Does turbidity or genetics account for morphological differences in eelgrass (Zostera marina) at two sites near Liberty Bay?

Renee Takesue1 , Jessie Lacy1 , and Sandra Talbot2

1 USGS, Coastal and Marine Geology, Santa Cruz, CA, jlacy@usgs.gov; 2 USGS, Alaska Science Center, Anchorage, AK, sandy_talbot@usgs.gov

Eelgrass is important to nearshore ecosystems because it provides habitat; it

modifies physical and sedimentary processes; it alters sediment geochemistry;

and it is a significant detrital carbon source. Because eelgrass is sensitive to

water column turbidity, nutrient loading, and low dissolved oxygen, it is consid-

ered an indicator of coastal water quality.

A site assessment of biological resources in and around Liberty Bay in April

2006 showed that, despite the presence of suitable beaches, there was no eel-

grass growing inside Liberty Bay north of Keyport and Lemolo. There was a

small fringing bed of Z. marina on the western shore of Point Bolin (PB) and an

extensive Z. marina meadow on Agate Passage near Sandy Hook (SH, Fig. 1).

Z. marina plants at PB were unusually small (Fig. 2) and were limited to a very

narrow depth range, < -1 m (MLLW). At low tide boat wakes impinged on the

shore and resuspended bottom sediment, increasing turbidity over the PB eel-

grass bed. Z. marina plants at SH were large (Fig. 2) and extended to depths of

at least –2 m (MLLW). Water movement was vigorous at SH due to strong tidal

currents in Agate Passage.

STUDY GOALS

PB SH

BACKGROUND

The rooted marine plant Zostera marina (eelgrass) shows a high degree of phenotypic plasticity in re-

sponse to environmental conditions. Our goal was to explore whether light limitation and/or genetics could

account for morphological differences between two populations of Z. marina growing near Liberty Bay.

Small Z. marina plants at PB. Large Z. marina plants at SH.

47.68

47.70

47.72

47.74

47.76

-122.68 -122.64 -122.60 -122.56

SLOB

Poulsbo

LibertyBay

Keyport

Lemolo

Agate

Pas

s

PointBolin

Bai

nbrid

ge Is

landSHPB

Figure 1. Map of Liberty Bay showing the locations of eel-grass beds and SLOB deployments.

Shoot density per 0.25 m2

0

50

100

150

200

PB SH

# of

sho

ots

2006

2007

Average leaf length

Leng

th (

cm)

2006

2007

PB SH

60

45

30

15

0

Average leaf width

0.0

0.2

0.4

0.6

0.8

1.0

Wid

th (

cm)

2006

2007

Figure 2. Z. marina measurements at Point Bolin (PB) and Sandy Hood (SH) during April 2006 and May 2007. (Left) Shoot densities per 0.25 m2, (Center) Average leaf lengths, (Right) Average leaf widths mea-

sured above the sheath. Error bars show ± 1 SE among individual plants.

HYPOTHESES (Fig. 3):

1. Small plant size and narrow depth range of Z. marina at PB reflected a sub-optimal light environment.

2. Wakes from Liberty Bay boat traffic created higher turbidity levels at PB compared to SH.

3. The difference in Z. marina plant size at PB and SH was genetically controlled.

EELGRASS

Growth# Shoots

Leaf lengthLeaf width

Leaf growth rate

Rhizomeextension

Reproduction# Flowering

shoots# Seeds

Seed sizeTiming

FrequencyDuration

PLANT

RESPONSE

PAR

GENETICVARIATION

SUSPENDEDMATTER

Organicparticles

Sedimentparticle

conc.

Bedsediment

Suspendedsediment

PhytoplanktonBacterioplankton

Epiphytes

Resuspension

PhotosynthesisAttenuation

PHYSICALENERGY

Waves

Wind -drivenBoat wakes

CurrentsTidal

Wind -driven

WATERCHEMISTRY

Nutrients

Oxygen?

SEDIMENTCHEMISTRY

Nutrients

Sul�desGrain size

(% �nes)

Organiccontent Sediment

redoxPore water exchange O2 renewal

Microbial Oconsumption

Porosity &

Permeability

Disturbance

Leaf uptake

Root uptake

SEDIMENTSOURCE

Terrestrial

Marine

DISSOLVEDINPUTS

Terrestrial

Marine

BIOOGICALUPTAKE

DISEASE

LIGHTNutrients

Tran

spor

t

Sort

ing

Transport

Transport

Temperature

Desiccation

Figure 3. Conceptual model of environmental and biological factors influencing eelgrass. Pa-rameters measured in this study are highlighted in yellow.

APPROACH

Field measurements- We measured turbidity and pressure with in situ sensors (self-logging optical back-

scatter sensors, or SLOBs) for one month in spring to determine whether resuspension of bottom sediment by

boat wakes created a sub-optimal light environment for eelgrass growth (Table 1). Plant metrics (shoot density,

leaf length, leaf width) were measured near the sensors to relate leaf size to turbidity levels. Leaf tissue were

collected for DNA genotyping to differentiate the two Z. marina populations and to measure the genetic diver-

sity. Sediment grain size distributions were measured as % weight of dry-sieved whole-φ size fractions.

Genetics- Eelgrass leaves were collected at 2 m intervals along a 100-m alongshore transect. Genetic di-

versity, degrees of clonality, and allelic richness of eelgrass beds were determined by comparing the polymor-

phism of ten DNA microsatellite loci characteristic of Z. marina (Reusch et al., 1999).

The SLOB at PB. The sensor window is 19 cm

above the bed.

Table 1. Summary of field measurements, observations, and samples. From 18-April to 16-May, sensors sampled at 1 Hz for 40 min-utes then turned off for 200 minutes throughout the deployment. On the last day (17-May), sensors sampled at 1 Hz for 56 minutes then turned off for 4 minutes throughout the day. The OBS were calibrated with sediment mixing experiments in the lab.

In situ sensors Period Depth, PB Depth, SHOptical backscatterPressureTemperature

mid-April to mid-Maymid-April to mid-Maymid-April to mid-May

19 cm above bed19 cm above bed19 cm above bed

33 cm above bed33 cm above bed33 cm above bed

Discrete observations, samples Period Depth, PB Depth, SHBoat wakesSuspended sediment concentration (n=5)Bed sediment (n=1)Z. marina depth rangesPlant metrics (n=10); genetics (n=50)

2-4 PM, 5/16/072-4 PM, 5/16/07

5/16/074/30/06

5/16-17/07

surface19 cm above bed-0.4 m (MLLW)0 to -0.6 m (MLLW)-0.4 m (MLLW)

surface33 cm above bed-0.6 m (MLLW)0 to -2 m (MLLW)-0.6 m (MLLW)

RESULTS

Bed sediment grain size- Bed sediment under the PB SLOB had a slightly larger mean grain size (φPB = 2.48) and was less well

sorted than bed sediment under the SH SLOB (φS = 2.60) (Fig. 4). The beach elevation at the PB SLOB was about 0.2 m higher

than that of the SH SLOB (Table 1).

0%

20%

40%

60%

80%

100%

>1 mm >500 um >250 um >125 um >63 um <63 um

Size class

% w

eigh

t>1 mm>500 um>250 um>125 um>63 um<63 um

PB bed sediment grain size distribution

% w

eigh

t

SH bed sediment grain size distribution

Size class>1 mm >500 um >250 um >125 um >63 um <63 um

0%

20%

40%

60%

80%

100%

Figure 4. Sediment grain size distributions of bed sediment below the SLOBs shown as % dry weight. (Left) PB. (Right) SH.

Chris Curran and Jessie Lacy do a final check of the Point Bolin SLOB. Water level was -0.4 m

MLLW.

Jessie Lacy and Greg Justin count eelgrass shoot densities at Sandy Hook. Picture taken at

low tide (-0.7 m MLLW). Arrow points to the SLOB.

RESULTS (cont’d)

Turbidity- Average water column turbidity was slightly higher over the PB eelgrass bed than the SH bed (Fig. 5). For example, average

suspended sediment concentration (SSC) over the last 21 hours of the deployment was 15.6±5.8 (1σ) mg/l over the PB eelgrass bed

and 9.0±4.5 (1σ) mg/L over the SH bed (based on 4-minute averages) (Fig. 6). The OBS sensors were positioned so as to measure tur-

bidity at the top of the canopy when eelgrass leaves were submerged and upright. Since Z. marina leaves were shorter at PB than at

SH, the PB OBS sensor was closer to the bed than the SH sensor (Table 1).

Figure 5. In situ sensor 10-minute averages from mid-April to mid-May at Point Bolin (PB, blue) and Sandy Hook (SH, green). (Top) water depth relative to the pressure sensor; (middle) SSC; (bottom) temperature.

12:00 16:00 20:00 0:00 4:00 8:00012345

m

Water depth above sensor

PBSH

12:00 16:00 20:00 0:00 4:00 8:000

10

20

30

40

50

SSC

, mg/

L

4−min ave SSC

12:00 16:00 20:00 0:00 4:00 8:008

10

12

14

16

degr

ee C

4−min ave temperature

71 yaM61 yaM

Figure 6. In situ sensor 4-minute averages on May 16 and 17 at Point Bolin (PB, blue) and Sandy Hook (SH, green). (Top) water depth relative

to the pressure sensor; (middle) SSC; (bottom) temperature.

Boat wakes- Boat wakes were apparent in the pressure

(water depth) record as an increase in variance lasting sev-

eral minutes (Fig. 7). The pressure record was ground-truthed

by documenting boat wake events during a 2-hour window on

May 16. Some boat wakes were followed by higher turbidity,

but there was more variability in the SSC records than could

be accounted for solely by boat wakes (e.g. Fig. 8).

Boat wakes break over the eelgrass bed and resuspend bottom sediment.

13:40 13:50 14:00 14:10 14:20 13:400

2

4

x 10−4

m2

30−sec variance of depth at Pt Bolin

May 16

Figure 7. Variance of water depth at PB increased during documented boat wake events (blue).

Genetics- The eelgrass bed at PB was genotypically richer than the bed at SH. There were 30 different genotypes at PB compared to only 12

genotypes at SH (n=50 plants sampled per site). Individual plants at PB were also more genetically diverse than those at SH. Eelgrass at SH pri-

marily reproduced clonally via rhizome branching while at PB eelgrass appeared to reproduce primarily by flowering and seedling establishment.

Small plants at PB could have been an annual population. An annual life cycle is a growth strategy in disturbed environments (Phillips et al., 1983).

17:00 17:10 17:20 17:300.6

0.7

0.8

0.9

1Water depth above sensor

m

17:00 17:10 17:20 17:300

0.5

1

1.5

2 x 10−3 30−sec variance of depth

m2

17:00 17:10 17:20 17:300

10

20

30

40

mg/

L

SSC (30−sec ave)

May 12

Figure 8. Boat wakes on May 12, apparent as increased water depth (top) and vari-ance (middle) did not always correspond

to inceased turbidity.

SIGNIFICANCE

Lower light levels could have accounted for smaller leaf sizes at PB.

Boat wakes were not the only cause of higher turbidity at PB. Other possible explanations: Turbidity increased with depth (PB sensor was 14 cm closer to the bed than SH sensor). Small eelgrass plants at PB were less effective in attenuating currents and waves. PB experienced more southerly wind waves than SH. Advection of sediment-laden turbid plumes from creeks near PB.

Pressure/OBS sensors were an effective way to monitor boat wake impinge-ment on the shore.

Higher geneotypic richness in the PB eelgrass bed, indicative of generative (flowering) rather than vegetative (clonal) reproduction, suggests a population subject to greater physical disturbance than at SH.

Phillips, R.C., Grant, W.S. and McRoy, C.P., 1983. Reproductive strategies of eelgrass (Zostera marina). Aquatic Botany, 16: 1-20.Reusch, T.B.H., Stam, W.T. and Olsen, J.L., 1999. Microsatellite loci in eelgrass Zostera marina reveal marked polymorphism within and among populations. Molecular Ecology, 8: 317-321.