Farm to Fish: Lessons from a Multi-Year Study on Agricultural Floodplain Habitat Ted Sommer,*1 Brian Schreier,1 J

Sponsored by the Delta Science Program and the UC Davis Muir Institute

RESEARCH Farm to Fish: Lessons from a Multi-Year Study on Agricultural Floodplain Habitat Ted Sommer,*1 Brian Schreier,1 J. Louise Conrad,2 Lynn Takata,2 Bjarni Serup,3 Rob Titus,3 Carson Jeffres,4 Eric Holmes,4 Jacob Katz5

ABSTRACT that seasonally flooded fields are highly Large areas of California’s historic floodplain productive, resulting in significantly higher have been separated from adjacent river levels of zooplankton and high Chinook channels by levee construction, allowing the Salmon growth rates as compared to the development of an extensive agricultural adjacent Sacramento River. We found similar industry. Based on successful partnerships results for multiple geographical areas in between agriculture and conservation groups the Central Valley, and in different cover to support migrating waterfowl, we examined types, such as non-rice crops and fallow whether seasonally flooded rice fields areas. Although field substrate type did not could be modified to provide off-channel detectably affect fish growth and survival, rearing habitat for juvenile Chinook Salmon connectivity with upstream and downstream Oncorhynchus tshawytscha. During winter and areas appeared to drive fish occupancy, spring of 2012-2017, we conducted a series because rearing young salmon were generally of experiments in Yolo Bypass and other attracted to inflow in the fields, and not regions of California’s Central Valley using all of the fish successfully emigrated off hatchery Chinook Salmon as a surrogate the fields without efficient drainage. In for wild Chinook Salmon, the management general, we faced numerous logistic and target for our project. Overall, we found environmental challenges to complete our research. For example, periodic unmanaged SFEWS Volume 18 | Issue 3 | Article 3 floods in the Yolo Bypass made it difficult to https://doi.org/10.15447/sfews.2020v18iss3art3 schedule and complete experiments. During severe drought conditions, we found that * Corresponding author: [email protected] managed agricultural habitats produced low 1 California Department of Water Resources Sacramento, CA 94236 USA and variable salmon survival results, likely 2 Delta Science Program, Delta Stewardship Council because of periodically high temperatures Sacramento, CA 95814 USA and concentrated avian predation. In 3 California Department of Fish and Wildlife addition, our project required substantial West Sacramento, CA 95605 USA landowner time and effort to install and 4 University of California Davis maintain experimental fields. Recent and Davis, CA 95616 USA future infrastructure improvements in Yolo 5 California Trout, Inc. San Francisco, CA 94104 USA Bypass could substantially improve options for experimental work and broaden efforts to enhance salmon habitat.

1 SAN FRANCISCO ESTUARY & WATERSHED SCIENCE VOLUME 18, ISSUE 3, ARTICLE 3

KEY WORDS wetlands are relatively productive habitats Chinook Salmon, Yolo Bypass, floodplain, (Dybala et al. 2017), helping California rearing, agriculture, management. meet some of its shorebird and waterfowl conservation objectives (Gardali et al. 2017; Strum et al. 2017). INTRODUCTION Worldwide, the loss of floodplain habitat The success of these avian-oriented has resulted in considerable reduction in programs raises the question of whether the productivity and diversity of large flooded fields can also be used as habitat river ecosystems (Winemiller et al. 2016). for native fishes and whether management California is no exception, as widespread practices can affect habitat quality. One levee construction separated historical location where this question is relevant is riverine and estuarine floodplain habitats Yolo Bypass (Figure 1), Sacramento River’s from adjacent river channels (Mount 1995; primary remnant floodplain with large areas Opperman 2012; Whipple et al. 2012). At the of agricultural lands (Sommer et al. 2001a, same time, there have been corresponding 2001b; Opperman 2012). Research over the declines in the distribution and abundance of past 2 decades has shown that seasonally a suite of native fishes (Moyle et al. 2011), an flooded lands support a suite of native fishes understandable outcome since a large body of and provide food web subsidies within and research provides evidence that floodplains to downstream habitats (Sommer et al. 1997, are high-quality habitat for multiple Central 2001a, 2001b; 2004; Jeffres et al. 2008). At Valley fishes (Sommer et al. 2001a, 2001b; the same time, these studies have shown Feyrer et al. 2006a, 2006b; Jeffres et al. that the Yolo Bypass is far from optimal 2008; Sommer et al. 2014). habitat because the landscape has been altered to drain relatively quickly, and is Construction of levees in the Central Valley often disconnected from the Sacramento has allowed the development of one of the River by levees and weirs that create major most extensive and productive agricultural passage problems for upstream migrating economies in the world (e.g., CDA 2018). Lands adult fishes, such as Chinook Salmon and that formerly seasonally flooded can now sturgeon (Sommer et al. 2014; Herbold et al. be farmed in most years. Recently, there 2018; Johnston et al. 2020). Several of these has been a growing interest in whether the issues will be addressed in coming years vast acreage of farmland that now occupies by proposed structural changes to Fremont formerly inundated floodplains adjacent to Weir at the upstream end of Yolo Bypass, river systems could be managed to improve and by the additional improvements to the their value to fish and wildlife (Suddeth floodplain’s water distribution system (USBR Grimm and Lund 2016; Katz et al. 2017). and CDWR 2019). However, the question A case in point is the Central Valley Joint remains whether changes to agricultural Venture Partnership, which promotes the use land management and infrastructure can of flooded rice fields as a means to increase provide reliable fish habitat that can increase available wetlands to support shorebirds and the growth and survival of juvenile native waterfowl migrating, overwintering, and fishes, and thereby (1) contribute to reversing breeding along the Pacific Flyway (Shuford their overall decline, (2) aid in the recovery and Dybala 2017). Conservation biologists of native fishes listed under the U.S. and working in partnership with local farmers California Endangered Species acts, (3) and have developed a system by which fields are increase the availability of fishery resources. reflooded after rice harvest to create seasonal wetland habitat for water birds during fall To help the overarching objective of and winter. These temporary agricultural providing reliable fish habitat, our team

2 SEPTEMBER 2020

Goose Club Sutter Bypass Ranch Study Area

Fremont Weir

Wallace Weir

Knaggs Ranch Study Area

Conaway Ranch Study Area Sacramento Weir Sacramento

Putah Creek Yolo Bypass Wildlife Area

Region Map

Yolo Bypass

Rio Vista San Francisco Bay 5 0 5 10 15 Dos Rios Study Area Kilometers

Figure 1 Location map for Yolo Bypass and related sampling areas. The Dos Rios study area is located along the San Joaquin River, south of the main area shown.

https://doi.org/10.15447/sfews.2020v18iss3art3 3 SAN FRANCISCO ESTUARY & WATERSHED SCIENCE VOLUME 18, ISSUE 3, ARTICLE 3

conducted a series of field studies during many of the lessons learned here should have 2012-2017. To test fish and food web at least some relevance to wild Sacramento responses within different land-management River Chinook Salmon. Finally, our project scenarios, we conducted our project on was separate from a number of other fish standard rice and winter wheat fields, management research projects in agricultural adjacent fallow lands, and rice fields parcels, such as current efforts to investigate with different harvest practices or other whether invertebrates grown on flooded experimental modifications. This work rice fields can be used as a food subsidy for yielded several publications that provided adjacent river channels (e.g., Cornwell and insight into habitat conditions in flooded rice Katz 2017). fields for fish and invertebrates (Conrad et al. 2016; Corline et al. 2017; Katz et al. 2017). The focus of our effort was on rearing habitat STUDY AREA for young Chinook Salmon, but this work The Yolo Bypass is a 24,000-ha, partially may also be relevant to other native fishes. leveed flood basin that is used to safely convey floodwaters away from Sacramento The goal of this paper is to summarize the Valley communities (Figure 1). The Yolo key lessons learned from 6 years of research Bypass contains a suite of habitats including on the feasibility of using farm fields as agricultural lands, managed wetlands, upland rearing habitat for juvenile Chinook Salmon habitat, and perennial ponds and channels, in the Yolo Bypass and other Central Valley with broad open-water tidal wetlands at locations. Our hope is that our summary will its downstream end where it joins the provide guidance to future researchers, as Sacramento-San Joaquin Delta (Sommer et well as inform managers as they evaluate al. 2005; Suddeth Grimm and Lund. 2016). potential management approaches. An The basin receives seasonal inflow from the important caveat is that our studies were Sacramento River, Colusa Basin (via Knights not intended as a proof of concept for any Landing Ridge Cut), Cache Creek, and Putah specific management actions. Rather, our Creek, as well as substantial perennial research was intended to examine some of tidal flow from the San Francisco Estuary the attributes that could reduce limitations via the lower Sacramento River at the to rearing conditions identified in early downstream end of the floodplain (Figure 1). research, and gain insight into some of The Yolo Bypass floods to various degrees the key considerations for potential future in approximately 80% of water years, but agricultural floodplain management. A inundation events are often relatively short second major caveat is that we had to rely (< 2 weeks) and sometimes driven entirely on juvenile hatchery Chinook Salmon as by inflow from the west-side tributaries. a surrogate for wild Chinook Salmon, our The most substantial flow events come from ultimate target for habitat restoration. We the Sacramento River, which enters the Yolo recognize that there are several potential Bypass via Fremont Weir and Sacramento differences in the behavior of hatchery Weir. However, in drought periods, such and wild Chinook Salmon (e.g., Davis et as during 2012-2015, there is little or no al. 2018). However, hatchery salmon were flooding. the only feasible alternative in this case since downstream migrating wild juvenile Chinook Salmon were mostly cut off from METHODS the Yolo Bypass because of extreme drought The general methods used for our field conditions. Nonetheless, hatchery salmon studies were described in detail by Katz et have been used successfully as a research al. (2017), Conrad et al. (2016), and Corline tool in many types of ecological studies, so et al. (2017). To improve the clarity of

4 SEPTEMBER 2020

Table 1 Summary of studies conducted during the multi-year investigation. See text for details.

Water year Year type Experiment Start date Duration Location Reporting 2012 Below Pilot study Jan 31 41 days Yolo Bypass Katz et al. 2017 normal Substrate preference Mar 13 14 days Yolo Bypass Conrad et al. 2016 telemetry experiment 2013 Dry Response to field 37 to 41 Corline et al. 2017; Feb 19 Yolo Bypass substrate (replicated fields) days Holmes et al. in prep Critically Depth & inflow preference California Trout et al. 2014 Feb 27 14 days Yolo Bypass dry telemetry experiment 2014 Yolo Bypass, Critically Geographical location California Trout et al. 2015 Feb 5 21 days Sutter Bypass, dry study 2015 Dos Rios Ranch Below California Trout et al. 2016 Flood extension Mar 13 21a days Yolo Bypass normal 2016 Lower trophic sampling Jan 19 122 days Yolo Bypass 2017 Wet Flood extension NA NAb Yolo Bypass

a. Late flood event made full flood extension experiment protocol unfeasible. Limited lower trophic wild fish sampling was conducted at Knaggs Ranch. b. Prolonged flooding precluded flood extension field containment. Limited lower trophic and wild fish sampling was conducted at Knaggs Ranch.

our multi-year synthesis, we summarize here the broad intent and general methods of each study year below and in Table 1. The general approach was to maintain shallow inundation (30-50 cm) in large (> 2 ha: Figure 2) single or replicated rice fields (9 � 0.8 ha each) in northern Yolo Bypass using water from either Knights Landing Ridge Cut or the Tule Canal (Katz et al. 2017). Field drains employed a plastic mesh live-car trap that allowed for fish to be captured and measured as they volitionally Figure 2 Photograph of example flooded field at Knaggs Ranch left the field before moving downstream. Their downstream emigration pathway was via Yolo Bypass during winter and early spring a drainage canal that discharged into the (Johnston et al. 2018; Takata et al. 2018). In Tule Canal/Toe Drain, a perennial channel 2015, additional flooded fields were studied that connects to the lowermost Sacramento on Conaway Ranch in Yolo Bypass, in Sutter River. The Tule Canal/Toe Drain represents Bypass, and at Dos Rios, a San Joaquin River a primary migration corridor for fish in floodplain. Managed flooding generally

https://doi.org/10.15447/sfews.2020v18iss3art3 5 SAN FRANCISCO ESTUARY & WATERSHED SCIENCE VOLUME 18, ISSUE 3, ARTICLE 3

occurred in February and March with a swimming throughout the flooded fields, duration of 3 to 6 weeks (Table 1), depending but mesh cages were also used as a tool primarily on magnitude and duration of to compare hatchery salmon growth and river and creek inflows to the bypasses, and survival across substrates in 2012 (Katz et al. opportunity for those waters to drain. 2017) or habitats (river, floodplain perennial channel, flooded fields) in 2016 and 2017. Sampling Approach For each year, we evaluated water 2012: Pilot Study quality (temperature, dissolved oxygen The initial study year was a pilot effort [DO], turbidity, pH), food web responses to evaluate whether managed flooding of (chlorophyll a, zooplankton), and fish growth a rice field could provide suitable habitat and condition (Corline et al. 2017; Katz et for juvenile salmon rearing, and to assess al. 2017). Water temperature in fields was associated growth and survival. A single recorded continuously at 10- to 15-minute 2-ha field contained a patchwork of four intervals with Onset HOBO® loggers, and agricultural substrate types, including disced a suite of other water-quality parameters (plowed soil with all vegetation removed), (DO, pH, conductivity, and temperature) short rice stubble (remains of rice plants left was measured and recorded using handheld in the field after harvest, in this case 5 cm and continuously installed multi-parameter high), high rice stubble (35 cm high), and sondes. We included plankton sampling fallow (weedy) vegetation. Approximately with the broad goal of characterizing the 10,200 juvenile salmon were released in communities and densities of phytoplankton the field, with a subset implanted with and zooplankton in the study fields. Because passive integrated transponder (PIT) tags, long-term monitoring of the Yolo Bypass so individuals could be identified, and includes weekly plankton sampling in both individual growth rates could be measured. the perennial Yolo Bypass channel of the Twenty PIT-tagged fish were also released in Toe Drain and the Sacramento River, we each of eight enclosures placed over patches could compare our experimental fields to of the different substrate types, to determine productivity across habitats. Because the if growth rates differed (Katz et al. 2017). study fields were shallow compared to canal and riverine channel environments, 2013: Comparison of Agricultural sampling methods had to be slightly modified Post-Harvest Substrates (Rice Stubble, compared to the Toe Drain and Sacramento Disced, or Fallow) River. As a result, we used hand-tosses of Substrates in flooded rice fields differ from a smaller 30-cm zooplankton net (153-μm those that juvenile salmon may encounter mesh), recording the length of the toss, and in natural floodplains or riverine systems. the relative percent of the net mouth that Thus, the goal of the second study year was was submerged during net retrieval. Detailed to investigate whether juvenile salmon had methods for zooplankton sampling are differential growth and survival rates across described in Corline et al. (2017). agricultural substrates, and whether they would preferentially use a specific substrate Fish used in the experiments were primarily type when given a choice. Our logic was that fall-run Chinook Salmon parr obtained understanding these responses could provide from Feather River Fish Hatchery; however, insight into whether some agricultural small numbers of wild Sacramento River practices provide more suitable salmon- Chinook Salmon were also studied in 2013 rearing conditions than others. (Feather River caught) and 2013, 2015, and 2016 (natural immigrants to flooded lands). To compare growth and survival rates across The majority of the study fish were free- rice stubble, disced, and fallow substrates, we

6 SEPTEMBER 2020

Fallow Stubble Disced Fallow Stubble Disced Stubble Fallow Disced

Figure 3 Schematic of nine replicated fields on Knaggs Ranch with substrate type and enclosure locations (black boxes)

created a series of nine 0.8-ha experimental congregate in areas near inflow and in fields with individual inlets and outlets, with deeper portions of fields. In 2014, we tested three replicates of each substrate (Figure 3). experimentally whether depth, proximity We placed approximately 4,600 hatchery- to inflow, or a combination of both affected origin juvenile Chinook Salmon in each field juvenile Chinook Salmon occupancy for 40 days and measured weekly during patterns. Similar to 2013, we employed PIT- the study period to estimate average growth tag technology in combination with large rates. field enclosures to track juvenile salmon movement. Specifically, we divided the study To examine substrate preference, we used field into three 18-m � 15-m enclosures. PIT-tag technology to track individual fish Within each enclosure, three 6-m � 15-m in two large circular enclosures (15.25 m panels were excavated into the substrate to diameter). In addition to examining the achieve three water depths: 0.45 m, 0.75 m, potential for preference among agricultural and 1.0 m. Panels were arranged so each substrates, this study also investigated depth was represented at each of three flow whether newer and smaller PIT tags (8-mm positions (upstream, midstream, downstream) tags compared to the traditional 12-mm tags) in a 3 � 3 factorial design. PIT antennae were were viable for detecting juvenile salmon placed in each depth panel, and PIT-tagged movements in these habitats. One enclosure juvenile salmon were released into each included three habitat treatments (fallow, enclosure, where they resided for 2 weeks. rice stubble, and disced), and the other served as a comparison with only the disced 2015: Comparison Across Different treatment. Each enclosure contained an array Geographical Areas (Sutter Bypass, Yolo of six rectangular PIT antennas arranged Bypass, San Joaquin River) in the same orientation. Fish remained in Until 2015, our studies had been restricted the enclosures for 14 days, during which to the Knaggs Ranch property in the occupancy data were collected. Detailed northern Yolo Bypass. We extended our methods can be found in Conrad et al. (2016). geographic scope during the fourth study year, investigating the food webs, juvenile 2014: Testing of Water Depth and Flow salmon survival, and juvenile salmon Characteristics of Fields growth on other agricultural properties, Observations during earlier study years and in other Central Valley floodplain-river indicated that juvenile salmon tended to systems. These additional properties were

https://doi.org/10.15447/sfews.2020v18iss3art3 7 SAN FRANCISCO ESTUARY & WATERSHED SCIENCE VOLUME 18, ISSUE 3, ARTICLE 3

Conaway Ranch adjacent to Cache Creek in the field. During the first week of flood the Yolo Bypass and south of the Knaggs extension, we held stocked hatchery salmon Ranch property, Goose Club Ranch adjacent and any entrained natural-origin salmon, to the Feather River in the Sutter Bypass, allowing us to estimate growth and survival and Dos Rios Ranch Preserve, located at rates upon drainage. Thereafter, we allowed the confluence of the Tuolumne and San salmon to leave fields if they chose to do Joaquin rivers (Figure 1). At each site we so. We outfitted field drains with a plastic installed three 6-m � 12-m enclosures. We mesh live-car trap, where we captured and measured, weighed, and placed 40 juvenile measured emigrating individuals before they salmon in each enclosure. After 3 weeks, we proceeded downstream. extracted, re-measured, and re-weighed fish, and analyzed gut contents to compare diets 2017: Lower Trophic Monitoring During across sites. We collected zooplankton and Extensive Flooding drift invertebrate samples at each location to In 2016, the attempt to test a “flood characterize the overall food web response to extension” concept was unsuccessful field inundation. because inundation occurred late in the season, resulting in unsuitably warm water 2016: Test of Feasibility of Using Agricultural temperatures for juvenile salmon in our Fields to Extend Natural Flood Events, experimental fields. We therefore made a Comparison of Different Habitats (flooded second attempt to conduct a flood extension Yolo Bypass, Tule Canal, Sacramento River), pilot in 2017 at Knaggs Ranch, Conaway Difference in Methodologies of Draining Ranch, and Swanston Ranch, and at a new Fields site in the Yolo Bypass Wildlife Area (YBWA) As an engineered floodplain, the Yolo Bypass located south of Interstate 80 between the is designed to drain efficiently. During cities of Davis and Sacramento (Figure 1). moderate inundation events, availability of Field infrastructure was identical to 2016, floodplain habitat can be brief—persisting with the YBWA utilizing flashboards to hold for a week or less. In 2016, our focus was water in similar fashion to Conaway and to test the feasibility of using agricultural Swanston ranches. As we describe below, infrastructure to extend the duration of a high flows made it infeasible to complete small to moderate flood event, increasing the the flood extension work, although we length of time flooded habitat was available were still able to conduct water-quality to fish. We called this idea “flood extension.” and food-web sampling, along with the use We planned similar studies in other study of experimental cages to evaluate salmon years, but extreme weather events prevented growth comparatively across experimental implementation (see below). Landowner sites. partners in the Yolo Bypass at Knaggs Ranch, Conaway Ranch, and Swanston Ranch Lessons Learned agreed to maintain shallow inundation for 3 to 4 weeks in a designated experimental Flooded Farm Fields Support High Biological field after a natural flood event. At Knaggs Productivity Ranch, the landowner made modest to Previous research has shown that inundated extensive modifications to the drainage Yolo Bypass floodplain habitat typically infrastructure to allow more control (using has substantially higher densities of a screw-type valve) over the drainage rate phytoplankton, zooplankton, and drift from the inundated field to the Toe Drain. At invertebrates than the adjacent Sacramento Swanston and Conaway ranches, inundation River across a suite of water year types was maintained with flashboards, which (Sommer et al. 2001a, 2001b; 2004; Goertler could be removed once it was time to drain et al. 2018). Our studies consistently showed

8 SEPTEMBER 2020

that managed inundation of agricultural When cages were used, salmon were PIT- fields supported statistically higher levels tagged to track individual fish growth rates of phytoplankton and invertebrates than within a specific habitat. We consistently the Sacramento River (Corline et al. 2017). found that salmon growth rates in cages Also notable was that phytoplankton placed in flooded in rice fields were higher and zooplankton densities in our flooded than growth rates for juvenile Chinook experimental fields in Yolo Bypass were Salmon of comparative life stage in any of higher than those measured during river- the adjacent riverine habitats and in other inundated flood events and in the Toe regions (Katz et al. 2017; Holmes et al., Drain, a perennial tidal channel (Corline et preprint under review). al. 2017; Katz et al. 2017). In addition, the invertebrate community in flooded rice fields Growth rates were also comparatively was completely dominated by zooplankton high when free-swimming salmon were (Corline et al. 2017), particularly Cladocera, introduced into larger-scale, 0.8-ha flooded whereas drift invertebrates such as Diptera agricultural fields. These studies were more were found in higher concentrations in representative than those using cages of how study sites at Conaway Ranch and Dos migrating salmon might use these habitats Rios. Drift invertebrates are often a more under natural flow events. For the multiple substantial part of the food web in natural years that free-swimming salmon were used flood events in Yolo Bypass (Sommer et al. (Table 1), they averaged a mean daily growth 2001a, 2001b, 2004; Benigno and Sommer rate of 0.98 mm d−1. Throughout all study 2009). Nonetheless, zooplankton densities years, caged salmon and free-swimming can be relatively high in Yolo Bypass (as salmon showed very similar growth rates compared to the Sacramento River) during within the same experimental study units, dry seasons and drought years (Frantzich et despite the fact that they likely experienced al. 2018; Goertler et al. 2018). The specific different micro-habitat conditions (e.g., reasons for these differences include longer velocity, prey, cover). This observation residence time and shallower depths in the suggests that our salmon growth results were Yolo Bypass than in adjacent perennial not influenced by cage effects, a well-known river channels (Sommer et al. 2004). Water issue in enclosure studies (Kellison et al. source also may have been important for 2003). quantity and composition of invertebrates, including zooplankton, since all the managed To better understand managed floodplain flooding work was conducted using water processes across the region, in 2015, from Knights Landing Ridge Cut, not the salmon were introduced in fields at a Sacramento River. variety of locations in the Central Valley with various vegetative substrates: Sutter Resulting Fish Growth Was Rapid Bypass (rice), three locations on the Yolo Given the high densities of prey in the Bypass (rice wetland), and Dos Rios Ranch flooded fields, along with the low metabolic at the confluence of the Tuolumne and San costs of maintaining position in a relatively Joaquin rivers (winter wheat). At all of the low-velocity environment, it is not surprising locations, juvenile Chinook Salmon grew at that growth rates of juvenile salmon were rates (range: 0.76–0.89 mm d−1) similar to comparatively high (e.g., Katz et al. 2017). those observed in experiments conducted This result was consistent across approaches at Knaggs Ranch in the Yolo Bypass during used: cages, enclosures open to the substrate, previous study years. These results suggest and free-swimming fish. that multiple geographical regions and substrate types can support high growth rates of juvenile Chinook Salmon.

https://doi.org/10.15447/sfews.2020v18iss3art3 9 SAN FRANCISCO ESTUARY & WATERSHED SCIENCE VOLUME 18, ISSUE 3, ARTICLE 3

Substrate Type Was Less Important for each substrate type (fallow, disced, and rice- Growth than Expected stubble; Figure 3), all receiving water from A key objective of our work in flooded Knights Landing Ridge Cut. Our hypotheses fields was to determine whether substrate for the replicated field study were that there type has a measurable influence on growth would be no differences in water quality, and survival of juvenile Chinook Salmon. invertebrate diversity, juvenile salmon Substrate and vegetation can be an important growth, or juvenile salmon survival among micro-habitat feature for young Chinook substrate types. Salmon (Tabor et al. 2011), so we posited that there could potentially be some difference Zooplankton levels increased over time, but in salmon performance across treatments. did not show a consistent effect of substrate In 2013, we examined this question across type on density or community structure different substrate types in two ways: (1) (Corline et al. 2017). However, the disced telemetry studies using PIT tags; and (2) treatment had lower average Daphnia pulex replicated fields. Both approaches indicated abundance in comparison to the fallow and that juvenile salmon did not have a clear stubble treatments, during the fifth week of preference for different substrates, and grew the experiment. Densities over all substrate at similar rates across substrates. treatments were much higher compared to open water or river habitats (Sommer et al. 2004). Telemetry We monitored the movements and use of Fish from the fallow fields grew a little PIT-tagged, hatchery-origin juvenile Chinook more slowly than fish from disced and Salmon for approximately 1 month over stubble fields, but the differences were not fallowed, disced, and rice stubble substrates statistically significant (Figure 4). Fish from in two circular enclosures to determine if all treatments grew at very high rates. there was any preferential use. One enclosure 100 included all three substrates, and one Disc Fallow contained only disced substrate (see Conrad 90 Stubble et al. 2016 for further detail). Juvenile salmon did not exhibit a habitat preference mm )

( 80

among these three choices. Although growth h t rates were slightly higher in the enclosure g Le n

that contained all three substrate types, r 70 o juvenile salmon grew at very high rates, F averaging 1.1 mm/d regardless of enclosure. These growth rates were higher than other 60 published growth rates for juvenile Chinook Salmon in the Yolo Bypass, and the region 50 generally (Sommer et al. 2001b). 0 3 4 5 6 Sample Week

Figure 4 Mean fork length (± standard error) of free- Replicated Fields swimming juvenile Chinook Salmon across substrate types We also observed comparably high over time. juvenile Chinook Salmon growth rates across treatments in large replicated field experiments. The 2013 study involved placing juvenile Chinook Salmon in nine fields, which contained three replicates of Across all fields, the average growth rate

10 SEPTEMBER 2020

was 0.93 mm/d, which is high compared to ranged from 0.0% to 29.3% (7.4% average) other published estimates from river channel in the replicated fields containing different habitats (e.g., Sommer et al. 2001b; Katz et al. agricultural substrates. This variability was 2017). likely unrelated to substrate type; instead, these low survival rates were most likely Flow and Connectivity are Critical a result of very dry conditions across Yolo Features of Juvenile Salmon Habitat on Bypass and the Central Valley, generally, Flooded Agricultural Fields when record drought conditions prevailed during 2012–2015, which affected water Behavior quantity and quality. In 2013, our replicated Throughout the 2012–2016 study period, field study likely presented one of the only we consistently observed that juvenile wetted floodplain areas for miles around, and Chinook Salmon were attracted to sources thus presented a prime feeding opportunity of inflow, and that this sometimes became for avian predators such as cormorants, the dominant factor in the distribution herons, and egrets. However, when the same of salmon on experimental fields or in set of fields was used in 2016, survival enclosures. In the previously described PIT- was much higher (ranging from 53.8% to tag observations in 2013, salmon in both 71.1%). This was generally a wetter period, enclosures positioned themselves nearest the avian predation pressure was reduced, and inflow, regardless of surrounding habitat we more efficiently opened the flash boards structure (Conrad et al. 2016). This result is (simple rice field drainage structures) to not surprising, given that juvenile stream facilitate faster drainage and fish emigration. salmonids commonly adopt and defend flow- Note, however, there were some differences oriented positions in stream environments in methodology among years, which may (e.g., Fausch 1984) for acquisition of drifting have contributed to survival variability. food resources. On flooded agricultural Taken together, these observations of free- fields, this orientation toward flow may not swimming salmon survival suggest that only be related to feeding behavior but may field drainage rate, and overall floodplain also serve to keep juvenile salmon in habitat habitat availability, are important factors for areas that are hydrologically connected and improving survival in managed agricultural have higher velocities. In fact, analyses floodplain habitats. of the environmental factors that predict movement of large groups of tagged juvenile Our observations of juvenile salmon Chinook Salmon in the Yolo Bypass found orientation to flow, and the importance of that drainage of flooded areas was a reliable efficient drainage on survival, reinforce predictor of fish emigration at downstream observations from natural floodplains that trapping stations (Takata et al. 2017). connectivity between perennial channel habitat and seasonal floodplain habitat is an essential attribute of river-floodplain systems Survival (Heiler et al. 1995). Connectivity of managed Although juvenile Chinook Salmon growth floodplain habitats to unmanaged habitats rates were consistently high across substrates in the river and floodplain is therefore a and study years, we observed highly variable foundational condition needed to allow survival of salmon, and available evidence volitional migration of juvenile salmon. from the studies suggests that this was Further research is needed to identify how to related, at least in part, to differences among provide sufficient connectivity to maximize years in drainage rates of the study fields rearing and migration opportunities for wild and habitat availability on the floodplain Chinook Salmon. at large. For example, survival in 2013

https://doi.org/10.15447/sfews.2020v18iss3art3 11 SAN FRANCISCO ESTUARY & WATERSHED SCIENCE VOLUME 18, ISSUE 3, ARTICLE 3

These Types of Studies Generate habitat was challenging. For example, we Numerous Challenges had hoped to test the idea of using rice Natural and managed floodplain habitat field infrastructure to extend the duration is subject to a variety of flow and of Yolo Bypass inundation events in an environmental conditions. Variation in attempt to approximate the longer-duration flow and temperature dictates when and events of more natural floodplains; that where managed agricultural habitats may be is, through flood extension. As noted by accessible and suitable for rearing salmonids, Takata et al. (2017), use of the Yolo Bypass with challenges during both wet and dry by wild Chinook Salmon is strongly tied years, as well as during warm periods. As to hydrology, and salmon quickly leave noted previously, survival in the replicated river-inundated floodplains once drainage fields was variable but generally low. We begins. We therefore reasoned that flooded associate these results with the effects of rice fields might provide an opportunity to extreme drought conditions that prevailed extend the duration of flooding beyond the during the core of our study from 2012 typical Yolo Bypass hydrograph. In 2015, through 2015. Although our field studies a flood extension study was planned but were conducted during a time of year not conducted because drought conditions when wild salmon have historically used precluded Sacramento River inflow via the Yolo Bypass floodplain (Takata et al. Fremont Weir. To test the flood extension 2017; Goertler et al. 2018), the extreme concept in 2016, we needed substantial drought made for warm water temperatures, landowner cooperation and assistance to and resulted in our study site being one install draining structures that allowed of the few inundated wetland locations maintenance of local flooding after high in the region. As such, avian predators flow events. Even then, we found it difficult were attracted to the experimental fields, to maintain water levels and field integrity exacerbating salmon mortality during during the tests. In our case, we were drainage. We observed high concentrations fortunate to have the cooperation of willing of cormorants, herons, and egrets on the landowners. Partnership with landowners experimental fields, and this concentration was key, and would be critical with any increased over the study period. As many future efforts to test the concept of flood as 51 wading birds (great blue heron, snowy extension. We also planned a similar test in and great egrets) and 23 cormorants were 2017, but high and long-duration flood flows noted during a single survey. The small prevented the study from occurring. Over scale of our project could have further the 6 years of study, except perhaps for 2013 exacerbated predation issues. This situation when we focused on other study priorities, highlights the importance of the weather- we never experienced ideal conditions to dependent, regional context of environmental adequately test the flood extension concept. conditions, which govern how and when We were either in a severe drought, during managed floodplains can be beneficial which the Yolo Bypass did not flood from the rearing habitats for juvenile salmon. Under river, or we experienced severe and sustained certain circumstances, flooded fields can flooding, which made it impossible to contain generate high salmon growth but in other flood waters within study fields. scenarios, these habitats can provide poor environmental conditions (e.g., exceed Based on these experiences, studying the thermal tolerances) for salmonids and/or concept of flood extension appears to become predation hot spots. depend on the occurrence of moderate flood events at the right time of year (December Even during wetter conditions, we found through March or April), provided fields are that management of agricultural floodplain appropriately designed to hold water and

12 SEPTEMBER 2020

allow efficient immigration and emigration transport staff and vehicles to move salmon of potentially large numbers of juvenile to our study site. salmon. However, significant outreach and communication is necessary with landowners While we were able to assess many important to maintain floodwaters on their fields biological metrics in our study, direct during the natural drainage period. Because measurement of the population-level effect these events cannot be predicted well of floodplain rearing on agricultural habitats ahead of time, these communications—and proved elusive. A traditional approach to availability of robust infrastructure—need addressing this question involves inserting to be constantly maintained even outside CWTs into very large numbers (e.g., 100,000s) the flood extension period. As suggested of experimental salmon and estimating the in the previous section, such potential population response from expanded CWT actions would need to be taken in a way recaptures in the ocean fisheries. Recoveries that maintains hydrologic connectivity of CWTs in adult salmon from experimental and salmon access, so that salmon can releases made in the Yolo Bypass have successfully locate potential managed generally been very low (Takata et al. 2017), habitats, use them for rearing, and then making it difficult to get a high level of successfully emigrate from them at the resolution with which to reliably compare appropriate time. Timing of such potential survival rates, including with values in the manipulations is critical because previous literature. Although CWT recoveries could sampling has shown that salmon quickly potentially be improved by increasing the emigrate from the floodplain during large- number of tagged salmon, the effort required scale drainage events (Sommer et al. 2005), even to collect a single data point would be leaving relatively low densities of salmon in substantial and is limited by the availability remaining ponded areas to potentially benefit of surplus hatchery salmon. A related issue from flood extension. is that it is difficult to design a survival experiment that provides a useful comparison Although our use of hatchery salmon gave us to other management strategies or migration more experimental options during drought corridors. For example, it is challenging to conditions, the use of these fish resulted in assess the incremental survival value of additional challenges. Our approach relied flooded agricultural habitat versus adjacent on a non-traditional use of hatchery salmon, perennial channels (i.e., Sacramento River, which required a suite of permits and Yolo Bypass Tule Canal). Telemetry can approvals to execute the project. As noted partially address this issue, but current above, the project coincided with a major acoustic tagging technology does not allow drought, so access to hatchery salmon was estimates of survival once smolts emigrate limited as a result of low salmon population from the estuary, and is also limited in the levels. In addition, use of hatchery salmon size of salmon that can be tagged. Ultimately, affected the time-period when we could addressing the question of population-level conduct experimental work. We were unable effects will likely depend on a combination to test salmon response to early season of measured field data, incorporating new flooding (e.g., January), because the hatchery methodologies for assessing survival to salmon were too small to receive coded-wire adulthood (e.g., isotopic and genetic tracers), tags (CWTs) as required under our permit and fish population models. conditions. Similarly, the timing of our work was affected by the availability of holding tanks at our partner hatchery (Feather River Fish Hatchery), and by the availability of

https://doi.org/10.15447/sfews.2020v18iss3art3 13 SAN FRANCISCO ESTUARY & WATERSHED SCIENCE VOLUME 18, ISSUE 3, ARTICLE 3

Infrastructure Improvements Could Hence, potential use of flooded agricultural Substantially Increase Research and fields as juvenile salmon rearing habitat Management Options should be evaluated in light of both a Our observations must be considered within modified hydrology and local land use the constraints of the infrastructure in the and infrastructure changes. Additional Yolo Bypass during the time of our field research is needed to address the efficacy work. In the intervening years since our and suitability of different potential water field studies, there has been a substantial sources, hydrology timing, connectivity with amount of progress in improving Yolo the Sacramento River, and related issues, Bypass infrastructure to support native such as the effects of operations on land use fishes. During 2017-2018, an inflatable and other species or life stages (e.g., upstream dam fish barrier and collection facility was migrating adult salmon). constructed at Wallace Weir at Knights Landing Ridge Cut (Figure 1). This facility can enhance potential water distribution CONCLUSIONS options for managed flooding studies under Based on over 2 decades of research, relatively low flow conditions when only California now has some of the best-studied Yolo Bypass tributary flows are available, temperate floodplains in the world. Work including sources from Colusa Basin, which conducted in the lowland floodplains of may not always have suitable water quality California’s Central Valley complements a for juvenile salmonids. It is important to growing body of literature from western note, however, that these local water sources North America that off-channel seasonal are not useful unless there is improved areas are important rearing habitat for connectivity with the Sacramento River, several species of salmonids (e.g., Swales allowing wild juvenile salmon to access and Levings 1989; Ogston et al. 2015). The seasonal habitat throughout the Yolo Bypass. programs of study in the Yolo Bypass and the To that end, the joint Environmental Impact Cosumnes River (e.g., Sommer et al. 2001a, Statement/Report was finalized in 2019 for 2001b; Jeffres et al. 2008) have demonstrated a project that will improve connectivity the ecological value of floodplain habitat between the Sacramento River and Yolo to Central Valley Chinook Salmon, and Bypass with a proposed notch in Fremont have contributed to California policies Weir (USBR and CDWR 2019). This proposed focused on multi-purpose projects that facility would allow managed flows at integrate flood-risk management objectives lower Sacramento River stages than the with the restoration of seasonal floodplain current weir structure, thereby increasing habitat (CDWR 2017). The Yolo Bypass is a the frequency and duration of seasonal particularly interesting example because it inundation, and providing improved access shows how flood management potentially to the floodplain from the Sacramento can be integrated with agriculture, wildlife, River fish migration corridor. This upgrade recreation, and fisheries, thereby balancing is required as a condition of the 2009 the interests of multiple stake-holder groups Biological Opinion for Salmonids for long- (Sommer et al. 2001a; Opperman 2012; term operation of the federal and state water Suddeth et al. 2016). Previous research on the projects (NMFS 2009). Yolo Bypass during river inundation events has consistently documented how access Our study did not specifically address to seasonal floodplain increases rearing these new facilities or their operations, and habitat and food availability, leading to how the concept of managed agricultural enhanced growth and life-history diversity floodplain habitat can be integrated into the of juvenile Chinook Salmon. The same primary purposes of these improvements. research has shown that the duration of

14 SEPTEMBER 2020

floodplain inundation is constrained by the As described in previous sections, flooded system’s design, so young salmon can be agricultural fields generate abundant food forced off the floodplain before they accrue resources that support high juvenile salmon maximum benefits, which is now being growth rates, but there were multiple addressed by the Fremont Weir Salmonid challenges (e.g., high water temperatures, Habitat Restoration and Fish Passage Project variable hydrology, predation, connectivity, (USBR and CDWR 2019). Hence, survival infrastructure, and fish entrainment) that of young salmon in the Yolo Bypass can be limited salmon survival and study execution. relatively high (Sommer et al. 2001a, 2001b; These observations suggest several targeted Takata et al. 2017; Johnston et al. 2018), but strategies aimed at improving the suitability anthropogenic features limit rearing benefits of managed agricultural habitat for salmon and create some stranding risks (Sommer et rearing. al. 2005; USBR and CDWR 2019). 1. Temperature problems may be reduced In this context, our work on agricultural by avoiding managed inundation of fields examined whether there were ways agricultural fields during warm spring that agricultural habitats—one of the periods. This does not mean that spring key anthropogenic features in the Yolo inundation is always undesirable, because Bypass and in other developed large-river some years have cool March and April floodplains globally—could be improved for periods, when habitat conditions could use by juvenile Chinook Salmon. As noted be highly suitable for juvenile salmon in the Introduction, our research effort rearing. was not intended as a proof of concept. Instead, we sought to better understand 2. Targeted inundation is questionable several critical habitat features in seasonally in drought situations when flooded inundated agricultural fields, and to identify fields appear more likely to draw avian some of the challenges that would need to predators, since shallow water foraging be addressed in future research and pilot habitat is limited elsewhere. Maintaining implementation projects. Furthermore, our deeper areas (e.g., 1 m or more) may help project should not be considered as advocacy to provide juvenile salmon with predator for managed agricultural habitats as a refugia during any water year type. substitute for more natural, unmanaged floodplain restoration projects. One of 3. Fields designed as fish habitat need our goals was to identify ways in which to have sufficient structural integrity managed flooding on agricultural fields in to support deeper water and flow. floodplains could potentially help mitigate Our observations suggest that typical for typically short inundation periods by production rice fields do not meet these extending natural flood events, allowing criteria. juvenile salmon to realize floodplain habitat benefits. We therefore still consider targeted 4. Finally, and most important, managed inundation of agricultural fields to be an fields would need to have high experimental approach that may in the connectivity with adjacent channels for future be complementary to demonstrated juvenile salmon to easily move on and salmon restoration tools, such as creation off the seasonal habitat under their own of natural habitats (e.g., floodplain, tidal volition. Simply put, the flooded fields wetlands, and riparian habitat), gravel are not viable rearing habitat unless wild rehabilitation, removal of migration barriers, juvenile Chinook Salmon have some flow supplementation, and water temperature way to find these locations, are able to management. find predator and temperature refugia as

https://doi.org/10.15447/sfews.2020v18iss3art3 15 SAN FRANCISCO ESTUARY & WATERSHED SCIENCE VOLUME 18, ISSUE 3, ARTICLE 3

necessary, and then safely emigrate by of managed extension of natural flood events the time water draw-down occurs. on agricultural fields is supported with evidence of not only high growth rates but As noted previously, fish entry into these also high rates of wild salmon entrainment, managed agricultural floodplain habitats suitable thermal conditions, and high in the Yolo Bypass could be facilitated survival rates of juvenile salmon, the next through natural flood events or perhaps new step would then be to move past the research infrastructure (e.g., Fremont Weir notch; and pilot stage to large-scale implementation, USBR and CDWR 2019). However, extension which would be a major effort. Developing of natural flood events remains an unproven a program in which extension of natural concept; moderate flood events timed in the flood events is one of the tools available to middle of the winter-spring rainy season managers to provide quality rearing habitat were uncommon during our study, and to juvenile salmon will require significant structural modifications to address needs investment in resources for coordination for enhanced flow connectivity and depth between agencies and landowners, scientific were not in place when these moderate flood support to provide for continued adaptive events did occur. Continued research on the management, and monetary investments potential value of extended natural flood in infrastructure in flood bypasses and events on agricultural habitat will need easements. these structural modifications (e.g., Fremont Weir notch) in place to adequately test this concept. A related issue is that managed ACKNOWLEDGEMENTS fields need to be located close to channels This research was supported by funding that are viable migration corridors. Long from the California Department of Water drainage ditch networks may not be viable Resources, US Bureau of Reclamation, and emigration corridors since they could have Metropolitan Water District, and by the high densities of predators and low water oversight of the Interagency Ecological quality. Contributing to the development Program. The effort would not have been of “predator hot spots” as juvenile salmon possible without the assistance of numerous emigrate from flooded agricultural fields staff from the California Department of would potentially negate the growth Water Resources, California Department of advantage that salmon can gain from these Fish and Wildlife, UC Davis, NOAA Fisheries, habitats. Future research should include Conaway Preservation Group, Swanston modeling that weighs the growth advantages Ranch, CalTrout, Pacific States Marine that flooded agricultural fields can confer Fisheries Commission, and the US Bureau against the focused predation risk that may of Reclamation. The Feather River Hatchery develop when large numbers of juvenile staff were essential to the study because they salmon leave the fields synchronously during drawdown.

As a next step, we advocate continued research and pilot projects to address the key issues and remaining scientific uncertainties we identified in our studies. Continued research will require active partnership with local landowners to achieve structural modifications on fields with prime locations near major migration corridors, before additional experimental work. If the concept

16 SEPTEMBER 2020

took measures outside of typical hatchery [CDWR] California Department of Water Resources. operations to provide experimental fish. We 2017. Central Valley Flood Protection Plan 2017 also owe particular thanks to John Brennan, Update. August 2017. [accessed 2020 Sep 03]. who provided expert guidance and access Sacramento (CA): CDWR. 210 p. Available from: to agricultural lands throughout the project. https://water.ca.gov/-/media/DWR-Website/Web- The manuscript was improved by comments Pages/Programs/Flood-Management/Flood-Planning- from Jeff Opperman and an anonymous and-Studies/Central-Valley-Flood-Protection-Plan/ reviewer. Files/2017-CVFPP-Update-FINAL_a_y19.pdf?la=en& hash=BCCCF94FABFBEF7AC0B7B9A9195F0148EBB AD20C REFERENCES Conrad JL, Holmes E, Jeffres C, Takata L, Ikemiyagi N, Benigno GM, Sommer TR. 2009. Just add water: Katz J, Sommer T. 2016. Application of passive sources of chironomid drift in a large river integrated transponder technology to juvenile floodplain. Hydrobiol. [accessed 2020 Jul salmon habitat use on an experimental agricultural 16];600:297-305. floodplain. N Am J Fish Manag. [accessed 2020 Jul https://doi.org/10.1007/s10750-007-9239-2 16];36:30–39. California Trout, UC Davis Center for Watershed https://doi.org/10.1080/02755947.2015.1111276 Sciences, California Department of Water Resources. Corline NJ, Sommer T, Jeffres CA, Katz J. 2017. 2014. The experimental agricultural floodplain Zooplankton ecology and trophic resources for habitat investigation at Knaggs Ranch on Yolo rearing native fish on an agricultural floodplain Bypass 2013-2014. [San Francisco (CA)]: US Dept. of in the Yolo Bypass California, USA. Wetlands Ecol the Interior, Bureau of Reclamation. 57 p. Manag. [accessed 2020 Jul 16];25:533–545 (2017). California Trout, UC Davis Center for Watershed https://doi.org/10.1007/s11273-017-9534-2 Sciences, California Department of Water Resources. Cornwell R, Katz J. 2017. How fish and farms can 2015. The effects of flow on volitional out-migration both survive in California. San Francisco Chronicle. of juvenile Chinook Salmon from Yolo Bypass rice October 25, 2017. [accessed 2020 Sep 03]. Available fields managed as agricultural floodplain habitat: a from: https://www.sfchronicle.com/opinion/ report of the Experimental Agricultural Floodplain openforum/article/How-fish-and-farms-can-both- Habitat Investigation. [San Francisco (CA)]: US Dept. survive-in-California-12306746.php of the Interior, Bureau of Reclamation. 32 p. Davis MJ, Woo I, Ellings CS, Hodgson S, California Trout, UC Davis Center for Watershed Beauchamp DA, Nakai G, De La Cruz SE. 2018. Sciences, California Department of Water Resources. Integrated diet analyses reveal contrasting trophic 2016. Report of the experimental agricultural niches for wild and hatchery juvenile Chinook floodplain habitat investigation 2015-2016. [San Salmon in a large river delta. Trans Am Fish Soc. Francisco (CA)]: US Dept. of the Interior, Bureau of [accessed 2020 Jul 16];147:818-841. Reclamation. 44 p. https://doi.org/10.1002/tafs.10088 [CDA] California Department of Agriculture. 2018. Dybala KE, Reiter ME, Hickey CM, Shuford WD, California agricultural exports 2017-2018. Available Strum KM, Yarris GS. 2017. A Bioenergetics from: https://www.cdfa.ca.gov/statistics/PDFs/2017- approach to setting conservation objectives for non- 18AgExports.pdf breeding shorebirds in California’s Central Valley. San Franc Estuary Watershed Sci. [accessed 2020 Jul 16];15(1). https://doi.org/10.15447/sfews.2017v15iss1art2 Fausch KD. 1984. Profitable stream positions for salmonids: relating specific growth rate to net energy gain. Can J Zool. [accessed 2020 Jul 16];1984, 62:441-451. http://doi.org/10.1139/z84-067

https://doi.org/10.15447/sfews.2020v18iss3art3 17 SAN FRANCISCO ESTUARY & WATERSHED SCIENCE VOLUME 18, ISSUE 3, ARTICLE 3

Feyrer F, Sommer T, Harrell W. 2006a. Managing Jeffres CA, Opperman JJ, Moyle PB. 2008. Ephemeral floodplain inundation for native fish: production floodplain habitats provide best growth conditions dynamics of age-0 splittail in California’s Yolo for juvenile Chinook Salmon in a California river. Bypass. Hydrobiol. [accessed 2020 Jul 16];573:213- Env Biol Fishes. [accessed 2020 Jul 16];83(4):449– 226 https://doi.org/10.1007/s10750-006-0273-2 458. https://doi.org/10.1007/s10641-008-9367-1 Feyrer F, Sommer T, Harrell W. 2006b. Importance of Johnston ME, Steel AE, Espe M, Sommer T, flood dynamics versus intrinsic physical habitat in Klimley AP, Sandstrom P, Smith D. 2018. Survival structuring fish communities: evidence from two of juvenile Chinook Salmon in the Yolo Bypass and adjacent engineered floodplains on the Sacramento the lower Sacramento River, California. San Franc River, California. N Am J Fish Manag. [accessed Estuary Watershed Sci. [accessed 2020 Jul 16];16(2) 2020 Jul 16];26:408-417. https://doi.org/10.15447/sfews.2018v16iss2art4 https://doi.org/10.1577/M05-113.1 Johnston ME, Frantzich F, Espe MB, Goertler P, Frantzich J, Sommer T, Schreier B. 2018. Physical and Singer G, Sommer T, Klimley AP. 2020. Contrasting biological responses to flow in a tidal freshwater the migratory behavior and stranding risk of slough complex. San Franc Estuary Watershed Sci. White Sturgeon and Chinook Salmon in a modified [accessed 2020 Jul 16];16(1). floodplain of California. Env Biol Fish. [accessed https://doi.org/10.15447/sfews.2018v16iss1/art3 2020 Jul 16];103:481–493 (2020). Gardali T, Marty JT, Yarris GS. 2017. The science https://doi.org/10.1007/s10641-020-00974-9 of setting conservation objectives for birds in Katz JVE, Jeffres C, Conrad JL, Sommer TR, California’s Central Valley: an introduction. San Martinez J, Brumbaugh S, Corline N, Moyle PB. Franc Estuary Watershed Sci. [accessed 2020 Jul 2017. Floodplain farm fields provide novel rearing 16];15(1). habitat for Chinook salmon. PLOS ONE. [accessed https://doi.org/10.15447/sfews.2017v15iss1art1 2020 Jul 16];12(6):e0177409. Goertler P, Jones K, Cordell J, Schreier B, Sommer T. https://doi.org/10.1371/journal.pone.0177409 2018. Effects of extreme hydrologic regimes on Kellison GT, Eggleston DB, Taylor JC, Burke JS. 2003. juvenile Chinook Salmon prey resources and diet An assessment of biases associated with caging, composition in a large-river floodplain. Trans Am tethering, and trawl sampling of summer flounder Fish Soc. [accessed 2020 Jul 16];147:287-299. (Paralichthys dentatus). Estuaries. [accessed 2020 Jul https://doi.org/10.1002/tafs.10028 16];26:64–71. https://doi.org/10.1007/BF02691694 Heiler G, Hein T, Schiemer F, Bornette G. 1995. Mount JF. 1995. California’s rivers and streams: the Hydrological connectivity and flood pulses as the conflict between fluvial process and land use. central aspects for the integrity of a river-floodplain Berkeley (CA): University of California Press). 359 p. system. River Res Appl. [accessed 2020 Jul 16]; Moyle PB, Katz JVE, Quiñones RM. 2011. Rapid 11:351-361. https://doi.org/10.1002/rrr.3450110309 decline of California’s native inland fishes: a Herbold B, Carlson SM, Henery R, Johnson RC, status assessment. Biol Conserv. [accessed 2020 Jul Mantua N, McClure M, Moyle PB, Sommer T 2018. 16];144(10):2414–2423. Managing for salmon resilience in California’s https://doi.org/10.1016/j.biocon.2011.06.002 variable and changing climate. San Franc Estuary [NMFS] National Marine Fisheries Service. 2009. Watershed Sci. [accessed 2020 Jul 16];16(2). Biological opinion and conference opinion on the https://doi.org/10.15447/sfews.2018v16iss2art3 long-term operations of the Central Valley Project Holmes EJ, Saffarinia P, Rypel AL, Bell-Tilcock MN, and State Water Project. Long Beach (CA): NMFS, Katz JV, Jeffres CA. Reconciling fish and farms: Southwest Region. Endangered Species Act Section Methods for managing California rice fields as 7 Consultation. [accessed 2020 Sep 02]. 845 p. salmon habitat. bioRxiv. Preprint under review. Available from: https://nrm.dfg.ca.gov/FileHandler. https://doi.org/10.1101/2020.08.03.234054 ashx?DocumentID=21473

18 SEPTEMBER 2020

Ogston L, Gidora S, Foy M, Rosenfeld J. 2015. Sommer T, Harrell WC, Feyrer F. 2014. Large-bodied Watershed-scale effectiveness of floodplain habitat fish migration and residency in a flood basin of restoration for juvenile Coho Salmon in the the Sacramento River, California, USA. Ecol Freshw Chilliwack River, British Columbia. Can J Fish Aq Fish. [accessed 2020 Jul 16];2014: 23:414–423 Sci. [accessed 2020 Jul 16];72:479-490. ht tps://doi.org/10.1111/ef f.12095 https://doi.org/10.1139/cjfas-2014-0189 Strum KM, Dybala KE, Iglecia MN, Shuford WD. Opperman JJ. 2012. A conceptual model for 2017. Population and habitat objectives for breeding floodplains in the Sacramento-San Joaquin Delta. shorebirds in California’s Central Valley. San Franc San Franc Estuary Watershed Sci. Estuary Watershed Sci. [accessed 2020 Jul 16];15(1). [accessed 2020 Jul 16];10(3). https://doi.org/10.15447/sfews.2017v15iss1art3 https://doi.org/10.15447/sfews.2012v10iss3art4 Suddeth Grimm R, Lund J. 2016. Multi-purpose Shuford WD, Dybala KE. 2017. Conservation objectives optimization for reconciliation ecology on an for wintering and breeding waterbirds in California’s engineered floodplain: Yolo Bypass, California. San Central Valley. San Franc Estuary Watershed Sci. Franc Estuary Watershed Sci. [accessed 2020 Jul [accessed 2020 Jul 16]; 15(1). 16];14(1). https://doi.org/10.15447/sfews.2017v15iss1art4 https://doi.org/10.15447/sfews.2016v14iss1art5 Sommer T, Baxter R, Herbold B. 1997. The Swales S, Levings CD. 1989. Role of off-channel ponds resilience of splittail in the Sacramento- in the life cycle of Coho Salmon (Oncorhynchus San Joaquin Estuary. Trans Am Fish Soc. kisutch) and other juvenile salmonids in the [accessed 2020 Jul 16];126:961-976. https://doi. Coldwater River, British Columbia. Can J Fish Aquat org/10.1577/1548-8659(1997)126<0961:ROSITS>2.3. Sci. [accessed 2020 Jul 16];46:232–242 CO;2 https://doi.org/10.1139/f89-032 Sommer TR, Harrell WC, Nobriga M, Brown R, Tabor RA, Fresh KL, Piaskowski RM, Gearns HA, Moyle PB, Kimmerer WJ, Schemel L. 2001a. Hayes DB. 2011. Habitat use by juvenile Chinook California’s Yolo Bypass: evidence that flood Salmon in the nearshore areas of Lake Washington: control can be compatible with fish, wetlands, effects of depth, lakeshore development, substrate, wildlife and agriculture. Fisheries. [accessed 2020 and vegetation. N Am J Fish Manag. [accessed 2020 Jul 16];26(8):6-16. https://doi.org/10.1577/1548- Jul 16]; 31:700-713. 8446(2001)026%3C0006:CYB%3E2.0. https://doi.org/10.1080/02755947.2011.611424 CO;2 Takata L, Sommer TR, Conrad JL, Schreier BM. Sommer TR, Nobriga ML, Harrell WC, Batham W, 2017. Rearing and migration of juvenile Chinook Kimmerer WJ. 2001b. Floodplain rearing of juvenile Salmon (Oncorhynchus tshawytscha) in a large river Chinook Salmon: evidence of enhanced growth and floodplain. Env Biol Fishes. [accessed 2020 Jul survival. Can J Fish Aquat Sci. [accessed 2020 Jul 16];100:1105–1120. 16];58(2):325-333. https://doi.org/10.1139/f00-245 https://doi.org/10.1007/s10641-017-0631-0 Sommer TR, Harrell WC, Mueller-Solger B, Tom B, [USBR and CDWR] US Bureau of Reclamation Kimmerer W. 2004. Effects of flow variation on and California Department of Water Resources. channel and floodplain biota and habitats of the 2019. Yolo Bypass Salmonid Habitat Restoration Sacramento River, California, USA. Aquat Cons: Mar and Fish Passage Project Environmental Impact Freshw Ecosys. [accessed 2020 Jul 16];14:247-261. Report. [Sacramento (CA)]: US Dept. of the Interior, https://doi.org/10.1002/aqc.620 Bureau of Reclamation. Available from: https:// Sommer T, Harrell W, Nobriga M. 2005. Habitat use www.usbr.gov/mp/nepa/nepa_project_details. and stranding risk of juvenile Chinook Salmon on a php?Project_ID=30484 seasonal floodplain. N Am J Fish Manag. [accessed 2020 Jul 16];25:1493-1504. https://doi.org/10.1577/M04-208.1

https://doi.org/10.15447/sfews.2020v18iss3art3 19 SAN FRANCISCO ESTUARY & WATERSHED SCIENCE VOLUME 18, ISSUE 3, ARTICLE 3

Whipple AA, Grossinger RM, Rankin D, Stanford B, Askevold RA. 2012. Sacramento-San Joaquin Delta historical ecology investigation: exploring pattern and process. Richmond (CA): San Francisco Estuary Institute-Aquatic Science Center. A Report of SFEI- ASC’s Historical Ecology Program. Publication #672. Available from: http://www.sfei.org/DeltaHEStudy Winemiller KO, McIntyre PB, Castello L, Fluet- Chouinard E, Giarrizzo T, Nam S, Baird IG, Darwall W, Lujan NK, Harrison I, et al. 2016. Balancing hydropower and biodiversity in the Amazon, Congo, and Mekong. Science. [accessed 2020 Jul 16];351(6269):128–129. http://doi.org/10.1126/science.aac7082