Ecosystem-Scale Selenium Modeling in Support of Fish and Wildlife Criteria Development for the -Delta Estuary, Administrative Report

Figures 1 through 12

U.S. Department of the Interior U.S. Geological Survey December, 2010

Find the full report and other attachments at http://www.epa.gov/region9/water/ctr 122°30’ 122°00’ 121°30’

Oil refineries Pumping plant Sacramento USGS Station 8.1 River Drainage basin Map Freeport location

CALIFORNIA Cache Slough

North Bay Aqueduct Suisun Napa Suisun Slough Montezuma River Slough Creek Nurse Rio Vista Slough 3 Mile Slough Honker Venice Island San Bay Big Break Mandeville Island Stockton ship Pablo Franks channel Bay Benicia Martinez Tract 38°00’ Antioch Chipps Contra Costa Stockton Island Contra Costa San Intake pump Smith’s Canal Joaquin Mildred Middle River Canal Central River Island Bay SWPBanks Bridge San CVP Delta Francisco Delta-Mendota Canal South San San Joaquin Francisco River Bay Vernalis

PACIFIC 0 10 km OCEAN 0 10 mi

Figure 1. Map of the San Francisco Bay-Delta Estuary. Suisun Bay and Carquinez Strait Sacramento/ Delta

Food web: C. amurensis Food web: aquatic insect to white sturgeon or to juvenile Chinook salmon diving ducks or steelhead trout

Inflow Refinery Yolo Bypass* (import) Impact of North internal Bay watershed input Se Se loads reduced loads Sacramento River by seasonally subjected watershed input high flows that to annual Se loads vary Sacramento River* accompany limit. with flow season their largest and water-year type. Se load inputs. Riverine Eastside streams* inputs as they mix with

streams seawater North Bay North • North Bay determine Se North Bay Aqueduct concentrations for the Bay. • Contra Costa

Regulated refinery Se loads* County

San Pablo elta Outflow Index = inflo San Joaquin Valley Grassland Bypass Project Net D w-dem • and annual restrictions apply Bay ay Suisun Bay to B w flo Regulated agricultural ut NDOI = [Delta inflow] – O [net Delta consumptive use] drainage Se load – [Delta exports] San Joaquin River San Joaquin River = NDOI watershed input Se loads depend on San Joaquin River recycle amount of river flow Clifton Court Forebay entering the Bay-Delta. Central Valley Project pumps • San Joaquin Delta Mendota Canal Valley

industry Se loads (?)* California Delta-Mendota Aqueduct Canal State Water Project pumps • Los Angeles South California Aqueduct • South Bay Bay

Demand (export) PACIFIC OCEAN

*Privately owned treatment works and other industries also may contribute Se loads

Figure 2. Conceptual details of sources of selenium, site-speci c food webs, and hydrodynamic connections for the Bay-Delta. 12,000 A agricultural Refinery and Agricultural 10,000 Se Loads

8,000

6,000 refinery

4,000 Se loads , lbs/year wet/wet 2,000 wet/wet

Proposed, observed wet/above normal data not available. 0 1985 1990 1995 2000 2005 2010

275,000 1997 B 1998 250,000 Bay-Delta transects November 5-6, 1997 225,000 June 16-17, 1998 October 7-8, 1998 200,000 April 13-14, 1999 November 4-5, 1999 175,000

150,000

125,000 1999 100,000

75,000 NDOI, cfs daily average per month 50,000

25,000

0 2000 2005 2010 Figure 3. Re nery and agricultural selenium loads (1985-2009) (A) and ow conditions (1996-2009) (B). A. Grassland Drainage Area or Grassland Bypass Project Se loads (1986-2008) 12,000

10,000

measured Se load, lbs Measured start of Grassland Bypass Project 8,000 (site B, San Luis Drain at Mud Slough)

6,000 target

Load target, lbs Se Grassland 4,000 Drainage Area target 3,296 lbs Se dry/below normal 2,000 measured 1,239 lbs Se 0 2009 1985 1990 1995 2000 2005 2010 Water year

B. Proposed Grassland Bypass Project Se load targets (2009-2019) 5,000 wet year proposed targets above normal year 4,000

dry/below normal year 3,000

2,000 Target, lbs Se

critically dry year

1,000

0 2008 2010 2012 2014 2016 2018 2020 Year

Figure 4. Grassland Bypass Project selenium loads (1997-2009) (A) and proposed selenium load targets (2009-2019) (B). San Francisco Bay-Delta Selenium Model composite source Se load (oil refinery, agricultural composite volume composite freshwater drainage through San Joaquin (Sacramento River = end-member Se concentration River or proposed San Luis ÷ San Joaquin River) (head of estuary) Drain, Sacramento River)

dissolved selenium phytoplankton, algae, bacteria, suspended particulate Transformation species (selenate, material and bed sediment (elemental Se, particulate partitioning (K ) selenite, organo-Se) d organo-Se, adsorbed selenite/selenate)

Bioaccumulation

Prey clam zooplankton amphipod

Trophic Transfer

sturgeon Predator diving ducks striped bass splittail splittail flounder

Effects

impaired reproduction, teratogenesis, selenosis

[adapted from Figure 2, Presser and Luoma (2006)]

Figure 5. San Francisco Bay-Delta Selenium Model from Presser and Luoma (2006). Ecosystem–Scale Selenium Model Source Volume (hydrodynamics) Receiving- Load and Water Hydrology Dissolved Se concentration speciation Environmental Dissolved Se speciation Partitioning (selenate, selenite, organo-Se) ECOLOGY K d Loss of individuals T N E M E G A N A M K S I R Population decline Transformation Particulate Se concentration Loss of species Particulate Se speciation Community change Bioavailability (phytoplankton, suspended material, biofilm, sediment)

TTFinvertebrate

Invertebrate Bioaccumulation Se concentration EFFECTS TTF TTF fish bird Reproduction in fish (deformity; larva and fry Se toxicity survival and Trophic Fish Wildlife guidelines growth) transfer Se Se concentration concentration Se protective Reproduction in reference dose birds (hatchability; teratogenesis; Processes chick survival and growth) Factors Se measurements [adapted from Figure 1, Presser and Luoma (2010)] Management guidelines Effects and ecology Predator/prey links

Figure 6. General ecosystem-scale selenium model from Presser and Luoma (2010). [Kd = empirically determined environmental partitioning factor between water and particulate material; TTF = biodynamic food web transfer factor between an animal and its food.] A. Quantifying Processes Environmental Foodweb partitioning biodynamics

• site hydrology • assimilation efficiency (AE) • speciation • ingestion rate (IR) • particulate type • efflux (ke)

Cprey = (AE) (IR) (Cparticulate)/ke

Cpredator = (AE) (IR) (Cprey)/(ke) TTF = (AE) (IR)/(k)

Cwater-column C particulate C invertebrate C fish or C bird

K d TTF invertebrate TTF fish TTF bird

C particulate C invertebrate C fish C bird = = = = C water-column C particulate C invertebrate C invertebrate

[adapted from Figure 2, B. Modeling Equations Presser and Luoma (2010)] Cinvertebrate = (Cparticulate) (TTFinvertebrate) Cfish = (Cinvertebrate) (TTFfish) Cbird = (Cinvertebrate) (TTFbird) Combined Equations Cfish = (Cparticulate) (TTFinvertebrate) (TTFfish) Cbird = (Cparticulate) (TTFinvertebrate) (TTFbird) If longer food web Cbird = (Cparticulate) (TTFinvertebrate) (TTFfish) (TTFbird) Cbird = (Cinvertebrate) (TTFfish) (TTFbird)

Translation to Water-Column Concentration

Cwater = (Cfish) ÷ [(TTFfish) (TTFinvertebrate) Kd]

Cwater = (Cbird) ÷ [(TTFbird) (TTFinvertebrate) Kd]

Figure 7. Quantifying processes (A) and modeling equations (B) for ecosystem-scale selenium modeling. Reproductive Effects

Diet exogenous Liver endogenous female adult bird Se source Se source female adult fish Liver endogenous Se source minor prehatch Diet exogenous exposure Se source egg maternally prehatch exposure egg Se mediated Se reproductive yolk-sac larvae effects failure of eggs to hatch embryo reduced production exposure during (most sensitive endpoint) deformity of viable eggs yolk sac adsorption* (hatchability not affected)

Se fry post-hatch chicks reproductive effects mortality deformity reduced growth

hatch into contaminated environment hatch into contaminated environment direct Se effects mortality liver reduced growth reduced growth mortality (survival) disease (growth rate in (little data available, first 5 days is best may be most sensitive juvenile direct Se predictor of post- endpoint) effects hatch survival) Winter Stress Syndrome (metabolic stress imposed over-winter mortality resulting by low temperature during in seasonal poisoning of young fish Se exposure increases toxicity of dietary Se)

* yolk proteins precursor, vitellogenin; yolk proteins lipovitellin and phosvitin (liver Se stored in yolk proteins until metabolized during embryonic and larval stages vitellogenesis = yolk proteins are synthesized in liver and transported to egg)

Chronic Effects in Adults

• compromised body condition (low body mass; edema) • oxidative stress (increased susceptibility to disease due to suppressed immune system) • decreased winter survival • decreased reproductive fitness (decreased breeding propensity; reduced recruitment) • behavioral impairment (missed breeding window; delayed timing of departure) • lowered saline tolerance and gill effects in fish

Figure 8. Ecotoxicology and e ects of selenium for sh and birds. Site-Specific Model Approach A. seaward site-specific fish predicted predicted predicted guidline*wb foodweb   TTFsturgeon  invertebrate  TTFC.amurensis  particulate  Kd  dissolved µg/g dw µg/g dw µg/g dw µg/L guideline

species- species- emperical sturgeon/ specific specific salinity- prey to particulate specific C. amurensis/ Se in Se in food predator Se in prey to prey partitioning Se in suspended predator for prey trophic C. amurensis biodynamic factor for dissolved particulate (sturgeon) (suspended material transfer trophic particulate dissolved and factor transfer material) suspended factor particulate Se

IR calculated from from calculated from measured suspended- calculated from USFWS measured particulate material (or transects of or assumed sturgeon muscle C) from transects and measured from Se; measured clam literature filtration rate; dissolved Se and literature Se (North Bay AE calculated from suspended matched dataset) measured particulate particulate Se for species from transects transects and literature bioavail- ability; ke assumed from literature for C. amurensis

B. landward site-specific fish predicted predicted predicted guidline*wb foodweb   TTFsalmon  invertebrate  TTFaquatic insect  particulate  Kd  dissolved µg/g dw µg/g dw µg/g dw µg/L guideline

species- species- emperical specific specific salinity-

Within specified salinity and residence time Chinook salmon/ prey to particulate Se in food specific insect/ Se in Se in prey predator to prey for prey partitioning Se in suspended predator aquatic trophic biodynamic (suspended factor for dissolved particulate (salmon) insect material transfer trophic particulate dissolved and factor transfer material) suspended factor particulate Se

from generic generic calculated USFWS field-derived field-derived from transects or assumed for fish for aquatic of measured from insect dissolved Se literature and suspended particulate Se for transects

*USEPA, USFWS, or assumed from literature; wb = whole body, dw = dry weight.

Figure 9. Site-speci c modeling approach for Bay-Delta seaward (A) and landward (B) locations. Se in duck tissues decreased body condition disease

A. delayed migration Diving duck Se effects model teratogenesis decreased missed decreased embryo survival breeding breeding and hatchability window propensity

reproductive impairment

reduced recruitment decreased winter survival

metabolized in 10 days spinal deformities egg yolk developing white severe edema sturgeon larvae B. mortality White sturgeon 2 years of deposition Se effects model maternal transport liver disease gill effects Se in adult white juvenile white low growth sturgeon tissues sturgeon reproductive females lowered saline -all seasons- tolerance

Se in salmonid tissues

growth/ C. susceptibility behavioral olfactory to disease development impairment impairment Juvenile salmonids impairment (downstream migration) Se effects model mortality reproductive impairment predation migration failure

reduced recruitment

Se in salmonid tissues Se sequestration in eggs

D. susceptibility olfactory behavioral developmental hatch, fry Salmonids to disease impairment impairment malformations survival (upstream migration) Se effects model mortality migration predation mating, failure (fishing spawning

reduced reproduction [adapted from Figure 7, reduced recruitment Presser et al. (in review)]

Figure 10. Selenium e ects models for diving ducks (A), white sturgeon (B), and juvenile salmonids (C and D). A. Seaward: dissolved Se clam-based food webs* Kd particulate Se

TTFclam = 17 clam (Corbula amurensis) TTFcrab = ? TTFbird egg = 2.6 TTFfish = 1.1 bird % clam in diet* fish % clam in diet* Dungeness crab lesser scaup 96 white sturgeon 41 surf scoter1 86 green sturgeon3 41 larvae juveniles adults greater scaup 81 Sacramento splittail 34 black scoter2 80 TTFfish = 1.1 TTF = ? white-winged scoter2 75 TTFbird egg = 2.6 California clapper rail 64 sturgeon humans fish-eating birds herring halibut sardine TTF = ? TTFbird egg = 2.6 sole TTF = ? TTFbird egg = 2.6 Chinook flounder bald eagle4 23% salmon rockfish 1see additional site-specific data in figure 12 skate seals 2infrequent visitor *USFWS (2008) shark sea lions 3assumed from white sturgeon 4also represents exposure of osprey

B. Landward: dissolved Se insect-based food webs* Kd particulate Se Other Delta food webs:

TTFinsect = 2.8 invertebrate TTF C. fluminea 2.8 aquatic insects copepod 1.35 (midge, corixid, damselfly, mayfly, etc.) amphipod 0.6–0.9 TTFfish = 1.1 mysid 1.3 TTFbird egg = 2.6 Daphnia 1.9 Chinook salmon juvenile fish 1.1–1.6 steelhead trout Siberian shrimp ? California black rail

TTFbird egg = 2.6 * TTFs shown are used in modeling. See text for additional information. Figure 11. Conceptual details of seaward clam-based food webs (A) and landward aquatic insect-based food webs (B) for modeling of the Bay-Delta. 2 4 3 1 Species at Risk: Exposure and Habitat Use Primary location Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov dissolved and particulates 1, 2, 3, 4 high-flow season low-flow season partitioning/transformation C. amurensis Se (µg/g dry wt.) spawning recovery 20 ppm data: 12/1995–11/1996 1, 2, 3 11 migratory birds surf scoter in Central Bay surf scoter in Suisun surf scoter a, b 1, 2, 3 over wintering (staging) and black scoter c white-winged scoter c migration north to breed greater scaup b lesser scaup b 1, 2, 3, 4

clapper rail* 1, 2, 3 breeding

bald eagle (assumed resident) 1, 2, 3 breeding upstream spawning white sturgeon (100 yrs)+ d 2, 3 2 year internal egg maturation

green sturgeon** (50 yrs)+ e 1, 2, 3 upstream spawning upstream spawning Chinook salmon* (3–7 yrs)+ migration upstream winter-run 1, 2, 3, 4 emigration to ocean emigration to ocean juveniles (4) Chinook salmon* (3–7 yrs)+ migration upstream spring-run emigration to ocean upstream spawning migration migration Central Valley steelhead 1, 2, 3, 4 emigration trout (winter run)**(4–7 yrs)+ juveniles (4) upstream spawning (ESU 12)++

emigration Central CA Coast steelhead 1, 2 migration trout (winter run)**(4–7 yrs)+ juveniles (2) upstream spawning (ESU 9)++

spawning in upper Delta Sacramento splittail*** 2, 3, 4 (5 yrs) spawning in estuary

delta smelt** (1 yr)+ 3, 4 spawning in mixing zone *listed as endangered a. 50% of surf scoter population overwinters in Bay c. infrequent visitor **listed as threatened Suisun Bay diet is 89% C. amurensis d. females require about two years for eggs to mature; small fraction ***listed as of concern San Pablo Bay diet is 71% C. amurensis move up rivers to spawn in late winter and early spring + e. more marine than white sturgeon, juveniles spend 2—3 years Typical lifespan Central Bay diet is V. philippinarum and soft-bodied prey in estuary before migrating to sea ++Evolutionarily Significant Unit (ESU) b. arrive in Canada in late May and immediately breed

Figure 12. Exposure and habitat use for predators at risk, patterns of C. amurensis selenium concentrations (1996), and hydrologic categories that may relate to environmental partitioning of selenium for the Bay-Delta.