Estuaries Vol. 21, No.1, p. 66-77 March 1998

Methane Distribution and Cycling in Tomales Bay, California1

FRANCIS J. SANSONE2

TERRI M. RUST

STEPHEN V. SMITH

Department of OceanographyUniversity of Hawaii1000 Pope RoadHonolulu, Hawaii 96822

ABSTRACT: Cycling of methane (CH.) in Tomales Bay, a 28-km' temperate estuary in northern California with rela­tively low inputs of organic carbon, was studied over a I-yr period. Water column CHi concentrations showed spatialand temporal variability (range = 8-100 nM), and were supersaturated with respect to the atmosp~ereby a factor of 2­37. Rates of net water column CHi production-oxidation were determined by in situ experiments, and were not foundto be significantly different from zero. Fluxes across the sediment-water interface,determined by direct measurementusing benthic chambers, varied from -0.1 ....mol m-' d- l to +16 ....mol m-' d- I (positive fluxes into water). Methaneconcentrations in the two perennial creeks feeding the bay varied annually (140-950 nM);these creeks were a significantCH. source to the bay during winter. In addition, maslrbalance calculations indicate a significant additional winter CHsource, which is hypothesized to result from storm-related runoff from dairy farms adjacent to the bay. Systemwide CH:budgets of the l6-km' inner bay indicate benthic production (110 mol d- ' ) anti atmospheric evasion (110 mol d- ' )dominated during summer, while atmospheric evasion (160 mol d- ' ) and runoff from dairy farms (90 mol d- ' ) dominatedduring winter. .

Introduction

The role of atmospheric methane (CH4 ) inEarth's radiative and chemical balance is well rec­ognized. Atmospheric CH4 levels have been in­creasing approximately 1% per year over the last200 yr (Rasmussen and Khalil 1981; Cicerone andOremland 1988; Etheridge et al. 1992; Lassey et al.1992), resulting in a contribution to the radiativeforcing of the atmosphere which is estimated to be20-30% of that caused by the increase in atmo­spheric carbon dioxide (Etheridge et al. 1992;Thompson et al. 1992). Despite their importancein assessing the present-day global CH4 budget,global and regional source inventories are not welldefined, particularly with respect to marine coastalenvironments. It is fairly well established that theopen ocean is responsible for -2% of the totalatmospheric CH4 input, although emissions frommore methane-rich marine environments could beconsiderably more important (Lambert andSchmidt 1993). However, data on coastal fluxes re­main sparse, with most authors referring to Eh­halt's (1974) estimate that -1-2% of the annualproduction of atmospheric CH 4 comes from"ocean shelf' environments (see Discussion, be­low).

1 School of Ocean and Earth Science and Technology Con­tribution No. 4508.

2 Corresponding author. Tele: 808 95&-8370; fax: 808 95&­7112; e-mail: sansonel(j,soest.hawaii.edu

© 1998 Estuarine Research Federation 66

Although dissolved CH4 concentrations are onthe order of 0.3-2.0 nM in open ocean surface wa­ters, they are typically higher and more variable incoastal regions (e.g., Atkinson and Richards 1967;Lamontagne et al. 1973; Rudd and Hamilton 1978;DeAngelis and Lilley 1987). Methane distributionin coastal environments is influenced by severalfactors: sediment and water-column productionand oxidation, diffusive and advective transport insediments, mixing with low concentration open­ocean seawater, input from runoff, and atmospher­ic evasion. At present, studies of these processeshave been limited to coastal environments char­acterized by rapid rates of organic matter deposi­tion, while our understanding of less organic-richenvironments remains speculative. This lack of spe­cific information on factors controlling sources,distributions, and sinks limits our understanding ofexisting global change processes and prevents con­fident predictions of anticipated global climatechanges.

The purpose of this study was to examine an es­tuary, Tomales Bay, California, which experiencesrelatively low inputs of organic matter to the sedi­ments, and identify the processes predominantlyinfluencing water column CH4 distribution over anannual cycle.

Study Location

Located approximately 50 km north of San Fran­cisco, Tomales Bay is a long and narrow (20 km X

Methane in Tomales Bay, California 67

N

!

CALIFORNIA

5J

• = sample station

oI

~..it"'; .'~.

~~~

~

Fig. 1. Location map of Tomales Bay, California. Station numbers (0-18) correspond to distance (km) from the mouth of thebay. Stations 0-8 are in the outer bay. and stations 10-18 are in the inner bay.

1.4 km), shallow (3 m average depth), submergedrift valley formed at the seaward end of the SanAndreas Fault (Fig. 1). Tomales Bay has a climatethat is basically bi-seasonal (Smith et aI. 1996; Fish­er et al. 1996; Smith and Hollibaugh 1997): sum­mer is generally hot, dry, windy, and experiencescoastal upwelling, whereas winter is cool, wet, lesswindy, and lacks upwelling.

The bay is divided into two hydrographically dis­tinct regions: inner bay and outer bay (Smith et aI.1987; Hollibaugh et aI. 1988). The outer bay oc­cupies the northern 8 km of the estuary (stations(Stn) 0-8) and has a water composition similar tothe adjacent coastal ocean, reflecting the rapid, tid­ally influenced exchange between coastal and out­er bay waters. Slow exchange between inner andouter bay, and freshwater input dominated by win­ter storms are the predominant hydrographic fea­tures of the inner bay (stations 10-18) (Smith etal. 1987, 1989, 1991). The water exchange time forthe inner bay is -40-50 d during the summer, and-10-20 d during the winter (Smith and Holli­baugh 1997). Dissolved oxygen concentrations

within the bay are generally greater than 200 11Myear-round. .

Freshwater runoff and coastal upwelling exhibitdistinct seasonal trends that are reflected in thebay's salinity, water temperature, and nutrient con­centrations, (Smith et aI. 1987, 1991). Low runoffin the summer months leads to an evaporation­dominated system in the inner bay, with salinitiesoften above oceanic values. During the period ofthis study, salinity ranged from 10 psu to 36 psu,and water temperature ranged from 5°C to 23°C.

Approximately 35% of the planktonic primaryproduction in the bay is estimated to reach thesediments, and virtually all of the stream-borneparticulate load is retained in the bay (Smith andHollibaugh 1997). Sediments in the bay consist ofsand and sandy mud in the outer bay and mud inthe inner bay; shoals are present at the mouth,head, and along the northeast shoreline. The sed­iment organic carbon content is about 0.5% (wt/dry-wt) in the outer bay and 1.5-2% in the innerbay (Plant 1995), and the sedimentation rate is -5mm yr I (Rooney 1995), yielding an organic mat-

68 F. J. Sansone at al.

ter burial rate of 25-100 g C m- 2 yr-I. Extensiveepifaunal and infaunal communities exist on theshoals as well as on the deeper mud bottom. Eel­grass (Zostera marina) communities cover approxi­mately 3.5 km2 of the subtidal (1-3 m) area andare most pervasive in the outer bay, covering 3 km2

(Spratt 1989; T. Moore personal communication).

MethodsWATER SAMPLING

Sample collection occurred every other monthfrom September 1990 through November 1991.Surface water samples for CH4 analysis were takenalong a transect running the length of the bay dur­ing each field trip (Fig. 1). In addition, during Sep­tember 1991, samples were collected from surfaceand mid-depth waters at other locations aroundthe bay (Fig. 2). Samples were taken over the sideof a boat with the aid of a small submersible pumpand lo-m length of Tygon tubing, which allowedthe water to be introduced directly into 60-ml bio­logical oxygen demand (BOD) bottles, minimizingcontact with the atmosphere. Immediately after­wards, 0.5 ml of an aqueous saturated HgCl2 solu­tion was introduced into each bottle, which wasthen sealed with a ground-glass stopper (coatedwith Apiezon-N grease) until analysis. Tilbrook(1991) demonstrated that HgCl2-preserved watersamples could be stored in this manner for at leastone year with no loss of CH4 ; all Tomales Bay sam­ples were analyzed within 40 d of collection.

Lagunitas and Walker creeks (Fig. 1), the twomajor freshwater sources, were sampled to quantityriver-derived CH4 inputs to the bay. The creek sam­ples were collected from relatively fast-moving andwell-aerated areas located immediately above theinfluence of high tide using 150-ec syringes withattached 8-em Tygon tubes; samples were pro­cessed as described above. United States Geologi­cal Survey stream gauge data were used to calcu­late flow rates (data and methods are summarizedby Smith et al. 1996).

BENTHIC FLUX MEASUREMENTS

Methane fluxes across the sediment-water inter­face were measured every other month with trip­licate in situ benthic chambers (e.g., Martens andKlump 1980) deployed at three unvegetated sites(as described by Dollar et al. 1991): Stn 16, Stn 10(located -200 m east of the water sampling site),and Stn 6 (located -300 m southeast of the watersampling site). The chambers were 0.5 m in di­ameter and incorporated internal stirring, andwere placed without disturbing the sediment. Dur­ing September 1991, benthic chambers were alsoset in an eelgrass bed near Stn 16 in order to com-

pare CH4 fluxes from a vegetated area with thosefrom the three unvegetated sites.

Mter chamber deployment, samples were col­lected by SCUBA diver 20 min and approximately24 h later by drawing chamber water into 150-ecsyringes via a stopcock-eontrolled sampling port.Following collection, samples for CH4 measure­ments were immediately transferred from the sy­ringes to 6o-ml BOD bottles via a 15-em length ofTygon tubing; the bottles were gently filled fromthe bottom with several bottle-volumes of sampleto minimize atmospheric exchange. The BOD bot­tles were preserved and sealed by the same pro­cedure described in the Water Sampling section,above.

The following equation can be used to calculatethe flux across the sediment-water interface:

where Jsed is the flux of CH4 across the sediment­water interface (j,Lmol m- 2 d- I), [CH4]i is the initialconcentration (j,LM), [CH4]r is the final concentra­tion (j,LM) , V is the volume of the benthic chamber(m3 ), A is the area of sediment covered by thechamber (m2), and t is the time of chamber de­ployment (h). The constancy of the flux was veri­fied by a 5o-h deployment in September at Stn 16with five samplings; during this experiment, fluxchamber CH4 concentrations increased linearlywith time (regression r2 = 0.969; Rust 1993).

METHANE PRODUCTION-OXIDATION MEASUREMENTS

Time series in situ experiments were conductedto determine water column net CH4 production­oxidation rates using a modification of the tech­nique of Sansone and Martens (1978). Duringeach experiment a total of twelve 25o-ml BOD bot­tles, wrapped in aluminum foil to eliminate pho­tosynthetic effects, were filled with bottom waternear Stn 12. Once filled with bay water, the bottles,with the exception of two which were used as time­zero controls, were secured inside a weighted crateand lowered to the bottom (5-6 m depth) near Stn12. During the months ofJuly and September, thesame experiment was also conducted near Stn 6 inthe outer bay.

Two bottles at a time were retrieved periodicallyover the 3-4 d following deployment. One of thebottles was used both for determination of tem­perature and dissolved oxygen (measured in thebottles with a YSI probe and meter) and for deter­mination of dissolved inorganic nitrogen (DIN).DIN was measured by filtering the water throughWhatman GFIC filters, storing it frozen in HCI­washed polyethylene bottles, and analyzing it byAutoAnalyzer (Smith et al. 1981). The water in the

Methane in Tomales Bay, Califomia 69

A

SECTION A-A'

o 2 3kmL-'__...1-.__-"-1__......'

Confours are methaneconcenfrations in nmolll

•••••• _- Mean low water

Ii.Om

5

10

15

8Om

5

SECTION S- S'

81

Fig. 2. Contour map of surface-water methane concentrations (nM) for September 1991. Cross sections A-A and B-B show verticalmethane distributions. Black dots indicate sample locations; stars indicate locations of the weather stations at Tomasini Point and themouth of Tomales Bay.

70 F. J. Sansone at al.

45 ..,---------------;=====;l 25___ Sin 16

-0- Sin 10- ..... sine

20"0-E- 15"0[>< 10:::>

LL

./1-"Glcco

~ 5 I '/

::!! /I

/

o

--0- Inner Bay___ Outer Bay

'/ \~ \

\-----+...... --5

30

10

35

40

::!!~ 25cco

~ 20

15

o J-__,_-1!..!9.!!90~1'-!!;99~1-,---r_-__,_--..,_--,__---.J 1990 1991

Fig. 3. Mean transect surface-water methane concentrationsin the inner and outer bays during 1990-1991. Error bars rep­resent 1 SD (n = 5).

Nov Jan Mar May Jul

Month

Sept Nov Sep Nov Jan Mar May Jul Sep Nov

Month

Fig. 4. Benthic fluxes of methane at the unvegetated sites,as determined with benthic chambers during 1990--1991. Errorbars represent 1 SD for replicate experiments (n = 3, with theexception of September 1990 and November 1991 [n = 2]).

other bottIe was prepared for CH4 analysis as de­scribed above under Water Sampling. Rates of netCH4 production~xidation in the water columnwere determined by least squares regression anal­ysis, assuming the accumulation and loss of CH4 tobe first~rder rate reactions with respect to CH4

concentration (e.g., Jannasch 1975).

METHANE ANALYSES

Methane concentration was determined by anon-line liquid stripping, cryo-trapping technique(modification of the method of Brooks et al. 1981)using a gas chromatograph equipped with a flameionization detector. The reproducibility of five suc­cessive injections of a 23 ppmv CH4 (in helium)standard was better than ± 1.5% (1 SD). The com­bined standard error of sampling and analysis, de­termined by twice collecting and analyzing five rep­licate water samples at Stn 12, was ± 5.6% (1 SD).

ResultsWATER COLUMN DISTRIBUTION

Surface CH4 concentrat.ions along the horizontaltransect and at other locations throughout the bayranged from 8 nM to 100 nM CH4 (Fig. 2). Mid­depth (3.5 m deep) water column concentrationsin September ranged from 14 nM to 31 nM CH4

and deep-water (8.5 m deep) concentrationsranged from 13 nM to 30 nM CH4 • Differencesbetween surface water and deep water CH4 con­centrations at anyone station were less than thestandard error of sampling and analysis (± 5.6%).Air-equilibrated concentrations, calculated fromthe temperature and salinity of each water sample(Weisenburg and Guinasso 1979), assuming an at-

mospheric CH4 mlxmg ratio of 1.70 ppmv (DIu­gokencky et al. 1994), were 2.1-2.7 nM. Saturationratios were determined by dividing the measuredCH4 concentration by the equilibrium concentra­tion; these calculations indicated that surface wa­ters were 2-37 times supersaturated with respect tothe atmosphere.

Inner and outer bay CH4 transects revealed apattern of consistently higher average CH4 concen­trations in the inner bay. Surface water CH4 con­centrations also showed variability during the year(Fig. 3). Methane concentrations were most uni-.form throughout the entire bay during the sum­mer months, May-July, corresponding with highCH4 input from the sediments and low stream in­put (discussed later). There was a noticeable dif­ference between inner bay and outer bay CH4 con­centrations in November 1990, and especially inMarch 1991. These sampling dates coincided withperiods of increased stream runoff following rain­storms.

BENTHIC FLUX

Temporal variability was also observed in thebenthic fluxes at each of the unvegetated samplingsites (Fig. 4). The largest fluxes occurred at Stn 16(11-16 ~mol m- 2 d- I ) during September and No­vember 1990, and at Stn 6 (11 ~mol m- 2 d- I ) andStn 10 (nearly 5 ~mol m- 2 d- J) in September 1991.The lowest benthic fluxes were measured in Janu­ary at Stn 6 (-0.0 ~mol m- 2 d- I ), in March at Stn10 (0.4 ~mol m- 2 d- J), and in May at Stn 16 (0.6~mol m- 2 d- J ). This trend is assumed to be sea­sonal, and is further supported by the observed

Methane In Tomales Bay, California 71

0

1000 b

800:::l:

ai 600l:

'".t:Gi::: 400

200

:l2"'e 600§~ 400

LL

Creek flow and the total runoff into the inner bay.Some of the intermittent streams carry fresh ma­nure from adjacent dairy farms into the bay duringwinter storms (R. Chambers personal communi­cation 1996), and we believe that CH4 from thesestreams may playa significant role in the inner bay,CH4 budget (see Discussion).

METHANE PRODUCTION-OXIDATION

Mean values of the net water column CH4 pro­duction-oxidation (Table 1) ranged from -970pmol I-I d- I to +290 pmoll- 1 d- I (positive valuesindicate net CH4 production). However, there wasno significant difference between sampling dates,and, except for the May measurement, the valueswere not significantly different from zero.

DiscussionThe variability of CH4 abundance, both spatially

and temporally, in Tomales Bay surface waters re­flects the interaction of several processes: chemical

1000 "T'=:a----,========n--+- Leg<rlitas Creek-0 - Total inner bay run""

800

200

Jan Mar May Jul Sep Nov

Month

Fig. 5. a) Flow rate of Lagunitas Creek and the total fresh­water runoff into inner Tomales Bay during 1991, estimatedaccording to the procedure of Smith et al. (1991). b) Methaneconcentrations in Lagunitas and Walker creeks during 1991;numbers next to symbols indicate the number of replicate sam­ples; error bars show the range of concentrations for duplicatesamples; dotted lines show extrapolated data used in flux cal­culations.

seasonality of dissolved nutrients released from thesediments (Dollar et al. 1991). The variation be­tween the three chambers at anyone samplingtime, a common attribute of benthic flux measure­ments (e.g., Martens and Klump 1980; Dollar et aI.1991), was presumably due to heterogeneity of thebenthos in the bay.

The benthic flux from eelgrass beds near Stn 16in September 1991 was 50 ~mol m-2 d- I. This fluxwas significantly greater than fluxes observed at theunvegetated sites (4-11 ~mol m-2 d- I). However,benthic fluxes were not measured in the eelgrassbeds during winter. We instead have estimated thewinter flux using the Arrhenius-type temperature de­pendence of CH4 production and emission previ­ously observed in other aquatic environments (e.g.,Sansone and Martens 1982; Sass et aI. 1991). Usingan activation energy (Ea ) value of 87 kJ mol- 1 re­ported by Sass et aI. (1991) for ricefield CH4 emis­sion, which is similar to values reported for a varietyof other aquatic environments (see references in Sasset al. 1991), the winter eelgrass flux can be estimatedby the following relationship:

EaIn F = - - + constant (2)RT

where F is the rate of CH4 flux from the sediment(~mol m-2 d- I), R is the ideal gas constant (8.31

J K-l mol-I), T is the temperature (OK), and theconstant is determined empirically (we used fluxdata for the eelgrass bed during September 1991[T = 18.6°C = 291.6°K]). Equation (2) yields anestimated winter (average T = 11.9°C = 284.9°K)eelgrass flux of 22 ~mol m- 2 d- I.

STREAM INPUT

Methane concentrations in Lagunitas and Walk­er creeks (140-950 nM) were 1-2 orders of mag­nitude higher than bay water column CH4 valuesdetermined during the course of this study, withthe exception of Stn 18 in March (Fig. 5b). Bothstreams displayed significant variability in CH4

abundance, with the highest values, 670 nM atWalker Creek and 950 nM at Lagunitas Creek, oc­curring in summer, and the lowest values occur­ring during winter. During all sampling periods,Lagunitas Creek CH4 concentrations were elevated'compared to those ofWalker Creek. Our values arewithin the 5-1,700 nM range reported for PacificNorthwest (USA) rivers (de Angelis and Lilley1987; Lilley et al. 1996).

Stream flows were maximal during the winter(Fig. 5a). In addition to the two perennial streams,there is substantial freshwater runoff into the bayduring the winter from intermittent streams. Forthe inner bay, this additional runoff can be seenin Fig. 5a as the difference between the Lagunitas

72 F, J. Sansone et al.

ATMOSPHERE (A)TABLE I. Net rates of methane production-oxidation (pmolI-I d- ' ) in waler column samples incubaled under in SiLU con­ditions during 1991. Positive values indicale nel meLhane pro­duction. Values are calculaled from linear regressions of time­course experimenlS using 6-7 samples colieCled over 3-4 d in­cubations; ranges are Lhe 95% confidence imervals; nd = nOldelermined. .

SUMMER

OUTER BAY &COASTAL OCEAN (E)

9t '3 INNER BAY 9". MISSINGSINK

Location

Station 12 Station 6Season Month (Inner Bay) (Outer Bay)

Wimer March -90 (:!:220) ndSummer May -970 (:!:680) nd

July -110 (:!:2.320) -280 (:!:1,250)Sepl. 290 (:!:400) -300 (:!:940)

WINTER

UNllEGETATED EELGRASSSEDIMENT (S) BEDS (5)

ATMOSPHERE (AI

UNllEGETATED EELGRASSSEDIMENT IS) BEDS IS)

Fig. 6. Diagram of inner Tomales Bay meLhane budgelS forsummer and winler. Values are seasonal means (mol do') :!: SLan­dard errors. Lelters in parenLheses indicate Lhe terms in Eq. (3).

greater than 1.5 m depth in Tomales Bay sedi­ments (Rust 1993), strongly suggests that bubbletransport is not a significant mechanism for trans­porting CH4 across the sediment-water interface.Evidence from seismic profiling of the sediment(S. V. Smith and J. T. Hollibaugh unpublisheddata) indicate that the depth of CH4 saturation isvariable but always occurs at sediment depths> 1m. It is believed that the primary mechanisms con­trolling the flux of CH4 across the sediment-waterinterface in Tomales Bay are diffusion and macro­faunal activity, as there was no visual evidence forresuspension of sediment or increased turbulenceat the sediment-water interface during windy pe­riods.

Animal-sediment interactions such as depositfeeding, burrowing, tube construction, and irriga­tion affect porewater transport, often significantlyaltering the geochemical characteristics of sedi­ments and overlying waters by increasing the fluxof certain dissolved constituents over that whichcan be accounted for by one-dimensional molec­ular diffusion (Hammond et al. 1975; Aller 1978,1980, 1984; Emerson et al. 1984; Hartman andHammond 1984). One-dimension'al diffusive CH4

fluxes, determined using concentrations measuredover the top 5-10 cm in cores retrieved at the sametime and site as chamber deployments, were 0-2.4

MISSINGSOURCE

94, 30INNER BAY

66124OUTER BAY &COASTAL OCEAN (E)

where V is the volume of the inner bay, C is theCH4 concentration in the bay water, S is the CH4

flux across the sediment-water interface, R is CH4

input from freshwater runoff, E is loss or gain dueto water exchange between the inner and outerbay, 0 is gain or loss due to water column CH4

production or oxidation, and A is loss due to at­mospheric evasion (Fig. 6). The budget is subdi­vided into summer (May-October) and winter(November-April). Note that outer bay exchangewith the coastal ocean is very fast (Smith et al.1987; Smith and Hollibaugh 1997), with a meanwater exchange time of -1 d, which prevents thebudgeting of the outer bay's nonconservative flux­es.

Sediment-Water Exchange

Exchange across the sediment-water interface incoastal aquatic environments is controlled by phys­ical mixing processes, bubble transport, diffusion,biological activity, and transport through emergentaquatic macrophytes (e.g., Martens and Berner1977; Dacey and Klug 1979; Martens and Klump1980; Hartman and Hammond 1984; Chanton andMartens 1988). The absence of observable bubblesand bubble tube structures in the upper sediment,in addition to CH4 saturation occurring only at

and biological reactions, exchange across the sed­iment-water interface, exchange across the air-wa­ter interface, freshwater input, and horizontal mix­ing. The relative importance of these processesmay vary seasonally (i.e., benthic production andmixing) as well as on a time scale of hours or days(i.e., storm events, wind, and tidal exchange).

SEASONAL BUDGET

A steady state mass balance equation can be writ­ten to describe a simple CH4 budget for the innerportion of Tomales Bay:

dCV- = S + R ± E ± 0 - A = 0 (3)

dt

f.Lmol m-"Ld- 1, or 56-100% less than chamber flux­es (Rust 1993).

To determine S in Eq. (3), seasonally averagedchamber fluxes were multiplied by the appropriateinner bay areas: 15.5 km2 for unvegetated sedi­ments and 0.5 km2 for eelgrass beds (Spratt 1989;T. Moore personal communication). The unvege­tated fluxes were estimated as means of all the flux­es measured at Sm 6 (sandy sediment), and Sm 10and Stn 16 (muddy sediment). This integrated thefluxes from the two unvegetated sediment types ata sandy:muddy ratio of 1:2, which has been previ­ously shown from nutrient budgets (Dollar et al.1991) to adequately reflect the bottom types pres­ent in the inner bay. The estimated summer fluxto the inner bay was 5.5 (± 0.7) f.Lmol m- 2 d- 1 andthe winter flux was 2.5 (± 0.5) f.Lmol m- 2 d- 1• Theuncertainty in benthic calculations reflects the spa­tial variability in chamber fluxes observed in trip­licate measurements at anyone site. This variationwas greatest during the summer when water tem­peratures were warmer and macrofaunal activitywas higher.

Although only a few measurements were ob­tained from sites vegetated with eelgrass, thesefluxes were considerably higher than those at un­vegetated sites. Approximately 3% of the inner baysupports eelgrass beds, which had a measured sum­mer flux of 50 f.Lmol m- 2 d- ' and an estimatedwinter flux of 21.5 f.Lmol m- 2 d- I • The relativelyhigh CH4 concentrations observed in the outer baynon-transect surface water samples in September1991 (Fig. 2) also indicate that the extensive outer­bay eelgrass beds (25% of the bottom as comparedwith 3% in the inner bay (Spratt 1989; T. Moorepersonal communication» can affect water col­umn CH4 concentrations. Note that CH4 inputsfrom Walker Creek, local runoff, and the coastalocean were unlikely sources at this time since riverflow is very low in summer, and mixing with thecoastal ocean waters would tend to decrease con­centrations rather than elevate them.

Water Column Production-Oxidation

Methane gain-loss due to net CH4 production­oxidation in the water column, 0, was examinedby in situ CH4 production-oxidation experiments(see above). However, because there was no signif­icant production or oxidation during the experi­ments (except for a small amount of oxidation inthe May experiment), this term is set to zero in themodel.

DIN concentrations measured during the oxi­dation experiments ranged from 0.5 f.LM to 5.2f.LM, generally increasing over the duration of eachexperiment; dissolved oxygen decreased, presum-

. ably due to respiration, but was always> 160 f.LM

Methane in Tomales Bay, Califomia 73

(Rust 1993). Together, these later data suggest thatthe microaerophilic bacteria responsible for watercolumn CH4 oxidation (e.g., Higgins et al. 1981)were inhibited by the oxygenated conditions in thewater column, as the ambient DIN concentrationsdid not reach the > 20-f.LM levels typically neces­sary to overcome the effects of oxygen inhibition(Rudd et al. 1976; Sansone and Martens 1978).The elevated dissolved oxygen levels also indicatethat the observed CH4 production must have oc­curred within oxygen-depleted microenviron­ments, such as in suspended particles (e.g., All­dredge and Cohen 1987; Sieburth 1987).

Freshwater Runoff

The seasonal CH4 inputs from freshwater runoff,R, were estimated by averaging six monthly fluxesfor each season. Fluxes (mol d- 1) were calculatedas the product of the monthly concentration(taken as the mean of Lagunitas Creek and WalkerCreek concentrations) and the monthly total run­off flow rate (Fig. 5).

As shown in Fig. 5, the maxima in the streamflow rates and the stream CH4 concentrations arepartially offset in time, resulting in a fivefold dif­ference in the seasonal fluxes (summer, 15 mold- I ; winter, 83 mol d- 1; Fig. 6) despite the tenfoldvariation in the flow rates. These stream fluxeswere 12% and 37% of the total summer and winterCH4 inputs, respectively, to the inner bay.

Inner Bay-Outer Bay Exchange

Values for the mean seasonal water mixing ratebetween the inner bay and outer bay (summer, 3X 106 m 3 d- 1; winter, 5 X 106 m 3 d- I ; Smith andHollibaugh 1997) allow estimation of the transportof dissolved CH4 out of the inner bay. These fluxesare computed by multiplying each of the abovemixing rates by the mean difference in the innerbay and outer bay water concentrations (Fig. 3) forthe appropriate season (summer, 2.9 nM; winter,13.2 nM). This yields a summer flux of 9 mol d- 'and a winter flux of 66 mol d- I (Fig. 6).

Air-Sea Exchange

Gas transfer across the air-sea interface has beenshown to be dependent upon the partial pressuredifference across the interface (e.g., Broecker andPeng 1974; Liss and Slater 1974; Hartmann andHammond 1984). Transfer can be determined bya general stagnant film equation (Liss and Slater1974) :

(4)

where J is the flux to the atmosphere (f.Lmol m- 2

d- ' ); K is the transfer velocity (m d- 1) (K = D/z,where D is the molecular diffusivity and z is the

74 F. J. sansone et al.

thickness of the stagnant film layer); c", is the gasconcentration in the surface seawater (fJ.M); andCeq is the gas concentration of seawater (fJ.M) atthe in situ temperature and salinity (data fromSmith and Hollibaugh 1997) when in equilibriumwith the atmosphere (1.7 ppmv; Dlugokencky et al.1994). calculated using the equilibrium relation­ship of Yamamoto et al. (1976).

Ceq can be calculat~d using the solubility equa­tion of Weisenberg and Guinasso (1979) and c",can be directly measured. It has been shown thatthe flux across the air-sea interface increases withincreasing wind speed (e.g., Liss 1973, 1983; Kan­wisher 1974) and can be related to K. Hartmannand Hammond (1985) have developed a predictiveequation for estuarine systems where gas exchangeis thought to be dominated by wind stress:

K = 34.6 R,. (Dm )O.5 (U IO )J.5 (5)

where R,. is the ratio of the kinematic viscosity ofpure water at 20°C to that of water at the temper­ature aod salinity of interest, Dm is the moleculardiffusivity of the gas at 20°C (cm2 S-I), and U lO isthe average wind speed (m S-I) at a height of 10m. Seasonally averaged wind speeds for TomalesBay (using Sierra Misco ALERT weather stations atTomasini Point and the mouth of Tomales Bay, Fig.2) were 1.6 m S-I and 2.6 m S-I for winter andsummer, respectively, yielding K values of 0.22 md- I and 0.55 m d- I •

The estimated K values can be used in Eq. (4),along with seasonally averaged surface water CH4

concentrations (Fig. 3), to calculate the followingfluxes to the atmosphere from the inner bay: 10.1mol m- 2 d- I and 6.75 mol m-2 d- I for summer andwinter, respectively. These values correspond to in­tegrated rates of 161 mol d- I and 108 mol d- I,respectively, for the whole inner bay.

Missing Sources and Sinks

The summer CH4 budget has an apparent "miss­ing sink" of 9 mol d- I , which is needed to closethe budget (Fig. 6). It is likely that this value issimply an artifact of the propagation of the errorsin the other terms in the budget. Alternatively, thenet water column oxidation measured at Stn 12 inMay (Table 1), but which was not included in thebudget, could also be part of this missing sink; thisoxidation rate, 970 ± 680 pmoll- I d- I, representsa seasonally averaged sink of 8 ± 6 mol d- J inte­grated seasonally over the entire inner bay (volume= 4.96 X 107 m 1 (Smith et aI., 1987».

In contrast, the winter budget has a much larger"missing source" of94 mol d- I (Fig. 6), again com­puted by mass balance of the measured positiveand negative fluxes into the bay. Storm-related sed­iment-water mixing is an unlikely source, as there

is essentially no methane in the upper 1 meter ofthe Tomales Bay sediments (Sansone et al. unpub­lished data) due to the high degree of bioturba­tion. It is likely that the additional CH4 source isprimarily the intermittent streams feeding the bay.During winter storms (see Stream Input, above),these streams are heavily contaminated with freshmanure and holding-pond overflow from adjacentdairy farms. Unfortunately, no. measurements weremade of the CH4 concentration in these streams.However, if these streams are indeed the missingsink in the winter budget, and assuming that theirflow is equal to the runoff into the inner bay notcoming from Lagunitas Creeks (Fig. 5a), theirmean CH4 concentration can be estimated by di­viding the missing source by the winter-average in­termittent stream flow into the inner bay (120 X101 m3 d -I, Fig. 5a). This calculation yields a valueof 990 nM, which is not an unreasonable· concen­tration given the high loading of manure in thesestreams, and is nearly identical to the May 1991CH4 concentration measured in Lagunitas Creek(Fig.5b).

Budget Summary

The inner bay budget (Fig. 6) indicates a fewobvious seasonal differences in the processes con­trolling CH4 levels in the 16-km2 inner bay. Sum­mer is dominated by benthic production (7 fJ.molm- 2 d- I) and atmospheric evasion (7 fJ.mol m- 2

d- I). These same summer processes were alsofound to be important to the summer CH4 budgetfor Cape Lookout Bight, North Carolina (Martensand Klump 1980). Conversely, atmospheric evasion(10 fJ.mol m- 2 d-I) and freshwater runoff (8 fJ.molm- 2 d- I ) dominated during winter.

IMPLICATIONS FOR LOCAL CARBON BUDGETS AND THE

GLOBAL METHANE CYCLE

The importance of benthic CH4 fluxes in the in­ner bay CH4 budgets invites a calculation of therole of benthic CH4 production in the sedimentrespiration of organic matter in Tomales Bay.Smith and Hollibaugh (1997) determined the lat­ter to be 51 mmol m-2 d- I , or -800,000 mol d- J

over the whole inner bay; in comparison, our re­sults indicate a mean inner bay benthic CH4 fluxof 80 mol d- I, or 0.01 % of the total benthic res­piration. This result, combined with the observedlow rates of water column CH4 production-oxida­tion, indicate water column CH4 cycling is only asmall part of the Tomales Bay carbon cycle.

On a larger scale, it might be assumed that ifTomales Bay, with its relatively low rate of organicmatter input (25-100 g C m-2 yr-I, see Study Lo­cation, above), is not representative of most estu­aries, it might instead be a reasonable proxy for

CH4-producing continental shelf sediments. Thisconclusion is based on a comparison with the-100-1800 g C m- 2 yr-J rate of organic mattersedimentation typical of estuaries sited at rivermouths and the < 30 g C m- 2 yr-J rate represen­tative of continental shelves (e.g., Muller and Suess1979; Heip et al. 1995). The low organic input ratein Tomales Bay is also reflected in the deep (200­250 cm) penetration of the organic matter oxidantsulfate in the inner bay sediments (Sansone et al.unpublished data). In comparison, sulfate deple­tion in organic-rich estuarine sediments will bereached within a few dm of the surface, whereasshelf sediments typically have sulfate penetrationto depths of several meters (J0rgensen 1983).

Thus we can use the data presented here to eval­uate Ehhalt's (1974) estiJIlate that the portion ofthe continental shelf with a depth less than 10 mcontributes between 0.7 X 1012 g CH4 yr-l and 14X 1014 g CH4 yr-J (or 4 X 1010 mol CH4 yr-l to 9X 1013 mol CH4 yr-l) to the atmospheric CH4 pool.He assumed that, as for freshwater lakes, only 1­10% of the earth's continental shelf less than 10m deep (1.4 X 106 km2 ) actually produces CH4 ,

and that these CH4-producing sediments have a re­lease rate similar to that of freshwater lakes (esti­mated as 50-100. g CH4 m -2 yr-l or 3.1-6.2 molm-2 yr-l). However, our results suggest this may bea substantial overestimate: we measured a mean an­nual production rate of only 80 mol d- I (1.8 mmolm-2 yr-l).

This suggests shallow continental shelf sedi­ments are likely to be less important in global CH4

cycling than previously thought. This conclusion isnot surprising in that freshwater sediments (which,as noted above, were used by Ehhalt (1974) to es­timate shelf fluxes) are much more likely than ma­rine sediments to release significant amounts ofCH4 to the water column (e.g., Bartlett and Harriss1993). This difference is due to the dominance ofsulfate reduction rather than CH4 production inthe decomposition of organic matter in the upperportions of marine sediments.

Finally, there is added uncertainty to these esti­mates due to the recently documented dischargeof methane-laden groundwater to nearshore waters(Bugna et al. 1996) and the presence of extensivemethane-rich, gas-charged sediments along conti­nental margins (e.g., Karisiddaiah and Veerayya1994). Clearly, there is a need for more data froma range of shallow shelf systems to unambiguouslydetermine the role of these environments on theglobal CH4 cycle.

Conclusions

The data presented here allow the cycling ofCH4 in Tomales Bay to be examined on a whole-

Methane in Tomales Bay, California 75

system basis. During summer the inner TomalesBay CH4 budget is dominated by sediment-seawaterfluxes and evasion to the atmosphere. In winterthe benthic fluxes are much smaller, export to theouter bay and coastal waters increases, and CH4

inputs from freshwater runoff become dominant.The principal uncertainties in the simple models

presented here are the errors associated with in­dividual fluxes. Also, the flux from vegetated sed­iments needs further study to be accurately as­sessed. Nevertheless, this study demonstrates theimportance of seasonality on the dynamics of thissystem, particularly with respect to the temporalvariability of terrestrial runoff and benthic fluxes.In particular, it underscores the importance· ofland-derived inputs of CH4 to estuarine systems. Fi­nally, it also illustrates the advantages of studyingan estuary with tractable topography and hydro­dynamics, in that it allows the assessment of whole­system dynamics, an objective that is difficult to ac­complish with more complicated systems.

ACKNOWLEDGMENTS

We thank Tim Hollibaugh, Nancy Craig, Randy Chambers,Mandy Joye, Josh Plant, Rich Muller, and Sue Vink for all theirhelp on numerous field expeditions; Ted Walsh for analysis ofdissolved inorganic nutrients; Randy Chambers for vivid first­hand observations of dairy farm organic matter inputs to theinner bay; and three anonymous reviewers for many helpful sug­gestions. This work was supported by National Science Foun­dation grant OCE-8914833 (to S. V. Smith andJ. T. Hollibaugh)and Office of Naval Research grant NOOOI4-94-1-0631 (to F. J.Sansone).

LITERATURE CITED

ALLDREDGE, A. L. AND Y. COHEN. 1987. Can microscale chemicalpatches persist in the sea? Microelectrode study of marinesnow, fecal pellets. Science 235:689--691.

ALLER, R. C. 1978. The effects of animal-sediment interactionson geochemical processes near the sediment-water interf;lce,p. 157-172. In M. L. Wiley (ed.), Estuarine Interactions. Ac­ademic Press, New York.

ALLER, R. C. 1980. Quantifying solute distributions in the bio­turbated zone of marine sediments by defining an averagemicroenvironment. Geochimica et Cosmochimica Acta 44: 1955­1966.

ALLER, R. C. 1984. The importance of relict burrow structuresand burrow irrigation in controlling sedimentary solute dis­tributions. Geochimica et Cosmochimica Acta 48:]929-1934.

ATKINSON, L. P. AND F. A. RICHARDS. 1967. The occurrence anddistribution of methane in the marine environment. Deep-SeaResearch 14:673--684.

BARTLETT, K. B. AND R. C. HARRISS. 1993. Review and assess­ment of methane emissions from wetlands. Chemosphere 26:261-320.

BROECKER, W. S. AND T. H. PE:-.IG. 1974. Gas exchange ratesberween air and sea. Tellus 26:2]-33.

BROOKS, J. M., D. F. REID, A.'iD B. B. BER."ARD. 1981. Methanein the upper water column of the Northwestern Gulf of Mex­ico. jaumal oj Geoph)'sical Research 86: 11,029-11,040.

Bl!GNA, G. c., J. P. CHA:>iTON, J. E. CABLE, W. C. BL'R.NETT, ~"DP. H. CABLE. 1996. The importance ofgroundwaterdischarg­es to the methane budgets of nearshore and continental shelf

76 F. J. Sansone et al.

waters of the northeastern Gulf of Mexico. Geochimica et Cos­mochimica Acta 60:4735-4746.

CHANTON,]. P. AND C. S. MARTENS. 1988. Seasonal variations inebullitive flux and carbon isotopic composition of methanein a tidal freshwater estuary. Global Biogeochemical Cycles 2:289­298.

CICERONE. R.]. AND R. S. OREI-ILAND. 1988. Biogeochemicalaspects of atmospheric methane. Global Biogeochemical C)'cles 2:299-327.

DACEY,]. W. H. AND M.]. KLUG. 1979. Methane emission fromlake sediment through water lilies. Science 203:1253-1255.

DE ANGELIS. M. A. AND M. D. LILLEY. 1987. Methane in surfacewaters of Oregon estuaries and rivers. Limnolo~ and Ocean­ography 32:71&-722.

DOLLAR. S.].. S. V. SMITH. S. M. VINK. S. OBRESSKl, A."D]. T.HOLLIBAUGH. 1991. Annual cycle of benthic nutrient fluxesin Tomales Bay. California, and contribution of the benthosto total ecosystem metabolism. Marine Ecolo~ Progress Series 79:115-125.

DWGOKENCKY, E.].. K. A. MASARlE, P. M. L>'NG, P. P. TA."S. 1.. P.STEEl.E. A."D E. G. NISBET. 1994. A dramatic decrease in thegrowth rate of atmospheric methane in the northern hemi­sphere during 1992. Geophysical &.search Leiters 21 :45-48.

EH~T. D. H. 1974. The atmospheric cycle of methane. Tel/u.s26:58-70.

EMERSON. S.. R. JAHNKE, AND D. HEGGIE. 1984. Sediment-waterexchange in shallow water estuarine sediments. Journal ofMa­rine Research 42:709-730.

ETHERIDGE. D. M., G. I. PEAR."IAN, AND P. ]. FRASER. 1992.Changes in tropospheric methane between 1841 and 1978from a high accumulation-rate Antarctic ice core. Tel/us 44B:282-294.

FISHER, D. T., S. V. SMITH. AND R. R. CHURCHILL. 1996. Simu­lation of a century of runoff across the Tomales watershed.Marin County. California. Journal of Hydrolo~ 186:253-273.

HA."IMOND. D. E., H. G. SIMPSON, A.'1D G. MATHIEU. 1975. Meth­ane and radon-222 as tracers for mechanisms of exchangeacross the sediment-water interface in the Hudson River Es­tuary, p. 119-132. In T. M. Church (ed.), Marine Chemistryin the Coastal Environment. ACS Symposium Series 18. Amer­ican Chemical Society. Washington. D.C.

HARTMA.". B. AND D. E. HAMMOl':D. 1984. Gas exchange ratesacross the sediment-water and air-water interfaces in southSan Francisco Bay. Journal ofGeophysical Research 89:3593-3603.

HARTMA...... B. AND D. E. HAMMOND. 1985. Gas exchange in SanFrancisco Bay. Hydrobiologia 129:59-68.

HElP. C. H. R.. N. K. GoOSEN. P. M.]. HERMA.",]. KROMKAMP,].]. MIDDELBURG, AND K. SoERA.ERT. 1995. Production and con­sumption of biological particles in temperate tidal estuaries.p. 1-149. In A. D. Ansell, R. N. Gibson, and M. Barnes (eds.),Oceanography and Marine Biology: An Annual Review. Ab­erdeen University Press. Aberdeen, Scotland.

HIGGINS. 1.].. D.]. BEST. R. C. HAMMO:"D, AND D. ScaTI'. 1981.Methane-oxidizing microorganisms. Microbiological Reviews 45:55&-590.

HOLLlBACGH,]. T.. B. E. COLE, S.]. DOl.LAR. S. W. HAGER. S. M.V''''K. W.]. KIM~IERER.S. OBREBSKI, S. V. SMITH, M. VALENTINO,A:"b T. W. WALSH. 1988. Tomales Bay, California: A macro­cosm for examining biogeochemical coupling at the land-seainterface. Eos 69:843-845.

jA:":"ASCH. H. W. 1975. Methane oxidation in Lake Kivu (cen­tral Africa). Limnolog)' and Oaanography 20:860-864.

j0RGE:"SE:". B, B. 1983. Processes at the sediment-water inter­face. p. 477-509. In B. Bolin and R. B. Cook (eds.), The MajorBiogeochemical Cycles and Their Interactions. Wiley. Chich­ester. United Kingdom.

KA....."'ISHER.]. 1974. On the exchange of gases between theatmosphere and the sea. Deep-Sea Research 10:860-864.

K-\RISIDDAIAH. S. M. AND M. VEERAYY>.. 1994. Methane-bearing

shallow gas-charged sediments in the eastern Arabian Sea: Aprobable source for greenhouse gas. Continental Shelf Research14:1361-1370.

LAMBERT, G. AND S. SCHMIDT. 1993. Reevaluation of the oceanicflux of methane: Uncertainties and long-term variation. Che­mosphere 26:579-589.

LAMONTAGNE, R. A.,]. W. SWiNNERTON, V.]. LINNENBOM, AND W.D. SMITH. 1973. Methane concentrations in various marineem;ronments. Journal of Geophysical Research 78:5317-5324.

LASSEY, K. R., D. C. Lo\\''E, M. R. MA."NING, A."D G. C. WAGHORN.1992. A source inventory for atmospheric methane in NewZealand and its global perspective. Journal of Geoph)'sical Re­search 97:3751-3765.

LILLEY, M. D.• M. A. DE ANGELIS. A.'1D E.]. OLSON. 1996. Meth­ane concentrations and estimated fluxes from Pacific North­west rivers. Mitteilungen - Intemationale Vereinigungfur Theoretis­che und Angewandli! Limnologie 25:187-196.

Llss, P. S. 1973. Processes of gas exchange across an air-waterinterface. Deep-Sea &.search 20:221-238.

Llss, P. S. 1983. Gas transfer: Experiments and geochemicalimplications, p. 241-298. In P. S. Liss and W. G. N. Slinn(eds.), Air-Sea Exchange of Gases and Particles. Reidel, TheNetherlands.

LIss, P. S. AND P. G. SLATER. 1974. Fluxes of gases across theair-sea interface. Nature 247:181-184.

MARTENS, C. S. AND R. A. BERNER. 1977. Interstitial water chem­istry of anoxic Long Island Sound sediments. I. Dissolved gas­es. Limnolo~ and Oceanography 22: I0-25.

MARTENS, C. S. AND.J. V. KLUMP. 1980. Biogeochemical cyclingin an organic-rich coastal marine basin. 1. Methane sedimelll­water exchanges. Geochimica et Cosmochimica Acta 44:471-490.

MOLLER. P.]. AND E. SUESS. 1979. Productivity, sedimentationrate. and sedimentary organic matter in the oceans--1. Or­ganic carbon preservation. Deep-Sea Research 26A:1347-1362.

PLA.,,<T.]. N. 1995. Effects of land use change on Tomales Bay.California: New ideas from old tracers. Masters Thesis. Uni­versity of Hawaii, Manoa. Hawaii.

RAsMUSSEN, R. A. AND M. A. K. ~IL. 1981. Atmosphericmethane (CH,): Trends and seasonal cycles. Journal of Gecrphysical Research 86:982&-9832.

ROONEY, ].]. 1995. Determination of sedimentation rates in.Tomales Bay. California. using a geographic information sys­tem. Masters Thesis. University of Hawaii. Manoa. Hawaii ..

RL'DD,]. W.. A. FURL'TANI. R. ]. FLETI', AND R. D. HAMILTO:".1976. Factors controlling methane oxidation of Shield Lakes:The role of nitrogen fixation and oxygen concentrations.Lim­nolo~ and Oceanography 21:357-364.

RUDD.]. W. AND R. D. HAMILTON. 1978. Methane cycling in aeutrophic shield lake and iL~ effects on whole lake metabo­lism. Limnology and Oceanography 23:337-348.

RUST. T. M. 1993. The relative importance of processes con­trolling the distribution of methane in Tomales Bay, Califor­nia. Masters Thesis. University of Hawaii. Manoa. Hawaii.

SAl\'SONE, F.]. A."D C. S. MARTE~S. 1978. Methane oxidation in. Cape Lookout Bight. North Carolina. Limnolog)' and Oceanog­

raph), 23:349-355.SAl\'SO"'E. F.]. A.'1D C. S. M.>,RTE~S. 1982. Volatile fally acid cy­

cling in organic-rich marine sediments. Geochimica et Cosmcrchimic.a Acta 46:1575-1589.

SASS. R. L.. F. M. FISHER. F. T. TL·R:"ER. AND M. F. jl·:\D. 1991.Methane emission from rice fields as influenced by solar ra­diation. temperature. and straw incorporation. Global Biogecrchemical Cycli!s 5:335-350.

SIEBl·RTH.]. Me!\:. 1987. Contrary habitats for redox specificprocesses: Methanogenesis in oxic waters and oxidation inanoxic waters. p. 11-3. In M. A. Sleigh (ed.). Microbes in theSea, Wiley and Sons, New York.

SMITH. S. v., W.]. KIMMERER. E. A. LAws. R. E. BROCK. A:\D T.W. WALSH. 1981. Kaneohe Bay sewage diversion experiment:

Perspectives on ecosystem responses to nutritional perturba­tion. Pacific Science 35:279-395.

SMITH, S. v., W. j. WIEBE, j. T. HOLLIBAUGH, S. j. DOllAR, S. W.HAGER, B. E. COLE, G. W. TRIBBLE, AND P. A. WHEELER. 1987.Stoichiometry of C, N, P, and Si fluxes in a temperate·dimateembayment. journal of Marine Research 45:427-460.

SMITH, S. V.,j. T. HOLLIBAUGH, S.j. DOLLAR, AND S. VINK. 1989.Tomales Bay, California: A case for carbon<ontrolled nitro­gen cycling. Limnology and Oceanography 34:37-52.

SMITH, S. V., j. T. HOLLIBAUGH, S. j. DOLLAR, AND S. VINK. 1991.Tomales Bay metabolism: C-N-P stoichiometry and ecosystemheterotrophy at the land-sea interface. Estuarine and CoastalShelf Science 33:223-257.

SMITH, S. v., R. M. CHAMBERS, AND j. T. HOLLIBAUGH. 1996.Dissolved and particulate nutrient transport through a coastalwatershed-estuary system. journal of Hydrology 176:181-203.

SMITH, S. V., AND j. T. HOLI.lBAUGH. 1997. Annual cycle andinterannual variability of ecosystem metabolism in a temper­ate climate embayment. Ecological Monographs 67:509-533.

Methane in Tomales Bay, California n

SPRATT, j. D. 1989. The distribution and density of eelgrass,Zostera marina, in Tomales Bay, California. California Fish andGame 75:204-212.

TILBROOK, B. 1991. Methane Sources and Sinks in UpperOcean Waters. Ph.D. Dissertation, University of Hawaii, Ho­nolulu, Hawaii.

THOMPSON, A. M., K. B. HOGAN, AND j. S. HOFFMAN. 1992. Meth·ane reductions: Implications for global warming and atmo­spheric chemical change. Atmospheric Environment :!6A:2665­2668.

WEISENBERG, D. A. AND N. L. GUINASSO. 1979. Equilibrium sol­ubilities of methane, carbon monoxide, and hydrogen in wa­ter and seawater. journal of Chemical Engineering Data 24:356­360.

YAMAMOTO, S., j. B. ALCAUSKAS, AND T. E. CROZIER. 1976. Sol·ubility of methane in distilled water and seawater. jouT7Ul1 ofChemical Engineering Data 21:7~0.

Receive~ f01" consideration, August 5, 1996Accepted f01" publication, june 18, 1997