SMOLT MONITORING IN THE LAGUNITAS CREEK WATERSHED 2017

SMOLT MONITORING IN THE LAGUNITAS CREEK WATERSHED 2017 Eric Ettlinger, Aquatic Ecologist Marin Municipal Water District, 220 Nellen Avenue, Corte Madera, CA 94925 (415) 945-1193 In association with the National Park Service and the Salmon Protection and Watershed Network (SPAWN) August 2017

Acknowledgements The Marin Municipal Water District (MMWD) would like to thank the Gallagher family for granting us access onto their property to conduct this monitoring. Cover image: Transitional steelhead smolt and PIT tag reader (photo by Brenna Fowler).

Table of Contents EXECUTIVE SUMMARY... 1 INTRODUCTION... 2 METHODS... 3 Lagunitas Creek Monitoring... 3 Olema Creek Monitoring... 4 San Geronimo Creek Monitoring... 5 Data Analysis... 5 RESULTS... 6 Lagunitas Creek Rotary Screw Trap... 6 PIT Tagged Coho... 9 Olema Creek Monitoring... 9 San Geronimo Creek Monitoring... 9 DISCUSSION... 9 Sampling conditions and effects on estimated smolt abundance and emigration timing... 9 Age and size composition of salmonids... 10 Smolt abundance trends and implications for winter survival... 11 REFERENCES... 12 List of Tables Table 1. Marking schedule at Lagunitas and San Geronimo Creek smolt traps... 4 Table 2. Estimated smolt emigration from Lagunitas Creek, 2006-2017... 6 Table 3. Salmonids captured in the Lagunitas Creek rotary screw trap by length and week... 7 Table 4. Non-salmonid fish captured in the rotary screw trap... 8 Table 5. PIT-tagged coho detections... 9 List of Figures Figure 1. Trap locations... 13 Figure 2. Lagunitas Creek smolt emigration and stream flow.... 14 Figure 3. Weekly trap efficiency and Lagunitas Creek flow... 15 Figure 4. Daily coho smolt captures at the Lagunitas Creek smolt trap, 2014-2017... 16 Figure 5. January water temperatures and coho smolt emigration timing... 17 Figure 6. Lagunitas Creek smolt estimates... 18 Figure 7. Coho emigration estimates and smolt lengths at the Lagunitas smolt trap... 19 i

EXECUTIVE SUMMARY Downstream migrating salmonid smolts were sampled using a rotary screw trap (RST) in lower Lagunitas Creek, near Point Reyes Station (Figure 1). This represents the 12 th consecutive year of smolt monitoring at that location. From mid-march through late May the trap was monitored cooperatively by staff and volunteers from the Marin Municipal Water District (MMWD) and the Watershed Stewards Program (WSP). A passive integrated transponder (PIT) tag antenna was also in operation at that location to detect PIT-tagged coho smolts. In addition, smolt traps were operated on Olema Creek by National Park Service (NPS) staff and on San Geronimo Creek by staff and volunteers of the Salmon Protection and Watershed Network (SPAWN). The winter and spring of 2016-17 was exceptionally wet and smolt monitoring was hampered by high stream flows from mid-march through mid-april. During the first 34 days of the survey the smolt trap was disabled for 17 days and only operating within target parameters for five days. As a result, the trap likely caught only six percent of migrating coho and 12% of steelhead smolts during this period, significantly increasing the uncertainty around smolt emigration estimates. A record 29,306 (±11,286) coho smolts emigrated past the Lagunitas Creek RST in 2017, which followed a record juvenile coho population estimate in 2016. The steelhead emigration was average, with 3,167 (±1,313) smolts migrating past the RST. The peak of the emigration for both species occurred during the week of April 10. The Chinook salmon emigration was the thirdhighest observed, at 2,224 (±425) smolts, and peaked during the second week of May. 1

INTRODUCTION Lagunitas Creek is a regionally important coastal stream for coho salmon (Oncorhychus kisutch) and steelhead trout (O. mykiss), although run sizes for both species are significantly reduced from historical numbers. Recent coho spawner estimates have averaged approximately 400 individuals, and available data suggest that steelhead runs are similar in size. Chinook salmon (O. tshawytscha) also spawn periodically in Lagunitas Creek and smolts have been observed in six of the last 12 years. MMWD has conducted annual smolt surveys on Lagunitas Creek since 2006, as well as in 1983, 1984 and 1985. Summer and fall electrofishing surveys for juvenile coho salmon and steelhead trout were conducted starting in Lagunitas Creek in 1970 and annually since 1993. Since 2012 coho fry captured during these surveys have been implanted with passive integrated transponder (PIT) tags. This represents one of the longest data records for juvenile salmonids in coastal streams of California. Surveys have been conducted cooperatively between MMWD, the California Department of Fish and Wildlife (CDFW), the National Park Service, the Marin Resource Conservation District, the Watershed Stewards Program, and the Salmon Protection and Watershed Network. Systematic coho adult spawner surveys began during the 1982-83 and 1983-84 spawning seasons, and have been conducted annually since 1995-96. Since the early 1980s, stream flows in Lagunitas Creek have been monitored daily by United States Geological Survey gages located in Samuel P. Taylor State Park and near Point Reyes Station. A separate gage is maintained by MMWD on San Geronimo Creek. Water temperature monitoring has been performed by MMWD since the early 1990s. Lagunitas Creek streambed conditions are monitored annually and salmonid habitat is quantified approximately every five years. This project is being conducted in collaboration with NPS and SPAWN, which are conducting similar monitoring surveys in Olema and San Geronimo Creeks, respectively. NPS has monitored salmonid smolt emigration from Olema Creek since 2004, and smolt monitoring was conducted on a tributary to Olema Creek between 1998 and 2004. Smolt monitoring on San Geronimo Creek has been conducted annually since 2006. Smolts leaving San Geronimo Creek, a tributary to Lagunitas Creek, are sampled again at the Lagunitas Creek RST, so the Lagunitas Creek smolt estimates include San Geronimo Creek smolts. Smolt monitoring in the Lagunitas Creek watershed is intended to answer the following questions: What are the trends in coho salmon and steelhead smolt abundance? What are salmonid overwinter survival rates, what factors influence those rates, and do those rates differ between subwatersheds? What are coho marine survival rates and how do these rates compare to other populations in the region? 2

METHODS Lagunitas Creek Monitoring A rotary screw trap (RST) with a five-foot diameter cone was installed on March 14, 2017 in lower Lagunitas Creek, approximately 2.1 miles above the Highway 1 Bridge in Point Reyes Station. The trap was situated in a pool directly downstream of a small bedrock cascade, and was in the same location as has been used since 2006. The bedrock cascade concentrates enough flow to operate the RST in the otherwise low gradient reach of the creek. The trap was operated through May 26, but was disabled or non-functional due to high stream flows for 14 days, non-operational as planned for an additional 14 days, and operating with the cone rotation speed outside the target range for nine days. For the entire 73-day season the trap was operated within target parameters for only 36 days. Trap function was visually inspected each day and the rotation speed of the trap cone was recorded daily. The trap was frequently moved toward or away from the cascade and/or laterally within the creek to maintain cone speeds of between three and eight revolutions per minute (RPM). Keeping the cone at or below eight RPM was difficult in the high stream flows until a breaking mechanism was installed in mid-april. Debris was removed daily from the live box and nearly filled the box on one occasion. A ½ -mesh screen in the live box allowed the smallest fry to escape from larger fish. Small gaps in the live box were created behind this screen to allow fry to escape, although some chose not to. Plywood baffles were installed in front of the cone on May 5 to increase the cone rotation speed and improve trap efficiency. Each day, captured fish were removed from the trap and identified by species. Salmonid smolts and parr were checked for marks such as fin clips, visually inspected for signs of smoltification, measured, weighed, allowed to recover, and then released downstream of the point of capture. Coho smolts were scanned for passive integrated transponder (PIT) tags, implanted the previous fall. Steelhead displaying characteristics intermediate between smolts and fry/parr (e.g. some loss of scales, some silver color, fading parr marks, etc.) were classified as transitional. These transitional steelhead could not be assumed to be emigrating and were not included in the smolt estimate. Coho were classified as fry, transitionals, or smolts based on the degree of smolt characteristic development. Young-of-the-year coho displaying smolt characteristics (e.g., silvery appearance) were classified as smolts. All Chinook were assumed to be emigrating and classified as smolts. Downstream migrating fry of all species that were less than 70 mm long were tallied into five-millimeter length bins and not weighed. Adult steelhead that appeared unspawned were released upstream of the bedrock cascade. Spawned steelhead (kelts) were immediately released off the trap. All fish not being marked (described below) were released into a separate pool downstream of the RST. In 2017 riffles downstream of the RST had been degraded by high flows, so fish were driven approximately 300 m downstream and released into a pool with abundant shelter. California freshwater shrimp were carried upstream of the bedrock cascade and released into an undercut bank with plentiful alder roots. The proportion of migrating fish captured each week (trap efficiency) was determined by recapturing previously marked fish. Up to ten smolts per species per day were given a fin clip unique to the 3

week and location of capture and were released approximately 500 m upstream. Some of these fish were subsequently recovered at the trap a second time and served as the basis for calculating trap efficiencies. Smolts smaller than 70 mm were not fin clipped. Marking of fish was coordinated with a simultaneous trapping effort conducted by SPAWN on San Geronimo Creek. Each monitoring effort used distinct fin clips to avoid duplicate marks from the two trapping locations (Table 1). Table 1. Marking schedule at Lagunitas and San Geronimo Creek smolt traps, 2017 Lower Lagunitas (MMWD) San Geronimo (SPAWN) Week Date Mark Applied Mark Applied 1 March 13 to March 19 lower caudal clip (LC) 2 March 20 to March 26 lower caudal clip anal clip (LCAC) 3 March 27 to April 2 No clip (away at conference) 4 April 3 to April 9 dorsal & lower caudal clip (DLC) dorsal & upper caudal clip (DUC) 5 April 10 to April 16 lower caudal clip (LC) upper caudal clip (UC) 6 April 17 to April 23 lower caudal clip anal clip (LCAC) upper caudal clip anal clip (UCAC) 7 April 24 to April 30 dorsal & lower caudal clip (DLC) dorsal & upper caudal clip (DUC) 8 May 1 to May 7 lower caudal clip (LC) upper caudal clip (UC) 9 May 8 to May 14 lower caudal clip anal clip (LCAC) upper caudal clip anal clip (UCAC) 10 May 15 to May 21 dorsal & lower caudal clip (DLC) dorsal & upper caudal clip (DUC) 11 May 22 to May 28 lower caudal clip (LC) upper caudal clip (UC) In addition to monitoring smolts with the RST, MMWD operated a PIT tag antenna upstream of the RST through part of the winter and spring of 2016-17. PIT tag monitoring was intended to estimate overwinter survival rates for coho tagged in different parts of the watershed. In 2016 coho between 55 and 65 mm were implanted with 8 mm full duplex tags (FDX) and fish 65mm were implanted with 12 mm half duplex tags (HDX). Fish with HDX tags could be detected at the antenna as well as at the screw trap where all captured fish were scanned with a handheld PIT tag detector. The antenna cannot detect FDX tags, and therefore FDX tags could only be detected at the screw trap. Unfortunately the antenna was damaged by high flows on February 7 and not repaired until late April. Olema Creek Monitoring A fyke/pipe trap was operated by NPS staff from March 15 through June 2. The trap was pulled for 11 days in March due to high flows and was operated for a total of 60 days. The trap design was based on traps used by CDFW on the Noyo River (Gallagher 2000). The trap was checked daily, and no more than 30 coho smolts (or up to 50% of the catch that day) were anesthetized with carbon dioxide and marked with PIT tags. Marked smolts were released immediately after recovering from anesthetization at a predetermined site approximately 100 m or at least three habitat units above the trap site. After being measured, all recaptured smolts and unmarked smolts were released immediately in low velocity areas below the trap. Studies using similar methods of marking and tagging have demonstrated little marking mortality (Greis and Letcher 2002) and a study using the same trapping methodology on five northwestern California streams revealed that 4

trap mortality was less than one percent for smolts and less than three percent for fry (Manning 2001). Salmonids were identified to species and life stage (fry, parr, smolt, or adult) prior to being measured. Fry are less than one year old and can be identified by the presence of distinct parr marks and small body size. Age 1+ steelhead and coho were separated into the following morphological categories: parr (smaller size, parr marks present) or smolt (faint or absent parr marks, silver body, deciduous scales, black fin margins). A random sub-sample of up to ten coho smolts and steelhead parr and smolts were measured to the nearest millimeter (fork length), and weighed to the nearest 0.01 g using an electronic scale. All fish that were anesthetized and marked were also measured and weighed. Any adult steelhead encountered in the trap were released downstream immediately without being measured. Random sub-samples of ten coho fry and ten steelhead fry were measured daily and individuals greater than 40 mm were weighed to the nearest 0.01 g using an electronic scale. Ten individuals of each non-salmonid species were also randomly selected, measured, and weighed. Sub-samples were obtained by taking blind scoops out of the holding bucket with a small aquarium dip net. After processing, each fish is placed in an aerated recovery bucket, keeping larger sculpin in separate buckets to avoid predation on smaller fish. Fish in the recovery bucket were monitored to ensure sedated fish recover fully before being released. In addition to smolt trapping, a PIT tag antenna array was installed upstream of the Olema Creek smolt trap (Figure 1). For more detailed descriptions of smolt trapping methods, please refer to SOP (standard operating procedure) 3, SOP 4, SOP 6, SOP 9, and SOP 11 of the San Francisco Bay Area Network Salmonid Monitoring Protocol version 4.0 (Reichmuth et al. 2010). San Geronimo Creek Monitoring SPAWN staff operated a fyke/pipe trap near the mouth of San Geronimo Creek from April 14 to May 26, for a total of 41 days. The trap was checked daily following protocols similar to those used by MMWD and NPS. A subset of coho and steelhead smolts were marked (Table 1) and released approximately 100 m upstream of the trap to estimate trap efficiency. Data Analysis The efficiency of the Lagunitas Creek rotary screw trap and populations of coho and steelhead smolts were estimated using Darroch Analysis with Rank Reduction (DARR) 2.0.2 software (Bjorkstedt 2005, 2010) from mark-recapture data. The DARR 2.0.2 software was developed to allow populations of downstream migrants to be estimated using mark-recapture data, particularly in small watersheds. This program applies a set of algorithms to stratified mark-recapture data to produce an abundance estimate while defining the variability in capture probability and the distribution of recaptured individuals within the strata. Mark-recapture data for steelhead was modified slightly for this analysis. During weeks two and four of the survey coho and steelhead smolts were marked and released shortly before the RST was disabled for four days. While some of these coho smolts were recaptured, none of the steelhead smolts were, causing DARR to calculate very low trap efficiencies for weeks two through five. The 29 steelhead marked in weeks two and four were not included in the marked subsample 5

for this analysis to avoid underestimating trap efficiency and inflating the overall steelhead smolt estimate. RESULTS Lagunitas Creek Rotary Screw Trap The Lagunitas Creek RST captured 5,550 coho smolts, 524 steelhead smolts, and 925 Chinook smolts in 2017. An estimated 29,306 coho smolts emigrated from Lagunitas Creek during the monitoring period (Table 2). We captured 45 young-of-the-year coho and 15 age 2+ coho smolts. The remaining coho catch was comprised of 1+ fish (1-2 years old) (Table 3). The highest estimated passage occurred during the week of April 10, with 8,787 coho smolts passing through and around the RST (Figure 2). The weekly trap efficiency for coho smolts varied from 6% to 100% (mean 34%) (Figure 3). The highest catch for a single day occurred on April 21 when 468 coho smolts were captured (Figure 4). Coho smolts averaged 111 mm fork length (FL) and weighed an average of 14.8 g. Table 2. Estimated smolt emigration from Lagunitas Creek, 2006-2017 Year Survey start date Survey end date Coho Steelhead Chinook Observed Estimated Observed Estimated Observed Estimated 2006 21 March 9 June 1,342 2007 15 March 30 May 611 2008 18 March 5 June 2,532 2009 10 March 5 June 3,150 2010 17 March 27 May 631 2011 1 April 20 May 1,684 2012 26 March 31 May 4,339 2013 19 March 7 June 4,942 2014 11 March 4 June 8,415 2015 19 March 9 June 7,373 2016 16 March 24 May 3,428 2017 14 March 26 May 5,550 5,946 (±1,570) 2,776 (±692) 6,101 (±780) 5,711 (±461) 2,129 (±480) 3,300 (±470) 8,315 (±1,372) 7,479 (±504) 15,055 (±1,974) 10,643 (±596) 9,719 (±2,225) 29,306 (±11,286) 308 475 449 646 651 829 251 684 448 814 371 524 6,949 (±6,133) 3,632 (±2,066) 1,134 (±259) 2,041 (±537) 3,867 (±1,419) 3,753 (±941) 1,991 (±1,252) 1,876 (±380) 1,720 (±478) 2,699 (±594) 4,396 (±3,099) 3,164 (±1,313) 237 504 775 2,445 0 0 0 0 0 0 0 0 0 0 0 0 1,229 2,005 191 925 2,011 (±241) 3,376 (±382) 833 (±370) 2,224 (±425) 6

Table 3. Salmonids captured in the Lagunitas Creek rotary screw trap by length and week, 2017. Smolt Monitoring in the Lagunitas Creek Watershed 2017 Coho Steelhead Week: 1 2 3 4 5 6 7 8 9 10 11 1 2 3 4 5 6 7 8 9 10 11 3/13 3/20 3/27 4/3 4/10 4/17 4/24 5/1 5/8 5/15 5/22 3/13 3/20 3/27 4/3 4/10 4/17 4/24 5/1 5/8 5/15 5/22 Dates 3/19 3/26 4/2 4/9 4/16 4/23 4/30 5/7 5/14 5/21 5/28 3/19 3/26 4/2 4/9 4/16 4/23 4/30 5/7 5/14 5/21 5/28 Length (mm) Age 0+ Age 0+ 20-24 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 25-29 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 6 0 0 0 0 9 30-34 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 35-39 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 40-44 0 0 0 0 0 1 2 0 0 0 0 0 0 0 0 0 0 3 1 0 2 0 6 45-49 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 0 1 3 7 50-54 Age 0 1+ 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 6 5 4 9 16 40 55-59 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 2 12 29 26 70 60-64 0 0 0 0 0 0 1 0 0 1 0 Age 0 1+ 0 0 0 0 0 1 0 3 37 16 57 65-69 0 0 Numbers 0 0 in bold 0 0 1 0 1 1 0 0 0 0 0 0 0 0 1 3 25 19 48 70-74 0 0 indicate 0 0that scales 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 6 9 16 75-79 0 0 were 0 collected 0 0 for 0 0 2 0 4 2 0 0 0 0 0 0 0 1 0 1 3 5 80-84 0 0 age 0analysis. 0 0 0 0 0 0 4 3 0 0 0 0 0 0 0 1 0 1 0 2 85-89 2 0 0 0 0 0 0 1 0 4 8 0 0 0 0 0 0 0 0 0 0 0 0 90-94 4 3 0 2 0 2 0 1 0 3 7 0 0 0 0 1 0 0 0 0 0 0 1 95-99 3 3 0 1 4 3 3 12 5 11 10 0 0 0 0 0 0 0 0 0 2 0 2 100-104 14 11 0 1 15 9 14 9 11 7 8 0 0 0 0 0 0 0 1 0 0 0 1 105-109 10 5 0 10 15 9 22 30 17 19 7 1 0 0 0 1 0 0 0 0 0 0 2 110-114 7 11 1 9 37 17 24 14 14 10 2 1 1 0 0 0 1 1 2 0 1 0 7 115-119 8 12 0 10 35 19 13 15 5 1 1 1 0 0 0 0 2 1 1 2 1 0 8 120-124 8 3 2 6 29 13 16 3 1 2 1 2 2 0 0 1 0 0 4 2 1 0 12 125-129 7 4 0 2 10 3 2 3 1 1 0 Age 0 2+ 0 0 0 0 0 1 1 0 1 0 3 130-134 4 0 0 1 4 4 1 0 0 0 0 0 2 0 0 0 0 1 4 1 2 0 10 135-139 0 2 1 1 6 0 0 0 0 0 0 2 0 0 0 1 0 1 6 1 0 0 11 140-144 Age 0 2+ 0 0 0 1 0 0 0 0 0 0 3 1 0 1 2 1 3 5 0 0 0 16 145-149 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 2 3 3 6 2 0 1 19 150-154 0 0 0 0 0 1 0 0 0 0 0 1 0 0 5 4 6 7 6 1 2 1 33 155-159 0 0 0 0 0 0 0 0 0 0 0 2 2 0 5 12 11 7 12 2 2 0 55 160-164 0 0 0 0 0 1 0 1 0 0 0 4 3 1 3 10 7 13 7 0 0 0 48 165-169 0 0 0 0 0 0 0 0 0 0 0 4 1 1 2 7 8 9 5 1 1 0 39 170-174 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 6 9 2 7 0 0 0 26 175-179 0 0 0 0 0 0 0 0 0 0 0 4 1 2 1 7 4 3 4 0 1 0 27 180-184 0 0 0 0 0 0 0 0 0 0 0 5 3 1 1 4 0 6 2 0 0 0 22 185-189 0 0 0 0 0 0 1 0 0 0 0 4 1 0 0 2 0 2 3 0 Age 0 3+ 0 12 190-194 0 0 0 0 0 0 0 0 0 0 0 3 1 0 1 0 1 1 1 1 1 0 10 195-199 0 0 0 0 0 0 0 0 0 0 0 2 3 0 0 0 0 1 1 0 1 0 8 200+ 12 5 1 1 1 1 1 0 2 0 1 25 Totals Age 0 0 0 0 1 0 2 4 1 2 15 20 6% 0 0 0 0 0 3 18 15 23 111 93 40% Age 1+ 67 52 3 42 149 79 95 90 54 66 49 93% 5 3 0 0 3 3 2 9 4 8 0 6% Age 2+ 0 2 1 1 7 2 1 1 0 0 0 2% 47 26 6 20 58 51 61 69 11 8 3 55% Chinook Week: 1 2 3 4 5 6 7 8 9 10 11 55-59 0 0 0 0 0 0 1 0 0 0 0 1 60-64 0 0 0 0 0 0 1 0 0 1 0 2 65-69 0 0 0 0 0 0 2 1 0 1 3 7 70-74 0 0 0 0 0 0 5 7 3 1 4 20 75-79 0 0 0 0 0 6 5 19 7 11 17 65 80-84 0 0 0 0 0 3 12 33 21 41 17 127 85-89 0 0 0 0 1 1 13 32 19 18 11 95 90-94 0 0 0 0 2 0 10 4 4 7 0 27 95-99 0 0 0 0 0 0 0 0 0 1 0 1 Totals 0 0 0 0 3 10 49 96 54 81 52 7

During the monitoring period, an estimated 3,164 steelhead smolts emigrated from Lagunitas Creek (Table 3). The peak of emigration occurred during the week of April 10, with an estimated 608 steelhead smolts (Figure 2). The highest catch for a single day occurred on April 21, when 39 steelhead smolts were captured. The weekly trap efficiency for steelhead smolts ranged from 9% to 33% (mean 21%) (Figure 3). For the season the steelhead catch was comprised of 40% fry, 6% age 1+ and 54% age 2+ and older steelhead (Table 3). Only one adult steelhead, apparently unspawned, was captured. Age 1+ steelhead were easily distinguished from YOY, but overlapped in size with Age 2+ steelhead. An indistinct break in the size distribution between 170 and 180 mm appeared to differentiate Age 3+ steelhead from younger fish. We classified 522 steelhead as fully-smolted, while 37 steelhead, ranging between 112 and 154 mm, were classified as transitional due to their mix of juvenile and smolt characteristics. Transition steelhead were not included in the emigration estimate, since some of these fish may not have emigrated. Steelhead smolts averaged 167 mm FL and weighed 51 g on average. We captured one 312 mm, heavily spotted O. mykiss that appeared to be a resident rainbow trout. An estimated 2,224 Chinook smolts emigrated from Lagunitas Creek, with a peak emigration of 726 during the week of May 8. The weekly trap efficiency for Chinook ranged from 23% to 84% (mean 41%) (Figure 3). Chinook smolts were 82 mm in length and weighed 6.5 g on average. A record number of certain non-salmonid fish species were also captured in the RST, including bluegill, goldfish, golden shiner, white crappie, and channel catfish (Table 4). A record number of bullfrog tadpoles (143) were also captured, as well as 231 CA freshwater shrimp, 57 crayfish, and one western pond turtle. Table 4. Non-salmonid fish captured in the RST Common Name Species Native? Catch Sculpin Cottus spp. Y 2270 CA roach Hesperoleucus symmetricus symmetricus Y 2165 Bluegill Lepomis macrochirus N 337 Black crappie Pomoxis nigromaculatus N 222 Goldfish Carassius auratus N 155 Threespine stickleback Gasterosteus aculeatus Y 147 Golden shiner Notemigonus crysoleucas N 95 White crappie Pomoxis annularis N 60 Sacramento sucker Catostomous occidentalis Y 34 Channel catfish Ictalurus punctatus N 22 Pacific lamprey Lampetra tridentata Y 17 Redear sunfish Lepomis microlophus N 5 Green sunfish Lepomis cyanellus N 4 Striped bass Morone saxatilis N 4 Common carp Cyprinus carpio N 2 8

PIT Tagged Coho During the summer and early fall of 2016 PIT tags were implanted into 677 coho fry. The first of these tags was detected at the PIT tag antenna on November 1 and 16 tags were detected before the antenna was damaged in early February. Only seven additional tags were detected at the antenna after it was repaired in late April. Two additional coho tagged in Lagunitas Creek were detected at the antenna on Olema Creek. Another 20 tagged coho were captured in the RST, for a total of 45 detections (6.6% of tagged coho). Approximately 300 coho (6% of all captured coho) were not scanned on three days during the survey due to the PIT-tag reader malfunctioning. Coho tagged in Lagunitas Creek were detected at a somewhat higher rate (7.7%) compared to coho from San Geronimo Creek (4.5%) and Devil s Gulch (3.7%). Table 5. PIT tagged coho detections Tag Type Tagging Location Fish Tagged in 2016 2017 Antenna Detections Antenna Detection Rate 2017 RST Detections Total Detections Total Detection Rate Lagunitas Creek 335 24 7.1% 10 34 10.1% HDX San Geronimo Cr. 37 0 0.0% 2 2 5.4% Devil's Gulch 20 0 0.0% 0 0 0.0% All 394 24 6.1% 12 36 9.1% Lagunitas Creek 147 N/A N/A 3 3 2.0% FDX San Geronimo Cr. 75 N/A N/A 3 3 4.0% Devil's Gulch 61 N/A N/A 3 3 4.9% All 283 N/A N/A 9 9 3.2% All All 677 N/A N/A 21 45 6.6% Olema Creek Monitoring The Olema Creek smolt trap captured 1,145 coho smolts and an estimated 5,698 (±1,151) coho emigrated from the creek (Michael Reichmuth, pers. comm.). This was highest estimate for the period of smolt monitoring, which began in 2004. The peak of the migration occurred during the week of April 25. Coho smolts averaged 118 mm in length. San Geronimo Creek Monitoring An estimated 1,136 (±184) coho smolts emigrated from San Geronimo Creek (Preston Brown, pers. comm.). Coho from San Geronimo Creek made up 3% of the total watershed emigration. DISCUSSION Sampling conditions and effects on estimated smolt abundance and emigration timing Taking advantage of the first moderate stream flows in months, the smolt trap was installed on March 14, which was one of the earliest start dates since trapping began in 2006. For the first two weeks the trap was disabled on weekends while WSP members were being trained. The trap was disabled again between March 29 and April 2 for the Salmonid Restoration Federation Conference, 9

and again for five days during the second week of April due to high flows. The trap was not operating reliably and full-time until April 18, just a few days before the peak coho and steelhead catches and the mid-point of the monitoring season. These disruptions to trap operation during the first five weeks hindered recaptures of marked salmonids, resulting in trap efficiency estimates of only 6% for coho and 9% of steelhead and significant uncertainty around smolt estimates. The 95% confidence interval of approximately 24,000-46,000 coho smolts (Figure 6) represents the most uncertainty around a smolt estimate in 12 years of monitoring. At the start of smolt trapping both coho and steelhead migrations appeared to be well on their way. Approximately 10% of both coho and steelhead smolts emigrated during the first week of monitoring, and catches of both species where unusually high for mid-march. Unfortunately we could not assess the magnitude of the early coho migration, before smolt trapping began, due to the inoperable PIT tag antenna. The first Chinook smolt was captured on April 16, and the trap recaptured 23% of marked Chinook that first week. High flows only disrupted trap operation for one day during the Chinook migration and trap efficiency for Chinook was relatively high. As we ve seen in previous years, recapture rates for Chinook were generally lower than recapture rates for coho, possibly indicating trap avoidance behaviors that differ between species. Water temperatures in January are a good predictor of when the first significant pulse of coho smolts will be observed 1 (Figure 7). The peak coho catch occurred on April 21, about a week later than what was expected. Age and size composition of salmonids Coho smolts were not only extraordinary for their abundance in 2017, but for their size. The average length of smolts was 111 mm, which while slightly below average, would not be expected with emigrations over 10,000 smolts (Figure 7). What s even more extraordinary is that when these coho were sampled as fry in 2016 they were among the smallest observed, averaging only 63 mm. This demonstrates very fast growth prior to smolting, with one PIT-tagged coho more than doubling in length during that period. Growth was likely fueled by terrestrial insects, either preyed upon on the floodplain or after they were washed into the creek during numerous over-bank flows. This abundant prey may also have enabled some coho to smolt in their first year. During the last two weeks of monitoring 22 YOY coho displaying smolt characteristics were observed. An additional eight YOY were classified as transitional, meaning they had some smolt features. Steelhead smolts also demonstrated extraordinary growth. YOY steelhead in 2015 were the smallest in over a decade, averaging only 54 mm long. This same year class averaged 160 mm long as smolts (i.e., tripled in length in a year and a half). All steelhead smolts (including both Age 2+ and 3+) averaged 169 mm long, which was almost average. 1 The first significant pulse of smolts was defined as the earliest of the top 5% of daily catches of coho smolts. 10

Smolt abundance trends and implications for winter survival The 2017 coho emigration was by far the largest migration on record for Lagunitas Creek, nearly doubling the previous record (Figure 9). It was also more than four times larger than the carrying capacity hypothesized in 2008 (Stillwater Sciences). Of the 41,000 juvenile coho estimated to reside upstream of the Lagunitas smolt trap in 2016, 70% survived through the winter. In Ettlinger et al. (2017) we proposed that smolt abundance was limited by the availability of flow refuge during moderate stream flows, based on a strong negative correlation between the number of days of moderate flows and smolt abundance. That correlation fell apart this year, when frequent storms resulted in many days of moderate and higher flows. In fact, every conceivable flow variable that was investigated (peak discharge, consecutive days of flow, intervals between flows, etc.) as well as biotic factors (fry length and distribution) failed to explain the extraordinary survival of coho through the winter of 2016-17. Habitat may have been available to coho that was unavailable in previous years, but it s unknown whether this additional habitat was a product of enhancement efforts, the prolonged drought (such as wood recruitment in the form of droughtstressed trees), or natural channel evolution. Coho abundance is increasing, and that trend may accelerate with the help of habitat enhancement efforts being implemented in 2017 and 2018. The goal of 52,000 coho smolts emigrating from Lagunitas Creek appears within reach. A population of that size, surviving in the ocean at typical rates, could produce runs of 2,600 adults, the current federal population recovery target. 11

REFERENCES Bjorkstedt, E. P. 2005. DARR 2.0: updated software for estimating abundance from stratified markrecapture data. NOAA Technical Memorandum NMFS-SWFSC-368. Bjorkstedt, E. P. 2010. DARR 2.0.2: DARR for R. http://swfsc.noaa.gov/textblock.aspx?division=fed&id=3346 Brown, P. (SPAWN). Personal Communication, August 3, 2017. Ettlinger, E., P. Doughty, V. Rogers and G. Andrew. 2016. Smolt monitoring in the Lagunitas Creek watershed 2015. Marin Municipal Water District. Gallagher, S. P. 2000. Results of the 2000 Steelhead (Oncorhynchus mykiss) Fyke Trapping and Stream Resident Population Estimations and Predictions for the Noyo River, California with Comparison to Some Historic Information. California State Department of Fish and Game, Steelhead Research and Monitoring Program, Fort Bragg, CA. Report FB-03-Draft, September 2000. 75pp. Greis, G., and B. H. Letcher. 2002. Tag retention and survival of age-0 Atlantic salmon following surgical implantation with passive integrated transponder tags. North American Journal of Fisheries Management 22: 219-222. Manning, D.J. 2001. Carrying capacity and limiting habitat analysis for coho salmon (Oncorhynchus kisutch) in streams of northwestern California. Humboldt State University: Masters Thesis, Arcata, California. Reichmuth, M. (NPS). Personal Communication, August 4, 2017. Reichmuth, M., B.J. Ketcham, D. Fong, S. Carlisle, E. Brown, and M. Koenen. 2010. Salmonid Monitoring Protocol for the San Francisco Bay Area Network: narrative and appendices version 4.0. Natural Resource Report NPS/SFAN/NRR 2010/202. National Park Service, Fort Collins, Colorado. Stillwater Sciences. 2008. Lagunitas limiting factors analysis; limiting factors for coho salmon and steelhead. Final Report. Prepared by Stillwater Sciences, Berkeley, California for Marin Resource Conservation District, Point Reyes Station, California. 12

a Rd ta l um ay sb Pt 1 ay hw Hig le ma To Re ye s -P e Lagunitas Creek Watershed Lagunitas Creek rotary screw trap, PIT tag antenna, and Point Reyes stream gage r asio Fran cis Drak e d Sir d yr Olema Creek PIT tag antenna lle Va Olema Creek smolt trap id g er San Bou ndar Francisco y Bay Ni c rm tf o Br ed Pacific Ocean Nicasio Reservoir Pla Ni c a s i o C Wate r sh Bl d v Golden Gate National Recreation Area Ol il's Dev em k ree La g u ac Samuel P. Taylor stream gage lc Gu Samuel P. Taylor State Park San Geronimo Creek smolt trap Larse n Cr eronimo Creek S an G ek ni ta s C re h Ke nt ke La Legend Hig Marin Municipal Water District hw Federal Lands 1 State Lands ay Marin County Open Space 0 0.5 1 2 3 4 5 Figure 1. Smolt monitoring locations in the Lagunitas Creek watershed. 13 6 Kilometers Marin Municipal Water District d r. re C ac 1 ay hw Hig Marin County Open Space Wo o

Smolts 10,000 9,500 9,000 8,500 8,000 7,500 7,000 6,500 6,000 5,500 5,000 4,500 4,000 3,500 3,000 2,500 2,000 1,500 Coho Estimate Coho Observations Steelhead Estimate Steelhead Observations Chinook Estimate Chinook Observations Flow 1,000 950 900 850 800 750 700 650 600 550 500 450 400 350 300 250 200 150 Mean Daily Stream Flow at the Point Reyes USGS Gage (cfs) 1,000 100 500 50-0 Figure 2. Lagunitas Creek smolt emigration and stream flow. 14

Trap Efficiency 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Trap Efficiency for Coho Trap Efficiency for Steelhead Trap Efficiency for Chinook Flow 1,000 950 900 850 800 750 700 650 600 550 500 450 400 350 300 250 200 150 100 50 - Mean Daily Stream Flow at the Point Reyes USGS Gage (cfs) Figure 3. Weekly trap efficiency and Lagunitas Creek stream flow. 15

Coho Smolts Captured (New and Recaptures) 500 450 400 350 300 250 200 150 Smoothed Mean 2006-2017 2017 2016 2015 2014 100 50 0 Date Figure 4. Daily coho smolt captures at the Lagunitas Creek smolt trap, 2014-2017 16

5/17 5/10 5/3 Date of First Coho Emigration Pulse 1 4/26 4/19 4/12 2017 R² = 0.7081 4/5 3/29 47.5 48.0 48.5 49.0 49.5 50.0 50.5 51.0 51.5 52.0 52.5 Mean January Water Temperature (F) 1 The first emigration pulse is defined as the earliest of the top 5% of daily coho smolt catches for that season. Figure 5. January water temperatures and coho smolt emigration timing, 2008-2017. 17

Coho population recovery target 50,000 40,000 Olema Creek Coho Lagunitas Creek Coho Lagunitas Creek Steelhead Lagunitas Creek Chinook Smolt Estimate 30,000 20,000 10,000-2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Figure 6. Lagunitas Creek smolt emigration estimates. Note: The coho recovery target assumes an ocean survival rate of at least 5%, resulting in 2,600 adult returns. 18

140 135 130 125 Mean smolt length (mm) 120 115 110 2017 105 100 R² = 0.5236 95 90-5,000 10,000 15,000 20,000 25,000 30,000 35,000 Coho smolt estimate Figure 7. Coho emigration estimates and smolt lengths at the Lagunitas Creek smolt trap. 19