The Contribution of Sacramento Valley Rice Systems to Methylmercury in the

Sacramento River

Christy Tanner1, Lisamarie Windham-Myers2, Jacob Fleck2, Kenneth Tate1, Stephen McCord3 and Bruce Linquist1

1. UC Davis, 2. US Geological Survey, 3. McCord Environmental

8th Biennial Bay-Delta Science Conference

October 29th 2014

Introduction

• Methylmercury (MeHg) is toxic and bioaccumulative

• Total Maximum Daily Load (TMDL) • 0.06 ng/L goal

• Produced in flooded soils

• 200,000 Ha of paddy rice in the Sacramento Valley

Objectives and Hypothesis

• Gather data from previous studies of MeHg concentrations

• Assess spatial and temporal patterns in MeHg concentration • Seasonal patterns?

• Differences between the Sacramento and Feather Rivers?

• Hypothesis: Rice drainage waters are an important source of MeHg to the Sacramento River • Are MeHg concentrations higher in ag drainage canals?

• What is the MeHg load from agricultural drainages?

Outline

• Methods and overview of the dataset

• Results

• Temporal patterns

• Spatial patterns

• Loads from agricultural drainages

• Conclusions

Ten years of Data Available

Program

Sampling period 1996-2007

# of sites

# of samples/site

USGS National Water Quality Assessment

5 23-29

Sacramento River Watershed Program

10 14-18

8 4

12 17-18

CALFED Bay-Delta Program 16 23-31

2.5 years

3.5 years

1.5 years

<1 year

3.5 years

Selected most relevant sites • Location with

respect to rice • Data

availability

Only three sites sampled in all studies Confluence sites sampled in one study each

Upstream Sacramento

Upstream Feather

Confluence

Agricultural Drains

Rice growing area

25 mi

Methods

• Un-balanced data set • Not all sites sampled in all years

• Mixed effects model • Random effects: site and year • Fixed effects: location and season

Outline

• Methods and overview of the dataset

• Results

• Temporal patterns • Spatial patterns

• Loads from agricultural drainages

• Conclusions

Long term trends

• 566 samples at selected sites

• Concentrations ranged from method detection limits to 1.98 ng/L

• 75% of samples >0.06 ng/L (TMDL goal)

0

0.5

1

1.5

2

2.5

1/96 6/01 12/06

Me

Hg

(ng

/L)

Date (month/year)

TMDL goal

Upstream Sacramento

Ag Drains

Upstream Feather

Confluence

‘97 Flood

Seasonal pattern

Seasonally Low MeHg

Concentrations

0

0.5

1

1.5

2

2.5

1/1 3/2 5/2 7/2 9/1 11/1

Me

Hg

(ng

/L)

Julian day

Upstream Sacramento

Ag Drains

Upstream Feather

Confluence

TMDL goal

Rice Growing Season

Outline

• Methods and overview of the dataset

• Results

• Temporal patterns

• Spatial patterns • Loads from agricultural drainages

• Conclusions

No difference between Sacramento and Feather Rivers

0.001

0.01

0.1

1

10

UpstreamSacramento

UpstreamFeather

UpstreamSacramento

UpstreamFeather

Me

Hg

(n

g/L

)

Maximum 75th percentile Median 25th percentile Minimum

Summer / Fall Winter / Spring

0.060 0.075 0.066 0.076

Winter/Spring MeHg concentrations are higher in Ag Drains

0.001

0.01

0.1

1

10

Ag Drains Sacramento R. Feather R. Confluence

Me

Hg

(ng

/L)

a b b

ab

Maximum 75th percentile Median 25th percentile Minimum

0.18

0.075 0.076 0.12

Summer/Fall: Same pattern but not significant

0.001

0.01

0.1

1

Ag Drains Sacramento R. Feather R. Confluence

Me

Hg

(ng

/L)

a a a a

Maximum 75th percentile Median 25th percentile Minimum

0.12 0.079

0.066 0.060

Location - Season Interaction

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

Summer Winter

Me

dia

n M

eH

g co

nce

ntr

atio

n

(ng

/L) Sacramento R.

Feather R.

Ag Drains

Confluence

Ag Drain Concentrations lower than Yolo Bypass rice study

Alpers CN, Fleck JA, Marvin-DiPasqualebe M, Stricker CA, Stephenson M, Taylor HE. Mercury cycling in agricultural and managed wetlands, Yolo Bypass, California: spatial and seasonal variations in water quality. Sci Total Environ 2014e;484:276–87.

0.001

0.01

0.1

1

10

100

winter summer

Me

Hg

(ng

/L)

TMDL goal (0.06 ng/L)

Outline

• Methods and overview of the dataset

• Results

• Temporal patterns

• Spatial patterns

• Loads from agricultural drainages • Conclusions

Load Calculations

• Load = Concentration x Flow

• Used Ratio estimator method:

Over all load = 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑜𝑏𝑠𝑒𝑟𝑣𝑒𝑑 𝑙𝑜𝑎𝑑𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑜𝑏𝑠𝑒𝑟𝑣𝑒𝑑 𝑓𝑙𝑜𝑤 * Over all flow

• Gaps in data: • No flow data for eastern Ag Drain

• Western Ag Drain water partially diverted to Yolo Bypass

• Limited data for confluence

? ?

X

MeHg load from Ag is small compared to loads in the Rivers

0

1

2

3

4

5

6

7

UpstreamSacramento

Upstream Feather Ag Drain (west)

MeH

g lo

ad (

g/d

ay)

Upstream Loads, all avaliable data

Error bars = RMSE

(Not sampled in earliest dataset)

Sum of upstream loads is similar to confluence load

0

10

20

30

40

upstream confluence upstream confluence

MeH

g Lo

ad (

g/d

ay) First Dataset Last Dataset

Upstream Sacramento Upstream Feather

Ag Drain (west) Confluence Error bars = RMSE

Load Estimates are consistent across studies

*Log scale

Conclusions

• Future studies should account for seasonal variation

• MeHg concentrations are elevated in agricultural drains • particularly in the winter

• MeHg concentrations were low compared to Yolo Bypass

• Measurable MeHg loads from agricultural drains were small compared to Loads in the Sacramento and Feather Rivers • Due to differences in flow

Acknowledgments

• Funding • Rice Research Board

• UC Davis Plant Sciences

• Data • Sacramento River Watershed Program

• J. Domagalski, National Water Quality Assessment

• C. Foe, CALFED Bay-Delta Program

• Coauthors

• Agroecosystems Lab

Precipitation

-5

-4

-3

-2

-1

0

1

0 0.2 0.4 0.6 0.8 1

Ln(M

eH

g (n

g/L

))

Valley Average Precipitation (in)

agConfluenceFeatherSacramento