table 14). The single tag detected in the Old River route was detected at the first site in that route, Old River East (ORE = site B1), and was not detected subsequently (Table 14). All tags detected within the study area came from the first release group. Of the 156 tags released at Durham Ferry in the second release group, 35 were detected at the receivers upstream of the release site, 12 were detected at the Durham Ferry receivers just downstream of the release site, and one was detected at both the Below Durham Ferry sites (BDF1 = site A3, BDF2 = site A4) (Table 14). The numbers of detections per site used in the Durham Ferry survival model were very similar to the overall counts of detections for sites downstream of Banta Carbona (Table 14 and Table 15).

A total of 158 Durham Ferry tags were classified as in a predator at some point during the study: 124 (19% of 643 released) from the first release group, and 34 (22%) from the second release group (Table 16). The majority of those classified as predators were first classified as such upstream of the study area; only 10 tags were classified as in a predator within the study area, all from the first release

39

group because no tags from the second release were detected within the study area. Of the 21 tags observed within the study area (i.e., at Mossdale or points downstream), 48% (10 tags) were classified as predators, and none were first classified as predators upstream of the study area. Of the 10 classified as predators within the study area, 5 were so classified upstream of the head of Old River (i.e., at sites MOS or HOR) (Table 16). Four tags were classified as predators at Lathrop or the second two of the predator removal study sites (SJL, RS5, or RS6); the single tag detected at the Old River East receiver was classified as in a predator upon departure from that site (Table 16). Approximately equal numbers of tags were first classified as in a predator upon arrival at a site (75 tags) as upon departure from the site (83 tags) (Table 16). Predator classifications on arrival were primarily due to unexpected travel time or extended regional or local residence times, and were most common at the receivers upstream of the Durham Ferry release site (Table 16). Predator classifications on departure were primarily due to long local and regional residence times, and were more common at sites between Durham Ferry and Mossdale (Table 16). When the detections classified as coming from predators were removed from the detection data, there were somewhat fewer detections available for survival analysis upstream of the study area, and no change in the number of detections available within the study area, although the counts at both the head of Old River (B0) and Lathrop (A7) declined by one tag (Table 13 - Table 15 vs. Table 17 - Table 19).

Medford Island Release A total of 491 acoustic-tagged Chinook salmon were released near Medford Island in late April and early May (Table 2, Table 13). Of these, 398 (81%) were detected at least once, including any predator-type detections. The majority of tag detections (325) occurred at the Medford Island receivers, which were located approximately 0.5 km downstream from the release site. A total of 43 tags (9% of those released) were detected upstream of Medford Island, whereas 92 tags (19%) were detected downstream of Medford Island, including detections the Old River receiver near the Old River mouth (OSJ = site B5) but excluding all other interior Delta receivers (Table 13). Of the tags detected downstream of Medford Island, 84 were detected in the San Joaquin River route, determined by detection at the San Joaquin Disappointment Slough receiver (SJD = site A13). A total of 48 tags were detected in the Interior Delta route, determined by detection at sites OSJ, MID, COL, the Highway 4 receivers, and the water export facility receivers (Table 13); COL is considered upstream of the Medford Island release site. For the survival model, 79 tags were assigned to the San Joaquin River route, and 44 were assigned to the Interior Delta route (Table 13).

40

Of the 491 tags released at Medford Island, 12 were detected as far upstream as SJS (A10), 35 were detected at MacDonald Island (A11), and 6 were detected in Turner Cut (F1) (Table 14). A total of 325 tags were detected at the Medford Island receivers. Twelve (12) tags were observed in Columbia Cut, 34 at the Middle River receivers at Middle River (MID = site C3), and 6 at OSJ (B5) (Table 14). Two tags were observed at the West Canal (WCL = site B3) receivers outside the Clifton Court Forebay; one of those two tags came from the lower San Joaquin River via Middle River (site C3) and was subsequently detected entering the Clifton Court Forebay (sites D1 and D2), whereas the other of the WCL tags came via Turner Cut and the Middle River receiver at Highway 4 (MR4 = site C2), and was not detected again after detection at WCL. No other tags were detected in the western Delta near the water facilities (Table 14). However, two other tags were detected at Middle River near Highway 4; neither of them was subsequently detected again. A total of 88 tags were detected at either Disappointment Slough (A13) or the Old River mouth receiver (B5) (Table 14). A total of 47 tags were detected at Jersey Point, and 10 detected at False River (all were also detected at Jersey Point). Thirty-five (35) tags were detected at Chipps Island (7% of the Medford Island release group), and 28 (6%) were detected at Benicia Bridge (Table 14).

When restricted to the detections used in the survival model, only one detection was used at False River (Table 15), because the other tags detected there were detected subsequently at Jersey Point. Likewise, only 298 tag detections were used for the Medford Island site (A12), because many tags were observed upstream after initial detection at Medford Island. Other numbers are similar to the total counts observed (Table 14 and Table 15).

A total of 106 Medford Island tags were classified as in a predator at some point during the study (22% of those released) (Table 16). The majority (74%) of those classified as predators were first classified as such at the Medford Island receivers, primarily due to long local or regional residence times. Of the 43 tags detected upstream of Medford Island, 16 (37%) were first classified as in a predator in that region (Table 16). Only 8 of the 92 (9%) tags observed downstream of Medford Island were first classified as a predator in that region (Table 16). None of the 8 tags first classified as in a predator downstream of Medford Island were later observed upstream of Medford Island, although one of these predator-type tags was later observed at the radial gates at the entrance to the Clifton Court Forebay. Four (4) of the 34 tags detected at the MID (site C3) receivers on Middle River were classified as a predator at that site (Table 16); none of these 4 were subsequently detected elsewhere. Of the 35 Medford Island tags observed at Chipps Island, none had been previously classified as in a predator,

41

although two were subsequently classified as in a predator at Benicia Bridge on account of long residence times. One of the 47 tags detected at Jersey Point was classified as in a predator (upon departure from that site; Table 16), but it was not later detected at Chipps Island.

When the detections classified as coming from predators were removed from the detection data, only one detection was removed from the San Joaquin River site near Disappointment Slough (SJD = A13), two detections from MID (C3) on Middle River, and two detections from Jersey Point (G1) (Table 14 vs. Table 18). Additionally, several detections were removed from sites upstream of MacDonald Island (Table 14 vs. Table 18). There were similarly small changes in the numbers of detections available for survival model (Table 15 vs. Table 19).

Tag-Survival Model and Tag-Life Adjustments Observed tag failure times ranged from 16.64 days to 44.97 days. Model fit was improved by right-censoring failure time data at 44.966 days; there were 11 tags with tag failure times > 44.966 days. Model fit comparisons using AIC to compare analyses that pooled over tag activation date and/or tag- life study resulted in selection of the pooled model (ΔAIC ≥ 15.4). Thus, a single tag survival model was fitted and used to adjust fish survival estimates for premature tag failure. The estimated mean time to failure from the pooled data was 40.0 days ( SE = 5.2 days) (Figure 10).

The complete set of detection data, including any detections that may have come from predators, contained detections that occurred after the tags began dying, although the majority of the detections at most sites occurred before many tags had died (Figure 11 and Figure 12). The sites with the latest detections were Durham Ferry Downstream, BDF2, and Banta Carbona from the Durham Ferry releases (Figure 11), and MacDonald Island, Columbia Cut, and Middle River at Middle River (site C3) from the Medford Island release (Figure 12). Some or all of these late detections may have come from predators. Without the detections classified as coming from predators, all detections from Durham Ferry releases occurred before day 10 (Figure 13), and all detections from the Medford Island release occurred before day 15 (Figure 14).

Tag-life corrections were made to survival estimates to account for premature tag failure as observed in the tag-life studies. For the full data set that included any detections that may have come from predators, all estimates of reach-specific tag survival were ≥ 0.94, out of a possible range of 0 – 1. Without the detections that were classified as coming from predators, the reach-specific tag survival

42

estimates were ≥ 0.996. Thus, there was little effect of either premature tag failure or corrections for tag failure on estimates of salmon reach survival for either release site.

Surgeon Effects The numbers of fish in each combination of release group and release site were evenly distributed across surgeon (Table 20). Additionally, for each surgeon, the number fish tagged was well- distributed across release group and site combination. A chi-squared test found no evidence of lack of independence of surgeon across release group and site ( χ 2 = 0.0025, df = 4, P = 1). The distribution of

tags detected at various detection sites was also well-distributed across surgeons and showed no

evidence of a surgeon effect on survival, route entrainment, or detection probabilities at these sites ( χ 2

= 8.745, df = 18, P = 0.9694; Table 21).

For Durham Ferry releases, estimates of cumulative survival to the Garwood Bridge receivers showed similar patterns among all surgeons downstream of the Durham Ferry Downstream receivers (DFD). Although Tagger B had higher estimated survival to site DFD, there was no difference in cumulative survival either to DFD (P=0.1398) or to the sites further downstream (P≥0.1833; Figure 15). For Medford Island releases, Tagger B had consistently lower cumulative survival to all sites, but the difference was not statistically significant (P≥0.3371; Figure 16). Analysis of variance found no effect of surgeon on individual reach survival estimates (P=0.8693). Furthermore, rank tests found no evidence of consistent differences in reach survival estimates for fish from different surgeons (P=0.3845).

Survival and Route Entrainment Probabilities Detection data of tagged Chinook salmon released at Durham Ferry in 2015 were sparse, especially in the Old River route and downstream of Lathrop in the San Joaquin River route. Model modifications necessary to fit the data are described in Statistical Methods: Parameter Estimation.

Release Group 1 For the mid-April release group, survival from the release site at Durham Ferry to the upstream ˆ  boundary of the study area at Mossdale was estimated at φAA1, 6 = 0.03 ( SE = 0.01) (Table 22). Of the 13 fish observed passing the head of Old River, all but one took the San Joaquin River route at that junction  (ψˆ A = 0.92; SE = 0.08). No tags were observed on the Rough and Ready Island receivers (RRI = site R1) at Burns Cutoff, and only two tags were observed at the Navy Drive Bridge receivers in the San Joaquin River (SJNB = site A9) near Burns Cutoff; the estimated probability of surviving from Lathrop to the Navy

43

 Drive Bridge receivers (i.e., estimate of the joint probability SSAA78) was 0.18 ( SE = 0.12). Only one fish was detected at the receivers at MacDonald Island (MAC = site A11), and none was detected on the Turner Cut junction (TCE/TCW = site F1); the estimated probability of surviving from Mossdale to the

Turner Cut junction (i.e., through the South Delta region in route A) was S A( SD) = 0.05 ( SE = 0.05),

assuming 100% detection probabilities at both MacDonald Island and Turner Cut (Table 22). Detections from the late April/early May release at Medford Island confirmed 100% detection probability at MacDonald Island. Additionally, although not modeled and estimated directly, all five Medford Island fish that were detected at Turner Cut were observed on both receivers at that site, suggesting 100% detection probability in Turner Cut. No fish from the first release at Durham Ferry was detected at ˆ either Chipps Island or Benicia Bridge ( SA = 0, assuming 100% detection at Chipps Island). Even had Chipps Island detection probabilities been lower, the point estimate of survival to Chipps Island through the San Joaquin River route would have remained 0. There was little or no difference in performance metric estimates when predator-type detections were included (Table 22 vs. Table 23; Appendix E.Table E2 vs. Table E3).

Only one fish was observed entering the Old River route at the head of Old River: ψˆ B = 0.08

( SE = 0.08;Table 22). This was too few fish with which to estimate survival from the Old River East receiver (ORE = site B1) to Chipps Island (SB ). No Durham Ferry fish were detected further in the Old

River route. Although survival to Chipps Island could not be estimated in the Old River route, the overall lack of detections at Chipps Island and Benicia Bridge yields the overall Delta survival estimate from

Mossdale to Chipps Island of S Total = 0, based on an assumption of 100% detection at Chipps Island. Even had Chipps Island detection probabilities been lower, the point estimate of total survival to Chipps Island would have remained 0.

The very low detection counts at ORE in the Old River route and SJNB and MAC in the San Joaquin River cast doubt on the detection probabilities at those sites. In particular, the single detections observed at ORE and MAC prevented estimation of detection probabilities at those sites. However, the consistently decreasing detection counts at sites leading to Mossdale and further downstream are ˆ  consistent with very low estimates of survival to Mossdale (φAA1, 6 = 0.03; SE = 0.01) and the Turner Cut

junction ( S A( SD) = 0.05; SE = 0.05).

44

Had the detection probability at ORE been 0.5 instead of the assumed value of 1.0, the estimated probability of entering the Old River route would have been 0.15 ( SE = 0.14) instead of the    ψ B = 0.08 ( SE = 0.08) using PB1 = 1. However, no additional information would be gained about survival within the Old River route or total survival through the Delta.

If the detection probabilities at both SJNB and MAC were 0.5 instead of 1.0, as estimated at SJNB using only two fish, and as assumed at MAC, then the reach-specific probability of survival from  Garwood Bridge to the Navy Drive Bridge receivers (SA8 ) would have been estimated at 1.00 ( SE   =0.47) instead of S A8 = 0.67 ( SE =0.27) as observed using PA9 = 1 and PA11 = 1. Likewise, the South

Delta survival in route A (i.e., Mossdale to MacDonald Island: SA( SD) ) would have been estimated at 0.11

( SE =0.11) instead of the observed estimate of S A( SD) = 0.05 ( SE =0.05). The estimate of the overall

survival to Chipps Island would have remained at 0, however.

Release Group 2 The late April/early May release group from Durham Ferry yielded few detections upstream of Banta Carbona, and no detections from Banta Carbona downstream (Table 15). The estimated probability of surviving from the release site to the Durham Ferry Downstream receivers (DFD = site A2)   was φ AA1, 2 = 0.04 ( SE = 0.02; Appendix E.: Table E2). It appears that more fish turned upstream from

   the release site to site DFU (A0) than downstream to DFD: φ AA1, 0 = 0.12 ( SE = 0.02). Both φ AA1, 2 and

   φ AA1, 0 were somewhat higher when predator-type detections were included: φ AA1, 2 = 0.08 ( SE = 0.02)

  and φ AA1, 0 =0.18 ( SE = 0.03) (Appendix E.: Table E3). However, the majority of fish released at Durham

Ferry in late April/early May were never detected (Table 14 and Table 18). It was not possible to estimate reach-specific survival probabilities downstream of DFD for this release group. However, under the assumption of 100% detection probability at Mossdale (which is consistent with data from the first release group), the probability of surviving from the Durham Ferry release site to Mossdale was ˆ  estimated as φAA1, 6 = 0 ( SE = 0), with or without predator-type detections (Table 22 and Table 23).

For fish released at Medford Island, most fish that were detected went past the San Joaquin ˆ  Receiver at Disappointment Slough (SJD = site A13; φMF, A 13 = 0.17; SE = 0.02); few were observed

45

ˆ  ˆ  turning upstream from the release site (φMF, A 11 = 0.05; SE = 0.01 and φMF,2 F = 0.02; SE = 0.01), and ˆ  even fewer passing the receivers in Old River near its mouth (OSJ = site B5; φMF,5 B = 0.01; SE < 0.01)

(Appendix E. : Table E2). None of the fish that turned upstream from the release site were subsequently observed at Chipps Island or Benicia Bridge: S AG11, 2 = 0 ( SE = 0), and S FG2, 2 = 0 ( SE = 0) (Table 22). Of the fish that went past Disappointment Slough, under half were observed at Chipps Island or Benicia

Bridge: S AG13, 2 = 0.42 ( SE = 0.06); from OSJ, S BG5, 2 = 0.32 ( SE = 0.18) (Table 22). There was no statistical difference in survival to Chipps Island via the two routes indicated by detection at SJD and OSJ, respectively (P=0.6239). Detection probability at Chipps Island was estimated at 0.93 ( SE = 0.05). ˆ  ˆ Transition probabilities to Jersey Point were estimated as φAG13, 1 = 0.55 ( SE = 0.06) from SJD, and φBG5, 1

= 0.42 ( SE = 0.23) from OSJ; the estimated transition probability from Jersey Point to Chipps Island was ˆ  φGG1, 2 = 0.77 ( SE = 0.07) (Table D2). Very similar estimates were observed when predator-type detections were included (Table 22 vs Table 23; Appendix E.: Table E2 vs Table E3). Detection data for the Medford Island release were not sparse, so no sensitivity analysis was performed.

Pooled Release Groups When pooled across release groups, estimates of model parameters were typically identical or nearly identical to those from the analyses of the individual release groups (Table 22, Table 23 and Appendix E.: Table E2 and Table E3). The result was anticipated because of the small Durham Ferry release in the second release group, which added few detections at sites in the Durham Ferry survival model.

Comparison between Release Groups Parameter estimates were significantly (family-wise α=0.10) different between release groups for parameters φAA1, 2 , φAA1, 6 , and φAA1, 0 (Table 24). For both φAA1, 2 and φAA1, 6 , the estimates for the first release group was higher than for the second release group (Table 24). San Joaquin River flows at Vernalis (Figure 17) and exports (Figure 18) were higher, whereas water temperatures were lower (Figure 19), for the first release group relative to the second release group.

Comparison between years Buchanan (2017) conducted a multi-year analysis of south Delta Chinook salmon survival data for study years 2010-2013. In Buchanan et al 2018, the multi-year analysis was updated to include the

46

2014 results where possible. In 2015, only 21 tags were detected at MOS, 11 at SJL, and 1 at ORE, which was too few tags to either assess model fit for 2015, or to update the models for 2015. Thus, no update was performed for 2015.

Travel Time For the mid-April release group, travel time from the Durham Ferry release site downstream to

Mossdale averaged approximately 1.4 days ( SE =0.22 days; Table 25a). The largest observed travel time to Mossdale was 4.5 days, omitting predator-type detections; when predator-type detections were included, the average travel time to Mossdale was 1.46 days ( SE =0.25 days) (Table 25b), and the

maximum observed was 16.8 days. Average travel time from Durham Ferry to Lathrop was 1.53 ( SE =0.36) days without predator-type detections (Table 25a), and just under 2 days with predator-type detections (Table 16b) for the mid-April release group; the maximum observed travel time to Lathrop was 4.0 days and 6.2 days with and without predator-type detections, respectively. In the San Joaquin River route, average travel times were successively longer to downstream sites, and the single tag detected at MacDonald Island was detected approximately 4 days after release at Durham Ferry (Table 25a, Table 25b). The single tag observed in the Old River route was detected at the ORE receivers approximately 4.9 days after release at Durham Ferry (Table 25a), although it had a second detection at ORE nearly 2.5 days later, when it was classified as predator (Table 25b).

Travel times from Durham Ferry for the late April/early-May release group were observed only to Durham Ferry Upstream (DFU), Durham Ferry Downstream (DFD), and the two “Below Durham Ferry” sites (BDF1 and BDF2), and only a single tag was observed at the latter two sites. Average travel time to the DFU site upstream of the release site was 0.25 days ( SE =0.04 days) without predator-type detections (Table 25a), and 0.49 ( SE =0.09) days with predator-type detections (Table 25b). Average

travel time to the DFD site just downstream of Durham Ferry was approximately 1.27 days ( SE =0.16 days) without predator-type detections, and 0.97 ( SE =0.17) days with predator-type detections (Table 25a, Table 25b); the difference was caused by the removal of six tags detected at DFD but previously classified as predators at DFU. The single tag detected at BDF1 and BDF2 arrived at both sites approximately 4 days after release (Table 25a). For the three sites downstream of Durham Ferry that had detections from both release groups, the second release group had longer travel times than the first release group (Table 25a, Table 25b).

47

Fish released at Medford Island took, on average, between approximately 0.5 and 0.7 days to move either downstream to the Disappointment Slough receivers (SJD) or Old River receivers (OSJ) or upstream to MacDonald Island (MAC); average travel time to Columbia Cut (COL) was slightly longer, at approximately 1 day (Table 25a). Travel times that included predator-type detections were slightly longer, on average (Table 25b). The average travel time from the Medford Island release site to Jersey

Point was approximately 2 days (1.95, SE =0.15) (maximum = 5 days), omitting the two tags classified as predators that were detected at Jersey Point (Table 25a); the two predator-type detections at Jersey Point were observed there 6 and 13 days after release at Medford Island, respectively. Average travel time to Chipps Island from the Medford Island release site was 3.64 ( SE =0.23) days, with or without predator-type detections (Table 25a, Table 25b); the maximum observed travel time to Chipps Island was 8.9 days. Average travel time to Benicia Bridge was 4.6 ( SE =0.30) days (Table 25a), with a maximum of 11.8 days.

Average travel times through reaches for Durham Ferry tags classified as being in juvenile Chinook salmon ranged from 0.06 days (approximately 1.5 hours) from Garwood Bridge (SJG) to Navy Drive Bridge (SJNB) (approximately 2.5 km), to 2.89 days from Mossdale (MOS) to the Old River East receivers (ORE) for the single tag observed making that transition (approximately 4.8 km) (Table 26a). Travel times from the San Joaquin River receivers near Lathrop (SJL) to Garwood Bridge (SJG) ranged from 0.7 days to 1.7 days (average = 1.1 days; ̴18 rkm) (Table 26a). Including the predator-type detections had little or no effect on average travel time through reaches, except for the reach from Mossdale to ORE, which had only one observation: including predators had the effect of increasing the travel time from 2.89 days to 5.36 days (Table 26b).

Average travel times through reaches for Medford Island tags classified as being in juvenile Chinook salmon ranged from 0.02 days (approximately 30 minutes) from the release site to the Medford Island receivers (MFE/MFW), approximately 0.5 km, to 3.95 days from the Disappointment Slough receivers (SJD) to False River (FRE/FRW; ̴15 rkm) (Table 26a). Only one tag was observed moving from SJD to FRE/FRW. Travel time from the SJD receivers to Jersey Point (JPT/JPE/JPW) ranged from 0.25 days to 4.11 days, and averaged 0.93 days ( SE =0.14 days) (Table 26a). Travel time from Jersey Point to

Chipps Island (MAT/MAE/MAW) ranged from 0.7 days to 4.7 days, and averaged 1.40 days ( SE =0.10 days) ( ̴25 rkm; Table 26a). Travel time from Chipps Island to Benicia Bridge averaged 0.89 days ( SE

48

=0.12 days), and ranged from 0.24 days to 2.95 days ( ̴18 rkm; Table 26a). There was little or no difference in average travel times from including detections from predator-type detections (Table 26b).

Mobile Monitoring There were a total of 316 individual tag codes detected by the 2015 South Delta Chinook Salmon Survival Study mobile monitoring surveys. Survey 1 detected 15 tags downstream of the Durham Ferry release site (see Durham Ferry Detail in Figure 20). Survey 2 detected a total of 52 codes—40 downstream and 12 upstream of the Durham Ferry release site (Figure 20). Three of the downstream codes detected during survey 2 were also detected on survey 1. Survey 3 detected a total of 105 tag codes (Figure 21). Forty-five of the detections during survey 3 occurred between the mouth of the Stanislaus River and the Durham Ferry release site and have corresponding GPS coordinates. The remaining 60 detections occurred after the equipment failure and passing the Durham Ferry release site, and do not have corresponding GPS coordinates. During survey 3 there were also 10 second detections: three codes that had been previously detected during survey 1 and seven codes that were detected during survey 2. Two of these detections had corresponding GPS codes, and both were detected approximately 1 km upstream of the release site. There was also one third detection that was detected in both surveys 1 and 2. Though this detection does not have corresponding GPS coordinates, it was detected approximately 10 m after passing the Durham Ferry release site. Survey 4 detected 21 codes, one that was also detected in survey 1 and one that was also detected in survey 2 (Figure 22). The second detections during survey 4 were 14.7 and 17.0 km downstream from the Durham Ferry release site. All codes detected during surveys 1–4 correspond to fish released at the Durham Ferry release site. All the codes detected during survey 5 correspond to fish that were released at the Medford Island release site. Survey 5 detected 139 tag codes, none of which were detected in previous surveys (Figure 23). However, within survey 5, 48 tags were detected a second time, and five tags were detected a third time. All the codes that were detected three times during survey 5 were detected within 400 m of the Medford Island release site. Eight-three percent of the second or third detections of each code were within 400 m of the first detections, and the second detections of the remaining 8 codes were between 450–1200 m of the first detections.

Discussion Survival in 2015 No detections of Durham Ferry tags were observed at Chipps Island in 2015. Delta survival was estimated at 0 for the first release group, under the assumption of 100% survival. The “Rule of Three”

49

(Van Belle, 2008, p. 49) gives a 95% upper bound on the joint probability of fish and tag survival and detection at Chipps Island of 0.1328. Assuming an average travel time through the Delta of 8 days (based on the averaged Delta travel time in 2012 from tag activation), the 95% upper bound on fish survival through the Delta is 0.1337 if Chipps Island detection probability was 1.0, or 0.1437 if the detection probability was 0.93 (the estimate for the Medford Island release group). The high upper bound reflects the low precision incurred by the low survival probability to Mossdale for the first release group (0.03) and the resulting low effective sample size with which to estimate survival to Chipps Island. The fact that survival was very low to Mossdale suggests that survival from Mossdale to Chipps Island was in fact close to 0, rather than the estimated upper bound.

For the second release group at Durham Ferry, no tags were detected within the study area, which was bounded upstream by the Mossdale receivers. Assuming an average travel time of 4 days from release at Durham Ferry to Mossdale (based on the travel time of the first release group), the “Rule of Three” gives a 95% upper bound on the probability of fish survival from release to Mossdale of 0.0191 if the Mossdale detection probability was 1.0 (as estimated from the first release group), or 0.0382 if the Mossdale detection probability was as low as 0.5. Either way, survival to Mossdale was very low for the second release group.

We were able to detect family-wise differences (α=0.10) in survival in some reaches (φAA1, 2 ,

φAA1, 6 , and φAA1, 0 ) between release groups. For both φAA1, 2 and φAA1, 6 , the estimates for the first release group were higher than for the second release group (Table 24). This result is expected as San Joaquin River flows were higher (Figure 17) and water temperatures were lower (Figure 19) for the first release group relative to the second release group. The transition probability of going from Durham

Ferry to upstream of Durham Ferry (φA1,A0 ) was higher for the second release group relative to the first (Table 24). Again this result may be in response to the higher water temperatures for the second release relative to the first, as the fish may have been seeking temperature refugia upstream.

 Survival from Medford Island to Chipps Island (SMF,G2) was estimated at only 0.08 ( SE =0.01;Table 22) for a distance of approximately 46 km. While low, the survival estimate includes the probability of survival immediately after release, when fish may be disoriented and have higher vulnerability to predation. Survival to Chipps Island from Disappointment Slough (approximately 8 km

downstream of Medford Island) was higher (0.42, SE =0.06), as was survival from the OSJ receivers in

50

Old River near its downstream junction with the San Joaquin River (0.32, SE =0.18) (Table 22). These survival probability estimates suggest a survival rate per km of 0.969 to 0.978, which translates to total survival probability of 0.06 to 0.14 through 91 km, the approximate travel distance from Mossdale to Chipps Island through the San Joaquin River. These estimates of survival rate per km are comparable to estimates from this region in previous years (Table 3-1 in Salmon Scoping Team 2017). Travel time through these regions in 2015 ranged from 1.4 days to 8.2 days (average = 2.74 days) from Disappointment Slough to Chipps Island, and from 1.7 days to 1.8 days (average = 1.74 days) from OSJ to Chipps Island; combined with survival estimates, these travel times suggest daily survival rates of 0.51 to 0.90 (mean = 0.52 to 0.73). This was considerably higher than the daily survival rate of 0.08 estimated for the reach from Durham Ferry to Mossdale for the first release group, using the average travel time of 1.4 days. The difference in survival between the two release sites may represent differences in temperature exposure (higher temperatures at Durham Ferry) or flow (higher flows at the Medford Island release site), possibly combined with other regional differences. The Disappointment Slough (SJD) and OSJ detection sites were either new in 2015 (SJD) or not usable for survival modeling in previous years (OSJ), and detections in this region (e.g., at Medford Island) have been sparse in previous years. Thus, it is not apparent if the differences observed between regions and release sites in 2015 were unique to this drought year, or if similar regional differences in survival exist in other years with less extreme flow and temperature conditions. Collecting more survival data in the region from Medford Island to Chipps Island in the future will help shed light on this question.

Mobile Monitoring Mobile monitoring can only be used to verify the presence of a tag at a specific location, but it cannot be used to verify the absence of a tag. Study tags present in the survey area during mobile monitoring may not have been emitting a code, or the codes they were emitting may not have been detected by the hydrophone. The batteries in the pooled tag life study indicated an estimated mean time to tag failure of 40.0 days, thus our mobile monitoring should have picked up the most of the tags if they were in the areas of our mobile monitoring, but some tags failed earlier (Figure 10) and those we would not have detected during our surveys. Even if a tag still had a functional battery, it is possible that tags were missed during a mobile monitoring survey, because tags can become buried in the sand and not detected by the hydrophone.

Mobile monitoring can also only provide information regarding the specific areas where surveys were conducted. For example, the CVP trashracks and Clifton Court Forebay have been shown to be

51

places of high tag density in the past (SJRGA 2013, Buchanan et al. 2015). However, since these locations were not included in the 2015 mobile monitoring surveys, no information was collected regarding post-study tag presence around these facilities (though data were provided by stationary receivers at these two specific locations).

Even when mobile monitoring data provide a positive confirmation of the location of a tag code, what this implies regarding the fate of the corresponding study fish is open to interpretation. It is possible that tag detections are actually false positives. Additionally, a detected tag may not be in a live study fish but lying on the bottom of the river, either from being shed from a live study fish or from being defecated by a predator. This scenario is likely if a tag was detected near the same location across multiple mobile monitoring dates. If a tag was detected at different points across the same or different mobile monitoring dates, this movement could indicate that the transmitter was in the digestive tract of a live predator or in a study fish that remained in the system longer than the expected time frame. It could also indicate some movement of the tag itself along the river bottom, depending on how far the distance between detections. In the future, visiting the same locations repeatedly over multiple surveys could provide more information regarding the last location a tag was detected. The more detections provided for a single tag may make interpretations regarding the ultimate fate of the tag less subjective, especially if the tag is shown to remain in the same location over multiple surveys.

Despite its limitations, mobile monitoring data can provide useful information, especially when used in conjunction with receiver detection data. It can be used to give more detailed detection data between receivers, which can be used to better pinpoint potential areas of high mortality. We did observe tags upstream of Durham Ferry (Figure 21), between Durham Ferry and Mossdale (Figure 22) as well as near the Medford Island release site (Figure 23) in mobile monitoring surveys, confirming the high mortality estimated in these reaches from the stationary receivers and suggesting specific areas of that mortality within these reaches in 2015. Summary Project Objectives Of the five specific project objectives, most could not be met in 2015 due to the poor survival ˆ  between Durham Ferry and Mossdale for both release groups: φAA1, 6 = 0.03, ( SE = 0.01) for the first ˆ  release group, and φAA1, 6 = 0 ( SE = 0) (assuming 100% probability of detection at Mossdale) for the second release group (Table 22). For the first objective (a), estimating survival from Durham Ferry and

52

Mossale through the Delta to Jersey Point to Chipps Island, was met for the first release group, if we assume 100% detection at Jersey Point and Chipps Island, but survival to Chipps Island from Mossdale was estimated to be 0 ( S Total ; Table 22) and no detections were made from the Durham Ferry releases at Jersey Point (Table 14). For the second release group, the objective was not met, as no Durham Ferry released fish were detected downstream of Banta Carbona, so survival could not be estimated to Jersey Point or Chipps Island. We were able to estimate the survival for tagged fish taking the San Joaquin ˆ route for the first release group ( SA = 0), but it could not be estimated for the second release group or for the Old River route (objective b).

We were able to identify reach-specific survival or transition probabilities in some locations of the Delta (objective c), primarily downstream of Medford Island, where detections at receiver sites were greater. Survival from Medford Island to Chipps Island was estimated at only 0.08 ( SE =0.01;Table 22), but that estimate includes any mortality associated with the newly released fish being temporarily disorientated after release at Medford Island. Transition probabilities from SJD and OSJ to Jersey Point ˆ  ˆ  were estimated as φAG13, 1 = 0.55 ( SE = 0.06) and φBG5, 1 = 0.42 ( SE = 0.23), respectively. The transition ˆ  probability from Jersey Point to Chipps Island (φGG1, 2 ) was 0.77 ( SE = 0.07 (Appendix E.: Table E2), and the transition probability from Chipps Island to Benicia Bridge (λ) was estimated at 0.74 ( SE = 0.07). It was helpful to get robust estimates of survival in these lower reaches of the Delta, as in the past we have had too few fish arriving at these downstream locations to get meaningful estimates of survival in these reaches.

We could not estimate the proportion of tagged fish entering Turner Cut or the factors influencing the proportion of tagged fish entering Turner Cut (Objective d), as none of the fish from the Durham Ferry releases were detected in Turner Cut. Only one tagged fish was detected entering Old River from the head of Old River, consistent with the low number of tagged fish arriving at Mossdale and presence of the HORB, so we were not able to estimate the proportion of tagged fish entering Old River either (Objective d). We were also unable to use the 2015 results in conjunction with past modeling results to assess the role and influence of flow and exports, a barrier at the head of Old River, mortality by reach, and the proportion of fish entering Turner Cut, on survival through the Delta (Objective e), due to the sparse detections at most locations.

53

We were able to evaluate the support for hypotheses relative to survival in the downstream reaches of the Delta. Consistent with our hypothesis (a), survival to Chipps Island was low for fish entering Columbia Cut (survival estimate = 0 for the fish released at Medford Island that entered Columbia Cut). Only six and 12 tagged fish from the Medford Island release were detected in Turner and Columbia cuts, respectively, and none of them was detected later at Chipps Island or Benicia, suggesting survival for fish diverted through these channels is low. However, this conclusion is based on a small number of fish entering these pathways so our conclusion is not robust. But these results are consistent with other years where tagged salmon observed entering Turner Cut were not detected at Chipps Island (SJRGA 2011, 2013).

No tagged fish from the Medford Island releases that entered Turner Cut or Columbia Cut were detected at the fish facilities (Clifton Court Forebay or at the CVP), which supports our hypothesis (b) that juvenile Chinook salmon diverted into these cuts are unlikely to survive to the fish facilities either. But again, there were relatively few tags that entered Turner or Columbia cuts of which to base this conclusion on, and transitions to the fish facilities were not modeled for Medford Island fish. However, there were several examples of tagged fish from the Medford release being diverted upstream towards the fish facilities. One tagged fish that entered Turner Cut reached the West Canal receiver site (WCL = site B3), which is near Clifton Court Forebay (Figure 4). The tag was not detected again after it was detected at WCL. Another 34 tags from the Medford release were detected in the Interior Delta at Middle River (MID; site C3, Figure 4,Table 14). Four of the 34 tags were classified as being in a predator at that site and not detected again, and one other of the 34 was later detected in Clifton Court Forebay (sites D1 and D2). No tagged fish were detected at the CVP (Table 14) but four tags were detected at the two Highway 4 receivers (Table 14). None of the 34 fish detected at MID (site: C3; Table 14) were detected at Chipps Island (φC3,G2; Appendix E.: Tables E2 and E3). Lastly, while there was no statistical difference in survival to Chipps Island via the two routes indicated by detection at SJD and OSJ, respectively (P=0.6239), the point estimate of the transition probability to Jersey Point was somewhat ˆ  less for those from the interior Delta route, OSJ ( φBG5, 1 = 0.42; SE = 0.23) than for the mainstem San ˆ  Joaquin River route, SJD ( φAG13, 1 = 0.55; SE = 0.06) (Appendix E.: Table E2).

Several tagged fish released at Medford Island were detected at Threemile Slough (16; Table 14), indicating that fish from the San Joaquin River can enter the Sacramento River at that point and potentially bypass Jersey Point.

54

Counter to Perry et al (2018) we found survival to be highest in the downstream reaches of the San Joaquin Delta (i.e. Jersey Point to Chipps Island) in 2015, with survival from Disappointment Slough or Old River at its mouth to Jersey Point somewhat lower. The lowest survival in 2015 was between Durham Ferry to Mossdale. This is much lower than in previous years, suggesting poor migration conditions for Durham Ferry released fish during both the releases there. Several tags were detected going upstream from the release site and may have been a result of fish trying to find cooler waters upstream, or a predator that passed through the predator filter. Flows were extremely low and water temperatures were near lethal levels (24oC; 76o F) for Chinook salmon (Brett et al 1982) (Figure 17, Figure 19) for the second release group (transported on April 28, and released April 29 and April 30). We continue to advocate for additional salmon survival studies and releases for the estimation of survival throughout a variety of flow and water temperature conditions. However, it appears that flow and water temperature conditions in 2015 were not sufficient to support Chinook salmon survival in the lower San Joaquin River and upper reaches of the south Delta. For survival estimates to be made in the Delta, survival between Durham Ferry and Mossdale must be great enough to allow a sufficient number of fish to reach the Delta. Delta survival estimates for both wild and hatchery stocks are important monitoring tools for further understanding of the factors affecting survival in and through the Delta.

55

References Brett, J.R., W.Clarke, and J.E. Shelbourn. 1982. Experiments on thermal requirements for growth and food conversion efficiency of juvenile chinook salmon, Oncorhynchus tshawytscha. Canadian Technical Report of Fisheries and Aquatic Sciences No. 1127.

Buchanan, R. P. Brandes, J. Ingram, M. Marshall, K. Nichols, D. LaPlante, D. Barnard, K.Towne and J.A. Israel (2018). 2014 South Delta Chinook Salmon Survival Study. Ed: P. Brandes, U.S. Fish and Wildlife Service. 217 pp.

Buchanan, R. (2017). Multivariate San Joaquin River Chinook Salmon survival investigation, 2010 – 2013. Prepared for U.S. Bureau of Reclamation, Sacramento, CA. October 6, 2017. 143 pp.

Buchanan, R., P. Brandes, M. Marshall, J. S. Foott, J. Ingram, D. LaPlante, and J. Israel (2015). 2012 South Delta Chinook Salmon Survival Study. Ed: P. Brandes, U.S. Fish and Wildlife Service.

Buchanan, R., P. Brandes, M. Marshall, K. Nichols, J. Ingram, D. LaPlante, and J. Israel (2016). 2013 South Delta Chinook Salmon Survival Study. Ed: P. Brandes, U.S. Fish and Wildlife Service.

Burnham, K. P., and D. R. Anderson (2002). Model selection and multimodel inference: A practical information-theoretic approach. 2nd edition. Springer. New York, NY. 488 pp.

Lady, J. M., and J. R. Skalski (2009). USER 4: User-Specified Estimation Routine. School of Aquatic and Fishery Sciences. University of Washington. Available from http://www.cbr.washington.edu/paramest/user/.

Li, T., and J. J. Anderson (2009). The Vitality model: A Way to understand population survival and demographic heterogeneity. Theoretical Population Biology 76: 118-131.

Liedtke, T.L., J.W. Beeman L.P. Gee (2012). A Standard Operating Procedure for the Surgical Implantation of Transmitters in Juvenile Salmonids. U.S. Geological Survey Open-File Report 2012-1267, 50 pp.

McCullagh, P., and J. Nelder (1989). Generalized linear models. 2nd edition. Chapman and Hall, London.

Perry, R. W. , A.C. Pope, J.G. Romine, P.L. Brandes, J.R. Burau, A.R. Blake, A.J. Ammann, and C.J. Michel. (2018). Flow-mediated effects on travel time, routing, and survival of juvenile Chinook salmon in a spatially complex, tidally forced river delta. Can. J. Fish. Aquat. Sci. https://doi.org/10.1139/cjfas-2017-0310

Peven, C. A. Giorgi, J. Skalski, M. Langeslay, A. Grassell, S. Smith, T. Counihan, R. Perry and S. Bickford (2005). Guidelines and Recommended Protocols For Conducting, Analyzing and Reporting Juvenile Salmonid Survival Studies in the Columbia River Basin. Final Draft. January 2005. 88 pp.

Salmon Scoping Team (2017). Effects of Water Project Operations on Juvenile Salmonid Migration and Survival in the South Delta. Volume 1: Findings and Recommendations. Prepared for Collaborative Adaptive Management Team. January 2017. 140 pp.

56

San Joaquin River Group Authority (2010). 2009 Annual Technical Report: On Implementation and Monitoring of the San Joaquin River Agreement and the Vernalis Adaptive Management Plan (VAMP). Prepared for the California Water Resources Control Board.

San Joaquin River Group Authority (2011). 2010 Annual Technical Report: On Implementation and Monitoring of the San Joaquin River Agreement and the Vernalis Adaptive Management Plan (VAMP). Prepared for the California Water Resources Control Board.

San Joaquin River Group Authority (2013). 2011 Annual Technical Report: On Implementation and Monitoring of the San Joaquin River Agreement and the Vernalis Adaptive Management Plan (VAMP). Prepared for the California Water Resources Control Board.

Seber, G. A. F. (2002). The estimation of animal abundance. Second edition. Blackburn Press, Caldwell, New Jersey.

Sokal, R. R., and Rohlf, F. J. (1995). Biometry, 3rd ed. W.H. Freeman and Co., New York, NY, USA.

Townsend, R. L., J. R. Skalski, P. Dillingham, and T. W. Steig (2006). Correcting Bias in Survival Estimation Resulting from Tag Failure in Acoustic and Radiotelemetry Studies. Journal of Agricultural, Biological, and Environmental Statistics 11: 183-196.

Van Belle, G. (2008). Statistical Rules of Thumb, 2nd ed. John Wiley & Sons, Inc., Hoboken, NJ, USA.

Vogel, D. A. (2010). Evaluation of acoustic-tagged juvenile Chinook salmon movements in the Sacramento-San Joaquin delta during the 2009 Vernalis Adaptive Management Program. Technical Report for San Joaquin River Group Authority. 72 p. Available http://www.sjrg.org/technicalreport/ (accessed 13 December 2011).

Vogel, D. A. (2011). Evaluation of acoustic-tagged juvenile Chinook salmon and predatory fish movements in the Sacramento-San Joaquin Delta during the 2010 Vernalis Adaptive Management Program. Technical report for San Joaquin River Group Authority. Available http://www.sjrg.org/technicalreport/ (accessed 13 December 2011).

57

Acknowledgements Funding for the project came from the Central Valley Project Improvement Act (CVPIA), with in- kind contributions from National Marine Fisheries Service (NOAA-Fisheries) and Department of Water Resources. Several individuals from a variety of agencies made this project possible. Those agencies and individuals who participated in the tagging and release components of the project were—USFWS: Brett Anderson, Judith Barkstedt, Karen Bobier, Jeffrey Cullen, Dustin Dinh, David Dominguez, Duncan Fry, Christopher Hart, Jack Ingram, Carlie Jackson, Katherine Jardine, Roxanne Kessler, David LaPlante, Jerrica Lewis, Tim Matt, Mike Marshall, Greg Nelson, Elizabeth Reynolds, Kandi Vargas, Phil Voong, Nick Wagner, William (Alex) Woolen; California Department of Water Resources: Rachel August; and USBR: Raymond Bark. USGS provided training for the surgeons (Theresa [Marty] Liedtke). USGS also installed, maintained and retrieved the receiver array: Chris Vallee, Norbert VanderBranden and Jon Burau, and pre-processed the data: Mike Simpson.

58

Figures

Figure 1: Tag being activated in VEMCO tag activator. Photo Credit: Jake Osborne/USFWS

(a) (b)

Figure 2: The transmitter validation station (a) at the Mokelumne River Hatchery during the 2015 Chinook Salmon Survival Study consisted of two VEMCO High Residency Receiver prototypes (circled), each attached to a 180 kHz hydrophone. The hydrophones were placed into recovery buckets to verify the tag codes emitted by tagged steelhead while fish recovered from surgery (b). Photo credits: USFWS.

59

Figure 3: Two of the smaller transport tanks mounted on a flatbed truck for transporting juvenile salmon in 2015. Photo credit: USFWS

60

Figure 4: Location of release sites and acoustic receivers used in the 2015 Chinook tagging study, with site code names (3- or 4-letter code) and model code (letter and number string). Site A1 is the release site at Durham Ferry. Site MF is the release site at Medford Island

61

Figure 5: Putting perforated holding can into “sleeve” prior to moving downstream for release. Photo credit/ Pat Brandes, USFWS

Figure 6: Moving perforated holding cans in “sleeves” downstream to release site at Durham Ferry Photo credit: Pat Brandes/USFWS

62

Release site

Holding site

Figure 7: Release location in relationship to holding cans at Durham Ferry for Chinook salmon in 2015. Photo credit: Pat Brandes/USFWS

63

Figure 8. Schematic of full 2015 release-recapture model for Durham Ferry releases, including the estimable parameters. Single lines denote single-array or redundant double-line telemetry stations, and double or triple lines denote dual-array or triple-array telemetry stations, respectively. Names of telemetry stations correspond to site labels in Figure 4.

64

Figure 9. Schematic of full 2015 release-recapture model for Medford Island releases, including the estimable parameters. Single lines denote single-array or redundant double-line telemetry stations, and double or triple lines denote dual-array or triple-array telemetry stations, respectively. Names of telemetry stations correspond to site labels in Figure 4.

65

Figure 10. Observed tag failure times from the 2015 tag-life study of VEMCO V4 tags, and fitted four-parameter vitality curve.

Figure 11. Four-parameter vitality survival curve, and the cumulative arrival timing of acoustic-tagged juvenile Chinook salmon from the Durham Ferry releases in 2015, including detections that may have come from predators.

66

Figure 12. Four-parameter vitality survival curve, and the cumulative arrival timing of acoustic-tagged juvenile Chinook salmon from the Medford Island release in 2015, including detections that may have come from predators.

Figure 13. Four-parameter vitality survival curve, and the cumulative arrival timing of acoustic-tagged juvenile Chinook salmon from the Durham Ferry releases in 2015, excluding detections that may have come from predators.

67

Figure 14. Four-parameter vitality survival curve, and the cumulative arrival timing of acoustic-tagged juvenile Chinook salmon from the Medford Island release in 2015, excluding detections that may have come from predators.

Figure 15. Cumulative survival from release at Durham Ferry to various points along the San Joaquin River route, by surgeon (tagger). Error bars are 95% confidence intervals.

68

Figure 16. Cumulative survival from release at Medford Island to various points along the route to Chipps Island, by surgeon (tagger). Cumulative survival to Benicia Bridge (BBR) is joint probability of survival to and detection at BBR. Error bars are 95% confidence intervals.

69

Figure 17: San Joaquin River flows at Vernalis (in cfs), for the first and second release groups of Chinook salmon in 2015.

Figure 18: Central Valley Project (CVP), State Water Project (SWP) and combined exports for the first and second release groups of Chinook salmon in 2015.

70

Figure 19: Water temperatures on the San Joaquin River at Lathrop in degrees C, for the first and second release groups of Chinook salmon in 2015.

71

150 km

Chipps Island Frank’s Tract Medford Island Release Site

5

Durham Ferry Detail

San Joaquin River Durham Ferry Release Site SWP CVP 1 2 Head of Old Mossdale Trawl River 4 0.5 km

Durham Ferry Release Site Two Rivers 3 Marina

Figure 20. The area along the San Joaquin River covered by five mobile monitoring surveys during the 2015 South Delta Chinook Salmon Survival Study. Survey 1 (green) was conducted on April 17, 2015, between the Durham Ferry release site and approximately 1 km downstream of the release site. Survey 2 (orange) was conducted on April 21, between approximately 1 km upstream and 1 km downstream from the Durham Ferry release site. Survey 3 (white) was conducted on May 8, began at the mouth of the Stanislaus River (at the Two Rivers Marina), and followed the San Joaquin River approximately 17.5 km downstream. Survey 4 (blue) was conducted on May 15 between the end of the third survey and the head of Old River, for about 13.5 km downstream. Survey 5 (yellow) was conducted on May 21, started approximately 1 km upstream from the Medford Island release site, went 10.6 km downstream, and ended midway into the eastern side of Frank’s Tract.

72

San Joaquin River

Durham Ferry Release Site

First detection Two Rivers Second detection Marina

Third detection Stanislaus River

Figure 21. Tag detections from the 2015 South Delta Chinook Salmon Survival Study mobile monitoring survey 3. Sixty detections were recorded past the Durham Ferry release site, but no GPS coordinates were obtained (purple track). The location of the third detection is approximate.

73

Head of Old Mossdale Trawl River

First detection Second detection

Figure 22. Tag detections from the 2015 South Delta Chinook Salmon Survival Study mobile monitoring survey 4.

74

Medford Island Release Site

Frank’s Tract

First detection Second detection Third detection

Figure 23. Tag detections from the 2015 South Delta Chinook Salmon Survival Study mobile monitoring survey 5.

75

Tables

76

Table 1. The tagging, transport, holding date and times, and numbers of Chinook salmon released during the 2015 South Delta Survival Study. Fish were released over a 24 hour period after being held for a minimum of 24 hours. The total number of mortalities encountered after transport, pre-release, or when assessed after 48 hours (dummy tags) are entered in parentheses Dummy- Release A Release B Release C Release D tagged Total fish Start Tagging Release Transport Transport Released tagged, Holding Date Site Tank Date; Time (A+B+C+D Date; Number Date; Number Date; Number Date; Number held and Date; Time +E+F+G+H) Time Released Time Released Time Released time Released assessed after 48 hours 4/15; 1 45 3 4/14/15; 4/14/15; 1457 1223-1350 1443 4/15; 4/14/2015 DF 2 161 (1) 36 3 2058 4/14/15; 4/14/15; 4/15; 4/16; 4/16; 3 9 45 26 (1) 9 1617-1720 1758 2058 0256 0901 4/16; 1 45 3 4/15/15; 4/15/15; 1458 1200-1313 1358 4/16; 4/15/2015 DF 2 162 36 3 2057 4/15/15; 4/15/15; 4/16; 4/17; 4/17; 3 9 30 42 9 1511-1615 1615 2057 0255 0900 4/17; 1 1505, 45 3 4/16/15; 4/16/15; 1506 1135-1252 1335 4/16/2015 DF 162 4/17; 4/18; 2 30 6 3 2105 0256 4/16/15; 4/16/15; 4/18; 4/18; 3 39 42 9 1445-1547 1625 0256 0856 4/18; 4/18; 1 29 (1) 14 (1) 9** 4/17/15; 4/17/15; 1458 2056 1145-1302 1347 4/18; 4/18; 4/17/2015 DF 2 160 (2) 30 6 12** 2056 0255 4/17/15; 4/17/15; 4/19; 4/19; 3 39 42 9** 1544-1648 1725 0255 0903 4/29; 3 (0 or 1 45 4/28/15; 4/28/15; 1459 1)*** 1052-1207 1342 4/29; 3 (0 or 4/28/2015 DF 2 151 (11) 32 (4) 2057 1)*** 4/28/15; 4/28/15; 4/29; 4/30; 4/30; 3 8(1) 43 (2) 23 (4) 9 (3)*** 1435-1536 1728 2057 0253 0859 **Fish given to Ken Nichols (CNFHC) for fish health assessment *** one dummy-tagged fish was missing from the holding can after the 48 hour holding period from among the group of 15; one dummy-tagged fish in transport 3 died during transport; one dummy-tagged fish from transport 1 or 2 died within about 5 hours; and two dummy-tagged fish from transport 3 were found dead after 48 hours.`

77

Table 2: Tagged Chinook salmon released near Medford Island (MI) for the last 3 days of releases in 2015. Fish were held at the hatchery for 24 hours prior to transport to the ferry dock near Medford Island. No holding at the release location was done. Dummy tagged fish were assessed after all fish from the truckload were released.

Release Release Release Release Release Release Release Release Total Transport A B C D E F G H Tagging Release Transport Released Number of Dummy- tagged Date; Date; Date; Date; Date; Date; Date; Date; Date; Date Site Tank (A+B+C+D fish assessed (date: time) Time Time; Time; Time; Time; Time; Time; Time; Time; +E+F+G+H) Number Number Number Number Number Number Number Number

4/30; 4/30; 1655, 1710, 1 1656; 3 (4/30/15: 1738) 1711; 4/30/15; 21 24 1313- 1445 4/30; 4/30; 4/29/2015 MI 162 1724, 2 1737; 3 (4/30/15: 1738) 1725; 15 21 4/30; 4/30; 4/30; 4/30; 4/30/15; 1847, 1857, 1908, 1920, 3 1640- 9 (4/30/15: 1922) 1848; 1858; 1909; 1921; 1730 21 21 21 18 5/1; 5/1; 5/1; 1755, 1 1806; 1816 3 5/1/15; 1756; 21 3 1434- 21 1620 5/1; 5/1; 2 1816; 1826; 3 4/30/2015 MI 162 18 18 5/1; 5/1; 5/1; 5/1/15; 5/1; 1928, 9 1857, 1918, 3 1750- 1908; 1929; 1858 1919; 1837 21 18 21 21

5/2; 5/2; 5/2; 1902, 1848 1855; 1903; 0; 9* 1 21 21 3 5/2/2015; 5/2; 1545- 5/2; 1902, 3 (5/2/15: 1910); 12* 2 1705 1909; 167 1903; 5/1/2015 MI 18 18 5/2/2015; 5/2; 5/2; 5/2; 5/2; 3 1548- 1927; 1934; 1941; 1949; 3 (5/2/15: 1950);

1641 21 21 20 24 9* *Dummy tagged fish held at the hatchery for 72 hours prior to fish health assessment by Ken Nichols (CNFHC)

78

Table 3: Water temperature and dissolved oxygen (DO) in transport tanks after loading, after transport, after transport prior to unloading (when different than that after transport) and in the river and the difference between the river and transport tank after transport for Chinook salmon releases in the south Delta in 2015.

Difference in Tank prior to water # unloading when temperature morts fish held in tank between just at release transport tank prior location prior to after transport to Transport Tank after loading Tank after transport unloading River and river release # morts after Loading Temp DO Temp DO Temp DO Temp DO Date Site Transport Tank time (˚C) (mg/L) (˚C) (mg/L) transport (˚C) (mg/L) (˚C) (mg/L) 13.5 11.83 14.7 12.09 0 16.6 11.05 4/14/2015 Durham Ferry 1 1 930 1.9 0 13.5 11.84 14.1 12.67 0 16.6 11.05 4/14/2015 Durham Ferry 1 2 930 2.5 0 13.9 12.82 14.6 13.08 0 17.0 11.16 4/14/2015 Durham Ferry 2 3 1342 2.4 1 925 13.4 11.84 14.6 12.18 0 15.4 11.27 4/15/2015 Durham Ferry 1 1 0.8 0 925 13.5 11.91 14.3 12.38 0 15.4 11.27 4/15/2015 Durham Ferry 1 2 1.1 0 1303 14.6 12.88 0 4/15/2015 Durham Ferry 2 3 15.8 13.13 16.8 11.31 1.0 0 930 13.6 11.87 0 4/16/2015 Durham Ferry 1 1 14.9 12.32 15.6 11.22 0.7 0 930 0 4/16/2015 Durham Ferry 1 2 13.6 11.94 14.9 12.83 15.6 11.22 0.7 0 1245 14.4 12.66 0 4/16/2015 Durham Ferry 2 3 15.6 13.01 17.3 11.32 1.7 0 923 13.7 11.91 2 4/17/2015 Durham Ferry 1 1 15.3 12.40 16.7 9.55 1.4 0 923 13.8 11.94 0 4/17/2015 Durham Ferry 1 2 15.1 12.36 16.7 9.55 1.6 0 1306 14.2 12.71 0 4/17/2015 Durham Ferry 2 3 15.7 13.13 18.0 9.69 2.3 0 AVG 13.8 12.18 15.0 12.63 16.5 10.81

4/28/2015 Durham Ferry 1 1 845 14.6 12.2 16.1 12.54 0 16.9 12.14 21.9 7.95 5.8 0 4/28/2015 Durham Ferry 1 2 845 14.7 12.4 16.1 12.34 0 16.9 12.34 21.9 7.95 5.8 4 4/28/2015 3a 16.9 0 7 Durham Ferry 2 1209 12.86 18.0 13.01 24.7 8.65 6.7 4/30/2015 Medford Island 1 1 1300 15.4 10.74 17.3 13.42 0 18.1 13.67 20.1 6.66 2.8 0 4/30/2015 Medford Island 1 2 1300 15.9 11.19 16.4 11.61 0 17.5 11.67 20.1 6.66 3.7 0 b 4/30/2015 Medford Island 2 3 1630 15.4 12.64 16.4 12.88 0 20.7 6.74 4.3 0

79

Table 3 (Continued)

Difference in water Tank prior to temperature unloading when between fish held in tank transport # morts at release tank after just prior location prior to transport and to Transport Tank after loading Tank after transport unloading River river release # morts after Loading Temp DO Temp DO Temp DO Temp DO Date Site Transport Tank time (˚C) (mg/L) (˚C) (mg/L) transport (˚C) (mg/L) (˚C) (mg/L) 5/1/2015 Medford Island 1 1 * 14.5 9.22 16.1 11.40 0 17.8 11.5 21.2 6.71 5.1 0 5/1/2015 Medford Island 1 2 * 14.5 9.2 16.2 12.50 0 18 13.34 21.2 6.71 5.0 0 5/1/2015 Medford Island 2 3 17.5 0 0 1745 10.72 18.0 11.13 21.2 6.58 3.2 5/2/2015 Medford Island 1 1 1500 17.4 11.47 17.4 11.26 0 18.7 11.13 22.3 7.19 4.9 0 5/2/2015 Medford Island 1 2 1500 18.1 12.25 18.5 12.53 0 19.3 12.22 22.3 7.19 3.8 0 5/2/2015 Medford Island 2 3 1515 16.2 10.03 17.4 11..64 0 18.6 10.77 21.1 6.74 3.7 0 AVG 15.9 11.24 17.0 12.24 21.6 7.14

80

Table 4: External characteristics assessed for Chinook salmon condition and short-term survival. Percent scale loss was also assessed on one side of the fish. Parameters were provided by T. Liedtke, USGS.

Character Normal Abnormal

High contrast dark dorsal surfaces and Low contrast dorsal surfaces and coppery colored Body color light sides sides

Fin hemorrhaging No bleeding at base of fins Blood present at base of fins

Eyes Normally shaped Bulging or with hemorrhaging

Dark beet red to cherry red colored Gill color Grey to light red colored gill filaments gill filaments

Vigor Active swimming (prior to anesthesia) Lethargic or motionless (prior to anesthesia)

81

Table 5: The parameters assessed during the necropsy of dummy-tagged and tag retention Chinook salmon during the 2015 South Delta Survival Study. The anterior and posterior sutures were scored separately and each was included in the composite score. Parameters were provided by T. Liedtke, USGS.

Composite Score Score Score Definition Parameter No signs of tag expulsion. I.e., no signs that the tag is being forced out through the incision or the 0 lateral body wall. Simple encapsulation may be present Some bulging or lateral pressure. I.e., some evidence that the tag is causing pressure on the incision Signs of tag 1 or the lateral body wall expulsion Expulsion process obvious or complete. I.e., the tag is obviously being forced out through the incision 2 or the lateral body wall, or the tag is already out

Suture 0 Present presence 1 Not present 0 Intact, suture encapsulates incision and avoids midline Suture 1 Not intact, suture encapsulates midline or does not encapsulate incision pattern 0 No irritation Mild (redness or Midline/ 1 swelling) C A ventral Suture Moderate (redness or irritation 2 Posterior swelling) Anterior (ranking per Severe (purulent D B A,B,C,D) 3 discharge) View looking down onto incision; suture entry and exit points 4 Ulceration

0 Completely closed, perfect apposition Incision 1 Incision partially open due to gape or overlap apposition 2 Incision completely open (>75%)

0 Healed Incision 1 Partially healed healing 2 Not healed

No organ damage present. i.e., no signs of damage either due to the surgery or the presence of the 0 tag. Tags can be adhered to organs as part of encapsulation process, but that does not constitute Organ damage Inclusion Some organ damage present. I.e., the suture captures, punctures, or entangles the pyloric caeca, 1 stomach, spleen, or intestine

Fungus 0 Yes present? 1 No

82

Table 6: Names and descriptions of receivers and hydrophones used in the 2015 Chinook Salmon tagging study, with receiver codes used in Figure 4, the survival model (Figures 8 –9), and in data processing by the United States Geological Survey (USGS). The primary release site was located at Durham Ferry; a secondary release site used for the second release was located near Medford Island. Average latitude and longitude are given for sites with multiple hydrophones.

Hydrophone Location Survival Data Processing Individual Receiver Name and Description Receiver Code Latitude Longitude Model Code Code San Joaquin River near Durham Ferry upstream of the release site, upstream 37° 41.138'N 121° 15.382'W DFU1 A0a 300982 San Joaquin River near Durham Ferry upstream of the release site, 37° 41.182'N 121° 15.399'W DFU2 A0b 301509 downstream San Joaquin River near Durham Ferry; release site (no acoustic hydrophone 37° 41.224'N 121° 15.722'W DF located here) San Joaquin River near Durham Ferry downstream of the release site, 37° 41.318'N 121° 16.564'W DFD1 A2a 300867/460084 upstream San Joaquin River near Durham Ferry downstream of the release site, 37° 41.320'N 121° 16.562'W DFD2 A2b 460085 downstream San Joaquin River below Durham Ferry 1 37° 43.283'N 121° 15.731'W BDF1 A3 460035 San Joaquin River below Durham Ferry 2 37° 43.072'N 121° 16.700'W BDF2 A4 460036 San Joaquin River near Banta Carbona, upstream 37° 43.657'N 121° 17.924'W BCAU A5a 300935 San Joaquin River near Banta Carbona, downstream 37° 43.700'N 121° 17.917'W BCAD A5b 460021 San Joaquin River near Mossdale Bridge, upstream 37° 47.503'N 121° 18.417'W MOSU A6a 300898 San Joaquin River near Mossdale Bridge, downstream 37° 47.552'N 121° 18.406'W MOSD A6b 300910 San Joaquin River upstream of Head of Old River, upstream 1 (not used in 37° 48.349'N 121° 19.057'W HORU1 B0c 450020 survival model) San Joaquin River upstream of Head of Old River, upstream 2 (not used in 37° 48.355'N 121° 19.121'W HORU2 B0a 301503/301508/4 survival model) 50023 San Joaquin River upstream of Head of Old River, downstream (not used in 37° 48.350'N 121° 19.182'W HORD B0b 301503/301505 survival model) San Joaquin River near Lathrop, upstream 37° 48.666'N 121° 19.177'W SJLU A7a 301504/301507 San Joaquin River near Lathrop, downstream 37° 48.697'N 121° 19.124'W SJLD A7b 301511/301512 Predator Removal Study Site 4 37° 49.116'N 121° 19.050'W RS4 N1 300870/301166 Predator Removal Study Site 5 37° 49.914'N 121° 18.730'W RS5 N2 300901/300872 Predator Removal Study Site 6 37° 51.080'N 121° 19.326'W RS6 N3 300924/300884 Predator Removal Study Site 7 37° 51.871'N 121° 19.418'W RS7 N4 300917/300879 Predator Removal Study Site 8 37° 53.266'N 121° 19.813'W RS8 N5 300878/300861

83

Table 6. (Continued)

Hydrophone Location Survival Data Processing Individual Receiver Name and Description Receiver Code Latitude Longitude Model Code Code Predator Removal Study Site 9 37° 54.347'N 121° 19.408'W RS9 N6 300871/300937 Predator Removal Study Site 10 37° 55.094'N 121° 19.236'W RS10 N7 300916/300914 San Joaquin River near Garwood Bridge, upstream 37° 56.108'N 121° 19.807'W SJGU A8a 300934/300979/ 450047 San Joaquin River near Garwood Bridge, downstream 37° 56.119'N 121° 19.827'W SJGD A8b 300873/300908/ 450048 San Joaquin River at Stockton Navy Drive Bridge, upstream 37° 56.798'N 121° 20.393'W SJNBU A9a 300877 San Joaquin River at Stockton Navy Drive Bridge, downstream 37° 56.806'N 121° 20.365'W SJNBD A9b 300889 Burns Cutoff at Rough and Ready Island, upstream 37° 56.416'N 121° 21.065'W RRIU R1a 300904 Burns Cutoff at Rough and Ready Island, downstream 37° 56.408'N 121° 21.076'W RRID R1b 300892 San Joaquin River Shipping Channel 37° 59.736'N 121° 26.424'W SJS A10 300943/300869/ 300932 San Joaquin River at MacDonald Island, upstream 38° 01.030'N 121° 27.688'W MACU A11a 300868/301158 San Joaquin River at MacDonald Island, downstream 38° 01.372'N 121° 27.930'W MACD A11b 300899/300859 San Joaquin River near Medford Island; release site (no acoustic hydrophone 38° 03.084'N 121° 30.462'W MF located here) San Joaquin River near Medford Island, east 38° 03.188'N 121° 30.689'W MFE A12a 300875/300866 San Joaquin River near Medford Island, west 38° 03.222'N 121° 30.790'W MFW A12b 300905/300863 San Joaquin River near Disappointment Slough, upstream 38° 05.490'N 121° 34.493'W SJDU A13a 300933/300897/ 300921 San Joaquin River near Disappointment Slough, downstream 38° 05.537'N 121° 34.515'W SJDD A13b 300930/300989/ 300922 Old River East, near junction with San Joaquin, upstream 37° 48.709'N 121° 20.134'W OREU B1a 300923/300900 Old River East, near junction with San Joaquin, downstream 37° 48.738'N 121° 20.134'W ORED B1b 300891/300864 Old River South, upstream 37° 49.231'N 121° 22.655'W ORSU B2a 300980 Old River South, downstream 37° 49.200'N 121° 22.669'W ORSD B2b 301154 West Canal, upstream (not used in survival model) 37° 50.783'N 121° 33.572'W WCLU B3a 300918 West Canal, downstream (not used in survival model) 37° 50.857'N 121° 33.601'W WCLD B3b 301501 Old River at Highway 4, upstream 37° 53.630'N 121° 34.026'W OR4U B4a 301510/301502

84

Table 6. (Continued)

Hydrophone Location Survival Individual Receiver Name and Description Receiver Code Data Processing Code Latitude Longitude Model Code Old River at Highway 4, downstream 37° 53.702'N 121° 33.990'W OR4D B4b 300991/300942 Old River at the San Joaquin River, upstream (closer to Old River mouth) 38° 03.752'N 121° 34.848'W OSJU B5a 300894/300985/300920 Old River at the San Joaquin River, downstream farther from Old River 38° 03.696'N 121° 34.917'W OSJD B5b 300909/300896/300939 mouth) Middle River Head, upstream 37° 49.471'N 121° 22.768'W MRHU C1a 300913 Middle River Head, downstream 37° 49.483'N 121° 22.808'W MRHD C1b 300990 Middle River at Highway 4, upstream 37° 53.767'N 121° 29.582'W MR4U C2a 301159/301157 Middle River at Highway 4, downstream 37° 53.807'N 121 °29.594'W MR4D C2b 300882/300938 Middle River at Middle River 38° 00.128'N 121° 30.712'W MID C3 300890/300902 Radial Gate at Clifton Court Forebay, upstream (in entrance channel to 37° 49.801'N 121° 33.397'W RGU1 D1a 300888 forebay), array 1 Radial Gate at Clifton Court Forebay, upstream, array 2 37° 49.776'N 121° 33.418'W RGU2 D1b 301161 Radial Gate at Clifton Court Forebay, downstream (inside forebay), 37° 49.816'N 121° 33.450'W RGD1 D2a 460009/300911 array 1 in dual array Radial Gate at Clifton Court Forebay, downstream, array 2 in dual array 37° 49.816'N 121° 33.450'W RGD2 D2b 460010/300881 Central Valley Project trashracks, upstream 37° 49.012'N 121° 33.506'W CVPU E1a 460012/460023/301164 Central Valley Project trashracks, downstream 37° 48.999'N 121° 33.536'W CVPD E1b 300981 Central Valley Project holding tank 37° 48.951'N 121° 33.548'W CVPtank E2 300907 Turner Cut, east (closer to San Joaquin) 37° 59.499'N 121° 27.296'W TCE F1a 450043/301162 Turner Cut, west (farther from San Joaquin) 37° 59.478'N 121° 27.328'W TCW F1b 450024/300876 Columbia Cut, upstream (closer to San Joaquin) 38° 01.636'N 121° 30.051'W COLU F2a 300895/300880 Columbia Cut, downstream (farther from San Joaquin) 38° 01.620'N 121° 30.102'W COLW F2b 300986/300931 San Joaquin River at Jersey Point, upstream 1 38° 03.460'N 121° 41.098'W JPT G1a 300954/300730/300948/ 300729 San Joaquin River at Jersey Point, east (upstream 2) 38° 03.381'N 121° 41.216'W JPE G1b 300712/300718/300728/ 300950 San Joaquin River at Jersey Point, west (downstream) 38° 03.325'N 121° 41.298'W JPW G1c 300721/300732/300715/ 300944/300727/ 300714/300717/300724

85

Table 6. (Continued)

Hydrophone Location Survival Individual Receiver Name and Description Receiver Code Data Processing Code Latitude Longitude Model Code False River, west (closer to San Joaquin) 38° 03.382'N 121° 39.870'W FRW H1a 300719/300956 False River, east (farther from San Joaquin) 38° 03.380'N 121° 39.824'W FRE H1b 300955/300722 Chipps Island (aka Mallard Island), upstream 1 38° 02.852'N 121° 55.833'W MAT G2a 301153/300887/300936/300941/ 300929/300883 Chipps Island (aka Mallard Island), east (upstream 2) 38° 02.885'N 121° 55.847'W MAE G2b 301156/300984/300915/300903/ 300862/300858 Chipps Island (aka Mallard Island), west (downstream) 38° 02.959'N 121° 56.035'W MAW G2c 300885/301165/300886/300928/30115/ 300860/300865/300912/300957/ 301452/300906/300940 Benicia Bridge 38° 02.440'N 122° 07.409'W BBR G3 301486-301493 Threemile Slough, south (not used in survival model) 38° 06.450'N 121° 41.042'W TMS T1a 300726/300723 Threemile Slough, north (not used in survival model) 38° 06.678'N 121° 40.990'W TMN T1b 300731/300720 Montezuma Slough, downstream (not used in survival model) 38° 04.287'N 121° 52.186'W MZTD T2b 301163 Spoonbill Slough, upstream (not used in survival model) 38° 03.315'N 121° 53.718'W SBSU T3a 300952 Spoonbill Slough, downstream (not used in survival model) 38° 03.326'N 121° 53.733'W SBSD T3b 300958

86

Table 7. Environmental monitoring sites used in predator decision rule and route entrainment analysis for 2015 Chinook Salmon study. Database = CDEC (http://cdec.water.ca.gov/) or Water Library (http://www.water.ca.gov/waterdatalibrary/).

Environmental Monitoring Site Data Available Detection Site Database Site Name Latitude (°N) Longitude (°W) River Flow Water Velocity River Stage Pumping Reservoir Inflow BDT 37.8650 121.3231 RS6, RS7, RS8 Yes Yes Yes No No Water Library CLC 37.8298 121.5574 RGU, RGD No No No No Yes CDEC CSE 38.0740 121.8501 MTZ No No Yes No No CDEC FAL 38.0554 121.6672 FRE/FRW Yes Yes Yes No No CDEC

GLC 37.8201 121.4497 ORS No No Yes No No Water Library HLT 38.0030 121.5108 COL, MID Yes Yes Yes No No CDEC MAL 38.0428 121.9201 MTZ, SBS, MAE/MAW No Yes Yesb No No CDEC MDB 37.8908 121.4883 MR4 No No Yes No No Water Library MDM 37.9425 121.5340 MR4 Yes Yes No No No CDEC MRU 37.8339 121.3860 MRH Yes Yes No No No Water Library MRZ 38.0276 122.1405 BBR No No Yes No No CDEC MSD 37.7860 121.3060 HOR, MOS Yes Yes Yes No No Water Library a ODM 37.8101 121.5419 CVP/CVPtank Yes Yes Yes No No CDEC OH1 37.8080 121.3290 ORE Yes Yes Yes No No Water Library OH4 37.8900 121.5697 OR4 Yes Yes Yes No No CDEC ORX 37.8110 121.3866 ORS Yes Yes No No No Water Library OSJ 38.0711 121.5789 OSJ Yes Yes Yes No No CDEC PRI 38.0593 121.5575 MAC, MFE/MFW, SJD Yes Yes Yes No No CDEC RMID040 37.8350 121.3838 MRH No No Yes No No Water Library ROLD040 37.8286 121.5531 RGU, RGD, WCL No No Yes No No Water Library RRI 37.9360 121.3650 SJS Yes Yes Yes No No Water Library SJG 37.9351 121.3295 RS9, RS10, SJG, SJNB, RRI Yes Yes Yes No No CDEC SJJ 38.0520 121.6891 JPE/JPW Yes Yes Yes No No CDEC SJL 37.8100 121.3230 SJL, RS4, RS5 No No Yes No No Water Library a = California Water Library was used for river stage. b = Used for river stage for SBS and MAE/MAW.

87

Table 7. (Continued)

Environmental Monitoring Site Data Available Detection Site Database Site Name Latitude (°N) Longitude (°W) River Flow Water Velocity River Stage Pumping Reservoir Inflow TRN 37.9927 121.4541 TCE/TCW Yes Yes Yes No No CDEC TRP 37.8165 121.5596 CVP/CVPtank No No No Yes No CDEC TSJ 38.0900 121.6869 TMS/TMN No No Yes No No Water Library TSL 38.1004 121.6866 TMS/TMN Yes Yes No No No CDEC VNS 37.6670 121.2670 DFU, DFD, BDF1, BDF2,BCA Yes No Yes No No CDEC WCI 37.8316 121.5541 RGU, RGD, WCL Yes Yes No No No Water Library a = California Water Library was used for river stage. b = Used for river stage for SBS and MAE/MAW.

88

Table 8a. Cutoff values used in predator filter in 2015. Observed values past cutoff or unmet conditions indicate a predator. Time durations are in hours unless otherwise specified. DF = Durham Ferry release site, MF = Medford Island release site. See Table 8b for Flow, Water Velocity, Extra Conditions, and Comment. Footnotes refer to both this table and Table 8b. Unobserved transitions are omitted. a Residence Time (hr) BLPS Near Migration Rateb, c Time since (Absolute No. of Cumulative Mid-field Far-field Field (km/hr) last visit (hr) value) No. of Visits Upstream Forays Detection Maximu Site Previous Site m Maximum Maximum Minimum Maximum Maximum Maximum Maximum Maximum DFU DF, DFD 1 2 15 0.2 (0.6f) 4 1 1 DFU 1 3 15 2 0 DFD DF, DFU 10 20 30 0.1 4 1 0 DFD 2 47 94 2 0 BDF1, BDF2, BCA 2 4 15 0.1 4 2 (1f) 1 BDF1 DF, DFD 15 30 45 0.05 4 1 0 BDF1 2 57 114 2 0 BDF2 2 4 45 0.05 4 2 1 BDF2 DF, DFD, BDF1 15 30 70 0.05 4 1 0 BDF2 2 57 114 2 0 BCA 2 4 45 0.05 4 2 1 f BCA DF, DFD, BDF2 15 30 60 (120 ) 0.1 4 1 0 BCA 0.2 55 110 2 0 MOS BCA 15 30 130 0.1 4 8 1 0 MOS 2 57 114 2 1 HOR 2 4 130 0.1 4 8 2 1 SJL HOR 10 25 50 0.1 4 20 8 2 0 e SJL 1 51 121 (20 ) 2 1 e RS4 1 10 176 (20 ) 0.1 4 20 8 2 1 RS4 SJL 5 15 151 0.1 4 15 8 2 0 a = Near field residence time includes up to 12 hours missing between detections, while mid-field residence time includes entire time lag between first and last detections without intervening detections elsewhere; far-field residence time includes all time from entry in region to arrival at and departure from current site. b = Approximate migration rate was calculated on most direct pathway. c = Missing values for transitions to and from same site: travel times must be 12 to 24 hours, unless otherwise specified under "Extra conditions." e = Condition at departure from previous site. f = See comments for alternative criteria. 89

Table 8a. (Continued)

Residence Timea (hr) BLPS Migration Rateb, c Time since (Absolute No. of Cumulative Near Field Mid-field Far-field Detection (km/hr) last visit (hr) value) No. of Visits Upstream Forays Site Previous Site Maximum Maximum Maximum Minimum Maximum Maximum Maximum Maximum Maximum RS4 RS5 0.5 10 283 (20e) 0.1 4 15 8 2 1 RS5 RS4 5 15 250 0.1 4 15 8 2 0 RS5 1 41 311 2 1 e RS6 0.5 10 360 (20 ) 0.1 4 15 8 2 1 RS6 RS5 10 25 360 0.1 4 15 8 2 0 e RS7 0.5 10 360 (20 ) 0.1 4 15 8 2 1 RS7 RS6 10 25 360 0.1 4 15 8 2 0 RS8 RS7 10 25 360 0.1 4 15 8 2 0 RS9 RS8 5 15 360 0.1 4 15 8 2 0 RS10 0.5 10 360 0.1 4 15 8 2 1 RS10 RS9 5 15 360 0.1 4 15 8 2 0 SJG RS10 12 24 360 0.1 4 15 8 2 0 SJNB SJG 6 12 360 0.1 4 15 8 2 0 SJS SJNB 2 4 8 0.1 4 24 8 2 0 SJS 1 30 58 2 2 MAC 5 10 204 0.1 4 24 8 2 3 TCE/TCW 5 10 102 0.1 4 24 8 2 3 MAC SJS 30 60 156 0.1 4 24 8 2 0 MAC 10 99 279 2 1 TCE/TCW 1 2 180 0.1 4 24 8 2 1 f MF, MFE/MFW 15 30 86 (360 ) 0.1 4 24 8 1 1 a = Near field residence time includes up to 12 hours missing between detections, while mid-field residence time includes entire time lag between first and last detections without intervening detections elsewhere; far-field residence time includes all time from entry in region to arrival at and departure from current site. b = Approximate migration rate was calculated on most direct pathway. c = Missing values for transitions to and from same site: travel times must be 12 to 24 hours, unless otherwise specified under "Extra conditions." e = Condition at departure from previous site. f = See comments for alternative criteria.

90

Table 8a. (Continued)

Residence Timea (hr) BLPS Migration Rateb, c Time since (Absolute No. of Cumulative Near Field Mid-field Far-field Detection (km/hr) last visit (hr) value) No. of Visits Upstream Forays Site Previous Site Maximum Maximum Maximum Minimum Maximum Maximum Maximum Maximum Maximum MAC SJD 15 30 60 0.1 4 24 8 1 1 TCE/TCW SJS 1 2 76 0.1 4 24 8 1 0 MAC 3 6 331 0.1 4 24 8 1 1 MID 1 2 4 0.5 1 24 8 1 1 MFE/MFW MAC 30 60 360 0.1 4 36 8 2 0 MF 30 60 120 0.01 36 1 0 MFE/MFW 15 104 360 3 1 COL 1 2 360 0.1 4 36 8 2 1 SJD 1 2 4 0.5 1 36 8 2 1 COL MAC 1 2 324 0.1 4 36 8 1 0 COL 1 32 360 2 1 f MF, MFE/MFW 3 6 38 (360 ) 0.1 4 36 8 1 1 MID 1 2 4 0.1 4 36 8 1 1 f f f SJD MAC, MF 40 (20 ) 80 (40 ) 160 (80 ) 0.1 4 36 8 1 0 f f f MFE/MFW 40 (20 ) 80 (40 ) 160 (80 ) 0.1 4 36 8 2 0 SJD 3 112 296 2 1 OSJ 3 6 82 0.1 4 36 8 1 1 MID 3 6 12 0.1 4 36 8 1 0 JPT/JPE/JPW 5 10 20 0.5 4 36 8 2 1 HOR MOS 15 30 140 0.1 4 8 1 0 HOR 2 57 114 2 1 a = Near field residence time includes up to 12 hours missing between detections, while mid -field residence time includes entire time lag between first and last detections without intervening detections elsewhere; far-field residence time includes all time from entry in region to arrival at and departure from current site. b = Approximate migration rate was calculated on most direct pathway. c = Missing values for transitions to and from same site: travel times must be 12 to 24 hours, unless otherwise specified under "Extra conditions." f = See comments for alternative criteria. 91

Table 8a. (Continued)

Residence Timea (hr) BLPS Migration Rateb, c Time since (Absolute No. of Cumulative Near Field Mid-field Far-field Detection (km/hr) last visit (hr) value) No. of Visits Upstream Forays Site Previous Site Maximum Maximum Maximum Minimum Maximum Maximum Maximum Maximum Maximum HOR SJL 2 4 140 (10e) 0.1 4 15 8 2 1 ORE HOR 10 20 40 0.1 4 20 8 1 0 ORE 1 46 106 2 1 WCL OR4, MR4 1 2 360 0.1 4 36 8 2 3 OR4 OR4 3 42 360 3 3 MID 5 10 360 0.1 4 36 8 1 3 MF, MFE/MFW, OSJ MID 1 2 4 0.1 4 36 8 1 0 SJD 1 2 338 0.1 4 36 8 1 1 MR4 TCE/TCW, MF 1 2 4 0.1 4 36 8 1 0 MID 1 2 271 0.1 4 36 8 1 1 MF, MFE/MFW, MID TCE/TCW 20 40 80 0.1 4 36 8 1 0 COL 20 40 80 0.1 4 36 8 2 1 MID 1 70 174 2 2 h i RGU/RGD WCL 24 (40 ; 80 ) 360 0.1 4 36 8 2 3 JPT/JPE/ SJD, OSJ, JPW MFE/MFW, MF 40 80 160 0.1 4 20 8 1 0 TMN/TMS 40 80 163 0.1 4 20 8 1 0 JPT/JPE/JPW 15 145 358 3 2 FRE/FRW 15 145 358 0.1 4 20 3 2 FRE/FRW OSJ 2 80 160 0.1 4 20 8 1 0 a = Near field residence time includes up to 12 hours missing between detections, while mid-field residence time includes entire time lag between first and last detections without intervening detections elsewhere; far-field residence time includes all time from entry in region to arrival at and departure from current site. b = Approximate migration rate was calculated on most direct pathway. c = Missing values for transitions to and from same site: travel times must be 12 to 24 hours, unless otherwise specified under "Extra conditions." e = Condition at departure from previous site. h = If returned to Forebay entrance channel from Clifton Court Forebay and most detections were at RGU (not RGD). i = If known presence at gates < 24 hours, or if present at RGU < 80% of total residence time before returning to Forebay entrance channel. 92

Table 8a. (Continued)

Residence Timea (hr) BLPS Migration Rateb, c Time since (Absolute No. of Cumulative Near Field Mid-field Far-field Detection (km/hr) last visit (hr) value) No. of Visits Upstream Forays Site Previous Site Maximum Maximum Maximum Minimum Maximum Maximum Maximum Maximum Maximum FRE/FRW FRE/FRW 2 106 290 3 2 JPT/JPE/JPW 2 110 298 0.1 4 20 3 2 TMN/TMS SJD, MF 4 8 16 0.2 4.5 20 8 1 0 TMN/TMS 1 51 91 2 2 JPT/JPE/JPW 4 8 205 0.2 4.5 20 8 2 2 SBS SBS 1 28 360 2 3 MAT/MAE/ MAW 1 2 360 0.1 4 20 8 2 3 MAT/MAE/ SJD, TMN/TMS, MAW JPT/JPE/JPW 40 200 360 0.1 4 20 8 1 (2f) 0 MAT/MAE/ MAW 10 50 360 2 3 BBR JPT/JPE/JPW 25 125 360 0.1 5.5 20 8 1 0 MAT/MAE/ MAW 25 125 360 0.1 5.5 20 8 2 0 BBR 5 25 360 2 3 a = Near field residence time includes up to 12 hours missing between detections, while mid-field residence time includes entire time lag between first and last detections without intervening detections elsewhere; far-field residence time includes all time from entry in region to arrival at and departure from current site. b = Approximate migration rate was calculated on most direct pathway. c = Missing values for transitions to and from same site: travel times must be 12 to 24 hours, unless otherwise specified under "Extra conditions." f = See comments for alternative criteria.

93

Table 8b. Cutoff values used in predator filter in 2015. Observed values past cutoff or unmet conditions indicate a predator. Time durations are in hours unless otherwise specified. DF = Durham Ferry release site, MF = Medford Island release site. Footnotes, Extra Conditions and Comment refer to both this table and Table 8a. Unobserved transitions are omitted.

Flowd (cfs) Water Velocityd (ft/sec) Detection Average during Site Previous Site At arrival At departuree At arrival At departuree transition Extra Conditions Comment DFU DF, DFD Alternate value if coming from DFD DFU Travel time < 20 DFD DF, DFU DFD Travel time < 20 BDF1, BDF2, Alternate value if coming BCA from BCA BDF1 DF, DFD, BDF2 BDF1 Travel time < 20 BDF2 DF, DFD, BDF1, BCA BDF2 Travel time < 20 BCA DF, DFD, Alternate value if coming BDF2 from DF BCA Travel time < 20 MOS BCA MOS < 14,000 < 2.7 Travel time < 20 HOR < 14,000 < 0.1 SJL HOR

SJL Travel time < 20 Next transition must be RS4 directed downstream RS4 SJL, RS5 RS5 RS4 RS5 Travel time < 20 d = Classified as predator if flow or velocity condition, if any, is violated. e = Condition at departure from previous site.

94

Table 8b. (Continued)

Flowd (cfs) Water Velocityd (ft/sec) Detection Average during Site Previous Site At arrival At departuree At arrival At departuree transition Extra Conditions Comment Next transition must be RS5 RS6 directed downstream RS6 RS5 Next transition must be RS7 < 1.3 directed downstream RS7 RS6 RS8 RS7 RS9 RS8 Next transition must be RS10 < 1.3 directed downstream RS10 RS9 SJG RS10 SJNB SJG SJS SJNB g g SJS < 0.3 (> -0.2) > -0.2 (< 0.3) Travel time < 20 MAC < 6000 < 0.3 TCE/TCW > -650 > -0.1 Detection time < 10 MAC SJS hours g g MAC < 0.3 (> -0.2) > -0.2 (< 0.3) TCE/TCW > -650 > -0.1 > -0.4 MF, SJD, Detection time < 10 Alternate value if coming MFE/MFW < 0.5 < 0.3 < 0.3 hours from MFE/MFW TCE/TCW SJS < 650 < 0.1 MAC < 650 < 17,000 < 0.1 < 0.3 < 0.2 MID > -650 < 5,000 > -0.1 < 0.2 d = Classified as predator if flow or velocity condition, if any, is violated. e = Condition at departure from previous site. g = High flow/velocity on departure requires low values on arrival (and vice versa).

95

Table 8b. (Continued)

Flowd (cfs) Water Velocityd (ft/sec) Detection Average during Site Previous Site At arrival At departuree At arrival At departuree transition Extra Conditions Comment Detection time < 2 MFE/MFW MAC, MF hours g g MFE/MFW < 0.2 (> -0.1) > -0.1 (< 0.2) -0.3 to 0.3 COL SJD < 12,000 < 12,000 < 0.2 < 0.2 < 0.3 MAC, COL, COL MID MF, Alternate value if coming MFE/MFW from MFE/MFW Alternate values if MAC, MF, Detection time < 3 average transition water SJD MFE/MFW > -7,000 > -0.1 -0.2 to 0.4 hours velocity outside range < 12,000 > -7,000 g g g g SJD (> -7,000) (< 12,000) < 0.2 (> -0.1) > -0.1 (< 0.2) OSJ > -7,000 > -5,000 > -0.1 > -0.2 MID > -7,000 > -0.1 JPT/JPE/JPW < 12,000 < 0.2 < 0.4 HOR MOS HOR < 14,000 Travel time < 20 SJL < 14,000 < 2 < 2 ORE HOR ORE Travel time < 20 Alternate values if WCL OR4, MR4 < 1500 NA (< 1500f) < 0.2 NA (< 0.2f) coming from MR4 < 4,000 > -4,000 g g g g OR4 OR4 (> -4,000) (< 4,000) < 0.5 (> -0.5) > -0.5 (< 0.5) MID < 4,000 < 0.5 d = Classified as predator if flow or velocity condition, if any, is violated. e = Condition at departure from previous site. g = High flow/velocity on departure requires low values on arrival (and vice versa).

96

Table 8b. (Continued)

Flowd (cfs) Water Velocityd (ft/sec) Detection Average during Site Previous Site At arrival At departuree At arrival At departuree transition Extra Conditions Comment MF, SJD, OSJ MFE/MFW < 5,000 < 0.2 MID > -2,500 > -0.1 Alternate value if coming MR4 TCE/TCW, MF < 3,000 < 0.2 < 0.1 (NAf) < 0.1 from MF MID < 3,000 < 0.2 < 0.1 MF, MID MFE/MFW < 5,000 < 0.2 TCE/TCW > -5,000 < 650 > -0.2 < 0.1 COL < 5,000 > -5,000 MID (> -5,000)g (< 5,000)g < 0.2 (> -0.2)g > -0.2 (< 0.2)g RGU/RGD WCL < 4,500 < 0.8 SJD, OSJ, MF, JPT/JPE/ MFE/MFW, JPW TMN/TMS Travel time < 50, JPT/JPE/JPW upstream foray < 25 km FRE/FRW Upstream foray < 25 km FRE/FRW OSJ FRE/FRW, JPT/JPE/JPW Upstream foray < 25 km TMN/TMS SJD > -23,000 > -0.4 MF, JPT/JPE/JPW g g g g TMN/TMS < 0 (> 0) > 0 (< 0) < 0 (> 0) > 0 (< 0) SBS SBS Travel time < 20 d = Classified as predator if flow or velocity condition, if any, is violated. e = Condition at departure from previous site. f = See comments for alternative criteria. g = High flow/velocity on departure requires low values on arrival (and vice versa).

97

Table 8b. (Continued)

Flowd (cfs) Water Velocityd (ft/sec) Detection Average during Site Previous Site At arrival At departuree At arrival At departuree transition Extra Conditions Comment MAT/MAE/ SBS MAW SJD, MAT/MAE/ TMN/TMS, Alternate value if coming MAW JPT/JPE/JPW > -0.2 Upstream foray < 25 km from JPT/JPE/JPW MAT/MAE/ MAW Upstream foray < 25 km BBR JPT/JPE/JPW Upstream foray < 25 km MAT/MAE/ MAW Upstream foray < 25 km BBR Upstream foray < 25 km d = Classified as predator if flow or velocity condition, if any, is violated. e = Condition at departure from previous site.

98

Table 9: Results of dummy tagged Chinook salmon evaluated at Durham Ferry after being held for 48 hours or at Medford Island (held at the hatchery for 24 hours prior to transport) after release of study fish as part of Salmon Survival Study in 2015. Only fish that were still alive after 48 hours were assessed for condition.

Examination Date, Mean (sd) Mean (sd) Normal No Fin Normal Eye Normal Gill Site Mortality Time Forklength (mm) scale loss % Body Color Hemorrhaging Quality Color

Durham Ferry 4/16/15, 0915 99.0 (4.7) 0/15 7.3 (5.6) 15/15 15/15 15/15 15/15

Durham Ferry 4/17/15, 0915 98.1 (5.0) 0/15 5.7 (1.8) 15/15 15/15 15/15 15/15

Durham Ferry 4/18/15, 0915 99.4 (4.6) 0/15 8.0 (3.2) 15/15 15/15 15/15 15/15

Durham Ferry 0/30*

Durham Ferry 4/30/15, 0915 96.9 (6.2) 2/12*** 5.0 (0) 10/10 10/10 7/10 10/10

Medford** 4/30/15, 1744 98.3 (3.8) 0/6 16.7 (10.3) 6/6 6/6 6/6 6/6

Medford** 4/30/15, 1923 95.7 (4.0) 0/9 10.6 (6.3) 9/9 9/9 9/9 9/9

Medford** 5/1/15, 1832 99.0 (4.7) 0/6 10.0 (4.5) 6/6 6/6 6/6 6/6

Medford** 5/1/15, 1935 100.2 (5.3) 0/9 18.9 (7.0) 9/9 9/9 9/9 9/9

Medford** 5/2/15, 1910 101.0 (6.2) 0/3 23.3 (2.9) 3/3 3/3 3/3 3/3

Medford** 5/2/15, 1954 106.0 (7.0) 0/3 16.7 (11.5) 3/3 3/3 3/3 3/3

MRH 0/30*

*Fish given to CA/NV Fish Health Center for further evaluation **Fish held at hatchery for 24 hours prior to release *** One fish died during transport, one died after transport, one fish was missing, while two were found dead after 48 hours

99

Table 10: Criteria for scoring and results of necropsy on dummy- tagged Chinook salmon assessed at Durham Ferry (DF) or Medford Island (MI) in 2015.

Score Potential Suture present? Incision Fungus Organ Peritoneal Signs of Composite apposition damage apposition expulsion score ANT POST

Min 0 0 0 0 0 0 0 0 Mid 1 1 1 0 1 1 1 6 Max 2 2 2 1 1 2 2 12

Date Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Time (Location) 4/16/15, 0915 (DF) 0.1 (0.3) 0.0 (0) 0.7 (0.8) 0 (0) 0.5 (0.7) 0.8 (0.6) 0 (0) 2.1 (1.3)

4/17/15, 0915 (DF) 0.3 (0.5) 0.0 (0) 0.7 (0.8) 0 (0) 0.3 (0.5) 0.7 (0.7) 0 (0) 1.9 (1.5)

4/18/15, 0915 (DF) 0.1 (0.3) 0.1 (0.4) 0.4 (0.6) 0 (0) 0.4 (0.6) 0.7 (0.7) 0 (0) 1.8 (1.3)

4/30/15, 0915 (DF) 0.1 (0.3) 0.3(0.5) 0.2 (0.4) 0.4 (0.5) 0.1 (0.3) 0.8 (0.8) 0 (0) 1.8 (1.6)

4/30/15, 1744 (MI) 0 (0) 0 (0) 0 (0) 0 (0) 0.3 (0.5) 0.7(0.5) 0 (0) 1.0 (0.6)

4/30/15, 1923 (MI) 0 (0) 0(0) 0.1(0.3) 0 (0) 0.3 (0.5) 1.0 (0) 0 (0) 1.4 (0.5)

5/1/15, 1832 (MI) 0.2 (0.4) 0.2 (0.4) 0.3 (0.5) 0 (0) 0.3 (0.5) 0.8 (0.8) 0 (0) 1.8 (1.8)

5/1/15, 1935 (MI) 0 (0) 0 (0) 0.2 (0.4) 0 (0) 0.2 (0.4) 0.9 (0.3) 0 (0) 1.3 (1.0)

5/2/15, 1910 (MI) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1.0 (0) 0 (0) 1.0 (0)

5/2/15, 1954 (MI) 0 (0) 0 (0) 0.3 (0.6) 0 (0) 0 (0) 0.7 (0.6) 0 (0) 1.0 (1.0)

100

Table 11. Scale loss and position of acoustic tag in Chinook salmon necropsied 31 days after tagging to assess tag retention during the 2015 South Delta Survival Study. Percent scale loss was assessed on the most compromised side of the fish; and the final position of the acoustic tag was identified relative to incision location; over incision, anterior to incision, or posterior to incision (0=no, 1=yes). Acoustic Tag Position Acoustic Tag Position Scale Scale Fish # Fish # Loss (%) Over Incision Anterior to Posterior to Loss (%) Over Incision Anterior to Posterior to (0/1) Incision (0/1) Incision (0/1) (0/1) Incision (0/1) Incision (0/1) 1 0 0 0 1 22 0 1 0 0 2 0 1 0 0 23 5 1 0 0 3 5 0 0 1 24 0 1 0 0 4 5 1 0 0 25 0 1 0 0 5 5 1 0 0 26 10 1 0 0 6 5 1 0 0 27 15 1 0 0 7 10 0 0 1 28 5 1 0 0 8 10 1 0 0 29 5 1 0 0 9 10 1 0 0 30 10 1 0 0 10 5 Unknown 31 20 0 0 1 11 10 0 0 1 32 15 1 0 0 12 5 1 0 0 33 10 1 0 0 13 5 0 0 1 34 10 1 0 0 14 15 0 0 1 35 10 1 0 0 15 5 1 0 0 36 25 1 0 0 16 0 1 0 0 37 0 0 1 0 17 10 0 0 1 38 0 1 0 0 18 15 1 0 0 39 0 1 0 0 19 5 Tag expelled 40 5 0 0 1 20 5 0 0 1 41 15 0 0 1 21 5 1 0 0 42 10 1 0 0

101

Acoustic Tag Position Acoustic Tag Position Scale Scale Fish # Fish # Loss (%) Over Incision Anterior to Posterior to Loss (%) Over Incision Anterior to Posterior to (0/1) Incision (0/1) Incision (0/1) (0/1) Incision (0/1) Incision (0/1) 43 0 1 0 0 66 0 1 0 0 44 10 1 0 0 67 5 1 0 0 45 10 0 0 1 68 15 1 0 0 46 5 1 0 0 69 10 1 0 0 47 10 1 0 0 Mean: 7.51 0.81 0.01 0.18 48 3 1 0 0 SD: 5.24 0.40 0.12 0.38 49 5 1 0 0 50 5 1 0 0 51 10 1 0 0 52 5 1 0 0 53 5 1 0 0 54 10 1 0 0 55 15 1 0 0 56 15 1 0 0 57 10 1 0 0 58 10 1 0 0 59 5 1 0 0 60 10 1 0 0 61 5 1 0 0 62 15 1 0 0 63 5 1 0 0 64 5 1 0 0 65 10 1 0 0

102

Table 12: Results of eight surgical parameters assessed on tag retention Chinook salmon held for 30 days after tagging during the 2015 South Delta Survival Study. The parameters, as outlined in Table 5, included 1) signs of tag expulsion (0=none, 1=some signs present, 2=tag expelled or partially expelled; 2) presence of the (1) anterior and (2) posterior suture (0=present, 1=untied, 2=not present); 3) Suture pattern (0= Intact, 1= Not intact, NA= Not available); 4) Suture irritation (1=Mild, 2 Moderate, 3 Severe, 4 Ulceration); 5) incision apposition (0=closed, good apposition, 1=partial gape or overlap, 2=completely open [>75%]); 6) incision healing (0=Healed, 1=Partially healed, 2 =Not healed); 7) organ inclusion (0=none, 1=yes); and (8) whether fungus was present (0=no, 1=yes).

Signs of Suture Presence Suture Pattern Suture Irritation Incision Incision Organ Tag (0/1) (0/1) (1/2/3/4) Fungus Fish # Apposition Healing Inclusion Expulsion (0/1) Anterior Posterior Anterior Posterior A B C D (0/1/2) (0/1/2) (0/1) (0/1/2) 1 0 0 0 0 0 2 1 2 0 0 1 0 0 2 0 0 0 0 0 2 2 1 1 0 1 0 0 3 0 0 0 0 0 1 2 2 2 0 0 0 0 4 0 0 0 0 0 1 1 1 1 0 0 0 0 5 0 0 0 0 0 1 2 1 1 0 0 1 0 6 0 0 0 0 0 2 2 2 2 0 0 0 0 7 0 0 0 0 0 2 2 3 3 0 0 0 0 8 0 0 0 0 0 1 1 1 1 0 0 0 0 9 0 0 0 0 0 1 1 1 1 0 0 0 0 10 0 0 0 0 0 1 2 2 3 0 0 0 0 11 0 0 0 0 0 1 1 1 1 0 0 0 0 12 0 0 0 0 0 0 0 0 0 0 0 0 0 13 0 0 0 0 0 1 1 1 1 0 0 0 0 14 0 0 0 0 0 2 3 2 3 0 0 0 0 15 0 0 0 0 0 1 2 0 1 0 0 1 0 16 0 0 0 0 0 1 2 1 2 0 0 0 0 17 1 0 0 0 0 1 2 3 3 0 0 0 0 18 0 0 0 0 0 1 1 1 2 0 0 0 0 19 2 0 1 1 N/A 1 1 0 3 0 1 0 0 20 0 0 0 0 0 0 2 3 3 0 0 0 0 21 0 0 0 0 0 1 1 1 1 0 0 0 0 22 0 0 0 0 0 1 0 0 0 0 0 0 0 23 0 0 0 0 0 0 0 1 0 0 0 0 0

103

Signs of Suture Presence Suture Pattern Suture Irritation Incision Incision Organ Tag (0/1) (0/1) (1/2/3/4) Fungus Fish # Apposition Healing Inclusion Expulsion (0/1) Anterior Posterior Anterior Posterior A B C D (0/1/2) (0/1/2) (0/1) (0/1/2) 24 0 0 0 0 0 0 0 0 0 0 0 0 0 25 1 0 0 0 0 2 0 0 0 0 0 0 0 26 0 0 0 0 0 1 1 1 1 0 0 0 0 27 0 0 0 0 0 0 0 0 0 0 0 0 1 28 0 0 0 0 0 0 0 0 0 0 0 0 0 29 0 0 0 0 0 0 0 0 0 0 0 0 0 30 0 0 0 0 0 2 0 2 0 0 0 0 1 31 0 0 0 0 0 0 0 0 0 0 0 0 0 32 0 0 0 0 0 2 0 3 3 0 1 0 1 33 0 0 0 0 0 2 0 0 0 0 0 0 0 34 0 0 0 0 0 3 3 3 3 1 2 0 1 35 0 0 0 0 0 2 1 1 1 0 0 1 0 36 0 0 0 0 0 0 0 0 0 0 0 0 0 37 0 0 0 0 0 1 0 1 0 0 0 1 0 38 0 0 0 0 0 0 0 0 0 0 0 0 0 39 1 0 0 0 0 1 1 1 1 0 0 0 0 40 0 0 0 0 0 1 1 1 1 0 0 0 0 41 0 0 0 0 0 0 0 3 0 0 0 0 0 42 0 0 0 0 0 3 0 0 0 0 0 0 0 43 0 0 0 0 0 0 0 2 0 0 0 0 0 44 0 0 0 0 0 0 0 2 2 0 0 0 0 45 0 0 0 0 0 2 2 0 0 1 1 0 0 46 0 0 0 0 0 1 1 1 1 0 0 0 0 47 0 0 0 0 0 2 1 1 1 0 0 0 0 48 0 0 0 0 0 1 1 1 1 0 0 0 0 49 0 0 0 0 0 0 0 2 2 0 0 0 1 50 0 0 0 0 0 0 0 0 0 0 0 0 0

104

Signs of Suture Presence Suture Pattern Suture Irritation Incision Incision Organ Tag (0/1) (0/1) (1/2/3/4) Fungus Fish # Apposition Healing Inclusion Expulsion (0/1) Anterior Posterior Anterior Posterior A B C D (0/1/2) (0/1/2) (0/1) (0/1/2) 51 0 0 0 0 0 1 0 0 0 0 0 0 0 52 0 0 1 0 N/A 0 0 0 0 0 0 0 0 53 0 0 0 0 0 0 1 0 0 0 0 0 0 54 0 0 0 0 0 1 1 0 0 0 0 0 1 55 0 0 0 0 0 0 0 1 1 0 0 0 0 56 0 0 0 0 0 1 1 0 1 0 0 0 0 57 0 0 0 0 0 1 1 0 0 0 0 0 1 58 0 0 0 0 1 2 2 0 0 0 0 0 0 59 0 0 0 0 0 0 0 0 0 0 0 1 0 60 0 0 0 0 0 0 0 0 1 0 0 0 0 61 0 0 0 0 1 1 1 0 0 0 0 0 0 62 0 0 0 0 0 1 0 1 1 0 0 0 0 63 0 0 0 0 0 0 0 3 3 0 0 0 0 64 0 0 0 0 0 1 1 1 0 0 0 0 0 65 0 0 0 0 0 1 1 2 2 0 0 0 0 66 0 0 0 0 0 0 0 1 0 0 0 0 0 67 0 0 0 0 0 0 0 0 2 0 0 0 0 68 0 0 0 0 1 2 1 3 3 0 0 0 0 69 0 0 0 0 0 1 0 0 0 0 0 0 0 Mean: 0.07 0.00 0.03 0.01 0.04 0.93 0.78 0.97 0.96 0.03 0.10 0.07 0.10 SD: 0.31 0.00 0.17 0.12 0.21 0.80 0.83 1.01 1.07 0.17 0.35 0.26 0.30

105

Table 13: Number of tags from each release group that were detected after release in 2015, including predator-type detections and detections omitted from the survival analysis.

Durham Ferry Releases Release Group 1 2 Total Number Released 643 156 799 Number Detected 361 40 401 Number Detected Downstream of Release Site 334 12 346 Number Detected Upstream of Study Area 361 40 401 Number Detected in Study Area 21 0 21 Number Detected in San Joaquin River Route 12 0 12 Number Detected in Old River Route 1 0 1 Number Assigned to San Joaquin River Route 12 0 12 Number Assigned to Old River Route 1 0 1

Medford Island Releases Release Group 1 2 Total Number Released 0 491 491 Number Detected 398 398 Number Detected Upstream of OSJ/SJD 332 332 Number Detected at Medford Island 325 325 Number Detected Upstream of Medford Island 43 43 Number Detected Downstream of Medford Island 92 92 a Number Detected in San Joaquin River Route 84 84 Number Detected in Interior Delta Route 48 48 Number Assigned to San Joaquin River Route 79 79 Number Assigned to Interior Delta Route 44 44 a = detected at SJD

106

Table 14. Number of tags observed from each release group at each detection site in 2015, including predator-type detections. Routes (SJR = San Joaquin River, OR = Old River, ID = Interior Delta) represent route assignment at the head of Old River for DF releases, and upon release for MF releases. Counts are summed (pooled) over all receivers in array and all routes unless otherwise noted. Route could not be identified for some tags. Durham Ferry Releases Survival Release Group Model Detection Site Site Code Code 1 2 Total Release site at Durham Ferry 643 156 799 Durham Ferry Upstream DFU A0 50 35 85 Durham Ferry Downstream DFD A2 290 12 302 Below Durham Ferry 1 BDF1 A3 47 1 48 Below Durham Ferry 2 BDF2 A4 124 1 125 Banta Carbona BCA A5 78 0 78 Mossdale MOS A6 21 0 21 Head of Old River HOR B0 19 0 19 Lathrop, Upstream SJLU A7a 12 0 12 Lathrop, Downstream SJLD A7b 11 0 11 Lathrop (Pooled) SJL A7 12 0 12 Predator Removal Study 4 RS4 N1 9 0 9 Predator Removal Study 5 RS5 N2 7 0 7 Predator Removal Study 6 RS6 N3 5 0 5 Predator Removal Study 7 RS7 N4 4 0 4 Predator Removal Study 8 RS8 N5 3 0 3 Predator Removal Study 9 RS9 N6 3 0 3 Predator Removal Study 10 RS10 N7 3 0 3 Garwood Bridge, Upstream SJGU A8a 3 0 3 Garwood Bridge, Downstream SJGD A8b 3 0 3 Garwood Bridge (Pooled) SJG A8 3 0 3 Navy Drive Bridge, Upstream SJNBU A9a 2 0 2 Navy Drive Bridge, Downstream SJNBD A9b 2 0 2 Navy Drive Bridge (Pooled) SJNB A9 2 0 2 Rough and Ready Island RRI R1 0 0 0 San Joaquin River Shipping Channel SJS A10 1 0 1 MacDonald Island Upstream MACU A11a 1 0 1 MacDonald Island Downstream MACD A11b 1 0 1 MacDonald Island (Pooled) MAC A11 1 0 1 Turner Cut TCE/TCW F1 0 0 0 Medford Island MFE/MFW A12 0 0 0 Columbia Cut COL F2 0 0 0 San Joaquin River Disappointment Slough (Pooled) SJD A13 0 0 0 Old River East, Upstream OREU B1a 1 0 1 Old River East, Downstream ORED B1b 1 0 1 Old River East (Pooled) ORE B1 1 0 1 Old River South ORS B2 0 0 0

107

Table 14. (Continued) Durham Ferry Releases Survival Release Group Detection Site Site Code Model Code Total West Canal WCL B3 0 0 0 Old River at Highway 4 OR4 B4 0 0 0 Old River at the San Joaquin Mouth OSJ B5 0 0 0 Middle River Head MRH C1 0 0 0 Middle River at Highway 4 MR4 C2 0 0 0 Middle River at Middle River MID C3 0 0 0 Radial Gates Upstream RGU D1 0 0 0 Radial Gates Downstream RGD D2 0 0 0 Central Valley Project Trashrack CVP E1 0 0 0 Central Valley Project Holding Tank CVPtank E2 0 0 0 Threemile Slough TMS/TMN T1 0 0 0 Jersey Point JPT/JPE/JPW G1 0 0 0 False River FRE/FRW H1 0 0 0 Montezuma Slough MTZ T2 0 0 0 Spoonbill Slough SBS T3 0 0 0 Chipps Island MAT/MAE/MAW G2 0 0 0 Benicia Bridge BBR G3 0 0 0

Medford Island Releases Survival Release Group Detection Site Site Code Model Code Total Release site at Medford Island 0 491 491 San Joaquin River Shipping Channel SJS A10 12 12 MacDonald Island Upstream MACU A11a 25 25 MacDonald Island Downstream MACD A11b 35 35 MacDonald Island (Pooled) MAC A11 35 35 Turner Cut, Upstream TCE F1a 6 6 Turner Cut, Downstream TCW F1b 6 6 Turner Cut (Pooled) TCE/TCW F1 6 6 Medford Island East MFE A12a 263 263 Medford Island West MFW A12b 279 279 Medford Island (Pooled) MFE/MFW A12 325 325 Columbia Cut, Upstream COLU F2a 11 11 Columbia Cut, Downstream COLD F2b 12 12 Columbia Cut COL F2 12 12 San Joaquin River Disappointment Slough, Upstream SJDU A13a 80 80 San Joaquin River Disappointment Slough, Downstream SJDD A13b 74 74 San Joaquin River Disappointment Slough (Pooled) SJD A13 84 84

108

Table 14. (Continued)

Medford Island Releases Survival Release Group Detection Site Site Code Model Code Total West Canal, Upstream WCLU B3a 2 2 West Canal, Downstream WCLD B3b 2 2 West Canal (Pooled) WCL B3 2 2 Old River at Highway 4, Upstream OR4U B4a 1 1 Old River at Highway 4, Downstream OR4D B4b 1 1 Old River at Highway 4 (Pooled) OR4 B4 1 1 Old River at the San Joaquin Mouth, Upstream OSJU B5a 5 5 Old River at the San Joaquin Mouth, Downstream OSJD B5b 6 6 Old River at the San Joaquin Mouth (Pooled) OSJ B5 6 6 Middle River at Highway 4, Upstream MR4U C2a 3 3 Middle River at Highway 4, Downstream MR4D C2b 3 3 Middle River at Highway 4 (Pooled) MR4 C2 3 3 Middle River at Middle River MID C3 34 34 Radial Gates Upstream #1 RGU1 D1a 1 1 Radial Gates Upstream #2 RGU2 D1b 1 1 Radial Gates Upstream (Pooled) RGU D1 1 1 Radial Gates Downstream #1 RGD1 D2a 1 1 Radial Gates Downstream #2 RGD2 D2b 1 1 Radial Gates Downstream (Pooled) RGD D2 1 1 Central Valley Project Trashrack CVP E1 0 0 Central Valley Project Holding Tank CVPtank E2 0 0 Threemile Slough, Upstream TMS T1a 15 15 Threemile Slough, Downstream TMN T1b 13 13 Threemile Slough (Pooled) TMS/TMN T1 16 16 Jersey Point Upstream 1 JPT G1a 34 34 Jersey Point East (Upstream 2) JPE G1b 42 42 Jersey Point West JPW G1c 44 44 Jersey Point: SJR Route JPE/JPW G1 41 41 Jersey Point: ID Route JPE/JPW G1 2 2 Jersey Point (Pooled) JPE/JPW G1 47 47 False River West FRW H1a 8 8 False River East FRE H1b 8 8 False River: SJR Route FRE/FRW H1 9 9 False River: ID Route FRE/FRW H1 1 1 False River (Pooled) FRE/FRW H1 10 10 Montezuma Slough MTZ T2 0 0 Spoonbill Slough (Pooled) SBS T3 0 0 Chipps Island Upstream 1 MAT G2a 28 28

109

Table 14. (Continued)

Medford Island Releases Survival Release Group Detection Site Site Code Model Code Total Chipps Island East (Upstream 2) MAE G2b 31 31 Chipps Island West MAW G2c 32 32 Chipps Island: SJR Route MAE/MAW G2 32 32 Chipps Island: ID Route MAE/MAW G2 2 2 Chipps Island (Pooled) MAE/MAW G2 35 35 Benicia Bridge: SJR Route BBR G3 24 24 Benicia Bridge: ID Route BBR G3 2 2 Benicia Bridge (Pooled) BBR G3 28 28

110

Table 15. Number of tags observed from each release group at each detection site in 2015 and used in the survival analysis, including predator-type detections. Counts are summed (pooled) over all receivers in array and all routes unless otherwise noted. Route could not be identified for some tags. Durham Ferry Releases Release Group Survival Detection Site Site Code Model Code 1 2 Total Release site at Durham Ferry 643 156 799 Durham Ferry Upstream DFU A0 39 28 67 Durham Ferry Downstream DFD A2 278 12 290 Below Durham Ferry 1 BDF1 A3 44 1 45 Below Durham Ferry 2 BDF2 A4 117 1 118 Banta Carbona BCA A5 74 0 74 Mossdale MOS A6 21 0 21 Head of Old River HOR B0 18 0 18 Lathrop, Upstream SJLU A7a 12 0 12 Lathrop, Downstream SJLD A7b 11 0 11 Lathrop (Pooled) SJL A7 12 0 12 Garwood Bridge, Upstream SJGU A8a 3 0 3 Garwood Bridge, Downstream SJGD A8b 3 0 3 Garwood Bridge (Pooled) SJG A8 3 0 3 Navy Drive Bridge, Upstream SJNBU A9a 2 0 2 Navy Drive Bridge, Downstream SJNBD A9b 2 0 2 Navy Drive Bridge (Pooled) SJNB A9 2 0 2 Rough and Ready Island RRI R1 0 0 0 San Joaquin River Shipping Channel SJS A10 1 0 1 MacDonald Island Upstream MACU A11a 1 0 1 MacDonald Island Downstream MACD A11b 1 0 1 MacDonald Island (Pooled) MAC A11 1 0 1 Turner Cut TCE/TCW F1 0 0 0 Old River East, Upstream OREU B1a 1 0 1 Old River East, Downstream ORED B1b 1 0 1 Old River East (Pooled) ORE B1 1 0 1 Jersey Point JPT/JPE/JPW G1 0 0 0 False River FRE/FRW H1 0 0 0 Chipps Island MAT/MAE/MAW G2 0 0 0 Benicia Bridge BBR G3 0 0 0

Medford Island Releases Survival Release Group Detection Site Site Code Model Code Total Release site at Medford Island 0 491 491 MacDonald Island Upstream MACU A11a 17 17 MacDonald Island Downstream MACD A11b 22 22 MacDonald Island (Pooled) MAC A11 22 22

111

Table 15. (Continued)

Medford Island Releases Survival Release Group Detection Site Site Code Model Code Total Medford Island East MFE A12a 225 225 Medford Island West MFW A12b 221 221 Medford Island (Pooled) MFE/MFW A12 298 298 Columbia Cut, Upstream COLU F2a 9 9 Columbia Cut, Downstream COLD F2b 10 10 Columbia Cut COL F2 10 10 San Joaquin River Disappointment Slough, Upstream SJDU A13a 74 74 San Joaquin River Disappointment Slough, Downstream SJDD A13b 69 69 San Joaquin River Disappointment Slough (Pooled) SJD A13 79 79 Old River at the San Joaquin Mouth, Upstream OSJU B5a 4 4 Old River at the San Joaquin Mouth, Downstream OSJD B5b 5 5 Old River at the San Joaquin Mouth (Pooled) OSJ B5 5 5 Middle River at Middle River MID C3 32 32 Jersey Point Upstream 1 JPT G1a 27 27 Jersey Point East (Upstream 2) JPE G1b 35 35 Jersey Point West JPW G1c 40 40 Jersey Point: SJR Route JPT/JPE/JPW G1 37 37 Jersey Point: ID Route JPT/JPE/JPW G1 2 2 Jersey Point (Pooled) JPT/JPE/JPW G1 43 43 False River West FRW H1a 0 0 False River East FRE H1b 1 1 False River: SJR Route FRE/FRW H1 1 1 False River: ID Route FRE/FRW H1 0 0 False River (Pooled) FRE/FRW H1 1 1 Chipps Island Upstream 1 MAT G2a 27 27 Chipps Island East (Upstream 2) MAE G2b 31 31 Chipps Island West MAW G2c 32 32 Chipps Island: SJR Route MAT/MAE/MAW G2 32 32 Chipps Island: ID Route MAT/MAE/MAW G2 2 2 Chipps Island (Pooled) MAT/MAE/MAW G2 35 35 Benicia Bridge: SJR Route BBR G3 24 24 Benicia Bridge: ID Route BBR G3 2 2 Benicia Bridge (Pooled) BBR G3 28 28

112

Table 16. Number of tags from each release group in 2015 first classified as in a predator at each detection site, based on the predator filter. Only sites with detections are included.

Durham Ferry Release Groups Classified as Predator Classified as Predator on Arrival at Site on Departure from Detection Site and Code Site Survival Detection Site Site Code 1 2 Total 1 2 Total Model Code Durham Ferry Upstream DFU A0 31 16 47 4 18 22 Durham Ferry Downstream DFD A2 13 0 13 22 0 22 Below Durham Ferry 1 BDF1 A3 4 0 4 1 0 1 Below Durham Ferry 2 BDF2 A4 4 0 4 20 0 20 Banta Carbona BCA A5 4 0 4 11 0 11 Mossdale MOS A6 2 0 2 0 0 0 Head of Old River HOR B0 1 0 1 2 0 2 Lathrop SJL A7 0 0 0 1 0 1 Predator Removal Study 4 RS4 N1 0 0 0 0 0 0 Predator Removal Study 5 RS5 N2 0 0 0 2 0 2 Predator Removal Study 6 RS6 N3 0 0 0 1 0 1 Predator Removal Study 7 RS7 N4 0 0 0 0 0 0 Predator Removal Study 8 RS8 N5 0 0 0 0 0 0 Predator Removal Study 9 RS9 N6 0 0 0 0 0 0 Predator Removal Study 10 RS10 N7 0 0 0 0 0 0 Garwood Bridge SJG A8 0 0 0 0 0 0 Navy Drive Bridge SJNB A9 0 0 0 0 0 0 San Joaquin River Shipping Channel SJS A10 0 0 0 0 0 0 MacDonald Island MAC A11 0 0 0 0 0 0 Old River East ORE B1 0 0 0 1 0 1 Total Durham Ferry Tags 59 16 75 65 18 83

Medford Island Release Groups Classified as Predator Classified as Predator on Arrival at Site on Departure from Detection Site and Code Site Survival Detection Site Site Code 1 2 Total 1 2 Total Model Code San Joaquin River Shipping Channel SJS A10 3 3 1 1 MacDonald Island MAC A11 2 2 5 5 Turner Cut TCE/TCW F1 0 0 1 1 Columbia Cut COL F2 0 0 4 4 Medford Island MFE/MFW A12 27 27 51 51 San Joaquin River Disappointment Slough (Pooled) SJD A13 0 0 4 4 West Canal WCL B3 0 0 0 0 Old River at Highway 4 OR4 B4 1 1 0 0 Old River at the San Joaquin Mouth OSJ B5 0 0 0 0

113

Table 16. (Continued)

Medford Island Release Groups Classified as Predator Classified as Predator on Arrival at Site on Departure from Detection Site and Code Site Survival Detection Site Site Code 1 2 Total 1 2 Total Model Code Middle River at Highway 4 MR4 C2 0 0 0 0 Middle River at Middle River MID C3 3 3 1 1 Radial Gates Upstream RGU D1 0 0 0 0 Radial Gates Downstream RGD D2 0 0 0 0 Jersey Point JPT/JPE/JPW G1 0 0 1 1 Chipps Island MAT/MAE/MAW G2 0 0 0 0 Benicia Bridge BBR G3 1 1 1 1 False River FRE/FRW H1 0 0 0 0 Threemile Slough TMS/TMN T1 0 0 0 0 Total Medford Island Tags 0 37 37 0 69 69

114

Table 17. Number of tags from each release group that were detected after release in 2015, including predator-type detections and detections omitted from the survival analysis.

Durham Ferry Releases Release Group 1 2 Total Number Released 643 156 799 Number Detected 332 24 356 Number Detected Downstream of Release Site 324 6 330 Number Detected Upstream of Study Area 332 24 356 Number Detected in Study Area 21 0 21 Number Detected in San Joaquin River Route 12 0 12 Number Detected in Old River Route 1 0 1 Number Assigned to San Joaquin River Route 11 0 11 Number Assigned to Old River Route 1 0 1

Medford Island Releases Release Group 1 2 Total Number Released 0 491 491 Number Detected 396 396 Number Detected Upstream of OSJ/SJD 330 330 Number Detected at Medford Island 323 323 Number Detected Upstream of Medford Island 42 42 Number Detected Downstream of Medford Island 91 91 Number Detected in San Joaquin River Routea 83 83 Number Detected in Interior Delta Route 46 46 Number Assigned to San Joaquin River Route 78 78 Number Assigned to Interior Delta Route 43 43 a = detected at SJD

115

Table 18. Number of tags observed from each release group at each detection site in 2015, excluding predator-type detections. Routes (SJR = San Joaquin River, OR = Old River, ID = Interior Delta) represent route assignment at the head of Old River for DF releases, and upon release for MF releases. Counts are summed (pooled) over all receivers in array and all routes unless otherwise noted. Route could not be identified for some tags. Durham Ferry Releases Survival Release Group Model Detection Site Site Code Code 1 2 Total Release site at Durham Ferry 643 156 799 Durham Ferry Upstream DFU A0 5 19 24 Durham Ferry Downstream DFD A2 282 6 288 Below Durham Ferry 1 BDF1 A3 44 1 45 Below Durham Ferry 2 BDF2 A4 122 1 123 Banta Carbona BCA A5 77 0 77 Mossdale MOS A6 21 0 21 Head of Old River HOR B0 17 0 17 Lathrop, Upstream SJLU A7a 12 0 12 Lathrop, Downstream SJLD A7b 11 0 11 Lathrop (Pooled) SJL A7 12 0 12 Predator Removal Study 4 RS4 N1 8 0 8 Predator Removal Study 5 RS5 N2 7 0 7 Predator Removal Study 6 RS6 N3 5 0 5 Predator Removal Study 7 RS7 N4 4 0 4 Predator Removal Study 8 RS8 N5 3 0 3 Predator Removal Study 9 RS9 N6 6 0 6 Predator Removal Study 10 RS10 N7 3 0 3 Garwood Bridge, Upstream SJGU A8a 3 0 3 Garwood Bridge, Downstream SJGD A8b 3 0 3 Garwood Bridge (Pooled) SJG A8 3 0 3 Navy Drive Bridge, Upstream SJNBU A9a 2 0 2 Navy Drive Bridge, Downstream SJNBD A9b 2 0 2 Navy Drive Bridge (Pooled) SJNB A9 2 0 2 Rough and Ready Island RRI R1 0 0 0 San Joaquin River Shipping Channel SJS A10 1 0 1 MacDonald Island Upstream MACU A11a 1 0 1 MacDonald Island Downstream MACD A11b 1 0 1 MacDonald Island (Pooled) MAC A11 1 0 1 Turner Cut TCE/TCW F1 0 0 0 Medford Island MFE/MFW A12 0 0 0 Columbia Cut COL F2 0 0 0 San Joaquin River Disappointment Slough (Pooled) SJD A13 0 0 0 Old River East, Upstream OREU B1a 1 0 1 Old River East, Downstream ORED B1b 1 0 1 Old River East (Pooled) ORE B1 1 0 1 Old River South ORS B2 0 0 0

116

Table 18. (Continued) Durham Ferry Releases Survival Release Group Detection Site Site Code Model Code Total West Canal WCL B3 0 0 0 Old River at Highway 4 OR4 B4 0 0 0 Old River at the San Joaquin Mouth OSJ B5 0 0 0 Middle River Head MRH C1 0 0 0 Middle River at Highway 4 MR4 C2 0 0 0 Middle River at Middle River MID C3 0 0 0 Radial Gates Upstream RGU D1 0 0 0 Radial Gates Downstream RGD D2 0 0 0 Central Valley Project Trashrack CVP E1 0 0 0 Central Valley Project Holding Tank CVPtank E2 0 0 0 Threemile Slough TMS/TMN T1 0 0 0 Jersey Point JPT/JPE/JPW G1 0 0 0 False River FRE/FRW H1 0 0 0 Montezuma Slough MTZ T2 0 0 0 Spoonbill Slough SBS T3 0 0 0 Chipps Island MAT/MAE/MAW G2 0 0 0 Benicia Bridge BBR G3 0 0 0

Medford Island Releases Survival Release Group Detection Site Site Code Model Code Total Release site at Medford Island 0 491 491 San Joaquin River Shipping Channel SJS A10 11 11 MacDonald Island Upstream MACU A11a 25 25 MacDonald Island Downstream MACD A11b 34 34 MacDonald Island (Pooled) MAC A11 34 34 Turner Cut, Upstream TCE F1a 5 5 Turner Cut, Downstream TCW F1b 5 5 Turner Cut (Pooled) TCE/TCW F1 5 5 Medford Island East MFE A12a 258 258 Medford Island West MFW A12b 277 277 Medford Island (Pooled) MFE/MFW A12 323 323 Columbia Cut, Upstream COLU F2a 10 10 Columbia Cut, Downstream COLD F2b 11 11 Columbia Cut COL F2 11 11 San Joaquin River Disappointment Slough, Upstream SJDU A13a 79 79 San Joaquin River Disappointment Slough, Downstream SJDD A13b 73 73 San Joaquin River Disappointment Slough (Pooled) SJD A13 83 83

117

Table 18. (Continued) Medford Island Releases Survival Release Group Detection Site Site Code Model Code Total West Canal, Upstream WCLU B3a 1 1 West Canal, Downstream WCLD B3b 1 1 West Canal (Pooled) WCL B3 1 1 Old River at Highway 4, Upstream OR4U B4a 1 1 Old River at Highway 4, Downstream OR4D B4b 1 1 Old River at Highway 4 (Pooled) OR4 B4 1 1 Old River at the San Joaquin Mouth, Upstream OSJU B5a 5 5 Old River at the San Joaquin Mouth, Downstream OSJD B5b 6 6 Old River at the San Joaquin Mouth (Pooled) OSJ B5 6 6 Middle River at Highway 4, Upstream MR4U C2a 3 3 Middle River at Highway 4, Downstream MR4D C2b 3 3 Middle River at Highway 4 (Pooled) MR4 C2 3 3 Middle River at Middle River MID C3 32 32 Radial Gates Upstream #1 RGU1 D1a 0 0 Radial Gates Upstream #2 RGU2 D1b 0 0 Radial Gates Upstream (Pooled) RGU D1 0 0 Radial Gates Downstream #1 RGD1 D2a 0 0 Radial Gates Downstream #2 RGD2 D2b 15 15 Radial Gates Downstream (Pooled) RGD D2 13 13 Central Valley Project Trashrack CVP E1 16 16 Central Valley Project Holding Tank CVPtank E2 32 32 Threemile Slough, Upstream TMS T1a 40 40 Threemile Slough, Downstream TMN T1b 42 42 Threemile Slough (Pooled) TMS/TMN T1 39 39 Jersey Point Upstream 1 JPT G1a 2 2 Jersey Point East (Upstream 2) JPE G1b 45 45 Jersey Point West JPW G1c 8 8 Jersey Point: SJR Route JPE/JPW G1 8 8 Jersey Point: ID Route JPE/JPW G1 9 9 Jersey Point (Pooled) JPE/JPW G1 1 1 False River West FRW H1a 10 10 False River East FRE H1b 0 0 False River: SJR Route FRE/FRW H1 0 0 False River: ID Route FRE/FRW H1 28 28 False River (Pooled) FRE/FRW H1 1 1 Montezuma Slough MTZ T2 1 1 Spoonbill Slough (Pooled) SBS T3 1 1 Chipps Island Upstream 1 MAT G2a 1 1

118

Table 18. (Continued)

Medford Island Releases Survival Release Group Detection Site Site Code Model Code Total Chipps Island East (Upstream 2) MAE G2b 31 31 Chipps Island West MAW G2c 32 32 Chipps Island: SJR Route MAE/MAW G2 32 32 Chipps Island: ID Route MAE/MAW G2 2 2 Chipps Island (Pooled) MAE/MAW G2 35 35 Benicia Bridge: SJR Route BBR G3 24 24 Benicia Bridge: ID Route BBR G3 2 2 Benicia Bridge (Pooled) BBR G3 28 28

119

Table 19. Number of tags observed from each release group at each detection site in 2015 and used in the survival analysis, excluding predator-type detections. Counts are summed (pooled) over all receivers in array and all routes unless otherwise noted. Route could not be identified for some tags. Durham Ferry Releases Release Group Survival Detection Site Site Code Model Code 1 2 Total Release site at Durham Ferry 643 156 799 Durham Ferry Upstream DFU A0 15 18 33 Durham Ferry Downstream DFD A2 275 6 281 Below Durham Ferry 1 BDF1 A3 42 1 43 Below Durham Ferry 2 BDF2 A4 117 1 118 Banta Carbona BCA A5 74 0 74 Mossdale MOS A6 21 0 21 Head of Old River HOR B0 17 0 17 Lathrop, Upstream SJLU A7a 11 0 11 Lathrop, Downstream SJLD A7b 10 0 10 Lathrop (Pooled) SJL A7 11 0 11 Garwood Bridge, Upstream SJGU A8a 3 0 3 Garwood Bridge, Downstream SJGD A8b 3 0 3 Garwood Bridge (Pooled) SJG A8 3 0 3 Navy Drive Bridge, Upstream SJNBU A9a 2 0 2 Navy Drive Bridge, Downstream SJNBD A9b 2 0 2 Navy Drive Bridge (Pooled) SJNB A9 2 0 2 Rough and Ready Island RRI R1 0 0 0 San Joaquin River Shipping Channel SJS A10 1 0 1 MacDonald Island Upstream MACU A11a 1 0 1 MacDonald Island Downstream MACD A11b 1 0 1 MacDonald Island (Pooled) MAC A11 1 0 1 Turner Cut TCE/TCW F1 0 0 0 Old River East, Upstream OREU B1a 1 0 1 Old River East, Downstream ORED B1b 1 0 1 Old River East (Pooled) ORE B1 1 0 1 Jersey Point JPT/JPE/JPW G1 0 0 0 False River FRE/FRW H1 0 0 0 Chipps Island MAT/MAE/MAW G2 0 0 0 Benicia Bridge BBR G3 0 0 0

Medford Island Releases Survival Release Group Detection Site Site Code Model Code Total Release site at Medford Island 0 491 491 MacDonald Island Upstream MACU A11a 20 20 MacDonald Island Downstream MACD A11b 24 24 MacDonald Island (Pooled) MAC A11 24 24

120

Table 19. (Continued)

Medford Island Releases Survival Release Group Detection Site Site Code Model Code Total Medford Island East MFE A12a 226 226 Medford Island West MFW A12b 238 238 Medford Island (Pooled) MFE/MFW A12 296 296 Columbia Cut, Upstream COLU F2a 8 8 Columbia Cut, Downstream COLD F2b 9 9 Columbia Cut COL F2 9 9 San Joaquin River Disappointment Slough, Upstream SJDU A13a 73 73 San Joaquin River Disappointment Slough, Downstream SJDD A13b 68 68 San Joaquin River Disappointment Slough (Pooled) SJD A13 78 78 Old River at the San Joaquin Mouth, Upstream OSJU B5a 4 4 Old River at the San Joaquin Mouth, Downstream OSJD B5b 5 5 Old River at the San Joaquin Mouth (Pooled) OSJ B5 5 5 Middle River at Middle River MID C3 31 31 Jersey Point Upstream 1 JPT G1a 25 25 Jersey Point East (Upstream 2) JPE G1b 33 33 Jersey Point West JPW G1c 38 38 Jersey Point: SJR Route JPT/JPE/JPW G1 35 35 Jersey Point: ID Route JPT/JPE/JPW G1 2 2 Jersey Point (Pooled) JPT/JPE/JPW G1 41 41 False River West FRW H1a 0 0 False River East FRE H1b 1 1 False River: SJR Route FRE/FRW H1 1 1 False River: ID Route FRE/FRW H1 0 0 False River (Pooled) FRE/FRW H1 1 1 Chipps Island Upstream 1 MAT G2a 27 27 Chipps Island East (Upstream 2) MAE G2b 31 31 Chipps Island West MAW G2c 32 32 Chipps Island: SJR Route MAT/MAE/MAW G2 32 32 Chipps Island: ID Route MAT/MAE/MAW G2 2 2 Chipps Island (Pooled) MAT/MAE/MAW G2 35 35 Benicia Bridge: SJR Route BBR G3 24 24 Benicia Bridge: ID Route BBR G3 2 2 Benicia Bridge (Pooled) BBR G3 28 28

121

Table 20. Number of juvenile Chinook salmon tagged by each surgeon in each release group and site during the 2015 Salmon Survival Study. DF = Durham Ferry release, MI = Medford Island release.

Release Site - Release Group Surgeon DF - 1 DF - 2 MI - 2 Total Tags A 214 52 164 430 B 215 52 164 431 C 214 52 163 429 Total Tags 643 156 491 1,290

122

Table 21. Release size, location, and counts of tag detections at key detection sites by surgeon in 2015, excluding predator- type detections. * = used in chi-square test of independence. a = pooled with other sites for chi-square test of independence.

Surgeon Detection Site A B C Release at Durham Ferry* 266 267 267 Below Durham Ferry 1 (BDF1)* 19 13 11 Below Durham Ferry 2 (BDF2)* 46 31 41 Banta Carbona (BCA)* 27 23 24 Mossdale (MOS)* 10 6 5 Head of Old River (HOR) 9 5 3 Lathrop (SJL) 5 4 2 Garwood Bridge (SJG) 1 1 1 Navy Drive Bridge (SJNB) 1 0 1 Old River East (ORE) 1 0 0 Release at Medford Island* 164 164 163 Medford Island (MFE/MFW)* 96 101 99 Disappointment Slough*a 28 22 28 Old River at San Joaquin (OSJ)*a 2 1 2 Jersey Point (JPT/JPE/JPW)* 16 11 14 Chipps Island (MAT/MAE/MAW)* 13 10 12 Benicia Bridge (BBR)* 10 8 10

123

Table 22. Performance metric estimates (standard error in parentheses) for tagged juvenile Chinook salmon released in the 2015 tagging study, excluding predator-type detections. Southern Delta ("SD") survival extended to MacDonald Island and Turner Cut in Route A. Population-level estimates were from pooled release groups.

Release Group Parameter 1 2 Population Estimate

ψAA NA NA NA

ψAF NA NA NA a ψA 0.92 (0.08) NA 0.92 (0.08) a ψB 0.08 (0.08) NA 0.08 (0.08) b SA 0 (0) NA 0 (0)

SB NA NA NA b STotal 0 (0) NA 0 (0) bc c SA(SD) 0.05 (0.05) NA 0.05 (0.05) b φA1,A6 0.03 (0.01) 0 (0) 0.03 (<0.01)

SA11,G2 NA 0 (0) 0 (0)

SF2,G2 NA 0 (0) 0 (0) d SA13,G2 NA 0.42 (0.06) 0.42 (0.06)

d SB5,G2 NA 0.32 (0.18) 0.32 (0.18)

SA12,G2 NA NA NA

SMF,G2 NA 0.08 (0.01) 0.08 (0.01) a = significant preference for route A (San Joaquin River Route) (α=0.05) for all release groups and for population estimate b = assumed 100% detection probability at downstream receiver site(s) c = assumed 100% detection probability at Turner Cut d = no significant difference between SJR route and Interior Delta route estimates (P = 0.6239)

124

Table 23. Performance metric estimates (standard error in parentheses) for tagged juvenile Chinook salmon released in the 2015 tagging study, including predator-type detections. Southern Delta ("SD") survival extended to MacDonald Island and Turner Cut in Route A. Population-level estimates were from pooled release groups.

Release Group Parameter 1 2 Population Estimate

ψAA NA NA NA

ψAF NA NA NA a ψA 0.92 (0.07) NA 0.92 (0.07) a ψB 0.08 (0.07) NA 0.08 (0.07) b SA 0 (0) NA 0 (0)

SB NA NA NA b STotal 0 (0) NA 0 (0) bc c SA(SD) 0.05 (0.05) NA 0.05 (0.05) b φA1,A6 0.03 (0.01) 0 (0) 0.03 (<0.01)

SA11,G2 NA 0 (0) 0 (0)

SF2,G2 NA 0 (0) 0 (0) d SA13,G2 NA 0.42 (0.06) 0.42 (0.06)

d SB5,G2 NA 0.31 (0.17) 0.31 (0.17)

SA12,G2 NA NA NA

SMF,G2 NA 0.08 (0.01) 0.08 (0.01) a = significant preference for route A (San Joaquin River Route) (α=0.05) for all release groups and for population estimate b = assumed 100% detection probability at downstream receiver site(s) c = assumed 100% detection probability at Turner Cut d = no significant difference between SJR route and Interior Delta route estimates (P = 0.6239)

125

Table 24. Estimates (standard errors in parentheses) of model survival and transition parameters by release group, and of the difference (∆) between release group estimates: ∆ = Release group 1 - Release group 2. P = P-value from two-sized z-test of ∆=0. Estimates were based on data that excluded predator-type detections. * = significant (positive) difference between release groups for family-wise α=0.10.

Parameter Release 1 Release 2 ∆ P

φA1,A0 0.02 (0.01) 0.12 (0.02) -0.09 (0.04) 0.0131* φA1,A2 0.61 (0.03) 0.04 (0.02) 0.57 (0.02) <0.0001* φA1,A6 0.03 (0.01) 0 (0) 0.03 (0.01) <0.0001* α = significant negative difference between release groups for family-wise α=0.10

126

Table 25a. Average travel time in days (harmonic mean) of acoustic-tagged juvenile Chinook salmon from release during the 2015 tagging study, excluding predator-type detections. Standard errors are in parentheses. Only sites with detections are included. See Table 25b for travel time from release with predator-type detections.

Durham Ferry Releases: Without Predator-Type Detections Release 1 Release 2 Pooled Detection Site and Route N Travel Time N Travel Time N Travel Time Durham Ferry Upstream (DFU) 15 0.34 (0.06) 18 0.25 (0.04) 33 0.29 (0.03) Durham Ferry Downstream (DFD) 275 0.09 (0.01) 6 1.27 (0.16) 281 0.09 (0.01) Below Durham Ferry 1 (BDF1) 42 0.39 (0.06) 1 3.97 (NA) 43 0.40 (0.07) Below Durham Ferry 2 (BDF2) 117 0.51 (0.04) 1 4.09 (NA) 118 0.51 (0.04) Banta Carbona (BCA) 74 0.79 (0.07) 0 NA 74 0.79 (0.07) Mossdale (MOS) 21 1.40 (0.22) 0 NA 21 1.40 (0.22) Head of Old River (HOR) 17 1.48 (0.26) 0 NA 17 1.48 (0.26) Lathrop (SJL) 11 1.53 (0.36) 0 NA 11 1.53 (0.36) Garwood Bridge (SJG) 3 3.03 (0.80) 0 NA 3 3.03 (0.80) Navy Drive Bridge (SJNB) 2 2.69 (0.74) 0 NA 2 2.69 (0.74) MacDonald Island (MAC) 1 4.05 (NA) 0 NA 1 4.05 (NA) Old River East (ORE) 1 4.89 (NA) 0 NA 1 4.89 (NA)

Medford Island Releases: Without Predator-Type Detections Release 1 Release 2 Pooled Detection Site and Route N Travel Time N Travel Time N Travel Time MacDonald Island (MAC) 24 0.53 (0.06) 24 0.53 (0.06) Medford Island (MFE/MFW) 296 0.02 (<0.01) 296 0.02 (<0.01) Disappointment Slough (SJD) 78 0.63 (0.08) 78 0.63 (0.08) Old River at San Joaquin (OSJ) 5 0.66 (0.23) 5 0.66 (0.23) Columbia Cut (COL) 9 1.04 (0.13) 9 1.04 (0.13) Middle River at Middle River (MID) 31 0.51 (0.04) 31 0.51 (0.04) Jersey Point (JPT/JPE/JPW), via SJD 35 2.02 (0.18) 35 2.02 (0.18) Jersey Point (JPT/JPE/JPW), via OSJ 2 1.81 (0.32) 2 1.81 (0.32) Jersey Point (JPT/JPE/JPW) 41 1.95 (0.15) 41 1.95 (0.15) False River (FRE/FRW), via SJD 1 5.34 (NA) 1 5.34 (NA) False River (FRE/FRW), via OSJ 0 NA 0 NA False River (FRE/FRW) 1 5.34 (NA) 1 5.34 (NA) Chipps Island (MAT/MAE/MAW), via SJD 32 3.72 (0.25) 32 3.72 (0.25) Chipps Island (MAT/MAE/MAW), via OSJ 2 3.02 (0.48) 2 3.02 (0.48) Chipps Island (MAT/MAE/MAW) 35 3.64 (0.23) 35 3.64 (0.23) Benicia Bridge (BBR) 28 4.59 (0.30) 28 4.59 (0.30)

127

Table 25b. Average travel time in days (harmonic mean) of acoustic-tagged juvenile Chinook salmon from release during the 2015 tagging study, including predator-type detections. Standard errors are in parentheses. Only sites with detections are included. See Table 25a for travel time from release without predator-type detections.

Durham Ferry Releases: With Predator-Type Detections Release 1 Release 2 Pooled Detection Site and Route N Travel Time N Travel Time N Travel Time Durham Ferry Upstream (DFU) 39 0.75 (0.14) 28 0.49 (0.09) 67 0.61 (0.08) Durham Ferry Downstream (DFD) 278 0.10 (0.01) 12 0.97 (0.17) 290 0.10 (0.01) Below Durham Ferry 1 (BDF1) 44 0.41 (0.07) 1 3.97 (NA) 45 0.42 (0.07) Below Durham Ferry 2 (BDF2) 117 0.55 (0.05) 1 4.09 (NA) 118 0.55 (0.05) Banta Carbona (BCA) 74 0.82 (0.08) 0 NA 74 0.82 (0.08) Mossdale (MOS) 21 1.46 (0.25) 0 NA 21 1.46 (0.25) Head of Old River (HOR) 16 1.70 (0.24) 0 NA 16 1.70 (0.24) Lathrop (SJL) 12 1.97 (0.24) 0 NA 12 1.97 (0.24) Garwood Bridge (SJG) 3 3.03 (0.80) 0 NA 3 3.03 (0.80) Navy Drive Bridge (SJNB) 2 2.69 (0.74) 0 NA 2 2.69 (0.74) MacDonald Island (MAC) 1 4.05 (NA) 0 NA 1 4.05 (NA) Old River East (ORE) 1 7.37 (NA) 0 NA 1 7.37 (NA)

Medford Island Releases: With Predator-Type Detections Release 1 Release 2 Pooled Detection Site and Route N Travel Time N Travel Time N Travel Time MacDonald Island (MAC) 22 0.65 (0.09) 22 0.65 (0.09) Medford Island (MFE/MFW) 298 0.02 (<0.01) 298 0.02 (<0.01) Disappointment Slough (SJD) 79 0.64 (0.08) 79 0.64 (0.08) Old River at San Joaquin (OSJ) 5 0.66 (0.23) 5 0.66 (0.23) Columbia Cut (COL) 10 1.14 (0.17) 10 1.14 (0.17) Middle River at Middle River (MID) 32 0.63 (0.08) 32 0.63 (0.08) Jersey Point (JPT/JPE/JPW), via SJD 37 2.10 (0.19) 37 2.10 (0.19) Jersey Point (JPT/JPE/JPW), via OSJ 2 1.81 (0.32) 2 1.81 (0.32) Jersey Point (JPT/JPE/JPW) 43 2.02 (0.16) 43 2.02 (0.16) False River (FRE/FRW), via SJD 1 5.34 (NA) 1 5.34 (NA) False River (FRE/FRW), via OSJ 0 NA 0 NA False River (FRE/FRW) 1 5.34 (NA) 1 5.34 (NA) Chipps Island (MAT/MAE/MAW), via SJD 32 3.72 (0.25) 32 3.72 (0.25) Chipps Island (MAT/MAE/MAW), via OSJ 2 3.02 (0.48) 2 3.02 (0.48) Chipps Island (MAT/MAE/MAW) 35 3.64 (0.23) 35 3.64 (0.23) Benicia Bridge (BBR) 28 4.62 (0.31) 28 4.62 (0.31)

128

Table 26a. Average travel time in days (harmonic mean) of acoustic-tagged juvenile Chinook salmon through the San Joaquin River Delta during the 2015 tagging study, excluding predator detections. Standard errors are in parentheses. Reaches beginning at sites with no detections are not shown. See Table 26b for reach travel time with predator-type detections.

Durham Ferry Releases: Without Predator-Type Detections Release 1 Release 2 Pooled Upstream Boundary Downstream Boundary N Travel Time N Travel Time N Travel Time Durham Ferry (Release) DFU 15 0.34 (0.06) 18 0.25 (0.04) 33 0.29 (0.03) DFD 275 0.09 (0.01) 6 1.27 (0.16) 281 0.09 (0.01) DFD BDF1 27 0.28 (0.06) 1 1.03 (NA) 28 0.29 (0.06) BCA 53 0.58 (0.07) 0 NA 53 0.58 (0.07) BDF1 BDF2 31 0.07 (0.01) 1 0.12 (NA) 32 0.07 (0.01) BDF2 BCA 63 0.19 (0.02) 0 NA 63 0.19 (0.02) BCA MOS 21 0.53 (0.08) 0 NA 21 0.53 (0.08) MOS HOR 17 0.13 (0.03) 0 NA 17 0.13 (0.03) SJL 11 0.23 (0.06) 0 NA 11 0.23 (0.06) ORE 1 2.89 (NA) 0 NA 1 2.89 (NA) HOR SJL 11 0.08 (0.02) 0 NA 11 0.08 (0.02) ORE 1 1.83 (NA) 0 NA 1 1.83 (NA) SJL SJG 3 1.11 (0.36) 0 NA 3 1.11 (0.36) SJG SJNB 2 0.06 (0.03) 0 NA 2 0.06 (0.03) RRI 0 NA 0 NA 0 NA SJNB SJS 1 0.29 (NA) 0 NA 1 0.29 (NA) MAC 1 0.35 (NA) 0 NA 1 0.35 (NA) TCE/TCW 0 NA 0 NA 0 NA MAC JPT/JPE/JPW 0 NA 0 NA 0 NA MAC MAT/MAE/MAW 0 NA 0 NA 0 NA ORE 0 NA 0 NA 0 NA MAC BBR 0 NA 0 NA 0 NA ORE 0 NA 0 NA 0 NA

129

Table 26a. (Continued)

Medford Island Releases: Without Predator-Type Detections Release 1 Release 2 Pooled Upstream Boundary Downstream Boundary N Travel Time N Travel Time N Travel Time Medford Island (Release) MAC 24 0.53 (0.06) 24 0.53 (0.06) MFE/MFW 296 0.02 (<0.01) 296 0.02 (<0.01) SJD 78 0.63 (0.08) 78 0.63 (0.08) OSJ 5 0.66 (0.23) 5 0.66 (0.23) COL 9 1.04 (0.13) 9 1.04 (0.13) MID 31 0.51 (0.04) 31 0.51 (0.04) MFE/MFW SJD 33 0.52 (0.10) 33 0.52 (0.10) OSJ 3 0.78 (0.15) 3 0.78 (0.15) MID 12 0.45 (0.04) 12 0.45 (0.04) COL MID 2 0.16 (0.10) 2 0.16 (0.10) SJD JPT/JPE/JPW 35 0.93 (0.14) 35 0.93 (0.14) OSJ 2 0.46 (0.16) 2 0.46 (0.16) SJD FRE/FRW 1 3.95 (NA) 1 3.95 (NA) OSJ 0 NA 0 NA JPT/JPE/JPW MAT/MAE/MAW 29 1.40 (0.10) 29 1.40 (0.10) MAC 0 NA 0 NA COL 0 NA 0 NA MID 0 NA 0 NA MAT/MAE/MAW BBR 26 0.89 (0.12) 26 0.89 (0.12)

130

Table 26b. Average travel time in days (harmonic mean) of acoustic-tagged juvenile Chinook salmon through the San Joaquin River Delta during the 2015 tagging study, including predator detections. Standard errors are in parentheses. Reaches beginning at sites with no detections are not shown. See Table 26a for reach travel time without predator-type detections.

Durham Ferry Releases: With Predator-Type Detections Release 1 Release 2 Pooled Upstream Boundary Downstream Boundary N Travel Time N Travel Time N Travel Time Durham Ferry (Release) DFU 39 0.75 (0.14) 28 0.49 (0.09) 67 0.61 (0.08) DFD 278 0.10 (0.01) 12 0.97 (0.17) 290 0.10 (0.01) DFD BDF1 28 0.32 (0.08) 1 1.03 (NA) 29 0.32 (0.08) BCA 53 0.61 (0.08) 0 NA 53 0.61 (0.08) BDF1 BDF2 31 0.08 (0.01) 1 0.12 (NA) 32 0.08 (0.01) BDF2 BCA 63 0.20 (0.02) 0 NA 63 0.20 (0.02) BCA MOS 21 0.54 (0.09) 0 NA 21 0.54 (0.09) MOS HOR 16 0.14 (0.03) 0 NA 16 0.14 (0.03) SJL 12 0.31 (0.09) 0 NA 12 0.31 (0.09) ORE 1 5.36 (NA) 0 NA 1 5.36 (NA) HOR SJL 12 0.10 (0.03) 0 NA 12 0.10 (0.03) ORE 1 4.31 (NA) 0 NA 1 4.31 (NA) SJL SJG 3 1.11 (0.36) 0 NA 3 1.11 (0.36) SJG SJNB 2 0.06 (0.03) 0 NA 2 0.06 (0.03) RRI 0 NA 0 NA 0 NA SJNB SJS 1 0.29 (NA) 0 NA 1 0.29 (NA) MAC 1 0.35 (NA) 0 NA 1 0.35 (NA) TCE/TCW 0 NA 0 NA 0 NA MAC JPT/JPE/JPW 0 NA 0 NA 0 NA MAC MAT/MAE/MAW 0 NA 0 NA 0 NA ORE 0 NA 0 NA 0 NA MAC BBR 0 NA 0 NA 0 NA ORE 0 NA 0 NA 0 NA

131

Table 26b. (Continued)

Medford Island Releases: With Predator-Type Detections Release 1 Release 2 Pooled Upstream Boundary Downstream Boundary N Travel Time N Travel Time N Travel Time Medford Island (Release) MAC 22 0.65 (0.09) 22 0.65 (0.09) MFE/MFW 298 0.02 (<0.01) 298 0.02 (<0.01) SJD 79 0.64 (0.08) 79 0.64 (0.08) OSJ 5 0.66 (0.23) 5 0.66 (0.23) COL 10 1.14 (0.17) 10 1.14 (0.17) MID 32 0.63 (0.08) 32 0.63 (0.08) MFE/MFW SJD 33 0.52 (0.10) 33 0.52 (0.10) OSJ 3 0.78 (0.15) 3 0.78 (0.15) MID 12 0.63 (0.13) 12 0.63 (0.13) COL MID 3 0.19 (0.09) 3 0.19 (0.09) SJD JPT/JPE/JPW 37 0.97 (0.14) 37 0.97 (0.14) OSJ 2 0.46 (0.16) 2 0.46 (0.16) SJD FRE/FRW 1 3.95 (NA) 1 3.95 (NA) OSJ 0 NA 0 NA JPT/JPE/JPW MAT/MAE/MAW 29 1.40 (0.10) 29 1.40 (0.10) MAC 0 NA 0 NA COL 0 NA 0 NA MID 0 NA 0 NA MAT/MAE/MAW BBR 26 0.90 (0.12) 26 0.90 (0.12)

132

Appendices Appendix A. Sample Size for 2015 Chinook Tagging Study

Prepared for:

Pat Brandes U.S. Fish and Wildlife Service Stockton, CA

Prepared by:

Rebecca Buchanan University of Washington Seattle, WA

2 December 2014

133

Summary of Recommendations

Assume

• Barrier is installed at head of Old River • Approximately 25% of fish reaching the Turner Cut Junction enter Turner Cut • The survival probability from Turner Cut to Chipps Island is 0.05 for fish entering Turner Cut

Single Release at Durham Ferry Estimating survival from Durham Ferry to Chipps Island

• If the overall detection probability at Chipps Island is 0.75: o If survival is moderate throughout the study area (i.e., 0.7 from release to Mossdale, and 0.03 from Mossdale to Chipps Island), then a Durham Ferry release of 500 should be sufficient to estimate survival to Chipps Island o If survival is low throughout the study area (i.e., 0.4 from release to Mossdale, and 0.02 from Mossdale to Chipps Island), then a Durham Ferry release of 1,500 will be required to estimate survival to Chipps Island. o If survival is very low throughout the study area (i.e., 0.2 from release to Mossdale, and <0.01 from Mossdale to Chipps Island), then it is unlikely that survival to Chipps Island will be estimable with any feasible release size. • If the overall detection probability at Chipps Island is 0.99: o Moderate survival: release of 150 at Durham Ferry o Low survival: release of 450 at Durham Ferry o Very low survival: not estimable Estimating survival from Mossdale to Turner Cut

• Estimating survival from Mossdale to Turner Cut is expected to be feasible with a release of 100 at Durham Ferry (300 if survival is very low).

Estimating survival from Turner Cut to Chipps Island

• Estimating survival from Turner Cut via either route requires releasing 100-450 at Durham Ferry if survival is moderate, and 175-800 at Durham Ferry if survival is low. • It will generally not be practical to estimate survival from Turner Cut if survival is very low.

Estimating survival from Mossdale to Chipps Island via the Old River route

• In the presence of a barrier at the head of Old River, it is likely that a reliable survival estimable will not be attainable in the Old River route with available sample sizes.

134

Minimum release size to estimate model parameters based on simulations using a single release at Durham Ferry under low survival scenario.

Minimum release size with

Parameter Definition Value PG2=0.75 PG2=0.99

SR Survival from DF to Chipps Island 0.008 1,500 450

SR0 Survival from DF to Mossdale 0.4 100 100

SA1 Survival from Mossdale to Turner Cut 0.3 100 100

SA2 Survival from Turner Cut to Chipps via SJR 0.07 800 250

SA Survival from Mossdale to Chipps via SJR 0.02 800 250

SB Survival from Mossdale to Chipps via Old River 0.07 1,500 1,000 (including facilities)

SF Survival from Turner Cut to Chipps via Turner Cut 0.05 800 250

ψA1 Route selection at HOR (stay in SJR) 0.98 100 100

ψA2 Route selection at Turner Cut (stay in SJR) 0.95 200 200

Minimum release size to estimate model parameters based on simulations using a single release at Durham Ferry under medium survival scenario.

Minimum release size with

Parameter Definition Value PG2=0.75 PG2=0.99

SR Survival from DF to Chipps Island 0.025 500 150

SR0 Survival from DF to Mossdale 0.7 100 100

SA1 Survival from Mossdale to Turner Cut 0.4 100 100

SA2 Survival from Turner Cut to Chipps via SJR 0.1 250 100

SA Survival from Mossdale to Chipps via SJR 0.035 250 100

SB Survival from Mossdale to Chipps via Old River 0.1 1,100 750 (including facilities)

SF Survival from Turner Cut to Chipps via Turner Cut 0.05 250 100

ψA1 Route selection at HOR (stay in SJR) 0.98 100 100

ψA2 Route selection at Turner Cut (stay in SJR) 0.95 100 100

Primary Release at Durham Ferry with Supplemental Release in SJR downstream of the OR flow split • Releasing a supplemental release in the San Joaquin River downstream of the head of Old River lowered the total necessary sample size to estimate survival from Durham Ferry to Chipps Island when survival was low or medium. • A supplemental release had little benefit in parameter estimation when survival was either very low or high.

135

Other Design Considerations for Sample Size • The detection probability at Chipps Island has a large effect on necessary sample sizes. Effort should be made to assure as high a detection probability at Chipps Island as is feasible.

2015 Study • With a single release of 648 fish at Durham Ferry, we can expect to estimate key regional survival parameters under all scenarios except those with low or very low survival downstream of Mossdale. • High detection probabilities at Chipps Island facilitate estimation of model parameters but do not fully compensate for very low survival probabilities at a release size of 648. • With a single release of 1,296 fish (pooled across 2 groups of 648) at Durham Ferry, parameters were estimable in most cases unless survival was very low. • Reasonable parameter estimates may be attainable under conditions of very low survival if detection probabilities at Chipps Island are very high (e.g., 0.99) and the two release groups of 648 are pooled.

136

Introduction This analysis updates previous sample size analyses for Chinook salmon acoustic tagging studies in the South Delta. It uses data simulation to determine the appropriate release size at Durham Ferry under a variety of conditions that range from very low survival to relatively high survival throughout the study area. Conditions with varying levels of survival in different regions of the study area are also considered. Primary focus is on estimating survival from Durham Ferry to Chipps Island both overall and in particular routes, based on either a single release at Durham Ferry or a primary release at Durham Ferry paired with a supplemental release downstream of the head of Old River in the San Joaquin River.

The 2015 tagging study is expected to make two releases of 648 fish at Durham Ferry. Parameters that can be reliably estimated using a release of 648 fish, or a pooled release of 1,296 (=2 × 648), are identified based on simulation results. The extensive simulations for different release sizes are provided for assistance in planning future studies.

Methods Detection data were simulated 5,000 times for a simplified survival model using a variety of parameter sets and candidate release sizes. For each simulated data set, parameter estimates were computed. For each parameter set and release size combination, the mean, maximum, and standard deviation of the sampling distribution of parameter estimates was computed, as well as the number of simulations in which each parameter was estimable. The minimum release size of those considered was identified that met several criteria on the estimability of parameters and the validity of the estimates.

Parameters estimated in the model were:

• Overall survival from Durham Ferry to Chipps Island (sR)

• Survival from Durham Ferry to the head of Old River (sR0) • Survival from the head of Old River to Chipps Island in both the San Joaquin River route and the

Old River route (sA, sB)

• Survival from the head of Old River to Turner Cut in the San Joaquin River (sA1) • Survival from the Turner Cut junction to Chipps Island in both the San Joaquin River route and

the Turner Cut route (sA2, sF)

• Route entrainment at the head of Old River and at Turner Cut (ψA1, ψA2) • Detection probabilities at the dual arrays at the 5 detection sites

Two types of release scenarios were considered: a single release at Durham Ferry, and a primary release at Durham Ferry paired with a supplemental release in the San Joaquin River between the head of Old River and Turner Cut.

Simulations were performed using parameters to represent a range of conditions, including similar levels of survival parameters (e.g., low or high) throughout all regions of the study area and also parameter combinations similar to those estimated in previous studies. Candidate release sizes ranged from 100 – 4,000 at Durham Ferry, and 15 – 596 at the supplemental release site (Table A1). In each

137

case, the total sample size (Durham Ferry release + supplemental release) was minimized for each parameter set.

Table A1. Release sizes considered for each survival scenario and release protocol.

Release Size Supplemental release? Durham Ferry Supplemental No 100, 150, 175, 200, 250, 300, 350, 400, 450, 500, 550, 600, 648, 700, 750, 800, 900, 1000, 1100, 1296, 1400, 1500, 1750, 2000, 3000, 4000 Yes 100, 150, 175, 200, 250, 300, 350, 400, 15, 25, 46, 48, 75, 96, 98, 125, 146, 450, 500, 550, 600, 700, 750, 800, 850, 148, 175, 196, 198, 246, 248, 296, 298, 900, 950, 1000, 1050, 1100, 1150, 346, 348, 396, 398, 446, 448, 473, 496, 1200, 1250 498, 546, 548, 596

Survival Model A reduced survival model was constructed based on the full survival model used in analysis of the 2012 Chinook salmon and steelhead studies. Survival is estimated from Durham Ferry to Mossdale (equated to the head of Old River in the reduced model) (SR0), from Mossdale/head of Old River to Chipps Island via the Old River route (SB) and to Turner Cut via the San Joaquin River (SA1), and from Turner Cut to

Chipps Island via both the San Joaquin River (SA2) and Turner Cut (SF). Route entrainment probabilities

are estimated at the head of Old River (ψA1) and Turner Cut (ψA2), and detection probabilities are modeled at all detection sites (dual arrays) (Figure A1). The total survival from the head of Old River to

Chipps Island via the San Joaquin River route (SA) and the overall survival from Durham Ferry to Chipps

Island via any route (SR) are defined based on the model parameters.

138

SA

SR

Figure A1. Schematic of model used in data simulations, with parameters: survival from Durham Ferry to Chipps Island (SR), survival from head of Old River/Mossdale in the San Joaquin River route (SA) and Old River route (SB), probability of survival from head of Old River/Mossdale to Turner Cut (SA1), survival from Turner Cut to Chipps Island in the San Joaquin River route (SA2) and Turner Cut route (SF), and the probabilities of remaining in the San Joaquin River at the head of Old River (ψA1) and at Turner Cut (ψA2). Other parameters are survival from Durham Ferry to the head of Old River/Mossdale (SR0), and detection probabilities on the upstream and downstream arrays at Lathrop (PA1a, PA1b), Old River (PB1a, PB1b), Turner Cut in the San Joaquin (PA2a, PA2b) and in Turner Cut (PF1a, PF1b), and at Chipps Island (PG2a, PG2b).

Parameters Sets Parameter values were selected based on the range of parameter estimates from the South Delta Chinook tagging studies in 2009 – 2012 (Table A2). This resulted in a wide range of parameter values considered. Exploring all possible permutations of the parameter values across the 6 survival parameters, 2 route entrainment parameters, and 10 detection probabilities was intractable. Instead, simulations were run on sets of parameters that were similar to parameter combinations observed in previous studies, or represented the extremes of possible survival combinations. Scenarios were selected to include very low, low, medium, and relatively high survival through the study area, as well as combinations with varying survival levels in different regions (“mixed”), which represented scenarios closest to the 2010, 2011, and 2012 study year results. The scenarios forming the basis of the results

139

and recommendations are identified in Table A3. Results for more extensive scenarios are provided in Appendix A1.

Table A2. Parameter values used in simulations. a = values were derived from other parameter values.

Parameter Definition Values a SR Survival from DF to Chipps Island 0.0004 – 0.127

SR0 Survival from DF to Mossdale 0.2, 0.4, 0.7, 0.9

SA1 Survival from Mossdale to Turner Cut 0.1, 0.3, 0.4, 0.5

SA2 Survival from Turner Cut to Chipps via SJR 0.02, 0.07, 0.10, 0.15, 0.16, 0.20 a SA Survival from Mossdale to Chipps via SJR 0.002 – 0.08

SB Survival from Mossdale to Chipps via Old River 0.01, 0.02, 0.05, 0.06, 0.07, 0.10, (including facilities) 0.15, 0.16, 0.20

SF Survival from Turner Cut to Chipps via Turner Cut 0.01, 0.05

ψA1 Route selection at HOR (stay in SJR) 0.4, 0.5, 0.6, 0.98

ψA2 Route selection at Turner Cut (stay in SJR) 0.75, 0.95

pA1a, pA1b Detection probabilities on upstream and downstream 0.85, 0.90 lines at SJR dual array at head of Old River

pA2a, pA2b Detection probabilities on upstream and downstream 0.85 lines at SJR dual array at Turner Cut

pB1a, pB1b Detection probabilities on upstream and downstream 0.85, 0.90 lines at Old River dual array at head of Old River

pF1a, pF1b Detection probabilities on upstream and downstream 0.85 lines at Turner Cut dual array

pG2a, pG2b Detection probabilities on upstream and downstream 0.5, 0.9 lines at Chipps Island dual array

Table A3. Parameter sets used in data simulations to estimate survival to Turner Cut and Chipps Island and route selection at Old River and Turner Cut. “Mixed” scenarios had mixes of low and high survival probabilities in different regions of the study area, whereas as other scenarios assumed similar-valued survival probabilities throughout the study area. Values of SR and SA were computed from values of other parameters. For each scenario, sF=0.05, ψA2=0.75, pG2a=pG2b=0.5 or 0.9 (equivalent to pG2=0.75 or 0.99), and pA2a = pA2b = pF1a = pF1b = 0.85.

Scenario SR0 SA1 SA2 SB ψA1 SR SA pA1a, pA1b, pB1a, pB1b 1: Very Low 0.2 0.1 0.02 0.01 0.98 0.0006 0.003 0.9 2: Low 0.4 0.3 0.07 0.07 0.98 0.008 0.020 0.9 3: Low Mixed 0.4 0.3 0.15 0.2 0.98 0.016 0.038 0.85 4: Medium 0.7 0.4 0.10 0.10 0.98 0.025 0.035 0.9 5: Medium Mixed 0.9 0.1 0.07 0.07 0.50 0.034 0.007 0.85 6: High Mixed 0.9 0.4 0.15 0.06 0.50 0.050 0.050 0.85 7: High 0.9 0.5 0.16 0.16 0.98 0.061 0.066 0.9

140

Criteria Four criteria were used to determine the minimum sample size necessary (Table A4). Criterion C1 was that the parameter was estimable in at least 95% of simulations. Parameters may be unestimable if the structure of the data is too sparse to estimate detection probabilities at detection sites bounding the reach in question, or if no tags are detected at the upstream boundary of the reach in question. Criterion C2 is that the parameter not be greater than 1.1 in 95% of the simulations. Because maximum likelihood estimates are (asymptotically) unbiased, a parameter whose true value is very high may have an estimate that is greater than 1. Estimates that are much greater than 1 raise suspicion on the entire set of parameter estimates. Criterion C3 is that the standard error on the parameter estimate not be greater than 0.10. The standard error is calculated as the standard deviation of the observed parameter estimates over all simulations. Criterion C4 is that the difference between the estimate and the true value of the parameter be small (≤0.05); the mean difference is calculated over all simulations.

Table A4. Criteria used for identification of minimum sample size necessary to estimate a model parameter, assuming 5000 simulations.

Criterion Definition C1 Parameter is estimable in at least 95% of simulations (4750 or more) C2 Probability estimate is not greater than 1.1 in 95% of simulations (4750 or more) C3 Standard error on parameter estimate is not greater than 0.10 C4 Difference between average of parameter estimates and true parameter value is not greater than 0.05

2015 Study The 2015 Chinook salmon tagging study is planning to release two groups of 648 fish at Durham Ferry. Parameters that can be reliably estimated using a release of 648 or of 1,296 (=2 × 648) were identified based on the simulations performed above.

141

Results Although a wide range of parameter values were considered for each survival scenario, recommendations are based on the following assumptions: approximately 25% of fish arriving at Turner

Cut enter Turner Cut (ψA2=0.75), and survival of Turner Cut fish to Chipps Island is SF =0.05. Most scenarios (scenarios 1 – 4, 7) also assume that the barrier is installed at the head of Old River (ψA1=0.98).

Results are detailed for the scenarios identified in Table A3. Results for all scenarios are detailed in Appendix A1.

Single Release at Durham Ferry, No Supplemental Release Under the very low survival scenario (scenario 1), in which all reaches have very low survival, and only

moderate detection probabilities at Chipps Island (pG2=0.75), only the upstream reaches are expected to have reliable survival estimates with sample sizes < 4,000 (Table A5). Estimating survival to Mossdale and to Turner Cut is expected to require 100 – 300 fish released at Durham Ferry. If detection probabilities are very high at Chipps Island, then all parameters except SR (survival from Durham Ferry to Chipps Island) are expected to be estimable using a release size of 4,000 (Table A5).

Under the low survival scenario (scenario 2) with overall detection probability at Chipps Island of

PG2=0.75, a release size of 100 – 200 is expected to be sufficient to estimate survival to Turner Cut and the route entrainment probabilities at the head of Old River and Turner Cut (Table A5). A release size of 800 is expected to be necessary to estimate survival from Turner Cut to Chipps Island, and up to 1,500 fish may be required to estimate survival all the way from Durham Ferry to Chipps Island. If the

detection probability at Chipps Island is very high (PG2=0.99), then only 450 fish are expected to be necessary to estimate survival from Durham Ferry to Chipps Island under the low survival scenario, and only 250 necessary to estimate survival from Turner Cut to Chipps Island. Regardless of the detection probability at Chipps Island, a release size of at least 1,000 is expected to be necessary to reliably estimate survival via the Old River route to Chipps Island in the presence of a physical barrier at the head of Old River (Table A5).

Under the low mixed survival scenario in which survival varies throughout the region but is low from

Durham Ferry to Chipps Island (scenario 3; SR=0.016), 150-400 fish are expected to be required to be released at Durham Ferry to estimate survival to Turner Cut and from Mossdale to Chipps Island via the San Joaquin River route. Estimating survival from Durham Ferry to Chipps Island, or from Turner Cut to Chipps Island via the Turner Cut route, is expected to require a release of 250-750, depending on the detection probability at Chipps Island. Estimating survival in the Old River route to Chipps Island is expected to be possible with 3,000 – 4,000 fish released at Durham Ferry (Table A6).

Under the medium survival scenario (scenario 4; SR=0.025), releases of 100-250 fish are expected to be sufficient to estimate survival in the San Joaquin River and from Turner Cut to Chipps Island, depending on the detection probability at Chipps Island (Table A6). A release of 500 at Durham Ferry is expected to be sufficient to yield a reliable estimate of survival from Durham Ferry to Chipps Island if the detection probability at Chipps Island is moderate (0.75).

142

The mixed survival scenarios yielding medium or high overall survival from Durham Ferry to Chipps Island (Table A7) suggest that a release of 100 – 350 will be sufficient to estimate survival all the way to Chipps Island, depending on detection probabilities at Chipps Island and overall survival in the San Joaquin River route. Lower survival requires more fish, generally. It is worth noting that both these scenarios assume high survival from Durham Ferry to Mossdale (SR0=0.9), which has been observed in some years but in contrast to preliminary survival estimates from 2014. These two scenarios both assume that the barrier is not installed at the head of Old River, which facilitates estimation of survival in the Old River route (Table A7).

If survival is relatively high throughout the entire study area (scenario 7, SR=0.060), then a Durham Ferry release of 200 would be sufficient to estimate all key parameters except for survival in the Old River route, regardless of the detection probability at Chipps Island. Estimating survival in the Old River route to Chipps Island is expected to require 750 to approximately 1,300 fish released at Durham Ferry if the barrier is installed at the head of Old River (Table A8).

Table A5. Minimum release size to estimate model parameters based on simulations using a single release at Durham Ferry under survival scenarios 1 (Very Low survival throughout study area) and 2 (Low survival throughout study area). NA = a release size of 4,000 fish may be insufficient to estimate the parameter.

1: Very Low Survival 2: Low Survival

Parameter True Value PG2=0.75 PG2=0.99 True Value PG2=0.75 PG2=0.99

SR 0.0006 NA NA 0.008 1,500 450

SR0 0.2 100 100 0.4 100 100

SA1 0.1 300 300 0.3 100 100

SA2 0.02 NA 4,000 0.07 800 250

SA 0.003 NA 4,000 0.02 800 250

SB 0.01 NA 4,000 0.07 1,500 1,000

SF 0.05 NA 4,000 0.05 800 250

ψA1 0.98 100 100 0.98 100 100

ψA2 0.75 1,296 1,296 0.75 200 200

143

Table A6. Minimum release size to estimate model parameters based on simulations using a single release at Durham Ferry under survival scenarios 3 (Low Mixed survival) and 4 (Medium survival throughout study area). NA = a release size of 4,000 fish may be insufficient to estimate the parameter.

3: Low Mixed Survival 4: Medium Survival

Parameter True Value PG2=0.75 PG2=0.99 True Value PG2=0.75 PG2=0.99

SR 0.016 750 250 0.025 500 150

SR0 0.400 100 100 0.7 100 100

SA1 0.300 100 100 0.4 100 100

SA2 0.150 400 175 0.1 250 100

SA 0.038 400 150 0.035 250 100

SB 0.200 4,000 3,000 0.1 1,100 750

SF 0.050 750 250 0.05 250 100

ψA1 0.980 100 100 0.98 100 100

ψA2 0.750 200 200 0.75 100 100

Table A7. Minimum release size to estimate model parameters based on simulations using a single release at Durham Ferry under survival scenarios 5 (Medium Mixed survival) and 6 (High Mixed survival). NA = a release size of 4,000 fish may be insufficient to estimate the parameter.

5: Medium Mixed Survival 6: High Mixed Survival

Parameter True Value PG2=0.75 PG2=0.99 True Value PG2=0.75 PG2=0.99

SR 0.034 350 150 0.050 250 100

SR0 0.900 100 100 0.900 100 100

SA1 0.100 150 150 0.400 100 100

SA2 0.070 450 250 0.150 250 150

SA 0.007 350 100 0.050 175 100

SB 0.070 350 100 0.060 175 100

SF 0.050 350 150 0.050 250 100

ψA1 0.980 100 100 0.980 100 100

ψA2 0.750 500 500 0.750 150 100

144

Table A8. Minimum release size to estimate model parameters based on simulations using a single release at Durham Ferry under survival scenario 7 (high survival throughout the study area). NA = a release size of 4,000 fish may be insufficient to estimate the parameter. 7: High Survival

Parameter True Value PG2=0.75 PG2=0.99

SR 0.06 200 100

SR0 0.9 100 100

SA1 0.5 100 100

SA2 0.16 150 100

SA 0.07 150 100

SB 0.16 1,296 800

SF 0.05 150 100

ψA1 0.98 100 100

ψA2 0.75 100 100

Primary Release at Durham Ferry with Supplemental Release In general, adding a supplemental release between the head of Old River and Turner Cut made little difference in the ability to estimate model parameters under any of the survival scenarios (Table A9-

Table A15). The primary exception was for the parameter SR, survival from Durham Ferry to Chipps Island. Adding a supplemental release lowered the total recommended sample size under the low, low mixed, and medium survival scenarios when detection probabilities at Chipps Island are not especially high (Table A10 – Table A12). The apparent lack of benefit from the supplemental release may reflect the combination of release sizes considered for the regional survival scenario simulations, or alternatively the low probabilities desired to be estimated and resulting difficulty in attaining estimates under any release size considered. On the other hand, releasing additional fish when survival probabilities are high has no apparent benefit to parameter estimation.

Table A9. Minimum release size at Durham Ferry (R1) and in San Joaquin River downstream of the head of Old River (R2) to estimate model parameters based on simulations Scenario 1: Very Low survival throughout study area. Minimum release sizes with

PG2=0.75 PG2=0.99 Parameter True Value R1 R2 R1 R2

SR 0.0006 NA NA NA NA

SR0 0.2 100 15 100 15

SA1 0.1 150 98 150 98

SA2 0.02 NA NA NA NA

SA 0.003 NA NA NA NA

SB 0.01 NA NA NA NA

SF 0.05 NA NA NA NA

ψA1 0.98 100 15 100 15

ψA2 0.75 1,100 75 1,100 98

145

Table A10. Minimum release size at Durham Ferry (R1) and in San Joaquin River downstream of the head of Old River (R2) to estimate model parameters based on simulations under Scenario 2: Low survival throughout study area.

Minimum release sizes with

PG2=0.75 PG2=0.99 Parameter True Value R1 R2 R1 R2

SR 0.008 300 548 450 15

SR0 0.4 100 15 100 15

SA1 0.3 100 15 100 15

SA2 0.07 300 548 450 15

SA 0.02 300 548 450 15

SB 0.07 NA NA 1,000 75

SF 0.05 300 548 450 15

ψA1 0.98 100 15 100 15

ψA2 0.75 200 15 200 15

Table A11. Minimum release size at Durham Ferry (R1) and in San Joaquin River downstream of the head of Old River (R2) to estimate model parameters based on simulations under Scenario 3: Low Mixed survival.

Minimum release sizes with

PG2=0.75 PG2=0.99 Parameter True Value R1 R2 R1 R2

SR 0.016 100 348 200 48

SR0 0.40 100 15 100 15

SA1 0.30 100 15 100 15

SA2 0.15 350 348 200 48

SA 0.038 100 348 200 48

SB 0.20 NA NA NA NA

SF 0.05 100 348 200 48

ψA1 0.98 100 15 100 15

ψA2 0.75 200 15 200 15

146

Table A12. Minimum release size at Durham Ferry (R1) and in San Joaquin River downstream of the head of Old River (R2) to estimate model parameters based on simulations under Scenario 4: Medium survival throughout study area.

Minimum release sizes with

PG2=0.75 PG2=0.99 Parameter True Value R1 R2 R1 R2

SR 0.025 150 248 150 15

SR0 0.7 100 15 100 15

SA1 0.4 100 15 100 15

SA2 0.1 150 248 150 15

SA 0.035 150 248 150 15

SB 0.1 1,200 396 800 25

SF 0.05 150 248 150 15

ψA1 0.98 100 15 100 15

ψA2 0.75 100 15 100 15

Table A13. Minimum release size at Durham Ferry (R1) and in San Joaquin River downstream of the head of Old River (R2) to estimate model parameters based on simulations under Scenario 5: Medium Mixed survival.

Minimum release sizes with

PG2=0.75 PG2=0.99 Parameter True Value R1 R2 R1 R2

SR 0.034 350 15 150 15

SR0 0.90 100 15 100 15

SA1 0.10 100 25 100 25

SA2 0.07 350 15 250 15

SA 0.007 350 15 150 15

SB 0.07 350 15 100 48

SF 0.05 350 15 150 15

ψA1 0.98 100 15 100 15

ψA2 0.75 500 25 500 25

147

Table A14. Minimum release size at Durham Ferry (R1) and in San Joaquin River downstream of the head of Old River (R2) to estimate model parameters based on simulations under Scenario 6: High Mixed survival.

Minimum release sizes with

PG2=0.75 PG2=0.99 Parameter True Value R1 R2 R1 R2

SR 0.050 150 98 100 15

SR0 0.90 100 15 100 15

SA1 0.40 100 15 100 15

SA2 0.15 250 98 150 15

SA 0.050 150 98 100 15

SB 0.06 150 98 100 15

SF 0.05 150 98 100 15

ψA1 0.98 100 15 100 15

ψA2 0.75 150 15 150 15

Table A15. Minimum release size at Durham Ferry (R1) and in San Joaquin River downstream of the head of Old River (R2) to estimate model parameters based on simulations under Scenario 7: High survival throughout study area.

Minimum release sizes with

PG2=0.75 PG2=0.99 Parameter True Value R1 R2 R1 R2

SR 0.06 100 98 100 15

SR0 0.9 100 15 100 15

SA1 0.5 100 15 100 15

SA2 0.16 150 98 100 15

SA 0.07 100 98 100 15

SB 0.16 1,500 400 900 48

SF 0.05 100 98 100 15

ψA1 0.98 100 15 100 15

ψA2 0.75 100 15 100 15

2015 Study Using a single release at Durham Ferry of 648 fish, we can expect to estimate the model parameters for all scenarios except those with low or very low survival downstream of Mossdale (Table A16, Table A17). In most cases, the average estimate is very similar to the true parameter value, but for some parameters, estimates are available in relatively few simulations (e.g., SA2 under the very low survival scenario, Table A16). Parameters with very small values tend to be overestimated in the cases when estimates are available (e.g., SR under the very low survival scenario, Table A16). In some simulations,

individual parameter estimates are far from the true values (e.g., SF under the very low survival scenario, Table A16). High detection probabilities at Chipps Island facilitate parameter estimation, but does not

148

compensate fully for very low survival parameters; even with nearly 100% detection probability at

Chipps Island, there were many simulations in which SR, SA2, SA, SB, and SF could not be estimated or had estimates of poor quality (e.g., far from true value) under the very low survival scenario (Table A17). However, the parameters were estimable in considerably more simulations under the low survival scenario when the Chipps Island detection probability was 0.99 (Table A17) than when it was 0.75 (Table A16).

If parameters are not estimable for individual release groups, it may be possible to pool data from two release groups for estimation. This gives a release size of 1,296. Under the very low survival scenario with a release group of 1,296, the observed bias in parameter estimates was considerably lower than with a release group of 648, but parameters with low values were not estimable in the majority of simulations unless the Chipps Island detection probability was very high (Table A18 and Table A19).

Overall, most or all model parameters are expected to be estimable using a release group of either 648 or 1,296, except if survival is very low. Increasing the detection probability at Chipps Island will make estimation more reliable, but will not completely compensate for the problems that may be encountered if survival is very low.

Table A16. Simulation results for release size of 648 at Durham Ferry, out of 5,000 simulations, assuming detection probabilities at Chipps Island are pG2a=pG2b=0.5 (pG2=0.75). Number of estimates = number of simulations (out of 5,000) that yielded estimate of the parameter. For each scenario, SF=0.05 and pA2a = pA2b = pF1a = pF1b = 0.85. Those results that violate criteria C1 – C4 are highlighted.

Distribution of Simulated Estimates

Scenario Parameter True Value Number of Minimum Mean Maximum SE Estimates

1: Very Low SR 0.0006 480 0.0015 0.0019 0.0069 0.0008

SR0 0.2 5000 0.13 0.20 0.26 0.02

SA1 0.1 4998 0.01 0.10 0.22 0.03

SA2 0.02 822 0.00 0.04 0.33 0.06

SA 0.003 822 0.006 0.01 0.03 0.01

SB 0.01 733 0.00 0.02 0.67 0.07

SF 0.05 760 0.00 0.11 1.33 0.19

ψA1 0.98 5000 0.91 0.98 1.00 0.01

ψA2 0.75 4998 0.25 0.76 1.00 0.14

149

2: Low SR 0.008 3717 0.002 0.009 0.05 0.005

SR0 0.4 5000 0.33 0.40 0.47 0.02

SA1 0.3 5000 0.18 0.30 0.41 0.03

SA2 0.07 4616 0.00 0.08 0.50 0.05

SA 0.02 4616 0.00 0.02 0.18 0.01

SB 0.07 4570 0.00 0.07 1.91 0.15

SF 0.05 4616 0.00 0.06 0.62 0.07

ψA1 0.98 5000 0.93 0.98 1.00 0.01

ψA2 0.75 5000 0.55 0.75 0.92 0.05

3: Low Mixed SR 0.016 4646 0.002 0.018 0.093 0.009

SR0 0.400 5000 0.32 0.40 0.47 0.02

SA1 0.300 5000 0.20 0.30 0.41 0.03

SA2 0.150 4964 0.00 0.16 1.16 0.08

SA 0.038 4964 0.00 0.040 0.22 0.02

SB 0.200 4869 0.00 0.21 2.35 0.26

SF 0.050 4964 0.00 0.05 0.82 0.07

ψA1 0.980 5000 0.94 0.98 1.00 0.01

ψA2 0.750 5000 0.54 0.75 0.91 0.05

4: Medium SR 0.025 4897 0.005 0.027 0.15 0.01

SR0 0.7 5000 0.63 0.70 0.76 0.02

SA1 0.4 5000 0.31 0.40 0.49 0.02

SA2 0.1 4997 0.01 0.10 0.69 0.04

150

SA 0.035 4997 0.003 0.037 0.21 0.01

SB 0.1 4994 0.00 0.11 1.55 0.14

SF 0.05 4997 0.00 0.05 0.41 0.04

ψA1 0.98 5000 0.95 0.98 1.00 0.01

ψA2 0.75 5000 0.63 0.75 0.97 0.03

5: Medium Mixed SR 0.034 3441 0.002 0.007 0.046 0.004

SR0 0.900 5000 0.85 0.90 0.95 0.01

SA1 0.100 5000 0.05 0.10 0.15 0.01

SA2 0.070 4327 0.00 0.07 0.47 0.06

SA 0.007 4327 0.00 0.007 0.05 0.005

SB 0.070 4327 0.00 0.08 0.83 0.10

SF 0.050 4327 0.00 0.05 0.74 0.08

ψA1 0.980 5000 0.96 0.98 1.00 0.01

ψA2 0.750 5000 0.54 0.75 0.94 0.06

6: High Mixed SR 0.050 4994 0.01 0.05 0.22 0.01

SR0 0.900 5000 0.85 0.90 0.94 0.01

SA1 0.400 5000 0.32 0.40 0.49 0.02

SA2 0.150 5000 0.06 0.15 0.40 0.04

SA 0.050 5000 0.02 0.05 0.12 0.01

SB 0.060 5000 0.00 0.06 1.05 0.09

SF 0.050 5000 0.00 0.05 0.24 0.04

ψA1 0.980 5000 0.96 0.98 1.00 0.01

151

ψA2 0.750 5000 0.63 0.75 0.85 0.03

7: High SR 0.06 5000 0.02 0.06 0.19 0.02

SR0 0.9 5000 0.95 0.90 0.94 0.01

SA1 0.5 5000 0.42 0.50 0.57 0.02

SA2 0.16 5000 0.04 0.16 0.33 0.03

SA 0.07 5000 0.02 0.07 0.14 0.01

SB 0.16 5000 0.00 0.17 1.02 0.14

SF 0.05 5000 0.00 0.05 0.24 0.03

ψA1 0.98 5000 0.96 0.98 1.00 0.01

ψA2 0.75 5000 0.65 0.75 0.85 0.03

Table A17. Simulation results for release size of 648 at Durham Ferry, out of 5,000 simulations, assuming detection probabilities at Chipps Island are pG2a=pG2b=0.9 (pG2=0.99). Number of estimates = number of simulations (out of 5,000) that yielded estimate of the parameter. For each scenario, SF=0.05 and pA2a = pA2b = pF1a = pF1b = 0.85. Those results that violate criteria C1 – C4 are highlighted.

Distribution of Simulated Estimates

Scenario Parameter True Value Number of Minimum Mean Maximum SE Estimates

1: Very Low SR 0.0006 1280 0.0015 0.0018 0.0069 0.0008

SR0 0.2 5000 0.15 0.20 0.26 0.02

SA1 0.1 4996 0.02 0.10 0.23 0.03

SA2 0.02 2199 0.00 0.04 0.50 0.06

SA 0.003 2199 0.00 0.006 0.03 0.01

SB 0.01 1955 0.00 0.02 1.00 0.09

SF 0.05 2026 0.00 0.10 2.00 0.19

ψA1 0.98 5000 0.93 0.98 1.00 0.01

152

ψA2 0.75 4997 0.14 0.76 1.00 0.14

2: Low SR 0.008 4921 0.002 0.008 0.025 0.004

SR0 0.4 5000 0.33 0.40 0.47 0.02

SA1 0.3 5000 0.20 0.30 0.41 0.03

SA2 0.07 4999 0.00 0.07 0.22 0.03

SA 0.02 4999 0.00 0.02 0.06 0.01

SB 0.07 4938 0.00 0.07 1.04 0.13

SF 0.05 4999 0.00 0.05 0.34 0.05

ψA1 0.98 5000 0.94 0.98 1.00 0.01

ψA2 0.75 5000 0.55 0.75 0.92 0.05

3: Low Mixed SR 0.016 5000 0.002 0.016 0.037 0.005

SR0 0.400 5000 0.33 0.40 0.48 0.02

SA1 0.300 5000 0.20 0.30 0.41 0.03

SA2 0.150 5000 0.01 0.15 0.36 0.05

SA 0.038 5000 0.004 0.038 0.09 0.01

SB 0.200 4919 0.00 0.20 1.05 0.20

SF 0.050 5000 0.00 0.05 0.32 0.05

ψA1 0.980 5000 0.94 0.98 1.00 0.01

ψA2 0.750 5000 0.54 0.75 0.90 0.05

4: Medium SR 0.025 5000 0.006 0.025 0.51 0.006

SR0 0.7 5000 0.63 0.70 0.76 0.02

153

SA1 0.4 5000 0.30 0.40 0.48 0.02

SA2 0.1 5000 0.02 0.10 0.22 0.03

SA 0.035 5000 0.01 0.035 0.07 0.01

SB 0.1 4997 0.00 0.10 0.70 0.11

SF 0.05 5000 0.00 0.05 0.19 0.03

ψA1 0.98 5000 0.95 0.98 1.00 0.01

ψA2 0.75 5000 0.60 0.75 0.88 0.03

5: Medium Mixed SR 0.034 4891 0.002 0.007 0.021 0.003

SR0 0.900 5000 0.85 0.90 0.95 0.01

SA1 0.100 5000 0.06 0.10 0.16 0.01

SA2 0.070 4992 0.00 0.07 0.24 0.04

SA 0.007 4992 0.00 0.007 0.02 0.003

SB 0.070 4992 0.00 0.07 0.50 0.08

SF 0.050 4992 0.00 0.05 0.44 0.06

ψA1 0.980 5000 0.96 0.98 1.00 0.01

ψA2 0.750 5000 0.52 0.75 1.00 0.06

6: High Mixed SR 0.050 5000 0.02 0.05 0.08 0.01

SR0 0.900 5000 0.85 0.90 0.94 0.01

SA1 0.400 5000 0.31 0.40 0.47 0.02

SA2 0.150 5000 0.06 0.15 0.26 0.03

SA 0.050 5000 0.02 0.05 0.10 0.01

SB 0.060 5000 0.00 0.06 0.51 0.07

154

SF 0.050 5000 0.00 0.05 0.19 0.03

ψA1 0.980 5000 0.95 0.98 1.00 0.01

ψA2 0.750 5000 0.62 0.75 0.95 0.03

7: High SR 0.06 5000 0.03 0.06 0.10 0.01

SR0 0.9 5000 0.85 0.90 0.94 0.01

SA1 0.5 5000 0.43 0.50 0.58 0.02

SA2 0.16 5000 0.07 0.16 0.25 0.03

SA 0.07 5000 0.03 0.07 0.11 0.01

SB 0.16 4997 0.00 0.16 0.81 0.12

SF 0.05 5000 0.00 0.05 0.19 0.03

ψA1 0.98 5000 0.96 0.98 1.00 0.01

ψA2 0.75 5000 0.64 0.75 0.83 0.03

Table A18. Simulation results for release size of 1,296 at Durham Ferry, out of 5,000 simulations, assuming detection probabilities at Chipps Island are pG2a=pG2b=0.5 (pG2=0.75). Number of estimates = number of simulations (out of 5,000) that yielded estimate of the parameter. For each scenario, SF=0.05 and pA2a = pA2b = pF1a = pF1b = 0.85. Those results that violate criteria C1 – C4 are highlighted.

Distribution of Simulated Estimates

Scenario Parameter True Value Number of Minimum Mean Maximum SE Estimates

1: Very Low SR 0.0006 837 0.0008 0.001 0.005 0.0006

SR0 0.2 5000 0.15 0.20 0.25 0.01

SA1 0.1 5000 0.04 0.10 0.21 0.02

SA2 0.02 1499 0.00 0.03 0.25 0.04

SA 0.003 1499 0.004 0.004 0.04 0.004

SB 0.01 1480 0.00 0.01 1.00 0.07

155

SF 0.05 1494 0.00 0.06 1.00 0.11

ψA1 0.98 5000 0.93 0.98 1.00 0.01

ψA2 0.75 5000 0.36 0.75 1.00 0.09

2: Low SR 0.008 4645 0.0008 0.009 0.06 0.005

SR0 0.4 5000 0.35 0.40 0.45 0.01

SA1 0.3 5000 0.23 0.30 0.38 0.02

SA2 0.07 4968 0.00 0.07 0.40 0.04

SA 0.02 4968 0.00 0.02 0.10 0.01

SB 0.07 4966 0.00 0.08 1.29 0.12

SF 0.05 4968 0.00 0.05 0.54 0.05

ψA1 0.98 5000 0.95 0.98 1.00 0.01

ψA2 0.75 5000 0.62 0.75 0.87 0.04

3: Low Mixed SR 0.016 4975 0.003 0.018 0.110 0.007

SR0 0.400 5000 0.35 0.40 0.45 0.01

SA1 0.300 5000 0.21 0.30 0.38 0.02

SA2 0.150 5000 0.02 0.15 0.74 0.05

SA 0.038 5000 0.004 0.039 0.18 0.01

SB 0.200 4999 0.00 0.20 1.27 0.17

SF 0.050 5000 0.00 0.05 0.39 0.04

ψA1 0.980 5000 0.95 0.98 1.00 0.01

ψA2 0.750 5000 0.60 0.75 0.88 0.04

156

4: Medium SR 0.025 5000 0.007 0.026 0.10 0.01

SR0 0.7 5000 0.66 0.70 0.75 0.01

SA1 0.4 5000 0.34 0.40 0.46 0.02

SA2 0.1 5000 0.03 0.10 0.24 0.03

SA 0.035 5000 0.01 0.035 0.07 0.01

SB 0.1 5000 0.00 0.10 0.71 0.09

SF 0.05 5000 0.00 0.05 0.18 0.03

ψA1 0.98 5000 0.96 0.98 0.99 0.005

ψA2 0.75 5000 0.66 0.75 0.84 0.02

5: Medium Mixed SR 0.034 4484 0.0008 0.008 0.043 0.004

SR0 0.900 5000 0.86 0.90 0.93 0.01

SA1 0.100 5000 0.07 0.10 0.14 0.01

SA2 0.070 4917 0.00 0.08 0.43 0.04

SA 0.007 4917 0.00 0.007 0.04 0.004

SB 0.070 4917 0.00 0.07 0.74 0.07

SF 0.050 4917 0.00 0.05 0.44 0.06

ψA1 0.980 5000 0.96 0.98 0.99 0.004

ψA2 0.750 5000 0.60 0.75 0.89 0.04

6: High Mixed SR 0.050 5000 0.02 0.05 0.12 0.01

SR0 0.900 5000 0.86 0.90 0.93 0.01

SA1 0.400 5000 0.35 0.40 0.46 0.01

SA2 0.150 5000 0.06 0.15 0.26 0.03

157

SA 0.050 5000 0.02 0.05 0.09 0.01

SB 0.060 5000 0.00 0.06 0.42 0.06

SF 0.050 5000 0.00 0.05 0.15 0.02

ψA1 0.980 5000 0.96 0.98 0.99 0.004

ψA2 0.750 5000 0.67 0.75 0.82 0.02

7: High SR 0.06 5000 0.03 0.06 0.11 0.01

SR0 0.9 5000 0.87 0.90 0.93 0.01

SA1 0.5 5000 0.45 0.50 0.56 0.02

SA2 0.16 5000 0.09 0.16 0.27 0.02

SA 0.07 5000 0.03 0.07 0.11 0.01

SB 0.16 5000 0.00 0.16 0.62 0.09

SF 0.05 5000 0.00 0.05 0.15 0.02

ψA1 0.98 5000 0.96 0.98 0.99 0.004

ψA2 0.75 5000 0.69 0.75 0.81 0.02

Table A19. Simulation results for release size of 1,296 at Durham Ferry, out of 5,000 simulations, assuming detection probabilities at Chipps Island are pG2a=pG2b=0.9 (pG2=0.99). Number of estimates = number of simulations (out of 5,000) that yielded estimate of the parameter. For each scenario, SF=0.05 and pA2a = pA2b = pF1a = pF1b = 0.85. Those results that violate criteria C1 – C4 are highlighted.

Distribution of Simulated Estimates

Scenario Parameter True Value Number of Minimum Mean Maximum SE Estimates

1: Very Low SR 0.0006 2318 0.0008 0.0011 0.0046 0.0006

SR0 0.2 5000 0.16 0.20 0.24 0.01

SA1 0.1 5000 0.04 0.10 0.17 0.02

158

SA2 0.02 3528 0.00 0.03 0.24 0.04

SA 0.003 3528 0.00 0.004 0.02 0.003

SB 0.01 3481 0.00 0.01 1.00 0.06

SF 0.05 3504 0.00 0.07 1.00 0.11

ψA1 0.98 5000 0.94 0.98 1.00 0.01

ψA2 0.75 5000 0.36 0.75 1.00 0.09

2: Low SR 0.008 5000 0.002 0.008 0.020 0.003

SR0 0.4 5000 0.35 0.40 0.46 0.01

SA1 0.3 5000 0.23 0.30 0.38 0.02

SA2 0.07 5000 0.00 0.07 0.16 0.02

SA 0.02 5000 0.00 0.02 0.04 0.01

SB 0.07 5000 0.00 0.07 0.61 0.08

SF 0.05 5000 0.00 0.05 0.21 0.04

ψA1 0.98 5000 0.96 0.98 1.00 0.01

ψA2 0.75 5000 0.59 0.75 0.87 0.04

3: Low Mixed SR 0.016 5000 0.005 0.016 0.031 0.004

SR0 0.400 5000 0.34 0.40 0.46 0.01

SA1 0.300 5000 0.22 0.30 0.38 0.02

SA2 0.150 5000 0.05 0.15 0.28 0.03

SA 0.038 5000 0.01 0.038 0.07 0.01

SB 0.200 4998 0.00 0.20 0.81 0.13

SF 0.050 5000 0.00 0.05 0.21 0.04

159

ψA1 0.980 5000 0.95 0.98 1.00 0.01

ψA2 0.750 5000 0.61 0.75 0.88 0.04

4: Medium SR 0.025 5000 0.01 0.025 0.04 0.004

SR0 0.7 5000 0.66 0.70 0.76 0.01

SA1 0.4 5000 0.34 0.40 0.48 0.02

SA2 0.1 5000 0.04 0.10 0.18 0.02

SA 0.035 5000 0.01 0.035 0.07 0.01

SB 0.1 5000 0.00 0.10 0.55 0.07

SF 0.05 5000 0.00 0.05 0.15 0.02

ψA1 0.98 5000 0.96 0.98 1.00 0.005

ψA2 0.75 5000 0.66 0.75 0.86 0.02

5: Medium Mixed SR 0.034 4997 0.0008 0.007 0.017 0.002

SR0 0.900 5000 0.86 0.90 0.93 0.01

SA1 0.100 5000 0.06 0.10 0.13 0.01

SA2 0.070 5000 0.00 0.07 0.22 0.03

SA 0.007 5000 0.00 0.006 0.02 0.002

SB 0.070 5000 0.00 0.07 0.33 0.05

SF 0.050 5000 0.00 0.05 0.32 0.04

ψA1 0.980 5000 0.96 0.98 0.99 0.004

ψA2 0.750 5000 0.59 0.75 0.89 0.04

6: High Mixed SR 0.050 5000 0.02 0.05 0.07 0.01

160

SR0 0.900 5000 0.86 0.90 0.94 0.01

SA1 0.400 5000 0.34 0.40 0.46 0.02

SA2 0.150 5000 0.09 0.15 0.23 0.02

SA 0.050 5000 0.03 0.05 0.07 0.01

SB 0.060 5000 0.00 0.06 0.34 0.05

SF 0.050 5000 0.00 0.05 0.15 0.02

ψA1 0.980 5000 0.96 0.98 0.99 0.004

ψA2 0.750 5000 0.67 0.75 0.82 0.02

7: High SR 0.06 5000 0.04 0.06 0.09 0.01

SR0 0.9 5000 0.87 0.90 0.94 0.01

SA1 0.5 5000 0.44 0.50 0.56 0.02

SA2 0.16 5000 0.10 0.16 0.23 0.02

SA 0.07 5000 0.04 0.07 0.10 0.01

SB 0.16 5000 0.00 0.16 0.48 0.08

SF 0.05 5000 0.00 0.05 0.12 0.02

ψA1 0.98 5000 0.96 0.98 0.99 0.004

ψA2 0.75 5000 0.69 0.75 0.82 0.02

Appendix A1 Simulation results are provided for all parameter sets considered. Results for scenarios with similar survival throughout the study area are in “Simulation output for Area-wide Survival Scenarios”, Tables A1A1 to A1A4. Results for scenarios with varying levels of survival in different regions of the study area are in “Simulation output for Mixed Survival Scenarios.”

161

Simulation output for Area-Wide Survival Scenarios

Table A1A1. Simulation and minimization output for area-wide survival scenarios (i.e., similar levels of survival throughout the study area) using a single release at Durham Ferry: minimum release size required to estimate parameters. For each scenario, pA1a = pA1b = pB1a = pB1b = 0.90 and pA2a = pA2b = pF1a = pF1b = 0.85. Survival scenario number does not correspond directly with scenarios in Table A3. Scenarios 11, 23, 35, and 47 here correspond to scenarios 1, 2, 4, and 7 in Table A3, respectively (highlighted). For scenarios where SR0=0.2, SA1, SA2, and SB are 0.1, 0.02, and 0.01, respectively. For scenarios where SR0=0.4, SA1, SA2, and SB are 0.3, 0.07, and 0.07, respectively. SR0=0.7, SA1, SA2, and SB are 0.4, 0.1, and 0.1, respectively. SR0=0.9, SA1, SA2, and SB are 0.5, 0.16, and 0.16, respectively.

True Parameter Values Minimum release size to estimate

Scenario sR sR0 sA sF ψA1 ψA2 pG2a=pG2b sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 1 0.00134 0.2 0.00175 0.01 0.4 0.75 0.5 NA 100 700 NA NA NA NA 150 3000 1 0.00134 0.2 0.00175 0.01 0.4 0.75 0.9 3000 100 750 3000 3000 3000 3000 150 3000 2 0.001356 0.2 0.00195 0.01 0.4 0.95 0.5 NA 100 550 NA NA NA NA 150 100 2 0.001356 0.2 0.00195 0.01 0.4 0.95 0.9 3000 100 550 3000 3000 3000 NA 150 100 3 0.00101 0.2 0.00175 0.01 0.6 0.75 0.5 NA 100 500 NA NA NA NA 150 2000 3 0.00101 0.2 0.00175 0.01 0.6 0.75 0.9 4000 100 500 4000 4000 4000 4000 150 2000 4 0.001034 0.2 0.00195 0.01 0.6 0.95 0.5 NA 100 400 NA NA NA NA 150 100 4 0.001034 0.2 0.00195 0.01 0.6 0.95 0.9 4000 100 400 3000 3000 3000 NA 150 100 5 0.000383 0.2 0.00175 0.01 0.98 0.75 0.5 NA 100 300 NA NA NA NA 100 1296 5 0.000383 0.2 0.00175 0.01 0.98 0.75 0.9 NA 100 300 NA NA NA NA 100 1296 6 0.000422 0.2 0.00195 0.01 0.98 0.95 0.5 NA 100 250 NA NA NA NA 100 100 6 0.000422 0.2 0.00195 0.01 0.98 0.95 0.9 NA 100 250 NA NA NA NA 100 100 7 0.00142 0.2 0.00275 0.05 0.4 0.75 0.5 NA 100 700 NA NA NA NA 150 3000 7 0.00142 0.2 0.00275 0.05 0.4 0.75 0.9 3000 100 700 3000 3000 3000 3000 150 3000 8 0.001372 0.2 0.00215 0.05 0.4 0.95 0.5 NA 100 550 NA NA NA NA 150 100 8 0.001372 0.2 0.00215 0.05 0.4 0.95 0.9 3000 100 550 3000 3000 3000 NA 150 100 9 0.00113 0.2 0.00275 0.05 0.6 0.75 0.5 NA 100 500 NA NA NA NA 150 2000 9 0.00113 0.2 0.00275 0.05 0.6 0.75 0.9 4000 100 500 3000 3000 3000 3000 150 2000 10 0.001058 0.2 0.00215 0.05 0.6 0.95 0.5 NA 100 400 NA NA NA NA 150 100 10 0.001058 0.2 0.00215 0.05 0.6 0.95 0.9 4000 100 400 3000 3000 3000 NA 150 100 11 0.000579 0.2 0.00275 0.05 0.98 0.75 0.5 NA 100 300 NA NA NA NA 100 1296

162

Table A1A1 (Continued) True Parameter Values Minimum release size to estimate

Scenario sR sR0 sA sF ψA1 ψA2 pG2a=pG2b sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 11 0.000579 0.2 0.00275 0.05 0.98 0.75 0.9 NA 100 300 4000 4000 4000 4000 100 1296 12 0.000461 0.2 0.00215 0.05 0.98 0.95 0.5 NA 100 250 NA NA NA NA 100 100 12 0.000461 0.2 0.00215 0.05 0.98 0.95 0.9 NA 100 250 NA NA NA NA 100 100 13 0.01944 0.4 0.0165 0.01 0.4 0.75 0.5 600 100 175 550 550 550 550 100 500 13 0.01944 0.4 0.0165 0.01 0.4 0.75 0.9 200 100 175 250 175 175 350 100 500 14 0.020016 0.4 0.0201 0.01 0.4 0.95 0.5 600 100 150 550 550 550 1750 100 100 14 0.020016 0.4 0.0201 0.01 0.4 0.95 0.9 200 100 150 175 175 175 1750 100 100 15 0.01516 0.4 0.0165 0.01 0.6 0.75 0.5 800 100 150 648 648 648 648 100 300 15 0.01516 0.4 0.0165 0.01 0.6 0.75 0.9 300 100 100 200 200 200 300 100 300 16 0.016024 0.4 0.0201 0.01 0.6 0.95 0.5 750 100 100 600 600 600 1296 100 100 16 0.016024 0.4 0.0201 0.01 0.6 0.95 0.9 250 100 100 200 200 200 1100 100 100 17 0.007028 0.4 0.0165 0.01 0.98 0.75 0.5 2000 100 100 900 900 1750 900 100 200 17 0.007028 0.4 0.0165 0.01 0.98 0.75 0.9 550 100 100 300 300 1000 300 100 200 18 0.008439 0.4 0.0201 0.01 0.98 0.95 0.5 1500 100 100 750 750 1500 900 100 100 18 0.008439 0.4 0.0201 0.01 0.98 0.95 0.9 450 100 100 250 250 1000 700 100 100 19 0.01992 0.4 0.0195 0.05 0.4 0.75 0.5 600 100 150 550 550 550 800 100 500 19 0.01992 0.4 0.0195 0.05 0.4 0.75 0.9 200 100 175 250 175 175 500 100 500 20 0.020112 0.4 0.0207 0.05 0.4 0.95 0.5 600 100 150 500 500 500 4000 100 100 20 0.020112 0.4 0.0207 0.05 0.4 0.95 0.9 200 100 150 200 175 175 3000 100 100 21 0.01588 0.4 0.0195 0.05 0.6 0.75 0.5 750 100 150 600 600 600 600 100 350 21 0.01588 0.4 0.0195 0.05 0.6 0.75 0.9 250 100 100 200 200 200 350 100 350 22 0.016168 0.4 0.0207 0.05 0.6 0.95 0.5 750 100 100 600 600 600 3000 100 100 22 0.016168 0.4 0.0207 0.05 0.6 0.95 0.9 250 100 100 175 175 175 1750 100 100 23 0.008204 0.4 0.0195 0.05 0.98 0.75 0.5 1500 100 100 800 800 1500 800 100 200 23 0.008204 0.4 0.0195 0.05 0.98 0.75 0.9 450 100 100 250 250 1000 250 100 200 24 0.008674 0.4 0.0207 0.05 0.98 0.95 0.5 1500 100 100 700 700 1750 1750 100 100 24 0.008674 0.4 0.0207 0.05 0.98 0.95 0.9 450 100 100 250 250 1000 1000 100 100

163

Table A1A1 (Continued) True Parameter Values Minimum release size to estimate

Scenario sR sR0 sA sF ψA1 ψA2 pG2a=pG2b sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 25 0.05068 0.7 0.031 0.01 0.4 0.75 0.5 250 100 150 250 200 200 250 100 200 25 0.05068 0.7 0.031 0.01 0.4 0.75 0.9 100 100 100 150 100 100 150 100 200 26 0.052696 0.7 0.0382 0.01 0.4 0.95 0.5 250 100 100 200 200 200 700 100 100 26 0.052696 0.7 0.0382 0.01 0.4 0.95 0.9 100 100 100 100 100 100 700 100 100 27 0.04102 0.7 0.031 0.01 0.6 0.75 0.5 350 100 100 250 250 250 250 100 150 27 0.04102 0.7 0.031 0.01 0.6 0.75 0.9 100 100 100 100 100 100 100 100 150 28 0.044044 0.7 0.0382 0.01 0.6 0.95 0.5 300 100 100 250 250 250 500 100 100 28 0.044044 0.7 0.0382 0.01 0.6 0.95 0.9 100 100 100 100 100 100 500 100 100 29 0.022666 0.7 0.031 0.01 0.98 0.75 0.5 550 100 100 300 300 1296 300 100 100 29 0.022666 0.7 0.031 0.01 0.98 0.75 0.9 175 100 100 100 100 800 100 100 100 30 0.027605 0.7 0.0382 0.01 0.98 0.95 0.5 450 100 100 250 250 1100 350 100 100 30 0.027605 0.7 0.0382 0.01 0.98 0.95 0.9 150 100 100 100 100 750 300 100 100 31 0.0518 0.7 0.035 0.05 0.4 0.75 0.5 250 100 100 250 200 200 350 100 200 31 0.0518 0.7 0.035 0.05 0.4 0.75 0.9 100 100 100 150 100 100 250 100 200 32 0.05292 0.7 0.039 0.05 0.4 0.95 0.5 250 100 100 200 200 200 1400 100 100 32 0.05292 0.7 0.039 0.05 0.4 0.95 0.9 100 100 100 100 100 100 1296 100 100 33 0.0427 0.7 0.035 0.05 0.6 0.75 0.5 300 100 100 250 250 250 250 100 150 33 0.0427 0.7 0.035 0.05 0.6 0.75 0.9 100 100 100 100 100 100 150 100 150 34 0.04438 0.7 0.039 0.05 0.6 0.95 0.5 300 100 100 200 200 200 1000 100 100 34 0.04438 0.7 0.039 0.05 0.6 0.95 0.9 100 100 100 100 100 100 750 100 100 35 0.02541 0.7 0.035 0.05 0.98 0.75 0.5 500 100 100 250 250 1100 250 100 100 35 0.02541 0.7 0.035 0.05 0.98 0.75 0.9 150 100 100 100 100 750 100 100 100 36 0.028154 0.7 0.039 0.05 0.98 0.95 0.5 450 100 100 250 250 1100 648 100 100 36 0.028154 0.7 0.039 0.05 0.98 0.95 0.9 150 100 100 100 100 750 450 100 100 37 0.10845 0.9 0.06125 0.01 0.4 0.75 0.5 150 100 100 250 100 100 150 100 150 37 0.10845 0.9 0.06125 0.01 0.4 0.75 0.9 100 100 100 150 100 100 100 100 150

164

Table A1A1 (Continued) True Parameter Values Minimum release size to estimate

Scenario sR sR0 sA sF ψA1 ψA2 pG2a=pG2b sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 38 0.11385 0.9 0.07625 0.01 0.4 0.95 0.5 150 100 100 200 100 100 450 100 100 38 0.11385 0.9 0.07625 0.01 0.4 0.95 0.9 100 100 100 100 100 100 450 100 100 39 0.090675 0.9 0.06125 0.01 0.6 0.75 0.5 150 100 100 175 100 150 100 100 100 39 0.090675 0.9 0.06125 0.01 0.6 0.75 0.9 100 100 100 100 100 100 100 100 100 40 0.098775 0.9 0.07625 0.01 0.6 0.95 0.5 150 100 100 150 100 150 300 100 100 40 0.098775 0.9 0.07625 0.01 0.6 0.95 0.9 100 100 100 100 100 100 300 100 100 41 0.056903 0.9 0.06125 0.01 0.98 0.75 0.5 250 100 100 150 150 1296 150 100 100 41 0.056903 0.9 0.06125 0.01 0.98 0.75 0.9 100 100 100 100 100 900 100 100 100 42 0.070133 0.9 0.07625 0.01 0.98 0.95 0.5 175 100 100 100 100 1296 200 100 100 42 0.070133 0.9 0.07625 0.01 0.98 0.95 0.9 100 100 100 100 100 900 200 100 100 43 0.11025 0.9 0.06625 0.05 0.4 0.75 0.5 150 100 100 250 100 100 250 100 150 43 0.11025 0.9 0.06625 0.05 0.4 0.75 0.9 100 100 100 150 100 100 150 100 150 44 0.11421 0.9 0.07725 0.05 0.4 0.95 0.5 150 100 100 175 100 100 900 100 100 44 0.11421 0.9 0.07725 0.05 0.4 0.95 0.9 100 100 100 100 100 100 700 100 100 45 0.093375 0.9 0.06625 0.05 0.6 0.75 0.5 150 100 100 175 100 100 150 100 100 45 0.093375 0.9 0.06625 0.05 0.6 0.75 0.9 100 100 100 100 100 100 100 100 100 46 0.099315 0.9 0.07725 0.05 0.6 0.95 0.5 150 100 100 150 100 150 648 100 100 46 0.099315 0.9 0.07725 0.05 0.6 0.95 0.9 100 100 100 100 100 100 500 100 100 47 0.061313 0.9 0.06625 0.05 0.98 0.75 0.5 200 100 100 150 150 1296 150 100 100 47 0.061313 0.9 0.06625 0.05 0.98 0.75 0.9 100 100 100 100 100 800 100 100 100 48 0.071015 0.9 0.07725 0.05 0.98 0.95 0.5 175 100 100 100 100 1296 400 100 100 48 0.071015 0.9 0.07725 0.05 0.98 0.95 0.9 100 100 100 100 100 800 300 100 100

165

Table A1A2. Simulation and minimization output for area-wide survival scenarios (i.e., similar levels of survival throughout the study area) using a primary release at Durham Ferry (R1) and a supplemental release in the San Joaquin River downstream of the head of Old River (R2): minimum release sizes required to estimate parameters. For each scenario, pA1a = pA1b = pB1a = pB1b = 0.90 and pA2a = pA2b = pF1a = pF1b = 0.85. Survival scenario number does not correspond directly with scenarios in Table A3. Scenarios 11, 23, 35, and 47 here correspond to scenarios 1, 2, 4, and 7 in Table A3, respectively (highlighted). For scenarios where SR0=0.2, SA1, SA2, and SB are 0.1, 0.02, and 0.01, respectively. For scenarios where SR0=0.4, SA1, SA2, and SB are 0.3, 0.07, and 0.07, respectively. SR0=0.7, SA1, SA2, and SB are 0.4, 0.1, and 0.1, respectively. SR0=0.9, SA1, SA2, and SB are 0.5, 0.16, and 0.16, respectively.

True Values Release size at Durham Ferry Supplemental release size

Scenario sR sR0 sA sF ψA1 ψA2 pG2a=pG2b sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 sR sR0 sA1 sA2 sA sB sF ψA1 ψA2

NA 100 150 NA NA NA NA 150 NA NA 15 248 NA NA NA NA 15 NA 1 0.00134 0.2 0.00175 0.01 0.4 0.75 0.5 NA 100 150 NA NA NA NA 150 NA NA 15 298 NA NA NA NA 15 NA 1 0.00134 0.2 0.00175 0.01 0.4 0.75 0.9 NA 100 150 NA NA NA NA 150 100 NA 15 198 NA NA NA NA 15 15 2 0.001356 0.2 0.00195 0.01 0.4 0.95 0.5 NA 100 150 NA NA NA NA 150 100 NA 15 198 NA NA NA NA 15 15 2 0.001356 0.2 0.00195 0.01 0.4 0.95 0.9 NA 100 175 NA NA NA NA 150 NA NA 15 148 NA NA NA NA 15 NA 3 0.00101 0.2 0.00175 0.01 0.6 0.75 0.5 NA 100 100 NA NA NA NA 150 NA NA 15 248 NA NA NA NA 15 NA 3 0.00101 0.2 0.00175 0.01 0.6 0.75 0.9 NA 100 175 NA NA NA NA 150 100 NA 15 98 NA NA NA NA 15 15 4 0.001034 0.2 0.00195 0.01 0.6 0.95 0.5 NA 100 175 NA NA NA NA 150 100 NA 15 98 NA NA NA NA 15 15 4 0.001034 0.2 0.00195 0.01 0.6 0.95 0.9 NA 100 150 NA NA NA NA 100 1150 NA 15 98 NA NA NA NA 15 75 5 0.000383 0.2 0.00175 0.01 0.98 0.75 0.5 NA 100 150 NA NA NA NA 100 1150 NA 15 98 NA NA NA NA 15 75 5 0.000383 0.2 0.00175 0.01 0.98 0.75 0.9 NA 100 100 NA NA NA NA 100 200 NA 15 98 NA NA NA NA 15 15 6 0.000422 0.2 0.00195 0.01 0.98 0.95 0.5 NA 100 100 NA NA NA NA 100 100 NA 15 98 NA NA NA NA 15 15 6 0.000422 0.2 0.00195 0.01 0.98 0.95 0.9 NA 100 150 NA NA NA NA 150 NA NA 15 298 NA NA NA NA 15 NA 7 0.00142 0.2 0.00275 0.05 0.4 0.75 0.5 NA 100 175 NA NA NA NA 150 NA NA 15 198 NA NA NA NA 15 NA 7 0.00142 0.2 0.00275 0.05 0.4 0.75 0.9 NA 100 150 NA NA NA NA 150 100 NA 15 198 NA NA NA NA 15 15 8 0.001372 0.2 0.00215 0.05 0.4 0.95 0.5

166

Table A1A2 (Continued).

True Values Release size at Durham Ferry Supplemental release size

Scenario sR sR0 sA sF ψA1 ψA2 pG2a=pG sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 sR sR0 sA1 sA2 sA sB sF ψA1 ψA2

2b NA 100 150 NA NA NA NA 150 100 NA 15 175 NA NA NA NA 15 15 8 0.001372 0.2 0.00215 0.05 0.4 0.95 0.9 NA 100 175 NA NA NA NA 150 NA NA 15 148 NA NA NA NA 15 NA 9 0.00113 0.2 0.00275 0.05 0.6 0.75 0.5 NA 100 200 NA NA NA NA 150 NA NA 15 125 NA NA NA NA 15 NA 9 0.00113 0.2 0.00275 0.05 0.6 0.75 0.9 NA 100 175 NA NA NA NA 150 100 NA 15 98 NA NA NA NA 15 15 10 0.001058 0.2 0.00215 0.05 0.6 0.95 0.5 NA 100 175 NA NA NA NA 150 100 NA 15 98 NA NA NA NA 15 15 10 0.001058 0.2 0.00215 0.05 0.6 0.95 0.9 NA 100 150 NA NA NA NA 100 1100 NA 15 98 NA NA NA NA 15 75 11 0.000579 0.2 0.00275 0.05 0.98 0.75 0.5 NA 100 150 NA NA NA NA 100 1100 NA 15 98 NA NA NA NA 15 98 11 0.000579 0.2 0.00275 0.05 0.98 0.75 0.9 NA 100 100 NA NA NA NA 100 250 NA 15 98 NA NA NA NA 15 15 12 0.000461 0.2 0.00215 0.05 0.98 0.95 0.5 NA 100 100 NA NA NA NA 100 250 NA 15 98 NA NA NA NA 15 15 12 0.000461 0.2 0.00215 0.05 0.98 0.95 0.9 600 100 200 600 600 600 600 100 450 15 15 15 15 15 15 15 15 25 13 0.01944 0.4 0.0165 0.01 0.4 0.75 0.5 200 100 175 250 200 200 200 100 450 15 15 25 15 15 15 15 15 15 13 0.01944 0.4 0.0165 0.01 0.4 0.75 0.9 300 100 175 300 300 300 300 100 150 298 15 15 298 298 298 298 15 15 14 0.020016 0.4 0.0201 0.01 0.4 0.95 0.5 200 100 175 200 200 200 200 100 150 15 15 15 15 15 15 15 15 15 14 0.020016 0.4 0.0201 0.01 0.4 0.95 0.9 550 100 150 550 550 550 550 100 350 198 15 15 198 198 198 198 15 15 15 0.01516 0.4 0.0165 0.01 0.6 0.75 0.5 250 100 150 250 250 250 250 100 300 15 15 15 15 15 15 15 15 25 15 0.01516 0.4 0.0165 0.01 0.6 0.75 0.9 350 100 150 350 350 350 350 100 100 298 15 15 298 298 298 298 15 15 16 0.016024 0.4 0.0201 0.01 0.6 0.95 0.5 200 100 150 200 200 200 200 100 100 48 15 15 48 48 48 48 15 15 16 0.016024 0.4 0.0201 0.01 0.6 0.95 0.9 550 100 100 550 550 NA 550 100 200 548 15 15 548 548 NA 548 15 15 17 0.007028 0.4 0.0165 0.01 0.98 0.75 0.5

167

Table A1A2 (Continued).

True Values Release size at Durham Ferry Supplemental release size

Scenario sR sR0 sA sF ψA1 ψA2 pG2a= sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 sR sR0 sA1 sA2 sA sB sF ψA1 ψA2

pG2b 500 100 100 500 500 1100 500 100 200 25 15 15 25 25 48 25 15 15 17 0.007028 0.4 0.0165 0.01 0.98 0.75 0.9 300 100 100 300 300 NA 300 100 100 548 15 15 548 548 NA 548 15 15 18 0.008439 0.4 0.0201 0.01 0.98 0.95 0.5 400 100 100 400 400 1000 400 100 100 48 15 15 48 48 96 48 15 15 18 0.008439 0.4 0.0201 0.01 0.98 0.95 0.9 300 100 175 350 300 300 300 100 500 298 15 15 298 298 298 298 15 15 19 0.01992 0.4 0.0195 0.05 0.4 0.75 0.5 200 100 175 250 200 175 200 100 450 15 15 15 15 15 15 15 15 15 19 0.01992 0.4 0.0195 0.05 0.4 0.75 0.9 200 100 175 250 200 200 200 100 150 398 15 15 398 398 398 398 15 15 20 0.020112 0.4 0.0207 0.05 0.4 0.95 0.5 175 100 175 175 175 175 175 100 150 15 15 15 25 15 15 15 15 15 20 0.020112 0.4 0.0207 0.05 0.4 0.95 0.9 150 100 150 200 150 150 150 100 300 548 15 15 548 548 548 548 15 25 21 0.01588 0.4 0.0195 0.05 0.6 0.75 0.5 250 100 150 250 250 250 250 100 350 15 15 15 15 15 15 15 15 15 21 0.01588 0.4 0.0195 0.05 0.6 0.75 0.9 175 100 150 175 175 175 175 100 100 498 15 15 498 498 498 498 15 15 22 0.016168 0.4 0.0207 0.05 0.6 0.95 0.5 200 100 150 200 200 200 200 100 100 48 15 15 48 48 48 48 15 15 22 0.016168 0.4 0.0207 0.05 0.6 0.95 0.9 300 100 100 300 300 NA 300 100 200 548 15 15 548 548 NA 548 15 15 23 0.008204 0.4 0.0195 0.05 0.98 0.75 0.5 450 100 100 450 450 1000 450 100 200 15 15 15 15 15 75 15 15 15 23 0.008204 0.4 0.0195 0.05 0.98 0.75 0.9 200 100 100 200 200 NA 200 100 100 548 15 15 548 548 NA 548 15 15 24 0.008674 0.4 0.0207 0.05 0.98 0.95 0.5 400 100 100 400 400 1100 400 100 100 25 15 15 25 25 15 25 15 15 24 0.008674 0.4 0.0207 0.05 0.98 0.95 0.9 250 100 150 250 250 250 250 100 200 15 15 15 15 15 15 15 15 15 25 0.05068 0.7 0.031 0.01 0.4 0.75 0.5 100 100 150 150 100 100 100 100 200 15 15 15 15 15 15 15 15 15 25 0.05068 0.7 0.031 0.01 0.4 0.75 0.9 200 100 150 200 200 200 200 100 100 48 15 15 48 48 48 48 15 15 26 0.052696 0.7 0.0382 0.01 0.4 0.95 0.5

168

Table A1A2 (Continued).

Scenario True Values Release size at Durham Ferry Supplemental release size

sR sR0 sA sF ψA1 ψA2 pG2a= sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 sR sR0 sA1 sA2 sA sB sF ψA1 ψA2

pG2b 100 100 150 100 100 100 100 100 100 15 15 15 15 15 15 15 15 15 26 0.052696 0.7 0.0382 0.01 0.4 0.95 0.9 300 100 100 300 300 300 300 100 150 15 15 15 15 15 15 15 15 15 27 0.04102 0.7 0.031 0.01 0.6 0.75 0.5 100 100 100 100 100 100 100 100 150 15 15 15 15 15 15 15 15 15 27 0.04102 0.7 0.031 0.01 0.6 0.75 0.9 250 100 100 250 250 250 250 100 100 25 15 15 25 25 25 25 15 15 28 0.044044 0.7 0.0382 0.01 0.6 0.95 0.5 100 100 100 100 100 100 100 100 100 15 15 15 15 15 15 15 15 15 28 0.044044 0.7 0.0382 0.01 0.6 0.95 0.9 100 100 100 100 100 1150 100 100 100 348 15 15 348 348 546 348 15 15 29 0.022666 0.7 0.031 0.01 0.98 0.75 0.5 150 100 100 150 150 800 150 100 100 15 15 15 15 15 15 15 15 15 29 0.022666 0.7 0.031 0.01 0.98 0.75 0.9 175 100 100 175 175 1100 175 100 100 198 15 15 198 198 546 198 15 15 30 0.027605 0.7 0.0382 0.01 0.98 0.95 0.5 150 100 100 150 150 800 150 100 100 15 15 15 15 15 15 15 15 15 30 0.027605 0.7 0.0382 0.01 0.98 0.95 0.9 200 100 150 250 200 200 200 100 200 48 15 15 48 48 48 48 15 25 31 0.0518 0.7 0.035 0.05 0.4 0.75 0.5 100 100 150 150 100 100 100 100 200 15 15 15 15 15 15 15 15 15 31 0.0518 0.7 0.035 0.05 0.4 0.75 0.9 150 100 150 150 150 150 150 100 100 98 15 15 98 98 98 98 15 15 32 0.05292 0.7 0.039 0.05 0.4 0.95 0.5 100 100 150 100 100 100 100 100 100 15 15 15 25 15 15 15 15 15 32 0.05292 0.7 0.039 0.05 0.4 0.95 0.9 250 100 100 250 250 250 250 100 150 25 15 15 25 25 25 25 15 15 33 0.0427 0.7 0.035 0.05 0.6 0.75 0.5 100 100 100 100 100 100 100 100 150 15 15 15 15 15 15 15 15 15 33 0.0427 0.7 0.035 0.05 0.6 0.75 0.9 250 100 100 250 250 250 250 100 100 25 15 15 25 25 25 25 15 15 34 0.04438 0.7 0.039 0.05 0.6 0.95 0.5 100 100 100 100 100 100 100 100 100 15 15 15 15 15 15 15 15 15 34 0.04438 0.7 0.039 0.05 0.6 0.95 0.9 150 100 100 150 150 1200 150 100 100 248 15 15 248 248 396 248 15 15 35 0.02541 0.7 0.035 0.05 0.98 0.75 0.5

169

Table A1A2 (Continued).

Scenario True Values Release size at Durham Ferry Supplemental release size

sR sR0 sA sF ψA1 ψA2 pG2a= sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 sR sR0 sA1 sA2 sA sB sF ψA1 ψA2

pG2b 150 100 100 150 150 800 150 100 100 15 15 15 15 15 25 15 15 15 35 0.02541 0.7 0.035 0.05 0.98 0.75 0.9 175 100 100 175 175 1200 175 100 100 198 15 15 198 198 396 198 15 15 36 0.028154 0.7 0.039 0.05 0.98 0.95 0.5 150 100 100 150 150 800 150 100 100 15 15 15 15 15 15 15 15 15 36 0.028154 0.7 0.039 0.05 0.98 0.95 0.9 100 100 100 250 100 100 100 100 150 25 15 15 25 25 25 25 15 15 37 0.10845 0.9 0.06125 0.01 0.4 0.75 0.5 100 100 100 150 100 100 100 100 150 15 15 15 15 15 15 15 15 15 37 0.10845 0.9 0.06125 0.01 0.4 0.75 0.9 100 100 100 175 100 100 100 100 100 15 15 15 15 15 15 15 15 15 38 0.11385 0.9 0.07625 0.01 0.4 0.95 0.5 100 100 100 100 100 100 100 100 100 15 15 15 15 15 15 15 15 15 38 0.11385 0.9 0.07625 0.01 0.4 0.95 0.9 150 100 100 175 150 150 150 100 100 15 15 15 15 15 15 15 15 15 39 0.090675 0.9 0.06125 0.01 0.6 0.75 0.5 100 100 100 100 100 100 100 100 100 15 15 15 15 15 15 15 15 15 39 0.090675 0.9 0.06125 0.01 0.6 0.75 0.9 100 100 100 150 100 150 100 100 100 25 15 15 25 25 25 25 15 15 40 0.098775 0.9 0.07625 0.01 0.6 0.95 0.5 100 100 100 100 100 100 100 100 100 15 15 15 15 15 15 15 15 15 40 0.098775 0.9 0.07625 0.01 0.6 0.95 0.9 200 100 100 200 200 NA 200 100 100 15 15 15 15 15 NA 15 15 15 41 0.056903 0.9 0.06125 0.01 0.98 0.75 0.5 100 100 100 100 100 950 100 100 100 15 15 15 15 15 15 15 15 15 41 0.056903 0.9 0.06125 0.01 0.98 0.75 0.9 150 100 100 150 150 NA 150 100 100 15 15 15 15 15 NA 15 15 15 42 0.070133 0.9 0.07625 0.01 0.98 0.95 0.5 100 100 100 100 100 900 100 100 100 15 15 15 15 15 15 15 15 15 42 0.070133 0.9 0.07625 0.01 0.98 0.95 0.9 100 100 100 200 100 100 100 100 150 15 15 15 48 15 15 15 15 15 43 0.11025 0.9 0.06625 0.05 0.4 0.75 0.5 100 100 100 150 100 100 100 100 150 15 15 15 15 15 15 15 15 15 43 0.11025 0.9 0.06625 0.05 0.4 0.75 0.9 100 100 100 200 100 100 100 100 100 15 15 15 15 15 15 15 15 15 44 0.11421 0.9 0.07725 0.05 0.4 0.95 0.5

170

Table A1A2 (Continued).

True Values Release size at Durham Ferry Supplemental release size Scenario

sR sR0 sA sF ψA1 ψA2 pG2a= sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 sR sR0 sA1 sA2 sA sB sF ψA1 ψA2

pG2b 100 100 100 100 100 100 100 100 100 15 15 15 15 15 15 15 15 15 44 0.11421 0.9 0.07725 0.05 0.4 0.95 0.9 100 100 100 175 100 150 100 100 100 48 15 15 48 48 48 48 15 15 45 0.093375 0.9 0.06625 0.05 0.6 0.75 0.5 100 100 100 100 100 100 100 100 100 15 15 15 15 15 15 15 15 15 45 0.093375 0.9 0.06625 0.05 0.6 0.75 0.9 100 100 100 150 100 150 100 100 100 48 15 15 48 48 48 48 15 15 46 0.099315 0.9 0.07725 0.05 0.6 0.95 0.5 100 100 100 100 100 100 100 100 100 15 15 15 15 15 15 15 15 15 46 0.099315 0.9 0.07725 0.05 0.6 0.95 0.9

100 100 100 150 100 NA 100 100 100 98 15 15 98 98 NA 98 15 15 47 0.061313 0.9 0.06625 0.05 0.98 0.75 0.5 100 100 100 100 100 900 100 100 100 15 15 15 15 15 48 15 15 15 47 0.061313 0.9 0.06625 0.05 0.98 0.75 0.9 150 100 100 150 150 NA 150 100 100 15 15 15 15 15 NA 15 15 15 48 0.071015 0.9 0.07725 0.05 0.98 0.95 0.5 100 100 100 100 100 900 100 100 100 15 15 15 15 15 46 15 15 15 48 0.071015 0.9 0.07725 0.05 0.98 0.95 0.9

171

Simulation output for Mixed Survival Scenarios

Table A1A3. Simulation and minimization output for mixed survival scenarios using a single release at Durham Ferry: minimum release size required to estimate

parameters. For each scenario, sF = 0.05, ψA2=0.75, and pA1a=pA1b=pA2a=pA2b=pB1a=pB1b=pF1a=pF1b=0.85. Survival scenario number does not correspond directly with scenarios in Table A3. Scenarios 7, 15, and 22 here correspond to scenarios 3, 5, and 6 in Table A3, respectively (highlighted).

True Parameter Values Minimum release size to estimate

Scenario sR sR0 sA1 sA2 sB ψA1 pG2a=pG2b sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 1 0.002167 0.7 0.1 0.02 0.02 0.98 0.5 NA 100 100 3000 3000 3000 NA 100 350 1 0.002167 0.7 0.1 0.02 0.02 0.98 0.9 1750 100 100 1000 1000 1000 1750 100 350 2 0.004102 0.2 0.3 0.07 0.07 0.98 0.5 3000 100 150 1750 1750 4000 3000 100 400 2 0.004102 0.2 0.3 0.07 0.07 0.98 0.9 1000 100 150 500 500 2000 1000 100 400 3 0.006993 0.9 0.1 0.07 0.07 0.98 0.5 1750 100 100 1000 1000 1000 1750 100 250 3 0.006993 0.9 0.1 0.07 0.07 0.98 0.9 550 100 100 300 300 450 550 100 250 4 0.007963 0.7 0.1 0.02 0.02 0.5 0.5 1500 100 175 1400 1400 1400 1500 100 648 4 0.007963 0.7 0.1 0.02 0.02 0.5 0.9 500 100 175 450 450 450 500 100 648 5 0.00895 0.2 0.3 0.07 0.07 0.5 0.5 1400 100 250 1296 1296 1296 1400 150 800 5 0.00895 0.2 0.3 0.07 0.07 0.5 0.9 450 100 250 400 350 350 450 150 750 6 0.0151 0.4 0.3 0.15 0.05 0.98 0.5 800 100 100 400 400 1296 800 100 200 6 0.0151 0.4 0.3 0.15 0.05 0.98 0.9 250 100 100 175 150 750 250 100 200 7 0.0163 0.4 0.3 0.15 0.2 0.98 0.5 750 100 100 400 400 4000 750 100 200 7 0.0163 0.4 0.3 0.15 0.2 0.98 0.9 250 100 100 175 150 3000 250 100 200 8 0.016325 0.2 0.5 0.2 0.1 0.98 0.5 750 100 175 700 400 4000 750 100 250 8 0.016325 0.2 0.5 0.2 0.1 0.98 0.9 250 100 150 250 150 3000 250 100 250 9 0.016725 0.2 0.5 0.2 0.2 0.98 0.5 750 100 175 700 400 NA 750 100 250 9 0.016725 0.2 0.5 0.2 0.2 0.98 0.9 250 100 175 250 150 NA 250 100 250 10 0.0175 0.4 0.3 0.15 0.05 0.5 0.5 700 100 150 750 500 500 700 100 400 10 0.0175 0.4 0.3 0.15 0.05 0.5 0.9 250 100 150 350 150 150 250 100 400 11 0.018125 0.2 0.5 0.2 0.1 0.5 0.5 700 100 350 1000 500 500 700 150 450 11 0.018125 0.2 0.5 0.2 0.1 0.5 0.9 250 100 300 500 150 150 250 150 450

172

Table A1A3 (Continued) True Parameter Values Minimum release size to estimate

Scenario sR sR0 sA1 sA2 sB ψA1 pG2a=pG2b sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 12 0.028125 0.2 0.5 0.2 0.2 0.5 0.5 450 100 350 900 350 550 450 150 450 12 0.028125 0.2 0.5 0.2 0.2 0.5 0.9 150 100 350 450 150 200 150 150 450 13 0.03265 0.4 0.5 0.2 0.1 0.98 0.5 400 100 100 300 200 1750 400 100 150 13 0.03265 0.4 0.5 0.2 0.1 0.98 0.9 150 100 100 150 100 1296 150 100 150 14 0.03345 0.4 0.5 0.2 0.2 0.98 0.5 350 100 100 350 200 4000 350 100 150 14 0.03345 0.4 0.5 0.2 0.2 0.98 0.9 150 100 100 150 100 3000 150 100 150 15 0.034425 0.9 0.1 0.07 0.07 0.5 0.5 350 100 150 450 350 350 350 100 500 15 0.034425 0.9 0.1 0.07 0.07 0.5 0.9 150 100 150 250 100 100 150 100 500 16 0.03514 0.7 0.4 0.15 0.06 0.98 0.5 350 100 100 175 175 700 350 100 100 16 0.03514 0.7 0.4 0.15 0.06 0.98 0.9 150 100 100 100 100 550 150 100 100 17 0.03625 0.4 0.5 0.2 0.1 0.5 0.5 350 100 150 550 250 250 350 100 250 17 0.03625 0.4 0.5 0.2 0.1 0.5 0.9 150 100 150 250 100 100 150 100 250 18 0.0385 0.7 0.4 0.15 0.06 0.5 0.5 350 100 100 300 250 250 350 100 175 18 0.0385 0.7 0.4 0.15 0.06 0.5 0.9 100 100 100 150 100 100 100 100 150 19 0.04518 0.9 0.4 0.15 0.06 0.98 0.5 300 100 100 150 150 550 300 100 100 19 0.04518 0.9 0.4 0.15 0.06 0.98 0.9 100 100 100 100 100 400 100 100 100 20 0.0468 0.9 0.4 0.15 0.15 0.98 0.5 300 100 100 150 150 1296 300 100 100 20 0.0468 0.9 0.4 0.15 0.15 0.98 0.9 100 100 100 100 100 800 100 100 100 21 0.0475 0.4 0.3 0.15 0.2 0.5 0.5 300 100 150 600 250 350 300 100 400 21 0.0475 0.4 0.3 0.15 0.2 0.5 0.9 100 100 150 350 100 100 100 100 400 22 0.0495 0.9 0.4 0.15 0.06 0.5 0.5 250 100 100 250 175 175 250 100 150 22 0.0495 0.9 0.4 0.15 0.06 0.5 0.9 100 100 100 150 100 100 100 100 150 23 0.05625 0.4 0.5 0.2 0.2 0.5 0.5 250 100 175 450 175 250 250 100 250 23 0.05625 0.4 0.5 0.2 0.2 0.5 0.9 100 100 150 250 100 100 100 100 250 24 0.073463 0.9 0.5 0.2 0.1 0.98 0.5 175 100 100 150 100 800 175 100 100 24 0.073463 0.9 0.5 0.2 0.1 0.98 0.9 100 100 100 100 100 600 100 100 100

173

Table A1A3 (Continued) True Parameter Values Minimum release size to estimate

Scenario sR sR0 sA1 sA2 sB ψA1 pG2a=pG2b sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 25 0.075263 0.9 0.5 0.2 0.2 0.98 0.5 175 100 100 150 100 1400 175 100 100 25 0.075263 0.9 0.5 0.2 0.2 0.98 0.9 100 100 100 100 100 1000 100 100 100 26 0.081563 0.9 0.5 0.2 0.1 0.5 0.5 175 100 100 250 100 100 175 100 100 26 0.081563 0.9 0.5 0.2 0.1 0.5 0.9 100 100 100 150 100 100 100 100 100 27 0.09 0.9 0.4 0.15 0.15 0.5 0.5 150 100 100 250 150 150 150 100 150 27 0.09 0.9 0.4 0.15 0.15 0.5 0.9 100 100 100 150 100 100 100 100 150 28 0.126563 0.9 0.5 0.2 0.2 0.5 0.5 100 100 100 200 100 150 100 100 100 28 0.126563 0.9 0.5 0.2 0.2 0.5 0.9 100 100 100 150 100 100 100 100 100

174

Table A1A4. Simulation and minimization output for mixed survival scenarios using a primary release at Durham Ferry (R1) and a supplemental release in the San Joaquin

River downstream of the head of Old River (R2): minimum release sizes required to estimate parameters. For each scenario, sF = 0.05, ψA2=0.75, and pA1a=pA1b=pA2a=pA2b=pB1a=pB1b=pF1a=pF1b=0.85. Survival scenario number does not correspond directly with scenarios in Table A3. Scenarios 7, 15, and 22 here correspond to scenarios 3, 5, and 6 in Table A3 (highlighted). True Values Release size at Durham Ferry Supplemental release size

Scenario sR sR0 sA1 sA2 sB ψA1 pG2a=pG2b sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 sR sR0 sA1 sA2 sA sB sF ψA1 ψA2

1 0.002167 0.7 0.1 0.02 0.02 0.98 0.5 NA 100 100 NA NA NA NA 100 350 NA 15 15 NA NA NA NA 15 15 1 0.002167 0.7 0.1 0.02 0.02 0.98 0.9 NA 100 100 NA NA NA NA 100 350 NA 15 15 NA NA NA NA 15 15 2 0.004102 0.2 0.3 0.07 0.07 0.98 0.5 550 100 150 550 550 NA 550 100 400 548 15 15 548 548 NA 548 15 15 2 0.004102 0.2 0.3 0.07 0.07 0.98 0.9 700 100 150 700 700 NA 700 100 400 146 15 15 146 146 NA 146 15 15 3 0.006993 0.9 0.1 0.07 0.07 0.98 0.5 1150 100 100 1150 1150 1150 1150 100 250 596 15 15 596 596 596 596 15 15 3 0.006993 0.9 0.1 0.07 0.07 0.98 0.9 550 100 100 550 550 550 550 100 250 15 15 15 15 15 15 15 15 48 4 0.007963 0.7 0.1 0.02 0.02 0.5 0.5 NA 100 150 NA NA NA NA 100 700 NA 15 15 NA NA NA NA 15 48 4 0.007963 0.7 0.1 0.02 0.02 0.5 0.9 450 100 100 450 450 450 450 100 700 48 15 48 48 48 48 48 15 48 5 0.00895 0.2 0.3 0.07 0.07 0.5 0.5 350 100 300 450 350 250 350 150 750 473 15 48 473 473 548 473 15 25 5 0.00895 0.2 0.3 0.07 0.07 0.5 0.9 400 100 300 400 400 400 400 150 750 15 15 48 15 15 25 15 15 25 6 0.0151 0.4 0.3 0.15 0.05 0.98 0.5 100 100 100 350 100 900 100 100 200 348 15 15 348 348 546 348 15 15 6 0.0151 0.4 0.3 0.15 0.05 0.98 0.9 250 100 100 250 250 850 250 100 200 15 15 15 15 15 46 15 15 15 7 0.0163 0.4 0.3 0.15 0.2 0.98 0.5 100 100 100 350 100 NA 100 100 200 348 15 15 348 348 NA 348 15 15 7 0.0163 0.4 0.3 0.15 0.2 0.98 0.9 200 100 100 200 200 NA 200 100 200 48 15 15 48 48 NA 48 15 15 8 0.016325 0.2 0.5 0.2 0.1 0.98 0.5 100 100 175 750 100 NA 100 100 250 148 15 25 148 148 NA 148 15 15 8 0.016325 0.2 0.5 0.2 0.1 0.98 0.9 200 100 175 250 200 NA 200 100 250 15 15 25 25 15 NA 15 15 15 9 0.016725 0.2 0.5 0.2 0.2 0.98 0.5 100 100 175 750 100 NA 100 100 250 148 15 15 148 148 NA 148 15 15 9 0.016725 0.2 0.5 0.2 0.2 0.98 0.9 200 100 175 250 200 NA 200 100 250 15 15 25 25 15 NA 15 15 15 10 0.0175 0.4 0.3 0.15 0.05 0.5 0.5 150 100 150 750 150 100 150 100 400 298 15 15 298 298 348 298 15 15 10 0.0175 0.4 0.3 0.15 0.05 0.5 0.9 200 100 150 350 200 200 200 100 400 15 15 15 15 15 15 15 15 15 11 0.018125 0.2 0.5 0.2 0.1 0.5 0.5 150 100 350 1150 175 100 150 150 450 148 15 15 148 148 148 148 15 15 11 0.018125 0.2 0.5 0.2 0.1 0.5 0.9 200 100 350 500 200 150 200 150 450 15 15 15 25 15 48 15 15 15 12 0.028125 0.2 0.5 0.2 0.2 0.5 0.5 175 100 350 900 175 100 175 150 450 98 15 15 98 98 248 98 15 15 12 0.028125 0.2 0.5 0.2 0.2 0.5 0.9 150 100 350 500 150 175 150 150 450 15 15 15 15 15 48 15 15 15

175

Table A1A4 (Continued) True Values Release size at Durham Ferry Supplemental release size

Scenario sR sR0 sA1 sA2 sB ψA1 pG2a=pG2b sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 sR sR0 sA1 sA2 sA sB sF ψA1 ψA2

13 0.03265 0.4 0.5 0.2 0.1 0.98 0.5 100 100 100 450 100 NA 100 100 150 125 15 15 125 125 NA 125 15 15 13 0.03265 0.4 0.5 0.2 0.1 0.98 0.9 100 100 100 150 100 1250 100 100 150 15 15 15 15 15 346 15 15 15 14 0.03345 0.4 0.5 0.2 0.2 0.98 0.5 100 100 100 450 100 NA 100 100 150 125 15 15 125 125 NA 125 15 15 14 0.03345 0.4 0.5 0.2 0.2 0.98 0.9 100 100 100 150 100 NA 100 100 150 15 15 15 15 15 NA 15 15 15 15 0.034425 0.9 0.1 0.07 0.07 0.5 0.5 350 100 100 350 350 350 350 100 500 15 15 25 15 15 15 15 15 25 15 0.034425 0.9 0.1 0.07 0.07 0.5 0.9 150 100 100 250 150 100 150 100 500 15 15 25 15 15 48 15 15 25 16 0.03514 0.7 0.4 0.15 0.06 0.98 0.5 100 100 100 175 100 1000 100 100 100 198 15 15 198 198 48 198 15 15 16 0.03514 0.7 0.4 0.15 0.06 0.98 0.9 100 100 100 100 100 550 100 100 100 15 15 15 15 15 15 15 15 15 17 0.03625 0.4 0.5 0.2 0.1 0.5 0.5 100 100 175 550 100 100 100 100 250 125 15 15 125 125 125 125 15 15 17 0.03625 0.4 0.5 0.2 0.1 0.5 0.9 100 100 175 250 100 100 100 100 250 15 15 15 15 15 15 15 15 15 18 0.0385 0.7 0.4 0.15 0.06 0.5 0.5 100 100 100 300 100 100 100 100 175 175 15 15 175 175 175 175 15 15 18 0.0385 0.7 0.4 0.15 0.06 0.5 0.9 100 100 100 150 100 100 100 100 175 15 15 15 15 15 15 15 15 15 19 0.04518 0.9 0.4 0.15 0.06 0.98 0.5 175 100 100 175 175 750 175 100 100 75 15 15 75 75 46 75 15 15 19 0.04518 0.9 0.4 0.15 0.06 0.98 0.9 100 100 100 100 100 400 100 100 100 15 15 15 15 15 25 15 15 15 20 0.0468 0.9 0.4 0.15 0.15 0.98 0.5 250 100 100 250 250 NA 250 100 100 15 15 15 15 15 NA 15 15 15 20 0.0468 0.9 0.4 0.15 0.15 0.98 0.9 100 100 100 100 100 850 100 100 100 15 15 15 15 15 46 15 15 15 21 0.0475 0.4 0.3 0.15 0.2 0.5 0.5 250 100 150 550 250 350 250 100 400 15 15 15 15 15 15 15 15 15 21 0.0475 0.4 0.3 0.15 0.2 0.5 0.9 100 100 150 350 100 100 100 100 400 15 15 15 15 15 15 15 15 15 22 0.0495 0.9 0.4 0.15 0.06 0.5 0.5 150 100 100 250 150 150 150 100 150 98 15 15 98 98 98 98 15 15 22 0.0495 0.9 0.4 0.15 0.06 0.5 0.9 100 100 100 150 100 100 100 100 150 15 15 15 15 15 15 15 15 15 23 0.05625 0.4 0.5 0.2 0.2 0.5 0.5 100 100 175 450 100 175 100 100 250 98 15 15 98 98 148 98 15 15 23 0.05625 0.4 0.5 0.2 0.2 0.5 0.9 100 100 175 250 100 100 100 100 250 15 15 25 15 15 15 15 15 15 24 0.073463 0.9 0.5 0.2 0.1 0.98 0.5 150 100 100 200 150 1050 150 100 100 15 15 15 15 15 75 15 15 15 24 0.073463 0.9 0.5 0.2 0.1 0.98 0.9 100 100 100 100 100 700 100 100 100 15 15 15 15 15 15 15 15 15 25 0.075263 0.9 0.5 0.2 0.2 0.98 0.5 150 100 100 200 150 NA 150 100 100 15 15 15 15 15 NA 15 15 15 25 0.075263 0.9 0.5 0.2 0.2 0.98 0.9 100 100 100 100 100 1100 100 100 100 15 15 15 15 15 15 15 15 15

176

Table A1A4 (Continued)

True Values Release size at Durham Ferry Supplemental release size

Scenario sR sR0 sA1 sA2 sB ψA1 pG2a=pG2b sR sR0 sA1 sA2 sA sB sF ψA1 ψA2 sR sR0 sA1 sA2 sA sB sF ψA1 ψA2

26 0.081563 0.9 0.5 0.2 0.1 0.5 0.5 100 100 100 300 100 100 100 100 100 48 15 15 48 48 48 48 15 15 26 0.081563 0.9 0.5 0.2 0.1 0.5 0.9 100 100 100 150 100 100 100 100 100 15 15 15 15 15 15 15 15 15 27 0.09 0.9 0.4 0.15 0.15 0.5 0.5 150 100 100 250 150 150 150 100 150 15 15 15 15 15 15 15 15 15 27 0.09 0.9 0.4 0.15 0.15 0.5 0.9 100 100 100 150 100 100 100 100 150 15 15 15 15 15 15 15 15 15 28 0.126563 0.9 0.5 0.2 0.2 0.5 0.5 100 100 100 200 100 150 100 100 100 15 15 15 25 15 15 15 15 15 28 0.126563 0.9 0.5 0.2 0.2 0.5 0.9 100 100 100 150 100 100 100 100 100 15 15 15 15 15 15 15 15 15

177

Appendix B Standard Operating Procedure for the 2015 South Delta Chinook Salmon Survival Study (3/5/2015) Materials Needed • Dissolved oxygen (DO) meter (e.g., YSI 85) • Acoustic tags and dummy tags (V-4) • VEMCO acoustic tag activator • VEMCO acoustic tag verification equipment (VR-100, VRHR prototype receivers) • Pill boxes for tag distribution • Chlorhexidine solution (Novalsan ; 30mL/L D-H2O) • Distilled or de-ionized water (D-H2O) • Aqui-S 20E (undiluted, directly from manufacturer) • Stress coat - stock concentration and 25% solution (250mL/L D-H2O) • Disinfectant solution (Virkon Aquatic or 70% ETOH) • 19 L bucket(s) marked at 10 L and clearly labeled ‘Anesthesia’ • 19 L buckets clearly labeled ‘Reject’ for fish not selected for tagging procedures • 19 L buckets clearly labeled “Lethal” for fish that need to be euthanized • 19 L buckets for post-surgical recovery of fish • Two gravity feed containers marked at 10 L, and connected by rubber tubing with in-line shut-off valves (one labeled ‘anesthesia’ and one labeled ‘freshwater’) • Designated syringes (5 mL) for measuring anesthetic and stress coat • Oxygen delivery system (cylinder, regulator, airline, air diffusers) for recovery buckets • Dip nets • Nitrile gloves (in all sizes) • Scale measuring to the nearest 0.1 g (weighing fish) • Scale measuring to the nearest 0.001 g (weighing tags) • Large plastic weigh boats or Tupperware container to weigh fish • Measuring board with ruler to the nearest millimeter • Surgical platform (cradle) • Autoclave • Trays for holding solutions used to disinfect surgical tools • Trays to rinse disinfected tools • Needle drivers (multiple sets) • Forceps (multiple sets) • Scalpel handle and blades (multiple sets) • Scissors (multiple sets) • Sutures: Vicryl plus 5-0 with an RB-1 needle • Spray bottles for disinfectant solution • Timer(s) • Sharps container • Datasheets, clipboards, and writing tools • Clip on tag labels to identify fish in recovery buckets • Clean rags for keeping tagging areas clean and dry • Aerators for bucket use (tagger recovery bucket, recovery at code out) Pre-tagging Activities • All acoustic tags will be weighed to the nearest 0.001 g • All acoustic tags need to be soaked a minimum of 24 hours prior to surgery in a saline solution to ensure that the tags are waterproof, and that the seals encapsulating the tags are functional (see the SOP on tag soak procedures) • Rinse, dry, and activate transmitters the day before they are to be implanted. Confirm operational status with the VEMCO tag activator and record the date and time when a tag is activated 178

Equipment Set up • Remove transport containers from the freezer and prepare them to receive tagged fish o Transport containers that leave the hatchery grounds and are delivered to the release site must be frozen for at least 24 h prior to being used again for the tagging operation. These details are outlined in the project Biosecurity Plan o When removing containers from the freezer, be sure to consult with the tagging coordinator to ensure that all containers undergo the minimum 24 h of exposure before they are removed and used • Prepare the transport truck to be able to circulate water through containers • Select and isolate fish for the first tag session o Use the crowder to concentrate source fish into a small section of the raceway to facilitate netting o Carefully net enough fish for the tag session into perforated 44 gallon cans positioned near the raceway wall o Monitor the density of fish in the holding cans to ensure they are not overcrowded and add additional cans as needed o Keep the holding cans covered to reduce disturbance and risk of fish loss back to the main raceway area o Move the crowder back to allow remaining fish sufficient space to reduce stress • Water temperatures during all aspects of the tagging operations cannot exceed 2 °C difference from the reference water source (for this study, the raceway where source fish are held) o Anesthesia buckets, gravity feed carboys, recovery buckets, and totes should not be filled until near the time they are needed to avoid warming o Anesthesia bucket and gravity feed carboys should be replaced regularly to prevent increasing water temperatures over time • Fill disinfection trays for surgical instruments with Novalsan • Fill rinse tray with de-ionized or distilled water • Fill pill boxes containing study tags with Novalsan and allow at least 20 minutes of contact time with the disinfectant. Following disinfection, thoroughly rinse transmitters in distilled or de-ionized water prior to implantation. Transmitters should only be handled by gloved hands or clean surgical instruments such as forceps following the disinfection step • Set up and calibrate scale, measuring board, and surgical platform • Fill gravity feed carboys with water from raceway o Add 1 ml Aqui-S 20E to the 10L of water in the anesthesia carboy and briefly agitate to ensure dispersal o The freshwater carboy is filled from the raceway and has no anesthesia added • Fill anesthesia bucket to 10 L line with water from source tank or raceway. Add 3 ml Aqui-S 20E and briefly agitate to ensure dispersal. Cover with a lid • Adding Aqui-S 20E to any container should be done carefully, with communication between the tagger and the assistant to avoid double dose or no dose outcomes • Retrieve a 5 gallon fish recovery bucket filled with water from the raceway that has been supersaturated with 130% to 150% oxygen. Add stress coat • Reference Tag and Tote inventory sheet and retrieve clip-on tag ID labels for recovery buckets to be used during tag operations • Check that a reject bucket has been filled with water from the source tank or raceway and is outfitted with an air bubbler • Check that a clearly labeled lethal bucket is ready for fish that need to be euthanized. This bucket should be positioned well away from the tagging stations to ensure that it is not confused with an anesthesia bucket • Start a tag data sheet and a daily fish reject tally datasheet for each tagging stations to account for fish that are handled but not tagged • The tagger should wear clean medical grade exam gloves during all procedures that involve handling fish Surgical Implantation of the Transmitter • Food should be withheld from fish for ~24 h prior to surgical implantation of the transmitter

Anesthetize Fish and Collect Morphometric Data • Net one fish from raceway and place directly into an anesthesia bucket 179

o Work from one holding container until it is empty to avoid stressing fish in all containers o Start a stopwatch immediately after the fish has been placed into the anesthesia bucket in order to track how long the fish is exposed to anesthesia o Place a lid on the bucket and deliver the bucket to a tagging station • Remove the lid after about 1 minute to observe the fish for loss of equilibrium. Keep the fish in the water for an additional 30-60 seconds after it has lost equilibrium o Time of sedation should normally be 2–4 minutes, with an average of about 3 minutes o If loss of equilibrium takes less than 1 minute or if a fish is in the anesthesia bucket for more than 5 minutes, reject that fish o If after sedating a few fish, if they are consistently losing equilibrium in more or less time than typical, the anesthesia concentration may need to be adjusted. This should only be done after consultation with the tag coordinator, and should be done in 0.5 ml increments. Concentration changes should be executed for all taggers simultaneously and recorded on the tagging datasheet • If a fish is unacceptable for tagging, place the fish in the reject bucket, inform the data recorder, and record it on the daily reject tally sheet • Record fish length, weight, and scale condition: o Start “air time” timer when a fish is removed from the anesthesia bucket o Transfer the fish to the scale and weigh to the nearest 0.1g . A fish is acceptable for tagging if it weighs at least 8.5 g, so that the tag burden does not exceed 5% of the weight of the fish. The transmitters used for this study are VEMCO brand, model V4, which weigh about 0.42 g in air o Transfer the fish to the measuring board. Measure fork length (FL) to the nearest mm o Check for any abnormalities and descaling . A fish is acceptable for tagging when it lacks deformations such as: abnormal color, gross anatomical deformations, damaged opercula with exposed gill filaments, gross scarring, bleeding scratches, any bulging eyes, gross signs of disease, any fungal infection, or any fin hemorrhaging . Scale condition is noted as Normal (N), Partial (P), or Descaled (D) and is assessed on the most compromised side of each fish. The normal scale condition is defined as loss of less than 5% of scales on one side of the fish. Partial descaling is defined as loss of 6–19% of scales on one side of the fish. Fish are classified as descaled if they have lost 20% or more of the scales on one side of the fish, and should not be tagged due to compromised osmoregulatory ability • Data must be vocally relayed to the recorder and the recorder should repeat the information back to the tagger to avoid miscommunication • Any fish dropped on the floor should be rejected. Fish dropped from the surgical platform to the table or working surface may be advanced through the tagging process or rejected based on the tagger’s evaluation of the fish. • The anesthesia containers should be emptied and remixed at regular intervals throughout the tagging operation to ensure the appropriate concentration and to avoid warming • The gravity feed containers should be monitored for volume and temperature and changed as needed to avoid inadequate volume to complete a surgery and significant warming (difference in water temperature from the raceway cannot exceed 2 °C

Transmitter Implantation • Place the fish into the surgical platform ventral side up • Anesthesia should be administered through the gravity feed tube as soon as the fish is on the surgery platform. Using the in-line valve, adjust the flow as needed so that the gilling rate of the fish is steady • Using a scalpel, make an incision approximately 5 mm in length beginning a few mm in front of the pelvic girdle. The incision should be just deep enough to penetrate the peritoneum, avoiding the internal organs. The spleen is generally near the incision point so pay close attention to the depth of the incision • Use forceps to open the incision to check that you did not damage any internal organs or cause excessive bleeding. If you observe damage or think you damaged an organ, do not implant the tag – reject that fish

180

• One scalpel blade can be used on about 5–7 fish. If the scalpel is pulling rough or making jagged incisions, it needs to be changed prior to tagging the next fish • Remove a disinfected transmitter from the pill box • Confirm the tube ID with the recorder and place the empty vial into the lid of the tray which holds the tags • Gently insert the tag into the body cavity and position it so that it lies directly beneath the incision and the ceramic head is facing forward. This positioning will provide a barrier between the suture needle and internal organs • Suture the incision with two to three interrupted stitches. Make note on the datasheet when three stitches are used, as two stitches is assumed to be the typical condition • Transfer the fish from the surgical platform to the appropriate recovery bucket with minimal handling by moving the platform as close as possible to the bucket or using a liner material to lift the fish for transfer o Immediately following surgery fish will be held in recovery containers that provide 130% to 150% DO for a minimum of 10 minutes o Holding time in recovery containers begins when the last fish is added to the container and will be monitored using a timer o Two recovery buckets are used for each group of three fish that will be transferred into one tote for transport to the release site. Call out the count of fish in the recovery buckets to the tagging assistant/recorder for confirmation. Put the lids back on the buckets. Once 3 fish are in the 2 buckets that make up a respective tote, attach the clipboard with tag datasheet to one of the two buckets and have the tagging assistant move the buckets to the tag verification staging area • Between surgeries the tagger should replace the instruments that were just used into the disinfectant bath. Each tagger will have at least 3 sets of surgical instruments to rotate through to ensure that tools get a thorough soaking in disinfectant between uses. Once disinfected, instruments should be rinsed in distilled or deionized water. Organic debris in the disinfectant bath reduced effectiveness so be sure to change the bath regularly Transmitter Verification • Obtain buckets and datasheet from tagging crew and start a timer for the 10 minute surgical recovery period • Gently place hydrophones attached to a VEMCO VR-100 or VRHR prototype receiver into each bucket • Watch the display on the VR-100 or computer for tag codes that appear on the monitoring screen. As tag codes are verified circle the tag code that is read on the display on the copy of the Tag and Tote provided to the tag verifier • Once all tags in a bucket have been verified, remove the hydrophone and secure the lid until the recovery period is complete • Once the 10 minute recovery period is complete, transfer the 2 buckets to an 18 gallon tote and confirm that all fish have recovered from anesthesia and are swimming normally. Move the tote to the truck loading area. If after the 10 minute recovery period, tag codes are not verified, continue to attempt verification by separating fish to one per bucket • If a tag does not code out, notify the tag coordinator and return the fish to the tagger who performed the surgery for tag extraction. Once the tag is removed, return to tag coordinator for a replacement tag to complete tag implantation Loading for Transport • Begin completion of fish loading, transport, and release data sheets • Fill hauling tank with water at same temperature as source tank and make sure the flow through system is established before notifying the tag coordinator that tagging can commence • Record temperature and DO in the transport tank • Bring buckets to the truck and check each for general fish condition and dead fish before placing into the tank. If a dead fish is found, notify the tag coordinator and return the fish to the tagger who performed the surgery for tag extraction. Once the tag is removed, return it to the tag coordinator so the tag code can be verified and a plan for reuse of the tag can be determined. The original entry should be crossed out in the data sheet with a comment of mort at loading • Call out the number of the bucket to the recorder and the number of fish in the bucket • Once all buckets have been loaded, confirm that the number of buckets matches the number that should be loaded and that there are no buckets remaining in the tagging area

181

• Secure the tank and tank lid for transport • Send previous days datasheets with transport crew (first transport truck) End of Session Activities • Validation of tag data and datasheet accuracy o Working together, each tagger and assistant team will review the transmitter tubes/serial numbers against the tag and tote inventory and the datasheets to verify that all of the transmitters provided for the session were implanted into study fish o The steps of the verification process should include reading the serial number on each tag tube, finding that serial number on the datasheet to confirm that it was implanted, and a simple count of the tags provided (as shown on the tag-tote inventory) vs. the tag tubes and data rows on the datasheets o Once the validation steps have been completed, both the tagger and the assistant initial the datasheet to confirm that the validation step has been completed • Review all datasheets and complete any missing information (e.g., tag end time, page numbers, validation initials) • Collect all datasheets, pill boxes, coin envelopes, and tag tubes and hand them in to the tagging coordinator • Organize tagging solutions and surgical instruments to be ready for the next tagging session End of Day Clean Up • At the end of each tagging day, wipe down or spray all surfaces with Virkon or 70% ETOH to disinfect • Use a toothbrush to remove all large organic debris from instruments, rinse them and dry them to prevent rust • Return all surgical instruments to the office for autoclaving • Make surgical tagging solutions as needed to be ready for the next tagging session • Inventory chemical solutions and tagging supplies (blades and suture) • Return any soiled rags to the office and have them washed • Rinse buckets with hose and place upside down to dry • Turn off oxygen cylinder General Fish Handling Reminders • Anesthesia and freshwater carboys and buckets should be filled just prior to tagging to avoid temperature changes and should be changed often. Check levels of carboys before each surgery to be certain that you will not run out of water during a surgery • USE CAUTION and COMMUNICATION when adding Aqui-S 20E to any container to avoid adding two doses or no doses to the container • Keep a lid on any bucket or tote that contains fish • Any fish dropped on the floor should be rejected. If a fish is dropped on the floor after it has been tagged, euthanize the fish, remove the tag, and place it into another fish • CAREFULLY HANDLE BUCKETS. Try not to bang them around, slam the handles, or otherwise handle in a rough manner as this can stress fish

182

Appendix C.

U.S. Fish & Wildlife Service

P ATHOGEN SCREENING AND GILL NA+/K+- ATPASE ASSESSMENT OF SOUTH DELTA CHINOOK AND STEELHEAD 2015 RELEASE GROUPS Ken Nichols

October 2015

US Fish and Wildlife Service California-Nevada Fish Health Center 24411 Coleman Fish Hatchery Rd Anderson, CA 96007 (530) 365-4271 Fax: (530) 365-7150 http://www.fws.gov/canvfhc/ 183

Summary The California-Nevada Fish Health Center performed an assessment of the health and smolt development of steelhead trout and Chinook salmon utilized in south Delta salmonid survival studies in March-May, 2015. A suite of pathogen assays and gill Na+/K+-ATPase activity measurements were performed on cohorts of acoustic tagged release groups to help explain any performance and survival differences during the studies. In both steelhead and Chinook health assessment groups, mortality over the holding period was low (0-4%) and no significant pathogen infections were detected. Differences in gill Na+/K+-ATPase activity were observed in both steelhead and Chinook sample groups; however the differences were small and likely would not affect migration or survival.

Recommended citation for this report is: Nichols, K. 2015. Pathogen screening and gill Na+/K+- ATPase assessment of south Delta Chinook and steelhead 2015 release groups. U.S. Fish & Wildlife Service, California-Nevada Fish Health Center, Anderson, CA. Accessible at: http://www.fws.gov/canvfhc.

Notice: The mention of trade names or commercial products in this report does not constitute endorsement or recommendation for use by the Federal government. The findings and conclusions in this report are those of the author and do not necessarily represent the views of the US Fish and Wildlife Service.

184

Background As a component of the south Delta salmonid survival studies examining reach-specific survival and distribution of migrating juvenile Chinook salmon and steelhead trout in the San Joaquin River and Delta, the California-Nevada Fish Health Center conducted a general pathogen screening and smolt physiological assessment. Steelhead trout were sampled in support of the 6-year Study required by the 2009 Biological Opinion on the Central Valley Project and State Water Project operations (RPA IV.2.2). The health and smolt development of the study fish can help explain their survival and migration performance during the studies. Similar pathogen screening and physiological assessments have been conducted on Chinook used in various studies in the south Delta since 1996. In the majority of these past studies, juvenile Merced River Hatchery Chinook were utilized, and occasionally significant infections with the myxozoan parasite Tetracapsuloides bryosalmonae, the causative agent of Proliferative Kidney Disease (PKD), were observed (Foott, Stone and Nichols 2007; Foott 2012). In 2014, the source for the juvenile Chinook changed to Mokelumne River Hatchery due to health concerns. No significant health issues were observed in the Mokelumne River fish in 2014 (Nichols 2014). Steelhead trout from Mokelumne River Hatchery have been assessed for these studies since 2010 and no significant health issues have been observed in these fish to date. Methods

Study Fish Both Chinook salmon and steelhead trout were obtained from the California Department of Fish and Wildlife Mokelumne River Hatchery. Health assessment groups were cohorts of acoustic tagged release groups and shadowed their tagged cohorts through handling, tagging (dummy tagged), transport, and in-river holding.

Steelhead – Groups of 24 yearling steelhead were sampled on 7 March, 28 March and 25 April, 2015. These groups were held in the San Joaquin River at the Durham Ferry release site for 48 hours prior to sampling.

Chinook – Groups of 30 juvenile Chinook salmon were sampled on 19 April and 4 May, 2015. The 19 April group was held in the San Joaquin River at the Durham Ferry release site for 48 hours before sampling. The 4 May group was held at the Mokelumne River Hatchery instead of the river due to elevated water temperatures at the release site. The 4 May Chinook group was held for 72 hours instead of 48 hours prior to sampling due to a schedule conflict.

Sample Collection Fish were euthanized, fork length was recorded, any abnormalities were noted and tissues were sampled for lab assays. A sample of kidney tissue was aseptically collected and inoculated onto brain- heart infusion agar for bacterial culture (USFWS and AFS-FHS 2014). A kidney tissue imprint was collected to screen for Renibacterium salmoninarum (the bacteria that causes bacterial kidney disease) by fluorescent antibody test (USFWS and AFS-FHS 2014). Kidney and spleen tissue were collected in 3 fish pooled samples for viral tissue culture (USFWS and AFS-FHS 2014). Gill tissue was collected to assess smolt development by gill Na+/K+-Adenosine Triphosphatase (gill ATPase) assay (McCormick 1993). For Chinook, samples of gill, liver, kidney and intestine tissues were collected from 12 fish from each group for histopathological examination (Humason 1979). In steelhead, samples of gill tissue were collected from all live fish for histopathological examination.

185

Results and Discussion

Fish Condition Steelhead – A total of 72 steelhead were examined (Table C1) and only one fish died over the 48 hour holding period. The mortality occurred in the 28 March group in a fish with moderate hemorrhaging at the suture site (Figure C1A), and it was likely this fish died due to complications following tagging. Pale gills were observed on one fish from the 28 March group, and significant scale loss (>50% of body) was observed in two fish from the 25 April sample group. Overall condition of the steelhead groups appeared good with no evidence to suspect survival differences.

Table C1. Fork length (FL), mortality, external abnormalities (Ext Abn), or internal abnormalities (Int Abn) in Steelhead health assessment groups. Observed external abnormalities included pale gills and scale loss.

Sample Date FL ±SE (mm) Mortality Ext Abn Int Abn

7 March 255 ±5 0/24 0/24 0/24

28 March 239 ±3 1/24 (4%) 1/23 (4%) 0/23

25 April 252 ±3 0/24 2/24 (8%) 0/24

Chinook – A total of 60 Chinook salmon were examined (Table C2) and there was no mortality in the health assessment groups over the holding period. Minor hemorrhaging (Figure C1B) or slightly pale gills (Figure C1C) were observed in 3 fish from the 19 April sample group. In the 4 May group, one fish had a small amount the liver extruding between the sutures with no hemorrhaging, and two other fish had minor hemorrhaging at the suture site. No evidence to suspect survival differences between Chinook groups due to fish condition was observed.

Table C2. Fork Length (FL), mortality, external abnormalities (Ext Abn), and internal abnormalities (Int Abn) in Chinook salmon health assessment groups. Observed external abnormalities included: pale gills, minor hemorrhaging and partly open sutures.

Sample Date FL ±SE (mm) Mortality Ext Abn Int Abn

19 April 102 ±0.9 0/30 3/30 (10%) 0/30

4 May 104 ±1.0 0/30 3/30 (10%) 0/30

Pathogen Screening Steelhead – No obligate bacterial or viral pathogens were detected in the 71 trout sampled. Other bacteria isolates (presumptive environmental contaminates due to field sampling conditions) were observed in 11% (8/71) of fish sampled. Minor parasitic infections were observed in 29% (20/70) of gill tissues examined by histopathology, with no associated lesion or other signs of impairment. Gill infections included: Capriniana piscium (presumptive ID, formerly known as Trichophrya) 23% (16/70); cyst-like xenoma due to an unidentified microsporidian 4% (3/70); and an infection of Ichthyophthirius 186

multifiliis 1% (1/70). None of these infections were likely to cause differences in survival between steelhead release groups.

Chinook – No obligate bacterial or viral pathogens were detected in 60 salmon sampled. Other bacterial isolates (presumptive environmental contaminates) were observed in 20% (12/60) of samples. No infections or signs of impairment were observed by histopathology of gill, liver, kidney or intestine tissues from 25 Chinook. No differences in survival due to infections would be expected in the Chinook groups,

A B C

Figure C1. Examples of external abnormalities including: (A) Moderate hemorrhaging at suture site of a steelhead; (B) minor hemorrhaging (lower) and normal (upper) sutures in Chinook; and (C) slightly pale gills in Chinook.

Gill Na+/K+-ATPase Activity Steelhead – Gill ATPase activity levels (μmol ADP*mg protein-1*hr-1) ranged from 0.3 to 6.9 in all groups (Figure C2). Median activity levels in the 7 March group were significantly lower than the later groups (Kruskal-Wallis test, P<0.001). Higher gill ATPase activity levels are associated with migrating smolts relative to their residual cohorts (Hanson et. al. 2011). In our experience, steelhead ATPase activity levels may increase over a short time period. While high levels may indicate salt-water readiness, lower values may have little biological significance due to this ability to rapidly change. Overall, the gill ATPase activity levels were low in all groups and the would not point to differences in migration or survival of the steelhead release groups.

Chinook – Gill ATPase activity levels ranged from 4.5 to 16.8 in all groups (Figure C3). Median activity levels in the 19 April group were higher than in the 4 May group. The modifications in holding conditions of the 4 May Chinook health assessment groups (mentioned above) and variability due to lab assay conditions likely contributed to the observed difference, and a direct comparison may not be useful. The majority of fish from both Chinook health assessment groups had activity levels consistent with smolts

187

(>6.7, CA-NV Fish Health Center unpublished data). Active migration in both release groups would be consistent with the observed activity levels.

b b

a

Figure C2. Boxplot of median gill ATPase activity in 2015 steelhead health assessment groups. Medians with the same letter are not significantly different (Kruskal-Wallis, P<0.001). Number of fish sampled for each group was 16.

a

b

Figure C3. Boxplot of median gill ATPase activity in 2015 Chinook health assessment groups. Medians with the same letter are not significantly different (Kruskal-Wallis, P=0.005). Number of fish sampled from each group was 16.

188

References Foott JS. 2012. FY2012 Technical Report: Pathogen screening and gill Na-K-ATPase assessment of juvenile Chinook salmon used in south delta acoustic tag studies. U.S. Fish & Wildlife Service California- Nevada Fish Health Center, Anderson, CA. Accessible at: http://www.fws.gov/canvfhc.

Foott JS, R Stone and K Nichols. 2007. Proliferative Kidney Disease (Tetracapsuloides bryosalmonae) in Merced River Hatchery juvenile Chinook salmon: mortality and performance impairment in 2005 smolts. California Fish and Game 93: 57-76.

Hanson KC, WL Gale, WG Simpson, BM Kennedy and KG Ostrand. 2011. Physiological characterization of hatchery-origin juvenile steelhead Oncorhynchus mykiss adopting divergent life-history strategies. Journal of Fish and Wildlife Management 2(1): 61-71.

Humason G L. 1979. Animal Tissue Techniques, 4th edition. W H Freeman and Co., San Francisco.

McCormick S D. 1993. Methods for Nonlethal Gill Biopsy and Measurement of Na+, K+- ATPase Activity. Canadian Journal of Fisheries and Aquatic Sciences. 50: 656-658.

Nichols, K 2014. Pathogen Screening and Gill Na+/K+- ATPase Assessment of South Delta Chinook and Steelhead 2014 Release Groups. U.S. Fish & Wildlife Service, California-Nevada Fish Health Center, Anderson, CA. Accessible at: http://www.fws.gov/canvfhc.

USFWS and AFS-FHS (U.S. Fish and Wildlife Service and American Fisheries Society-Fish Health Section). 2014. Standard procedures for aquatic animal health inspections. In AFS-FHS. FHS blue book: suggested procedures for the detection and identification of certain finfish and shellfish pathogens, 2014 edition. Accessible at: http://afsfhs.org/bluebook/b

189

Appendix D. Water temperatures (0C) in transport tanks during transport from Mokelumne River Hatchery to the holding location at Durham Ferry or release site at Medford Island for Chinook salmon during the spring of 2015.

Chinook Transport Tank Water Temperature (Transport 1) 4/14/2015

Tank 1 Tank 2 16

C ° 15

14 Temperature Temperature 13 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 Transport Time in Minutes

Chinook Transport Tank Water Temperature (Transport 2) 4/14/2015 Tank 3

16 C ° 15

14

Temperature Temperature 13 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 Transport Time in Minutes

190

Chinook Transport Tank Water Temperature (Transport 1) 4/15/2015

Tank 1 Tank 2 16

C ° 15

14 Temperature Temperature 13 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 Transport Time in Minutes

Chinook Transport Tank Water Temperature (Transport 2) 4/15/2015 Tank 3 16

C ° 15

14 Temperature Temperature 13 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 Transport Time in Minutes

191

Chinook Transport Tank Water Temperature (Transport 1) 4/16/2015

Tank 1 Tank 2

16 C ° 15

14 Temperature Temperature 13 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 Transport Time in Minutes

Chinook Transport Tank Water Temperature (Transport 2) 4/16/2015

Tank 3 16

C ° 15

14 Temperature Temperature 13 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 Transport Time in Minutes

192

Chinook Transport Tank Water Temperature (Transport 1) 4/17/2015

Tank 1 Tank 2

16 C ° 15

14 Temperature Temperature 13 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 Transport Time in Minutes

Chinook Transport Tank Water Temperature (Transport 2) 4/17/2015 Tank 3

16 C ° 15

14

Temperature Temperature 13 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 Transport Time in Minutes

193

Chinook Transport Tank Water Temperature (Transport 1) 4/28/2015

Tank 1 Tank 2

19 C ° 17

15

Temperature Temperature 13 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 Transport Time in Minutes

Chinook Transport Tank Water Temperature (Transport 2) 4/28/2015 Tank 3

19 C ° 17

15

13 Temperature Temperature 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 Transport Time in Minutes

194

Chinook Transport Tank Water Temperature (Transport 1) 4/30/2015

Tank 1 Tank 2

19 C ° 17

15

Temperature Temperature 13 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 Transport Time in Minutes

Chinook Transport Tank Water Temperature (Transport 2) 4/30/2015

Tank 3

19 C ° 17

15

Temperature Temperature 13 1 6 11 16 21 26 31 36 41 46 51 56 61 Transport Time in Minutes

195

Chinook Transport Tank Water Temperature (Transport 1) 5/1/2015 Tank 1 Tank 2

19 C ° 17

15

Temperature Temperature 13 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 Transport Time in Minutes

Chinook Transport Tank Water Temperature (Transport 2) 5/1/2015

Tank 3

19 C ° 17

15

Temperature Temperature 13 1 6 11 16 21 26 31 36 41 46 51 56 61 Transport Time in Minutes

196

Chinook Transport Tank Water Temperature (Transport 1) 5/2/2015 Tank 1 Tank 2

19 C ° 17

15

Temperature Temperature 13 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 Transport Time in Minutes

Chinook Transport Tank Water Temperature (Transport 2) 5/2/2015 Tank 3

19 C ° 17

15

13 Temperature Temperature 1 6 11 16 21 26 31 36 41 46 51 56 61 Transport Time in Minutes

197

Appendix E. Survival Model Parameters

Table E1. Definitions of parameters used in the release-recapture survival model. DF = Durham Ferry model, MF = Medford Island model. Parameters used only in the full model(s) but not in the implemented model(s) are noted.

Parameter Definition

SA2 Probability of survival from Durham Ferry Downstream (DFD) to Below Durham Ferry 1 (BDF1) (DF)

SA3 Probability of survival from Below Durham Ferry 1 (BDF1) to Below Durham Ferry 2 (BDF2) (DF)

SA4 Probability of survival from Below Durham Ferry 2 (BDF2) to Banta Carbona (BCA) (DF)

SA5 Probability of survival from Banta Carbona (BCA) to Mossdale (MOS) (DF)

SA6 Probability of survival from Mossdale (MOS) to Head of Old River (HOR) (DF)

SB0 Probability of survival from Head of Old River (HOR) to Lathrop (SJL) or Old River East (ORE) (DF)

SA6,AB Probability of survival from Mossdale (MOS) to Lathrop (SJL) or Old River East (ORE) (DF); = SA6*SB0

SA7 Probability of survival from Lathrop (SJL) to Garwood Bridge (SJG) (DF) Probability of survival from Garwood Bridge (SJG) to Navy Drive Bridge (SJNB) or Burns Cutoff (RRI) receivers

SA8 (DF) Probability of survival from Garwood Bridge (SJG) to MacDonald Island (MAC) or Turner Cut (TCE/TCW) (i.e.,

SA8(SD) through the South Delta region) (DF)

SA8,G2 Overall survival from Garwood Bridge (SJG) to Chipps Island (MAT/MAE/MAW) (DF)

SA9 Probability of survival from Navy Drive Bridge (SJNB) to San Joaquin Shipping Channel (SJS); full DF model only

SA9,AF Probability of survival from Navy Drive Bridge (SJNB) to MacDonald Island (MAC) or Turner Cut (TCE/TCW) (DF)

SA9,G2 Overall survival from Navy Drive Bridge (SJNB) to Chipps Island (MAT/MAE/MAW)

SR1 Probability of survival from Burns Cutoff (RRI) to San Joaquin Shipping Channel (SJS); full DF model only Probability of survival from Burns Cutoff (RRI) to MacDonald Island (MAC) or Turner Cut (TCE/TCW); full DF

SR1,AF model only

SR1,G2 Overall survival from Burns Cutoff (RRI) to Chipps Island (MAT/MAE/MAW); full DF model only Probability of survival from San Joaquin Shipping Channel (SJS) to MacDonald Island (MAC) or Turner Cut

SA10 (TCE/TCW); full DF model only

SB1 Probability of survival from Old River East (ORE) to Chipps Island (MAT/MAE/MAW) (DF)

SB1,G2 Probability of survival from Old River East (ORE) to Chipps Island (MAT/MAE/MAW) (DF); = SB1

φA1,A0 Joint probability of moving from Durham Ferry release site upstream toward DFU, and surviving to DFU (DF)

φA1,A2 Joint probability of moving from Durham Ferry release site downstream toward DFD, and surviving to DFD (DF) Joint probability of moving from MacDonald Island (MAC) toward Jersey Point (JPT/JPE/JPW) and False River

φA11,GH (FRE/FRW), and surviving to JPT/JPE/JPW or FRE/FRW; full DF model only Joint probability of moving from MacDonald Island (MAC) toward Jersey Point (JPT/JPE/JPW), and surviving to

φA11,G1 JPT/JPE/JPW (DF) Joint probability of moving from MacDonald Island (MAC) toward Chipps Island (MAT/MAE/MAW), and

φA11,G2 surviving to MAT/MAE/MAW (MF) Joint probability of moving from Medford Island (MFE/MFW) toward the San Joaquin River near

φA12,A13 Disappointment Slough receivers (SJD), and surviving to SJD; full MF model only Joint probability of moving from Medford Island (MFE/MFW) toward Old River at San Joaquin River (OSJ), and

φA12,B5 surviving to OSJ; full MF model only Joint probability of moving from Medford Island (MFE/MFW) toward Middle River at Middle River (MID), and

φA12,C3 surviving to MID; full MF model only

198

Table E1. (Continued) Parameter Definition Joint probability of moving from the San Joaquin River near Disappointment Slough receivers (SJD) toward Jersey Point (JPT/JPE/JPW) and False River (FRE/FRW), and surviving to JPT/JPE/JPW or FRE/FRW; full MF model

φA13,GH only Joint probability of moving from the San Joaquin River near Disappointment Slough receivers (SJD) toward

φA13,G1 Jersey Point (JPT/JPE/JPW), and surviving to JPT/JPE/JPW; full MF model only Joint probability of moving from Old River at San Joaquin River (OSJ) toward Jersey Point (JPT/JPE/JPW) and

φB5,GH False River (FRE/FRW), and surviving to JPT/JPE/JPW or FRE/FRW; full MF model only Joint probability of moving from Old River at San Joaquin River (OSJ) toward Jersey Point (JPT/JPE/JPW), and

φB5,G1 surviving to JPT/JPE/JPW; full MF model only Joint probability of moving from Middle River at Middle River (MID) toward Chipps Island (MAT/MAE/MAW),

φC3,G2 and surviving to MAT/MAE/MAW; full MF model only Joint probability of moving from Turner Cut (TCE/TCW) toward Jersey Point (JPT/JPE/JPW) and False River

φF1,GH (FRE/FRW), and surviving to JPT/JPE/JPW or FRE/FRW; full DF model only Joint probability of moving from Turner Cut (TCE/TCW) toward Jersey Point (JPT/JPE/JPW), and surviving to

φF1,G1 JPT/JPE/JPW (DF) Joint probability of moving from Columbia Cut (COL) toward Middle River at Middle River (MID), and surviving

φF2,C3 to MID; full MF model only Joint probability of moving from Columbia Cut (COL) toward Chipps Island (MAT/MAE/MAW) and surviving to

φF2,G2 MAT/MAE/MAW (MF) Joint probability of moving from Jersey Point (JPT/JPE/JPW) toward Chipps Island (MAT/MAE/MAW) and

φG1,G2 surviving to MAT/MAE/MAW (MF and full DF model) Joint probability of moving from Medford Island release site toward MacDonald Island (MAC), and surviving to

φMF,A11 MAC (MF) Joint probability of moving from Medford Island release site toward Medford Island receivers (MFE/MFW), and

φMF,A12 surviving to MFE/MFW; full MF model only Joint probability of moving from Medford Island release site toward San Joaquin River receivers near

φMF,A13 Disappointment Slough (SJD), and surviving to SJD (MF) Joint probability of moving from Medford Island release site toward Old River at San Joaquin River (OSJ), and

φMF,B5 surviving to OSJ (MF) Joint probability of moving from Medford Island release site toward Columbia Cut (COL), and surviving to COL

φMF,F2 (MF) ψ ψA1 Probability of remaining in the San Joaquin River at the head of Old River; = 1 - B1 (DF)

ψA2 Probability of remaining in the San Joaquin River at Burns Cutoff; 1 - ψR2; full DF model only

ψA3 Probability of remaining in the San Joaquin River at Turner Cut; 1 - ψF3; full DF model only

ψB1 Probability of entering Old River at the head of Old River; = 1 - ψA1 (DF)

Probability of entering Burns Cutoff at its upstream junction with the San Joaquin River; = 1 - ψA2; full DF model

ψR2 only

ψF3 Probability of entering Turner Cut at its junction with the San Joaquin River; = 1 - ψA3; full DF model only

Probability of moving downriver in the San Joaquin River at the Jersey Point/False River junction; = 1 - ψH1; full

ψG1 DF and MF models only

ψH1 Probability of entering False River at the Jersey Point/False River junction; = 1 - ψG1; full DF and MF model only

PA0a Conditional probability of detection at DFU1 (DF)

PA0b Conditional probability of detection at DFU2 (DF)

PA0 Conditional probability of detection at DFU (either DFU1 or DFU2) (DF)

PA2a Conditional probability of detection at DFD1 (DF)

PA2b Conditional probability of detection at DFD2 (DF) 199

Table E1. (Continued) Parameter Definition

PA2 Conditional probability of detection at DFD (either DFD1 or DFD2) (DF)

PA3 Conditional probability of detection at BDF1 (DF)

PA4 Conditional probability of detection at BDF2 (DF)

PA5 Conditional probability of detection at BCA (DF)

PA6 Conditional probability of detection at MOS (DF)

PA7a Conditional probability of detection at SJLU (DF)

PA7b Conditional probability of detection at SJLD (DF)

PA7 Conditional probability of detection at SJL (either SJLU or SJLD) (DF)

PA8a Conditional probability of detection at SJGU (DF)

PA8b Conditional probability of detection at SJGD (DF)

PA8 Conditional probability of detection at SJG (either SJGU or SJGD) (DF)

PA9a Conditional probability of detection at SJNBU (DF)

PA9b Conditional probability of detection at SJNBD (DF)

PA9 Conditional probability of detection at SJNB (either SJNBU or SJNBD) (DF)

PA10 Conditional probability of detection at SJS; full DF model only

PA11a Conditional probability of detection at MACU (DF, MF)

PA11b Conditional probability of detection at MACD (DF, MF)

PA11 Conditional probability of detection at MAC (either MACU or MACD) (DF, MF)

PA12a Conditional probability of detection at MFE; full MF model only

PA12b Conditional probability of detection at MFW; full MF model only

PA12 Conditional probability of detection at MFE/MFW (either MFE or MFW); full MF model only

PA13a Conditional probability of detection at SJDU; full MF model only

PA13b Conditional probability of detection at SJDD; full MF model only

PA13 Conditional probability of detection at SJD (either SJDU or SJDD) (MF)

PB0 Conditional probability of detection at HOR (DF)

PB1a Conditional probability of detection at OREU (DF)

PB1b Conditional probability of detection at ORED (DF)

PB1 Conditional probability of detection at ORE (either OREU or ORED) (DF)

PB5a Conditional probability of detection at OSJU (MF)

PB5b Conditional probability of detection at OSJD (MF)

PB5 Conditional probability of detection at OSJ (either OSJU or OSJD) (MF)

PC3 Conditional probability of detection at MID; full MF model only

PF1a Conditional probability of detection at TCE; full DF model only

PF1b Conditional probability of detection at TCW; full DF model only

PF1 Conditional probability of detection at TCE/TCW (either TCE or TCW); full DF model only

PF2a Conditional probability of detection at COLU (MF)

PF2b Conditional probability of detection at COLD (MF)

PF2 Conditional probability of detection at COL (either COLU or COLD) (MF)

PR1a Conditional probability of detection at RRIU; full DF model only

PR1b Conditional probability of detection at RRID; full DF model only 200

Table E1. (Continued)

Parameter Definition

PR1 Conditional probability of detection at RRI (either RRIU or RRID); full DF model only

PG1a Conditional probability of detection at JPT; full DF, MF models only

PG1b Conditional probability of detection at JPE; full DF, MF models only

PG1c Conditional probability of detection at JPW; full DF, MF models only

PG1 Conditional probability of detection at JPT/JPE/JPW (DF, MF)

PG2a Conditional probability of detection at MAT; full DF, MF models only

PG2b Conditional probability of detection at MAE; full DF, MF models only

PG2c Conditional probability of detection at MAW; full DF, MF models only

PG2 Conditional probability of detection at MAT/MAE/MAW (DF, MF)

PH1a Conditional probability of detection at FRW; full DF, MF models only

PH1b Conditional probability of detection at FRE; full DF, MF models only

PH1 Conditional probability of detection at FRE/FRW; full DF, MF models only Joint probability of surviving from Chipps Island to Benicia Bridge and detection at Benicia Bridge (MF, full DF λ model only)

λA6 Joint probability of surviving from MOS to Benicia Bridge and detection at Benicia Bridge; = STotalλ (DF)

λA8 Joint probability of surviving from SJG to Benicia Bridge and detection at Benicia Bridge; = SA8,G2λ (DF)

λA9 Joint probability of surviving from SJNB to Benicia Bridge and detection at Benicia Bridge; = SA9,G2λ (DF)

λB1 Joint probability of surviving from ORE to Benicia Bridge and detection at Benicia Bridge; = SB1,G2λ (DF)

λA13 Joint probability of surviving from SJD to Benicia Bridge and detection at Benicia Bridge; = SA13,G2λ (MF)

λB5 Joint probability of surviving from OSJ to Benicia Bridge and detection at Benicia Bridge; = SB5,G2λ (MF) Joint probability of surviving from Medford Island release site to Benicia Bridge and detection at Benicia Bridge;

λMF = SMF,G2λ (MF)

201

Table E2. Parameter estimates (standard errors in parentheses) from survival model for tagged juvenile Chinook salmon released in 2015, excluding predator-type detections. Parameters without standard errors were estimated at fixed values in the model. Population-level estimates are from pooled data from both release groups. Some parameters were not estimable because of sparse data. Release Group Parameter 1 2 Population Estimate

SA2 0.44 (0.05) 0.43 (0.05)

SA3 0.79 (0.08) 0.80 (0.07)

SA4 0.54 (0.05) 0.53 (0.05)

SA5 0.28 (0.05) 0.28 (0.05)

SA6 0.81 (0.09) 0.81 (0.09)

SB0 0.71 (0.11) 0.71 (0.11)

SA6,AB 0.57 (0.11) 0.57 (0.11)

SA7 0.27 (0.13) 0.27 (0.13)

SA8 0.67 (0.27) 0.67 (0.27) ab b SA8(SD) 0.33 (0.27) 0.33 (0.27) a SA8,G2 0 0

SA9

SA9,AF

SA9,G2

SR1

SR1,AF

SR1,G2

SA10

SB1

SB1,G2

φA1,A0 0.03 (0.01) 0.12 (0.03) 0.04 (0.01)

φA1,A2 0.61 (0.03) 0.04 (0.02) 0.50 (0.03)

φA11,GH

φA11,G1

φA11,G2 0 0

φA12,A13

φA12,B5

φA12,C3

φA13,GH

φA13,G1 0.55 (0.06) 0.55 (0.06)

φB5,GH

φB5,G1 0.42 (0.23) 0.42 (0.23)

φC3,G2

φF1,GH

φF1,G1

φF2,C3 a = assumed 100% detection probability at downstream receiver site(s) b = assumed 100% detection probability at Turner Cut

202

Table E2. (Continued) Release Group Parameter 1 2 Population Estimate

φF2,G2 0 0

φG1,G2 0.77 (0.07) 0.77 (0.07)

φMF,A11 0.05 (0.01) 0.05 (0.01)

φMF,A12

φMF,A13 0.17 (0.02) 0.17 (0.02)

φMF,B5 0.01 (<0.01) 0.01 (<0.01)

φMF,F2 0.02 (0.01) 0.02 (0.01)

ψA1 0.92 (0.08) 0.92 (0.08)

ψA2

ψA3

ψB1 0.08 (0.08) 0.08 (0.08)

ψR2

ψF3

ψG1

ψH1

PA0a 0.73 (0.13) 1 0.90 (0.06)

PA0b 0.67 (0.14) 1 0.87 (0.06)

PA0 0.91 (0.07) 1 0.99 (0.01)

PA2a 1

PA2b 1

PA2 0.70 (0.04) 1 0.70 (0.04)

PA3 0.24 (0.04) 0.25 (0.04)

PA4 0.85 (0.04) 0.85 (0.04)

PA5 1 1

PA6 1 1

PA7a 1 1

PA7b 0.91 (0.09) 0.91 (0.09)

PA7 1 1

PA8a 1 1

PA8b 1 1

PA8 1 1

PA9a 1 1

PA9b 1 1

PA9 1 1

PA10 c PA11a 1 0.83 (0.08) 0.84 (0.07) c PA11b 1 1 1 c PA11 1 1 1

PA12a

PA12b c = assumed; could not be estimated due to sparse data 203

Table E2. (Continued) Release Group Parameter 1 2 Population Estimate

PA12

PA13a

PA13b

PA13 0.91 (0.04) 0.91 (0.04)

PB0 1 1 c c PB1a 1 1 c c PB1b 1 1 c c PB1 1 1

PB5a 0.80 (0.18) 0.80 (0.18)

PB5b 1 1

PB5 1 1

PC3 c c PF1a 1 1 c c PF1b 1 1 c c PF1 1 1

PF2a 0.89 (0.10) 0.89 (0.10)

PF2b 1 1

PF2 1 1

PR1a

PR1b

PR1

PG1a

PG1b

PG1c c PG1 1 0.84 (0.06) 0.84 (0.06)

PG2a

PG2b

PG2c c PG2 1 0.93 (0.05) 0.93 (0.05)

PH1a

PH1b

PH1 λ 0.74 (0.07) 0.74 (0.07)

λA6 0 0

λA8 0 0

λA9

λB1

λA13 0.31 (0.05) 0.31 (0.05)

λB5 0.24 (0.13) 0.24 (0.13)

λMF 0.06 (0.01) 0.06 (0.01) c = assumed; could not be estimated due to sparse data 204

Table E3. Parameter estimates (standard errors in parentheses) from survival model for tagged juvenile Chinook salmon released in 2015, including predator-type detections. Parameters without standard errors were estimated at fixed values in the model. Population-level estimates are from pooled data from both release groups. Some parameters were not estimable because of sparse data. Release Group Parameter 1 2 Population Estimate

SA2 0.45 (0.05) 0.43 (0.05)

SA3 0.76 (0.08) 0.77 (0.08)

SA4 0.54 (0.05) 0.53 (0.05)

SA5 0.28 (0.05) 0.28 (0.05)

SA6 0.86 (0.08) 0.86 (0.08)

SB0 0.72 (0.11) 0.72 (0.11)

SA6,AB 0.62 (0.11) 0.62 (0.11)

SA7 0.25 (0.13) 0.25 (0.13)

SA8 0.67 (0.27) 0.67 (0.27) ab b SA8(SD) 0.33 (0.27) 0.33 (0.27) a SA8,G2 0 0

SA9

SA9,AF

SA9,G2

SR1

SR1,AF

SR1,G2

SA10

SB1

SB1,G2

φA1,A0 0.06 (0.01) 0.18 (0.03) 0.08 (0.01)

φA1,A2 0.63 (0.03) 0.08 (0.02) 0.53 (0.03)

φA11,GH

φA11,G1

φA11,G2 0 0

φA12,A13

φA12,B5

φA12,C3

φA13,GH

φA13,G1 0.57 (0.06) 0.57 (0.06)

φB5,GH

φB5,G1 0.42 (0.23) 0.42 (0.23)

φC3,G2

φF1,GH

φF1,G1

φF2,C3 a = assumed 100% detection probability at downstream receiver site(s) b = assumed 100% detection probability at Turner Cut

205

Table E3. (Continued) Release Group Parameter 1 2 Population Estimate

φF2,G2 0 0

φG1,G2 0.74 (0.07) 0.74 (0.07)

φMF,A11 0.05 (0.01) 0.05 (0.01)

φMF,A12

φMF,A13 0.18 (0.02) 0.18 (0.02)

φMF,B5 0.01 (<0.01) 0.01 (<0.01)

φMF,F2 0.02 (0.01) 0.02 (0.01)

ψA1 0.92 (0.07) 0.92 (0.07)

ψA2

ψA3

ψB1 0.08 (0.07) 0.08 (0.07)

ψR2

ψF3

ψG1

ψH1

PA0a 0.75 (0.07) 1 0.86 (0.04)

PA0b 0.90 (0.05) 1 0.95 (0.03)

PA0 0.98 (0.02) 1 0.99 (<0.01)

PA2a 1

PA2b 1

PA2 0.69 (0.04) 1 0.69 (0.04)

PA3 0.24 (0.04) 0.25 (0.04)

PA4 0.85 (0.04) 0.85 (0.04)

PA5 1 1

PA6 1 1

PA7a 1 1

PA7b 0.92 (0.08) 0.92 (0.08)

PA7 1 1

PA8a 1 1

PA8b 1 1

PA8 1 1

PA9a 1 1

PA9b 1 1

PA9 1 1

PA10 c PA11a 1 0.77 (0.09) 0.78 (0.09) c PA11b 1 1 1 c PA11 1 1 1

PA12a

PA12b c = assumed; could not be estimated due to sparse data 206

Table E3. (Continued) Release Group Parameter 1 2 Population Estimate

PA12

PA13a

PA13b

PA13 0.91 (0.04) 0.91 (0.04)

PB0 1 1 c c PB1a 1 1 c c PB1b 1 1 c c PB1 1 1

PB5a 0.80 (0.18) 0.80 (0.18)

PB5b 1 1

PB5 1 1

PC3 c c PF1a 1 1 c c PF1b 1 1 c c PF1 1 1

PF2a 0.90 (0.09) 0.90 (0.09)

PF2b 1 1

PF2 1 1

PR1a

PR1b

PR1

PG1a

PG1b

PG1c c PG1 1 0.84 (0.06) 0.84 (0.06)

PG2a

PG2b

PG2c c PG2 1 0.93 (0.05) 0.93 (0.05)

PH1a

PH1b

PH1 λ 0.74 (0.07) 0.74 (0.07)

λA6 0 0

λA8 0 0

λA9

λB1

λA13 0.31 (0.05) 0.31 (0.05)

λB5 0.23 (0.13) 0.23 (0.13)

λMF 0.06 (0.01) 0.06 (0.01) c = assumed; could not be estimated due to sparse data 207

208