DISTRIBUTION, ABUNDANCE, AND IMPLICATIONS FOR CONSERVATION OF WINTER WATERBIRDS ON ,

JOHN P. KELLY AND SARAH L. TAPPEN, AudubonCanyon Ranch, Cypress Grove Preserve,Marshall, California 94940

TomalesBay, on the Marin Countycoast, is one of California'slargest and least disturbedestuaries, yet little has been publishedon the statusof its waterbirdsor their vulnerabilityto environmentalthreats. Shuford et al. (1989) reportedthe resultsof waterbirdcensuses of up to 10 consecutive years in the adjacentPoint Reyes area between 1965 and 1982 but providedlimited informationon waterbirduse of TomalesBay. Other anecdotalaccounts and aerial survey work over the last30 yearssuggest that TomalesBay supports unusually high abundances of somewaterbird species (Moffitt1943, Grinnelland Miller 1944, Calif.Dept. Fishand Game unpubl. data,U.S. Fishand Wildlife Service unpubl. data, National Audubon Society ChristmasBird Counts)and is worthyof protection(National Oceanic and AtmosphericAdministration 1987, Neubacheret al. 1995). In thispaper, we presentresults from sevenwinters of baywidewaterbird censuseson TomalesBay, compareresults with other studiesto providea historicaland geographicalperspective on the importanceof TomalesBay to waterbirds,evaluate species distributions within the bay with regardto possiblehabitat relationships and the importanceof particularhabitat areas, and, on the basisof theseevaluations, identify needs for conservationof winterwaterbird populations on TomalesBay. We addressthese objectives with regardto all waterbirdspecies associated with TomalesBay and immediatelyadjacent ponds and marshes, with the exceptionof sandpipers (Scolopacidae),plovers (Charadriidae), and large and medium-sizedgulls (Laridae;Table 1). Most gullswere excludedfrom the studybecause of difficultiesin countinglarge numbersof moving birds and identifying species.Sandpipers, plovers, and gullswere alsodifficult or impossibleto observefrom surveyboats because these speciesoften concentratedon exposedtidal flats or in alternativehabitats away from the bay.Abundances of wintershorebirds and gulls on TomalesBay have been reported elsewhere (Kelly 1993, Kellyet al. 1996).

STUDY AREA

The longstraight expanse of TomalesBay floodsthe lower20 km of the fault-generated Olema Valleyon the centralCalifornia coast, about 45 km northwestof SanFrancisco (Figure 1; Galloway1977). The baydiffers from other, generallyshallower, Pacific coast estuaries and lagoonsin havinga muchgreater area of open water at low tide and thereforemore waterbird habitatthrough the tidalcycle. Approximately 91% of the bay's28.5-km z areais subtidal. With hillssurrounding the narrow(1-2 km)bay, and little fetch whenwinds blow from directions other than northwest,surface waters remain relativelycalm duringwinter. Tomales Bay contains37 eelgrass(Zostera marina)beds covering a total area of 392 hectares,predominanfiy in the Western Birds 29:103-120, 1998 103 •••••••00•00•00 •00•0•00000•000000•00000

0•0•0•0•0•0•0•0000••0• •000•0••000•000000•00000

•••0•0•0•00••0• ••0•00•0•000•000000•00000

•0•00•0••0• •000•00••0• ...... 000000000•0•00•0000000000•00

00000•000•00•00000000•0•00 00•0•0•00••0•0••0•0000

00•0•••0••••0•0•0

00•00•0•000•00000000•000000•

••••••••oooooo•

00•000•0•000•000•00000•00000•

00•00•00•0•00000••000•

•••000••0•••000• •0•0••00•0•0•00•00•000•

•••00••0••0•0000• •0•0••00••0•00•0000000 WINTER WATERBIRDS ON TOMALES BAY, CALIFORNIA

;and Point

Walker Tom's Point

oint

Pacific Routes of Ocean ObservationBoats

PointReyes NationalSeashore Point

0 2 4 km I,,,, I I Lagunita., Creek

Figure1. TomalesBay studyarea. Shadingsdenote winter waterbird count sections; parallellines denote three boat transects. Supplementary counting points (circled) are (1) Walker Creek, (2) LivermoreMarsh, (3) MillertonGulch to Bivalve,(4) Bivalve,and (5) Inverness. northernhalf of the bayfrom Pelican Point northward to Tom'sPoint (Figure 1, Spratt 1989). Water depthsin the bay averageabout 3 m below mean lowerlow water (MLLW) and vary from intertidal shoals along the eastshore

106 WINTER WATERBIRDS ON TOMALES BAY, CALIFORNIA to areasup to 20 m deepin the mainchannel along the west shore. In general, sedimentsgrade from primarily fine to coarsesand in the northernreaches of the bay to muddiersubstrates in southernportions (Daetwyler 1966). The two primarypoints of freshwaterinflow, and the largesttidal fiats and marshes,are at LagunitasCreek at the southend of the bay and Walker Creek near the north end (Figure 1). Numerous smaller perennial and ephemeralstreams enter the bay alongits perimeter,each associated with a small,generally insular delta marsh. Rainfall is highlyseasonal, with 95% of annualrain fallingfrom October throughApril and 55% from December throughFebruary, based on 42-year averagesat CypressGrove Preserve. Spatialand temporal variations in the bay'ssalinity are influencedby variably high levelsof freshwaterinflow duringwinter, low flows in summer,and constraintson tidalexchange with the oceanimposed by the linearshape of the bay and the narrow mouth at its northwestend. Thereforesalinities in the southernreaches of the bay are highlyvariable, ranging from nearly freshafter heavywinter runoff to slightlyhypersaline in summer,whereas regulartidal mixing at the north end of the bay maintainssalinities that consistentlyreflect those of the outercoastal waters (Hollibaugh et al. 1988). The differencebetween mean high and mean low tidesis about1.1 m, with an averageannual maximum tide swingof about2.5 m, basedon National Oceanic and AtmosphericAdministration predictions. Tidal currentsare strongestalong the west shorewithin about6 km of the northwestend but are generallygentle.

METHODS

We conducted21 (mean3/yr, minimum1, maximum4) baywidecounts of waterbirdson TomalesBay overthe sevenwinters from 1989-90 to 1995- 96. Eachyear, consecutive counts were made at 2- to 3-weekintervals from mid-Decemberto lateFebruary each year. Each baywide survey encompassed simultaneouscounts conducted from three 17- to 21-ft (5.3- to 6.6-m) Boston Whalersor similarboats. The boatstraveled in formation,following parallel 18-kmtransects from MillertonPoint near the southend of the baynorth to the SandPoint Buoy (Figure 1). Boatsfollowing the east-shoreand west-shore transectseach carried one team of observers.The midbayboat carriedtwo (portand starboard)teams. Each team consistedof two competentwaterbird observers,using binoculars, and one data recorder.Observers on the east- shoreboat countedall waterbirdsin a transectbounded by the path of their boat and the eastshore. Similarly, observers on the west-shoreboat counted all waterbirdsin a transectbounded by their boat and the west shore. Observerson the midbayboat countedall waterbirdsbetween the other two boats.Drivers of the east-and west-shoreboats continually adjusted their positionsvisually to maintain roughlyequal distancesbetween boats and transectboundaries. Therefore, shoreline transects each represented approxi- matelya fourthof the arealextent of the bayand the midbaytransect covered approximatelyhalf of the bay'sextent. We maintainedcontinual radio communication among observation boats to facilitateadjustments of interboat and boat-to-shoredistances and to clarify counts of confusingbirds or flocks along the transectborders.

107 WINTER WATERBIRDS ON TOMALES BAY, CALIFORNIA

Observerson each boat countedonly those birdsthat passedsouthward throughan imaginaryvertical plane defined by the bowsof the threeboats. The boatstraveled at velocitiesof about4 knots,but occasionally we stopped to countdense aggregations of birds.Birds flushing ahead of the moving boatswere not counteduntil they eventuallypassed southward through the observationplane. Birdspassing northward over or aroundthe boatswere subtractedfrom the counts,but this seldomoccurred as long as boats maintaineda constantvelocity. Countswere generallyconducted during rising neap tides, with at leasta 4.5-ft tideheight (MLLW) when the boatsarrived at the shallowWalker Creek Delta/Tom'sPoint area, abouttwo hoursafter starting each count. We began the countsbetween 0930 and 1000 to avoidearly morning fog and rough afternoonseas, and each count was completed in about3.5 hours.Conditions duringcounts were generally calm (Beaufort scale 0-2). In areasthat could not be reachedby boats,we conductedsupplementary counts immediately prior to the boatcounts, using telescopes from shore(Figure 1). Duringeach count, we partitionedthe rawtransect data into four sections: (1) southof TomasiniPoint; (2) TomasiniPoint to PelicanPoint; (3) Pelican Pointto Tom'sPoint; (4) Tom'sPoint to SandPoint (Figure1). The sections correspondto intervalsdistinguished by estuarineconditions that might influencethe composition,distribution, or availabilityof foodfor waterbirds. Suchconditions include differences in waterturbidity and color(pers. obs.), eelgrassdistribution (Spratt 1989), substratetexture (Daetwyler 1966), water temperature, freshwater inflow, and ocean-water exchange (Hollibaughet al. 1988). Therefore,data from eachcount were partitioned into 12 subsets,delineated by threetransects each divided into four sections. Supplementalshore-based data were pooled with the appropriatesection and transect counts. We recorded all waterbirds in the shallows at the extremesouth end of the bay(south of MillertonPoint) as occurring in either east-shore or west-shore transects. We identifiedwaterbirds to specieswhenever possible. Undifferentiated loons,Horned/Eared grebes,Western/Clark's grebes, cormorants, scaup, and scoterswere allocatedto speciesin proportion to the number of identifiedbirds of thosespecies, within each transect section on eachcount day, if the numberof identifiedindividuals exceeded the numberof undiffer- entiatedindividuals and the numberof identifiedindividuals was greater than 50 for grebesor 100 for cormorants,scaup, or scoters.Generally, only very small numbers of birds were undifferentiated in the field. To facilitatecomparisons among countareas of differentsizes, we trans- formed speciesabundances within each countarea (sectionwithin transect) intodensities (birds/km•). We modeleddifferences in waterbirddensities using a mixed-modelanalysis of covariance(ANCOVA), designatingsection and transectas fixedeffects and year (winter season) as a randomeffect. Species thatcommonly exploit freshwater habitats in winter(43% of speciesanalyzed) couldexhibit intraseasonal fluctuations as a resultof variablerainfall patterns and correspondingchanges in the availabilityof freshwaterwetlands in this region (Shufordet al. 1989). Furthermore,late migrantsof some species couldarrive during the wintercensus period. Therefore, we useddays (since 1 Decemberwithin each winter) as a covariateto controlfor possibledifferences

108 WINTER WATERBIRDS ON TOMALES BAY, CALIFORNIA relatedto the intraseasonaltiming of counts.Species densities were log- transformedprior to analysisto improvethe normalityof the data.For species with significantsection or transecteffects on densities,we usedthe Tukey procedurefor multiplepairwise comparisons to determinewhich section or transect means differed from others (Dixon 1992). We also conducted pairwisecomparisons of proportionalspecies diversity, based on the Shannon index (Magurran 1988), and speciesrichness, but becausesections and transectsvaried in extent,we limitedour interpretationof differencesto those thatwere not confounded by possible species-area effects. Finally, we looked for possibletemporal trends by examiningpartial regression coefficients of baywideabundances on year and days.

RESULTS

We observed51 speciesof waterbirdsduring our surveys on TomalesBay. The total number of waterbirdsaveraged 21,943 individualsper count [standarderror (SE) 647, min. 14,842, max. 25,553]. Mean waterbird densityon TomalesBay was 770 birds/kmz (SE 22.7). Variabilityin total numbersof waterbirdswas primarilya functionof the numbersof the Surf Scoter,Bufflehead, Greater Scaup, Ruddy Duck, and Black Brant (Figure 2, Table 1). Waterfowl(Anatidae) accounted for 51.0% of the speciesand 85.3% of the total individualsobserved; the Surf Scoter,Bufflehead, and Greater Scaup together accountedfor 70.0% of the total waterbirds. Surface-feedingducks, such as the NorthernPintail and Mallard,typically occurredin only low numbers. All waterbirdspecies for which our data were adequatefor parametric analysisselected habitats within Tomales Bay nonrandomly with respect to the transectsor sectionsused in thisstudy (P < 0.05; Table2). Area countswere generallynot biasedby the effectsof observationboats on movementsof waterbirdsbecause birds rarely flushed across count-area boundaries. Results forthe Greater Scaup and Black Scoter included a slighttransect bias because theyoccasionally flushed from east-shore to midbaywaters before they were counted,but they did not flushcross section boundaries. In general,waterbird densitieswere highestalong the eastshore and betweenPelican Point and Tom's Point (Figure2, Table 2). However,several species preferred other areas(Figure 2, Tables1 and 2). For example,Pacific and Red-throatedloons andWestern Grebes occurred in significanfiygreater densities in the midbay transect and in the section from Tomasini Point to Pelican Point. Horned and Earedgrebes concentrated in the middletwo sectionsof the bay.Red-necked Grebesoccurred almost exclusively north of PelicanPoint, especially in the deep channelnear the west shore (Table 1 and pers. obs.).Black Brant concentratedin the northernthird of the bay,with highest densities between PelicanPoint and Tom's Point (Figure 2, Table2). In contrast,Greater Scaup concentratedsouth of PelicanPoint, especially along the eastshore north of TomasiniPoint. Black Scoters also concentrated along the eastshore north of TomasiniPoint (Table1 and pers. obs.). Becausespecies diversity and richnessvaried positively with sectionsize, we couldnot determineif sectiondifferences were independent of possible species-areaeffects. We also foundno significantdifferences in propor-

109 WINTER WATERBIRDS ON TOMALES BAY, CALIFORNIA

RelativeDensity (birds/km 2)

Species Sections Transects

2 3 4 W M E

Red-throated Loon

Pacific Loon

Common Loon

Horned Grebe

Eared Grebe

Western Grebe

Clark's Grebe

Double-crested Cormorant

Brandt's Cormorant

Pelagic Cormorant

Black Brant

Amedcan Wigeon

Greater Scaup

Surf Scoter

White-winged Scoter

Common Goldeneye

Buffiehead

Red-breasted Merganser

Ruddy Duck

American Coot

Forster's Tern

Waterbird Total

0% 100% 0% 100%

Figure 2. Relativewaterbird densities by sectionand transect(Figure I) and mean baywideabundances (standard errors) of waterbirdson TomalesBay. Sections:1, southof TomasiniPoint; 2, TomasiniPoint to PelicanPoint; 3, PelicanPoint to Tom's Point; 4, Tom's Point to Sand Point. Transects:W, west shore; M, midbay; E, east shore.Differences among section and transectdensities are analyzedin Table 2.

11o WINTER WATERBIRDSON TOMALES BAY, CALIFORNIA

Table 2 Effectsof Section,Transect, and Year on WaterbirdDensities in Tomales Bayø

Section Transect

Species ANCOVAb 1 2 3 4 E.shore MidbayW. shore

Red-throated Loon S**,T A C B B A B A PacificLoon S**,T*, ST** A B A A A B A Common Loon S**,ST** A B B B Horned Grebe S**,T*, ST**,TY A C B, C B B B A Eared Grebe S**,T**, Y*, ST**,SY* A C C B A C B Western Grebe S**,T**, Y**, ST** C D B A B B A Clark's Grebe S**,T*, Y, ST**,SY* B B A A B B A Double-crested Cormorant S**,T**, ST** B C B,C A A A A Brandt's Cormorant S**,T**, ST** A B B B A A A Pelagic Cormorant S**,T**, ST** A B C C A A B Black Brant S**,T*, ST**, SY* A B D C A B A, B American Wigeon S**,T**, ST** B A, B B A B A A GreaterScaup S**,T*, Y, ST** B, C C A, B A A A A Surf Scoter S**,ST** A B B B White-winged Scoter S**,T**, ST**, TY A B C C B B A Common Goldeneye S**,T*, Y*, ST** A B B, C C B A B Buffiehead S**,T**, Y*, ST** B A C B B A A Red-breasted Merganser S**,ST**, TY A B B C RuddyDuck S**,T**, ST** C B C A C A B American Coot S**,T**, ST**, TY* A A B A C A B Forster's Tern S, T*, TY** A A A A B B A Waterbird Total S**,T**, ST** A B C A, B C B A

aEachsignificant main effect for section or transectis followed with multiple pairwise comparisons: means with the sameletter are not significantlydifferent (Tukey procedure, experimentwise P > 0.05). Significantdifferences among means are ranked:A < B < C < D. Sectionand transect preferences, suggestedby significantlyhighest means and means that do notdiffer from significantly highest means, are in boldface.See Figure1 for sectionand transect locations. bMixed-modelanalysis of covariance.Covariate, days since 1 Decemberwithin year (winter season); letter indicatesF ratiosignificant at P < 0.05, *P < 0.01, **P < 0.001.

tional speciesdiversity among transects(P > 0.10). However,species richnesswas significantly greater along the east shore (mean 31.0, SE 0.94, n = 21) than in eitherthe midbay(mean 24.4, SE 0.82, n = 21) or west- shore(mean 24.4, SE 0.62, n = 21) transects--inspite of the largerarea representedby the midbaytransect (experimentwise P < 0.01). We foundsignificant seven-year population trends (P < 0.05) reflecting annualincreases for the Common Loon (mean annualincrease, b = 15.8),

111 WINTER WATERBIRDS ON TOMALES BAY, CALIFORNIA

EaredGrebe (b -- 38.3), WesternGrebe (b -- 93.6), BlackBrant (December/ Januaryonly, to excludenorthward migrants: b -- 231.6), CommonGolden- eye (b -- 15.6), Red-breastedMerganser (b -- 17.0), and AmericanCoot (b -- 65.0). Partialregression coefficients revealed slight but significant (P < 0.05) 7-yeardeclines for the WhitePelican (b -- -2.1) andWhite-winged Scoter (b -- -3.9) and a largeannual decrease for the Surf Scoter(b: -515.9). However, the significantregression for SurfScoter resulted from relatively high numbers in 1989-90 (mean 10,470; SE 347.6, n -- 2); thereafter,abundances remainedat significantlylower levels (mean 6315; SE 1601, n -- 19; t -- 8.1, P < 0.001) with no annualtrend (P > 0.05). Similarly,the dramatictrend in winteringBlack Brant resulted primarily from an earlyincrease, from 1989- 90 (mean 17.5, SE 5.5, n -- 2) to 1991-92 (mean 1342; SE 356.5, n: 2); December/Januaryabundances apparently stabilized at this level for the remainderof the study(mean 1275; SE 120.0, n: 12). Regularlyoccurring species with the greatestvariation from year to year were the PacificLoon [coefficientof variation(CV) 0.83], Pied-billedGrebe (CV 1.02), White Pelican(CV 1.67), Brown Pelican(CV 0.88), Bonaparte's Gull (CV 0.90), and Forster'sTern (CV 0.97). In contrast,we observed relativelystable annual abundancesof the Red-throatedLoon (CV 0.37), Double-crestedCormorant (CV 0.30), Buffiehead(CV 0.25), andRuddy Duck (CV 0.36). We foundsignificant intraseasonal trends in the abundancesof only a few species.The WhitePelican, Bonaparte's Gull, Buffiehead, and Ameri- canWigeon declined significantly as winter progressed, while the EaredGrebe and White-wingedScoter increased significantly (P < 0.05).

DISCUSSION TomalesBay Populations The numberof waterbirdspecies we observedon TomalesBay wassimilar to that normally occurringin other wetlandsalong the outer Marin and SonomaCounty coast (Shuford et al. 1989). Densitiesof manyspecies were alsosimilar to thosereported for otherestuaries in the PointReyes area, but waterbirdabundances were dramatically higher, presumably because of the greaterextent of TomalesBay in comparisonto other sites(cf. Page et al. 1983, Shuford et al. 1989). In additionto the high waterbirdnumbers documentedby our study,Tomales Bay alsosupports 11,000 to 18,000 winteringshorebirds (Kelly 1993) and tensof thousandsof winteringgulls (Kellyet al. 1996 and unpubl.data). TomalesBay is probablyof statewidesignificance for some waterbird species.Population estimates derived from the aerialMidwinter Waterfowl Surveyduring the sameperiod (U.S. Fishand WildlifeService, unpubl. data) are not directlycomparable with the data we collectedfrom boats on TomalesBay. Waterbirdabundances recorded from boats are basedon relativelythorough observation, frequently using actual counts rather than estimates,and are probablycloser to the actual numbersof birds than abundancesrecorded from aerial surveys (Conant et al. 1988). Comparisons of our Januaryboat count data with JanuaryU. S. Fishand WildlifeService (USFWS)aerial surveys of TomalesBay, availableonly for 1990, 1991, and 1993, suggestthat aerial countsunderestimated numbers of Bufflehead,

112 WINTER WATERBIRDS ON TOMALES BAY, CALIFORNIA scoters,and Black Brant by 49.5%, 41.5%, and 17.4%, respectively(see alsoConant et al. 1988). Usingthese differences to adjustcomparisons with statewideresults from 1989-90 to 1995-96 reported by the USFWS MidwinterWaterfowl Survey, we estimatethat winter populationsof the Bufflehead,scoters, and BlackBrant (December/Januaryonly) on Tomales Bay represent12.3%, 6.4%, and 30.8%, respectively,of statewidepopula- tions.We emphasizethat theseestimates are roughapproximations at best and shouldbe interpretedcautiously. In general, our data suggestthat TomalesBay supportsnumbers of Buffiehead,scoters, and Black Brant comparableto totalsfor all estuariesand lagoonsin Mendocino,Humboldt, and Del Norte countiescombined. Tomales Bay may be the singlemost importantestuary, with the exceptionof San FranciscoBay, southof the ColumbiaRiver for winteringBufflehead (cf. USFWS MidwinterWaterfowl Surveyunpubl. data; Monroe 1973, Barnhartet al. 1992). On the basisof aerialsurveys conducted from 1968 to 1970 by the CaliforniaDepartment of Fish and Game and the Point ReyesBird Observatory,Smail (1972) reported high countsof scoters(5788), Buffiehead(7502), and scaup (1437) and overall waterfowl abundances(mean 6786, max. 19,997), providingsome evidencethat waterbirdnumbers on TomalesBay were similarto those observedin our study. In a statewidestudy of seabird abundances,Briggs et al. (1987) identifiedTomales Bay as an important area of concentrationfor the Red-throated, Pacific, and Common loons, Eared and Horned grebes,and scoters. AudubonChristmas Bird Count (CBC) resultsfor TomalesBay prior to 1989 providerough estimates at bestfor waterbirdspecies, because of the near impossibilityof estimatingbaywide abundances from the singleobser- vationboat usedon thosecounts. However, specieswith restricteddistribu- tions in the bay may be monitoredmore accuratelyby CBC data. For example,comparisons of our resultswith CBC datasuggest that the number of AmericanCoots has declined by an orderof magnitudesince the 1970s, a patternthat is alsoevident elsewhere in the Point Reyesarea (Shufordet al. 1989). The mediannumber of Redheadsrecorded on the CBC from 1970 to 1981, primarilyfrom nearthe mouthof WalkerCreek on TomalesBay, was 376 (min. 1, max. 784) but has sincedwindled to 9 (min. 0, max. 15). Since 1989, CBC data for TomalesBay have been derivedfrom Decembercounts conducted for thisstudy. We haveassumed, except where noted otherwise, that our datarepresent "wintering"populations only, but abundancesof some speciesmay have swelledor declinedbecause of arrivalsor departuresof late migrantsor birds with intraseasonallocal or regionalmovements. The significantintraseasonal declineswe detectedin the Buffieheadand AmericanWigeon may have resultedfrom emigrationassociated with midwinterincreases in rainfalland correspondingincreases in the availabilityof seasonaland interiorwetlands (Shufordet al. 1989). In contrast,significant intraseasonal increases in Eared Grebesconformed to patternsreported in San FranciscoBay salt ponds (Swarthet al. 1982) andother wetlands in the PointReyes area (Shuford et al. 1989), possiblyresulting from a protractedfall migration (Jehl 1988). In general,the eastshore of the TomalesBay, with shallowshoals and mudfiats,headlands and protectedcoves, supports greater abundances and

113 WINTER WATERBIRDS ON TOMALES BAY, CALIFORNIA more speciesof waterbirdsthan eitherthe midbayor westshore. Neverthe- less,preferred habitats of variousspecies occur throughout the bay (Figure 2, Table 2). Waterbirddistributions on TomalesBay are associatedwith strongwinter salinity and temperaturegradients, differences in water circu- lation, turbidity, and exchange rates with nearshore coastal waters (Hollibaughet al. 1988), distributionof eelgrassbeds (Spratt 1989), and a wide rangeof bottomsubstrates (Daetwyler 1966). The mostdiverse and abundantconcentrations of waterbirdsin the bay occurbetween Pelican Point and Tom'sPoint, where eelgrassbeds are the mostconcentrated and estuarinecirculation is enriched by the tidal delivery of nutrientsand planktonfrom the outer ocean,as well as by the inflowof freshwaterand nutrientsfrom Walker Creek, i.e., where habitat diversityand nutrient suppliesare greatest (Figure 1; Hollibaughet al. 1988, Smithand Hollibaugh 1997). The qualityand extentof inflow from Walker Creek dependson managementof a 189-km2 watershed,which is dominatedby agricultural grasslands,coastal scrub, mixed evergreenforest, and oak woodland,and on managedwater releasesfrom the SoulajuleReservoir (Storm 1972, Madrone Associates1976).

ConservationImplications The importanceof winteringareas in the dynamicsof waterbirdpopula- tionsis suggestedby the potentiallycrucial role these areas play in courtship or pairingand deposition of fat storesused later as energyfor reproduction (Heitmeyerand Fredrickson1981, James 1989, Baldassarreand Bolen 1994). Winterhabitat quality has been linked to annualsurvival, recruitment, and reproductivesuccess of waterfowl(Raveling 1979, Haramiset al. 1986, Ravelingand Heitmeyer 1989, Baldassarreand Bolen 1994). Eelgrass,the mostabundant large marine plant in TomalesBay (Hardwick 1973), providescrucial winter food for BlackBrant, surface-feedingducks, and other waterfowl(Yocum and Keller 1961, Baldassarreand Bolen 1994). Our resultsshow a recentincrease in the useof TomalesBay by wintering BlackBrant, with abundanceswell abovewinter averages of approximately 200 birdsin the 1980s (Calif.Dept. Fishand Game unpubl.data). Eelgrass also supportsa rich estuarinefauna that providesadditional food for waterbirds(Day et al. 1989). Althoughthe distributionof eelgrassin Tomales Bay has been relativelystable (Spratt 1989, Moore and Mello 1995), it is sensitiveto changesin salinity,turbidity, and temperature(Day et al. 1989, Baldassarreand Bolen 1994) andthus potentially vulnerable to reductionsin the qualityor quantityof freshwater supplied by WalkerCreek. In TomalesBay, eelgrassprovides winter spawninghabitat for approxi- mately5500 tons(20-year average) of PacificHerring (Clupeaharengus palla$i; Moore and Mello 1995). From Decemberthrough March, scoters, Bufflehead,scaup, Black Brant, goldeneyes,American Coots, and other speciesof waterbirdsheavily exploit the roe of PacificHerring (Hardwick 1973, Bayer1980, Briggset al. 1987, Haegele1993), whileadult herring are consumedby loons, large grebes, and cormorants(Palmer 1962). High rates of kleptoparasitismby gullscan limit the foragingactivities of waterbirdsfeeding on recentlydeposited herring eggs (Bayer 1980). Such disturbanceis probablyenhanced by the dailyarrival at WalkerCreek delta

114 WINTER WATERBIRDS ON TOMALES BAY, CALIFORNIA of severalthousand gulls from a locallandfill near Cotati, SonomaCounty (Kellyet al. 1996). If somewaterbirds depend on herring(or herringroe) in winter,overhar- vestby the commercialherring fishery could limit their populations. Because annualharvest quotas are basedon estimatesof spawningbiomass from the previousyear, which fluctuate considerably, the fisherycan over- or underes- timate the proportionof herringbeing harvested (Suer 1987, Moore and Mello 1995). Well-informedecosystem management will requirefocused researchon the link between Pacific herring and waterbirds,accurate forecastsof spawningbiomass, and cautiousharvest quotas to guidethe commercialherring fishery. Interestingly,scoters occur in large numberssouth as well as north of Pelican Point, rather than concentratingover the dense eelgrassareas (availableto spawningherring) to the north.Also, Greater Scaup concentrate in the southernportion of the bay. However,herring normally spawn as far southas TomasiniPoint, suggestingthat the herring'shabit of spawningin eelgrassis influencedby a preferencefor reducedsalinities that couldcause them to shiftsouthward when the flow of freshwater into the bay is reduced (Mooreand Mello 1995). Alternatively,the spatialresponses of waterbirdsto irregularherring runs may not be adequatelysampled by our counts.Scoters and scaupalso feed on molluscsand other invertebrates(Bellrose 1976), suggestingthat bottomsediments at the southernend of the bay (Daetwyler 1966) provideimportant foraging substrates for thesespecies. Black Scoters, near the southernend of their Pacificwinter range, are uncommonto rare elsewherealong the Californiacoast (Grinnell and Miller 1944, McCaskieet al. 1979, Garrettand Dunn 1981). Our studyindicates that annual concentra- tionsof approximately70 (max.180) BlackScoters are regular along the east shore just north of Tomasini Point, reflectingsimilar concentrations,of approximately60 birds,observed just below Highway One in thisarea each winter since1961 (R. Stallcuppers. comm.)The cobblesand gravellyand shellyshoals that dominatethe bottom substratesin this area (Daetwyler 1966) areconsistent with preferred foraging substrates reported elsewhere for thisspecies (Bordage and Savard1995). During winter, salinities,turbidity, sedimentation, and water exchange between the bay and the outer ocean vary greafiy with the extent of freshwaterrunoff, especially near LagunitasCreek (Hollibaugh et al. 1988). Becauseof the linearconfiguration of the bay,the availabilityof preyspecies may be especiallysensitive to anthropogenicand other terrogenousinflu- ences.For example,the longwater-residence times of up to fourmonths in the southernpart of the bay duringdrought or summer(Hollibaugh et al. 1988) implythat the baymay retaincontaminants for longperiods of time, potentiallyallowing toxic materialsderived from watershedrunoff to enter foodwebs or becometrapped in bay sediments.Most diving ducks depend on benthicprey, which are in turn vulnerableto sedimentcontamination fromthe watershed(Cain 1988). CommonLoons and other fish-eating birds avoidhighly turbid water, presumably because it interfereswith prey detec- tion (Pehrsson1984, Haney 1990). Extendedwater-residence times be- tweenperiods of peakwinter runoff interact with watershed erosion, leading to extendedperiods of highturbidity. Thus some waterbirds that concentrate

115 WINTER WATERBIRDSON TOMALES BAY, CALIFORNIA in the shallowsouthern end of the bay or near the mouthof WalkerCreek could be limited by high turbidity generated by developmentor other activitiesin the watershed.Rafts of Western and Clark's grebesusually concentrateover deeper areas (pers. obs.), possibly to avoidturbid runoff by foragingin pocketsof clearersaline water near the bottom(Pehrsson 1984), althoughthey may alsoreflect concentrations of prey fishes. In a substantialportion of the bay, commercialaquaculture of oysters (Crassostreagigas) and baymussels (Nlytilus edulis) has introduced various racks,floats, and bagsthat alterwaterbird habitat on the surfaceand in the water columnand may alterthe structureof bottomsubstrates where diving ducksfeed (Kellyet al. 1996). To our knowledge,scientific information on the effects of such structures on waterbird habitat does not exist. In winter,waterbirds occupy open habitatswhere they generallyrely on flightas a responseto disturbanceby predatorsor humans.Even diving birds suchas loons and small grebes frequenfiy rely on flightto escapedisturbance by approachingboats, althoughlarge grebesescape by diving. Flight is energeticallyexpensive in birds,with costsas high as 15 timesthe basal metabolicrate (King 1974). Optimal-foragingtheory (Krebset al. 1983) providesa frameworkfor understandingdisturbance, by predictingthat birds must compensatefor increasedlevels of disturbance,either by increasing their rate of food intakeor flyingto other,less profitable but lessdisturbed areas to feed. Many winteringwaterbirds must also accumulatefat and protein reservesto overrideperiods of low food availabilityand to prepare for springmigration (Davidson and Rothwell1993, Baldassarreand Bolen 1994). If feedingopportunities are alreadyrestricted, or waterbirdsare close to their energythreshold, direct disturbance by humanscould have consid- erableenergetic effects on individualbirds, leading to abandonmentof an area or starvation(Davidson and Rothwell 1993). We observeddramatic increasesin sport fishing, kayaking,and other recreationalboating on Tomales Bay during the course of this study, especiallywhere waterbirdsconcentrate in the northernthird of the bay. Eachboat is associatedwith a radiusof disturbancedetermined by its speed of movementand the sensitivityof birdspecies. Fast-moving power boats, motorizedpersonal watercraft ("jet skis"), and aggregations of boatssuch as occur around popular fishing spots create larger zones of disturbance (Burger1981; pers.obs.). A few boatsdistributed across an eelgrassbed can prevent most waterbirdsfrom usingthe area. Sea kayaks,unlike other recreationalboats, tend to travel in loose groupsthat displacewaterbirds from marshedges, shallow coves, and sandbars. Low-flying aircraft occa- sionallytrack the linear shapeof the bay, resultingin a wider band of disturbancethan observedaround boats (R. Stallcuppers. comm.). Estuarinewaterfowl may be seriouslyaffected by evenoccasional distur- banceduring key partsof the feedingcycle. American Wigeon feeding in eelgrassbeds return to feed if disturbedearly in the (tidal)feeding cycle but abandonthe site altogether,until the next tidal cycle,if disturbedafter the eelgrassis exposed by the tide(Fox et al. 1993). BlackBrant, which also feed on eelgrassin TomalesBay, showsimilar avoidance responses to increased presenceof humans(Stock 1993). In TomalesBay, huntingdisturbance is concentratednear the mouths of Lagunitasand Walker creeks;because

116 WINTER WATERBIRDS ON TOMALES BAY, CALIFORNIA theseare the onlytwo largeareas of coastalmarsh habitat, similar areas are not availablefor refugeelsewhere on the bay. We did not measureflight distances,return rates,or other behaviorsof birdsthat flushedoccasionally in responseto the movementof our observationboats. The array of potentialconflicts between human activities and wintering waterbirdson TomalesBay impliesthe likelihoodof cumulativeor synergis- tic effects.Long-term protection of waterbirdpopulations on TomalesBay may dependon broad-basedplanning, in whichdiverse agencies coopera- tivelyconsider overall management effects on abundancesand distributions of waterbirds.We recommendthat management efforts focus on controlling boattraffic and other forms of directhuman disturbance, protecting eelgrass beds,assessing the effectsof herringharvest and aquaculture on waterbirds, and monitoringhuman land useand other processesin the watershedthat may affectthe qualityof waterbirdhabitats in the estuary.

SUMMARY

We analyzed patterns of abundance and distribution of wintering waterbirdsin TomalesBay, California,on the basisof 21 baywidewinter surveysover 7 years from 1989-90 to 1995-96. Total waterbirdabun- dancesaveraged 21,943 (min. 14,842, max. 25,553), excludingshorebirds and gulls.Mean waterbirddensity was 770 birds/km2 with the greatest concentrationsalong the eastshore and betweenPelican and Tom's points. The Surf Scoter, Bufflehead,and Greater Scaup together accountedfor 70% of the total waterbirds. Aerial count data from the U.S. Fish and WildlifeService Midwinter Waterfowl Survey underestimated the numbersof waterbirdsin comparisonto our countsmade from boats.Such differences adjustedfor, our abundanceestimates for scotersand Buffleheadappear similarto otherabundances reported in the 1960s and 1970s. Comparedto other coastalwetlands in California, TomalesBay providesparticularly importantwinter habitat for the Red-throatedand Commonloon, Earedand Hornedgrebe, Black Brant, Surf and Black scoter. Except for San Francisco Bay, TomalesBay may providethe most importantwinter habitatfor the Buffiehead on the Pacific coast south of the Columbia Riven Waterbird distributionson TomalesBay are associatedwith a dynamicarray of habitat conditions.Increasing human activitiesin TomalesBay and its watershed interactwith theseprocesses and conditionsand suggestseveral concerns for conservationof winter waterbirds.Protection of waterbirdpopulations will requireparticular attention to the controlof humandisturbance, protec- tion of eelgrassbeds, the possibleeffects of the herring fishery and aquaculture,and managementof processesin the watershed.

ACKNOWLEDGMENTS

The followingobservers provided expertise in thefield: Sarah Allen, Nancy Angelesco, Bob Baez,Rosilyn Bazurto, Gay Bishop,John Boyd, Tom Bradner,Brian Bullick, Ken Burton, Hugh Cotter, Diana Creber,Walt Creber,Rigdon Currie, Sam Dakin, Bob DeLeo, John Dillon,Joe Drennan,Caroline Dutton, Steve Engel, Gayanne Enquist, Jules Evens, Gary Falxa, Katie Fehring, David Ferrera, Virginia Fletcher,Nicole Gallagher,Jim Gallagher, Gayle Garman, Felicia Guest, Bruce Hamilton, Eithne Haran,

117 WINTER WATERBIRDS ON TOMALES BAY, CALIFORNIA

Daphne Hatch, Holly Heinzmann,LaRee Holmes,Jim Holt, MaggieHynes, Gwen Jensen,Mary Ellen King, RichardKirschman, Laura Leek, William LeGro, Michele gapes, LorraineMacKenzie, Kay MacMahon,Aspen Mayers, John McDonagh,Ellen McKnight,Margaret McClune, Olive Mills, Jan Mofet, Dan Murphy,Muriel Niehaus, Terry Nordbye,Karen Paull, John Peterson,Ray Peterson,Linda Petrulias,Yvonne Pierce,Karen Porteous,Jamie Ross,Helena Russell,Ellen Sabine, Jim Sarraco,Ann Schofield,Dave Schurr, Craig Scott, Elaine Senf, David Shuford, Brian Simon, Dabne Smith, Hillary Smith, Joe H. Smith, Craig Solin, Anne Spencer,Rich Stallcup,Judy Temko, StephenThai, Janet Theissen,Wayne Thompson,Forest Tomlinson, Brett Walker,Tanis Walters, Katherine Warner, Penny Watson, Sophie Webb, Tom White, Diane Williams,David Wimpfheimer, and ChrisWood. We thank the following:Tom Baty for his excellentleadership and assistancein planning,coordinating, and conductingthe counts;Richard Allen, Tom Baty, Robert Cardwell,George Curth, Sam Dakin, RickGrieve, Richard Plant, Erich Rienecker, and the InvernessYacht Club for volunteeringtheir boatsand skillsas boat operators;Bill Beal, Grant Fletcher,David Hartje, Kirk Hastings,Pat Garmy, and Skip Schwartzfor operatingthe AudubonCanyon Ranch boat; David Shufordand Rich Stallcupfor helpfulsuggestions on censusmethods and excellentcomments on an earlierdraft of thispaper; and Stephen F. Baileyand Tim Manolisfor helpfulsuggestions on improving the final manuscript.This paper is a contributionof AudubonCanyon Ranch.

LITERATURE CITED

Baldassarre,G. A., and Bolen, E.G. 1994. WaterfowlEcology and Management. Wiley, New York. Barnhart,R. A., Boyd,M. J., and Pequegnat,J. E. 1992. The Ecologyof Humboldt Bay,California: An EstuarineProfile. U.S. Fishand Wildlife Service Biol. Rep. 1. Bayer,R. D. 1980. Birdsfeeding on herringeggs at the YaquinaEstuary, Oregon. Condor 82:193-198. Bellrose, F. C. 1976. Ducks, Geese, and Swans of North America. Stackpole, Harrisburg,PA. Bordage,D., andSavard, J. L. 1995. BlackScoter, in the Birdsof NorthAmerica, No. 177 (A. Poole and F. Gill, eds.).Acad. Nat. Sci., Philadelphia. Briggs,K. T., Tyler,W. B., Lewis,D. B., and Carlson,D. R. 1987. Birdcommunities at sea off California: 1975-1983. Studies Avian Biol. 11. Burger,J. 1981. The effectof humanactivity on birdsat a coastalbay. Biol. Cons. 21:231-241. Cain, B. W. 1988. Winteringwaterfowl habitat in Texas:Shrinking and contami- nated, in Waterfowl in Winter (M. W. Weller, ed.), pp. 583-607. Univ. Minn. Press,Minneapolis. Conant,B., King,J. G., Trapp,J. L., andHodges, J. I. 1988. Estimatingpopulations of duckswintering in southeastAlaska, in Waterfowlin Winter (M. W. Weller, ed.), pp. 541-552. Univ. Minn. Press,Minneapolis. Daetwyler,C. C. 1966. Marine geologyof TomalesBay, central California. Univ. Pacific,Pac. Mar. Sta. Res. Rep. (6):1-169. Davidson,N., and Rothwell,P. 1993. Human disturbanceto waterfowlon estuaries: The conservationand coastalmanagement implications of currentknowledge. WaderStudy Group Bull. 68: 97-105. Day, J. W., Jr., Hall, C. A. S., Kemp, W. M., Yafiez-Arancibia,A. 1989. Estuarine Ecology.Wiley, New York. Dixon,W. J. 1992. BMDP StatisticalSoftware Manual. Univ. Calif. Press, Los Angeles.

118 WINTER WATERBIRDSON TOMALES BAY, CALIFORNIA

Fox, A.D., Bell, D. V., and Mudge,G. P. 1993. A preliminarystudy of the effectsof disturbanceon feedingWigeon grazingon Eel-grassZostera. Wader Study GroupBull. 68:67-71. Galloway, A. J. 1977. Geology of the Point Reyes Peninsula,Marin County, California. Calif. Div. Mines Geol. Bull. 202. Garrett, K., and Dunn, J. 1981. Birdsof SouthernCalifornia: Status and Distribution. Los AngelesAudubon Soc., Los Angeles. Grinnell,J., and Miller, A. H. 1944. The distributionof the birdsof California.Pac. Coast Avifauna 27. Haegele,C. W. 1993. Seabirdpredation of pacificherring, Clupea pallasi, spawnin British Columbia. Can. Field Nat. 107:73-82. Haramis, G. M., Nichols, J. D., Pollock, K. H., and Hines, J. E. 1986. The relationshipbetween body massand survivalof winteringCanvasbacks. Auk 103:506-514. Haney, J. C. 1990. Winter habitatof CommonLoons on the continentalshelf of the United States. Wilson Bull. 102:253-263. Hardwick,J. E. 1973. Biomassestimates of spawningherring, Clupea harengus pallasi, herring eggs, and associatedvegetation in TomalesBay. Calif. Fish Game 59:36-61. Heitmeyer, M. E., and Fredrickson,L. H.. 1981. Do wetlandsconditions in the MississippiDelta hardwoodsinfluence Mallard recruitment? Trans. N. Am. Wild. Nat. Res. Conf. 46:44-57. Hollibaugh,J. T., Cole, B. E., Dollar,S. J., Hager,S. W., Vink, S. M., Kimmerer,W. J., Obrebski,S., Smith, S. V., Valentino,M., and Walsh, T. W. 1988. Tomales Bay, California:A macrocosmfor examiningbiogeochemical coupling at the land-sea interface. Eos 69:843-845. James,R. A. Jr. 1989. Mate feedingin winteringWestern Grebes. J. FieldOrnithol. 60:358-360. Jehl, J. R., Jr. 1988. Biologyof the Eared Grebe and Wilson'sPhalarope in the nonbreedingseason: A studyof adaptationsto salinelakes. Studies Avian Biol. 12. Kelly, J.P. 1993. Fall, early-winter,late-winter, and spring 1992/93 shorebird censuseson TomalesBay, California. ACR Project Rep. 89-4-4, Audubon CanyonRanch, 4900 ShorelineHwy., StinsonBeach, CA 94970. Kelly, J.P., Evens,J. G., Stallcup,R. W., and Wimpfheimer,D. 1996. Effectsof aquacultureon habitatuse by winteringshorebirds in TomalesBay, California. Calif. Fish Game 82:160-174. King,J. R. 1974. Seasonalallocations of time and energyresources in birds,in Avian Energetics(R. A. Paynter,Jr., ed), pp. 4-85. NuttallOrnithological Club, Cambridge. Krebs,J. R., Stephens,D. W., and Sutherland,W. J. 1983. Perspectivesin optimal foraging,in Perspectivesin Ornithology(A. H. Brushand G. A. Clark,Jr., eds.), pp. 165-215. CambridgeUniv. Press,Cambridge, England. MadroneAssociates. 1976. EnvironmentalImpact Report on the SoulajuleProject. Marin MunicipalWater District, 220 Nellen Ave., Corte Madera, CA 94925. Magurran,A. E. 1988. EcologicalDiversity and Its Measurement.Princeton Univ. Press,Princeton, N.J. McCaskie,G., DeBenedictis,P., Erickson,R., and Morlan,J. 1979. Birdsof Northern California:An AnnotatedField List, 2nd ed. Golden Gate Audubon Soc., Berkeley. Moffitt, J. 1943. Twelfth annual Black Brant censusin California. Calif. Fish Game 29:19-28.

119 WINTER WATERBIRDS ON TOMALES BAY, CALIFORNIA

Monroe, G. 1973. The NaturalResources of HumboldtBay. Calif. Dept. Fish & Game Coastal Wetland Ser. 6. Moore,T. O., andMello, J. J. 1995. Pacificherring, Clupea pallasi, studies and fishery managementin TomalesBay, 1992-93, withnotes on HumboldtBay and Crescent City arealandings. Calif. Dept. Fish& GameMar. Res.Div. Admin.Rep. 95-5. NationalOceanic and AtmosphericAdministration. 1987. Gulf of the Farallones NationalMarine Sanctuary Management Plan (J. DobbinAssociates, Inc., eds.). U.S. Dept. Commerce,NOAA, Marine and EstuarineMgmt. Div., 1825 Con- necticutSt. NW, Washington,D.C. 20235. Neubacher,D., DelrOsso,J., Livingston,D., and Murray, R. 1995. TomalesBay/ BodegaBay WatershedBoundary Study. Point ReyesNational Seashore, Marin and SonomaCounties, California. U.S. Govt. PrintingOffice 1995-685-430. Page,G. W., Shuford,W. D., Evens,J. G., and Stenzel,L. E. 1983. The distribution and abundanceof aquaticbirds in wetlandsof the Point Reyesto Bodegaarea. Report to the Point Reyes-FarallonesMarine Sanctuary.Point Reyes Bird Observatory,4990 ShorelineHwy., StinsonBeach, CA 94970. Palmer, R. S. 1962. Handbookof North American Birds, vol. 1. Yale Univ. Press, New Haven, CT. Pehrsson,O. 1984. Diving duck populationsin relation to their food supplies,in Coastal Waders and Wildfowl in Winter (P. R. Evans and J. D. Goss-Custard, eds.),pp. 101-116. CambridgeUniv. Press,Cambridge, England. Raveling,D. G. 1979. The annualcycle of body conditionof Canada Geesewith specialreference to controlof reproduction.Auk 96: 234-252. Raveling,D. G., and Heitmeyer,M. E. 1989. Relationshipsof populationsize and recruitment of Pintails to habitat conditionsand harvest. J. Wildlife Mgmt. 53:1088-1103. Shuford,D. W., Page, G. W., Evens,J. G., and Stenzel,L. E. 1989. Seasonal abundanceof waterbirdsat Point Reyes:A coastalCalifornia perspective. W. Birds 20:137-265. Small,J. 1972. The birdsof TomalesBay, in TomalesBay EnvironmentalStudy (R. Corwin, ed.), pp. 124-134. ConservationFoundation, Washington, D.C. Smith, S. V., and Hollibaugh,J. T. 1997. Annualcycle and interannualvariability of ecosystemmetabolism in a temperate climate embayment. Ecol. Monogr. 67:509-533. Spratt, J. D. 1989. The distributionand densityof eelgrass,Zostera marina, in TomalesBay, California.Calif. Fish& Game 75:204-212. Stock, M. 1993. Studieson the effectsof disturbanceson stagingBrent Geese:A progressreport. Wader Study Group Bull. 68:29-34. Storm, D. R. 1972. Hydrologyand waterquality aspects of the TomalesBay Study, in TomalesBay EnvironmentalStudy (R. Cotwin,ed.), pp. 84-101. Conserva- tion Foundation,Washington, D.C. Suer,A. L. 1987. The Herringof SanFrancisco and Tomales Bays. Ocean Res. Inst., San Francisco. Swarth,C. W., Akagi, C., and Metropulos,P. 1982. The distributionpatterns and ecologyof waterbirdsusing the Coyote Hills salt ponds.Report to the San FranciscoBay NationalWildlife Refuge, P.O. Box 524, Newark,CA 94560. Yocom, C. F., and Keller, M. 1961. Correlationof food habitsand abundanceof waterfowl,Humboldt Bay, California.Calif. Fish Game 47:41-54.

Accepted 13 January 1998

120