GEOPHYSICAL RESEARCH LETTERS, VOL. 38, L02604, uodO,1029/2010GL045810, 2011
Nearshore chlorophyll-a events and wave-driven transport
Erika E, McPhee-Shaw,' Karina J, Nielsen,2 John L Largier,3 and Bruce A Menge4
ReceiverllO OcLuber 2010; revised 6 December 2010; accepLed 15 December 2010: publisherl21 January 2011.
llJ
Contlnuous records of temperature
alld
chlorophyll-a
rchl-a1fluorescence were used to characterize ph,10plankton
variability' at t\vo shallo\v intertidal stations on the northern
California coast. ChI-a records from spring and summer
2007 alld 2008 were characterlzed by distinct peaks
p(""rsisting for 1.5 to five days. 111CSC peaks n..vn.:scnt bloom
events. which oilen coincided behveen the h"vo sites even
though they are separated by ~150 kIT!' Blooms did not
appear to be directly forced by individual upwelling
l-lJisodcs. \Vhilc some events \\I(,.,'n; assot:latcu 'with Tcv(,.,'rsals
of upwelling-favorable winds~ there was a stronger
relationship between chI-a peaks and peaks in surface swell
height. This relationship was dominant in spring and early
summer, but no longer evident by .Iuly. We suggest that
these llearshore chI-a peaks are an accumulation of
phytoplankton caused by convergence of onshore wave
transport against the impermeable coastal boundary.
Scaling arguments sho"\-v this mechanism to be consistent
'with observed chI-a illcrease rates. This mechanism has not
been previously considered as a forcing for blooms at rocky
coasts~ and it may have significant implications for
understanding coastal productivity, larval dispersal, and
llearshore \vatl-'1- quality. Citation: McPhee-Shmv, E. E., K. J.
Nielsen, .I. L. Largier, and n. A. Menge (2011), Nearshore
chlorophyll-a cvcnttl am1 wavc-urivcll tmlltlpOit. Geophys. Res.
Lett., 38, L02G04, doi: 10,1029/201 OGL04581 0,
1. introduction
[2] Understanding chI-a variability in shallow intertidal
l-'1wiromnents is extremely impOliant: phytoplanktOll arc the
base of the foud chain tor filter feeders, alld bluom eVl-'11ts
may be directly linked to shellfish toxicity and other effects
on higher trophic levels, Coastal upwelling along the west
coast of North America in the California Current Ecosystem
dlives high regiunal plimary productivity and productive
fisheries. However, a direct connection between discrete.
intraseasoml upwelling events and phytoplankton blooms
at similar time scales is far from clear, and individual bloom
eVl-'11ts can be markedly deeoupled from upwellillg l Wieters
et aI., 2003 j. A pattl-'111 of lueally maximum chI-a c<'}1lfilled
to a relatively thin band near the coast has been documented in many locations [TVieters e{ aI., 2003; Kim et al.,
I\-foss Landing \-farine Lahoratories, San .Io~e: State: Lnive:rsit)l.
Landing. California, CSA.
~Depart111ent of BIOlogy, Sonoma State Univenity, Rohne:rt Park,
California. USA.
~Dodega ),·:larine Laboratory, UniversilY ofCa1ifurnia, Davis, I30dega
1'1o~s
BaY~DC~1p'iajOrt"lnl,ie'I"tUOS[AZ'
0010_"')',
..
"- Oregon Slale University. Corvallis, Oregon,
USA.
Copyright 2011 by Lhe American Geophysical Cnion.
0094-8276il1/2010GL045810
2009J yet questiuns remain cuneellling the causes alld valiability ofthis band, Twice-weekly samples from the southem
California inner shelf demonstrated that this near-coast feature was consistently amplified compared to chl-a offshore
un the shel( but ullrelated to upwelling (cold temperatures)
ur offshore '\-Villds lKim et at., 2009J. Intertidal recruitml-'11t
of mussel larvae has been shO\vn to vary dramatically from
larval numbers outside of the surfzone, implying important
dynamics benveen the iooer shelf and the coastline that we do
110t yot understarld l Rilav et al., 2008; Shan!c,' et 01., 2010 j,
31 The event scale of blooms is similar to the synoptic, or
"weather band," scale that characterizes a good portion of
coastal dynamics in a system that is fundamentally forced by
wind. DYllamics in this category include upwelling/relaxatiun
cycles, coastal eddies, surface swell, alld even raillfalL It is
reasonable to h)rpothesize that high chl-a events persisting
several days are related to such dynamics, and indeed continental shelf studies have pointed out links benveen retentive
circulati<'}1l pattellls and e01Wl-'1-genee of planktOll le.g.,
ROlfghan et a/', 2005; Ryan et d, 2008], \\hle recent work
(A J Lucas et aL, The green ribbon: 'vlulti-scale physical
control of phytoplankton productivity and community sttucture OVl-'1- a llanuw cuntilll-'11tal shelf, submitted tu Limnology
and ()ceanOb'l-aphy, 2(10) shows that illtelllal wave mixillg
may account for the observed nearshore chlorophyll maximum in southern California and other locations characterized
by sha11O\v stt-atification, an alternative mechanism must be
fOUlld tu explain nearshore maxima in unstratified watl-'1"S
such as those off northem Califomia, We observed a correspondence betw-een many chl-a blooms and energetic surface
s"\-vell events, and suggest convergence of wave-driven banspOli causing phytoplanktOll accumulation at the coastal
buundary as an additional, previously ullc<'}1lsidered, mechanism forcing nearshore phytoplankton blooms,
r
2. Methods
[4] Temperature and fluorescence \vere measured every
15 minutes at two sites un the nurthelll Calitolllia cuast
between May arld late August of 2007 arld 2008. irlStllJments \·vere affixed to rocky intertidal benches~ a\·vay from
tide pools, just above mean lower low "\-vater (MLL\V) and
fully exposed to nearshore waters at both locations. The first
site was at Bodega Marine Laboratory, facing west/southwest
on Bodega Head at 383 I 87 c N, 123,0742 c W (henceforth
B'vlL) and the second at a west-facing site, Kibesillah Hill
(henceforth KH), at 39,5999°N, 123,7889°W, approximately
20 km northwest of FOli Bragg (Figure 1). Both sites are
located un a stretch uf rocky cuastlille expused to the Opl-'11
ocean and are not sheltered by a local headland or
embayment.
[5] Temperature "\-vas measured \vith Onset Temperature
TidBit v2 Data Loggl-'1-S (accuracy 0.2°;;C, time COllStallt
5 minutes), Fluorescence was measured with a WET Labs
L02604
1 of 5
L02604
MCPHEE-SHAW ET AL: NEARSHORE CHLOROPHYLL-A EVENTS
Figure 1. Map of study sites.
ECO fluorometer (Ex/Em: 470/695 mu). Raw fluorescence
was calibrated to in situ chI-a (p.giL) (a proxy for phytoplankton biomass) from extracted chl-a (extraction methods
described by Wieters et al. [20031) collected in 3 replicate
sample bottles every 2 weeks during deployment periods.
The fluorometer vvindow ,vas cleaned hi-monthly to prevent
biofouling. ChI-a records \",ere de-spiked and statistical
outli(""rs (>2 SD from the mCall for the time sl-'lics) were
removed. Records corresponding to a tide height <1 III above
the fluorometer height were removed. Instruments at both
sites were deployed ~0.5 m above MLLW, )~elding data
COVl-'l-agc over typically 60% of each day, leaving ample
data for charactl-'lizillg events \\~th time scales of days.
Intertid al temperature was slightly warmer than, but very
,veIl cOlTelated vvith, temperature measured at 4-m depth at
a mOOling 1.2 km offshorc of Bodcga Hcad dUlil1g summcr
2007. A least-squarcs fit bct\vccn mooring alld BML intcrtidal temperature for 01 June to 15 September 2007 gives
TUllcrlidal - 0.899 • TmooruHI. + 0.471, with R2 - 0.73. No
improvement vvas obtained-by accounting for time of solar
day or limitil1g to highcr tidcs. Thc high con-c1ation, and thc
close visual match ben"veen offshore and intertidal temperature (Figure 20, red and black lines) demonstrate that intertidal temperatures were strongly controlled by the same
mcchanisms that control tcmpl-'raturc on thc shclt~ alld that
thcsc shallow rccords can bc uscd to c1mractl-'1izc occanic
conditions. Hourly \vind speed and direction and significant
Vi/ave height and direction were obtained from National Data
Buoy Center (NDBC) buoys 46014 (near Point Arena,
(39.195K c N 123.9694°W)), 46013 (11 oar Bodega Bay,
38.2419°N 123.3006°W), and NDBC 46042 (near Monterey
Bay, 36.7886 c N In.4042 C W).
3. Observations
[6] \Vater temperature, vvind vectors, and significant vvave
height sq uarcd, H;ig (proporti 011 al to wa vc l-'11 l-'1'gy), arc ShO\Vl1
L02604
\vith chI-a records at both stations in figure 2. Background
chI-a levels Wl-'1'C typkally 5 to 10 I"g L- 1, punctuatcd by
pcaks whcrc ch1-a was clevatcd to bctwccn 20 alld 30, and
sometimes >50, I'g L- 1 (Figures 2d, 2e, 2i, and 2j). lildividual peaks, marked vvith letters, Vi/ere defined by rapid
chi-a increases to more than one standard deviation
(~8 I"g LI) abovc levels of thc prcvious fcw days alld
remaining elevated for a period of at least one day. Some
peaks lasted just over a day, most were 1.5 to' 3 davs
duration, but the longest bloom persisted up to 6 days.
ChI-a valiancc im.,Tcascd in July relativc to May alld car1y
JUI1C, palikularly at thc B\!1 L sitc in 2008, making it difficult to distinguish bloom events from rapid, \vithin-day
chi-a fluctuations during the late summer.
[7] Winds were predominantly upwelling-favorable (south!
southcastward) throughout thc study (Figurcs 2b and 2g) alld
the \"vater temperatures remained cold, betw'een 8 and 12°C,
until at least mid-July in both years, reflecting the strong
seasonal signal of coastal upvvelling. There was no relationship bctwcl-'11 individual bloom CVl-'11tS and distinct cold
cvcnts that might indkatc growth spun-cd by a pu1sc of
newly upwelled nutrients. Temperature varied at time scales
quite different from the scales of chI-a variability. Various
instances of \vater temperature falling and rising from maxima to mil1ima and back Wl-'1'C gradual, typkally takil1g about
10 to 12 days, in contrast to the more abmpt rise and fall of
chl-a within several days. Although unrelated to cold temperatures in a day-to-day sense, most of the blooms in our
study wcrc obscrvcd duril1g upwcl1ing conditions and rc1atively stcady upwc111ng winds. Upwcllil1g winds wcrc punctuated by brief relax....1tion events (\"vind \"veakening or
reversing for 2 days or less). Several bloom events appeared
associated \vith wind relaxation, e.g., the 01 to 02-July
inccption ofa l<.mg-livcd bloom at BML (cvcnt k, Figurc 2c)
and a shOlicr bloom at KH (CVl-'11t t~ Figurc 2d) follo\\~ng a
brief relaxation on 29 June 2007 and weak \,rinds for several
days. Also, event d in 2007 and events c, m, h, and p in 2008,
either coincided with or immediately followed significant
wind rcvcrsals. SCVl-'11 of 28 total blooms idl-'11tificd coincided with wind relaxation, while 21 (75%) coincided with
upwelling conditions, suggesting blooms were not associated v·lith v-/ind relaxations, but possibly with upv~'elling
favorabk winds (X' ~ 5.19, df ~ I, P ~ (Uln7).
l RJ Of thc oCCanObTJ.-aphic paramctl-'1's invcstigatcd, surfacc
swell energy showed the strongest relationship to chi-a
events. In 2007, events a, b, c, d, e, h, i, j, 1<, and I coincided
v·lith relative peaks in H~ig. In 2008, chI-a events a, b, c, d, f, g,
i, j, k, 1, m, n, and p cO-occulTcd with pcaks in surfacc wavc
height. Although wave direction is not ShOW1~ almost all of
the bloom events occurred during northwest swell (300° to
320°) typical for spring and summer in this region. The initial
lisc of ch I-a consistcntly followcd thc initial risc of H~ig with
a lag of typically 5 to 8 (but ranging from 0 to 21) hours.
vVith few exceptions, the wave-related bloom events were
limited to May and June. The wave-to-chl-a association
seemed to disappear in the late summer. Of the 28 bloom
cvcnts idcntificd, 23 (82%) wcrc associatcd with c1cvatcd
waves, and a test of independence benveen peaks in H;ig
and blooms provides supporting evidence for association at
the event-to-event level (X' ~ 14.26, df ~ I, P ~ 0.0002).
Howcvcr, thcrc Wl-'1'C wavc CVl-'11tS without con-cspondil1g
ncarshorc blooms, and although thc risc of H;io- and chI-a
were often surprisingly simull"lneOUS, magninJde did not
2 of 5
MCPHEE-SHAW ET AL: NEARSHORE CHLOROPHYLL-A EVENTS
L02604
L02604
2007
~'--' 40 jd) KH Chll a
~20
,::=~~~~~~~~==~~==~==~
~~ 40
E' 20
Ie] BML Chll,.
L--~~~~--O~5~11~8-2CO~5~"~8~~8~~~D~7~~D'~/l~7~~O'~i2§7~~07~~~7~~07~/l~7~~07~~~7!J
2008
i~
If) Temper ture&: KH
lue), BML fred)
u 12
~
10
8
'";;;;;~;;;;~;;;;~;;;;~;;;;~;;==~====~====~====~
20~{gJ Wind V&ctOI'S, ND C46013 (black) .and N BC46014 <gray)
Figure 2. Records from 01 May to 31 July 2007. (a) Near-surface water temperatures at the KH intertidal site (blue), the
BML intertidal site (red), and BML mOUling oftshore of the BVI L intertidal site (black). (b) Wi"d vectors fi'om NDBe buoys
46013 (black) and 46014 (grey). A northward vector points upward, eastward is to the right. (c) Significant wave height
squared (H~jg. m') at NDBC buoys 46014 (blue). 46013 (green), and 46042 (red). Chl-a at (d) KH and (e) BML. Individual
IS-minute records are plotted as small blue dots. The larger green dots are the 25-hour average chi-a centered around the
highest tide of each day, which captuf(;s the greatest contillUOUS Spall ofin-\vatcr cov(""ragc. (f-j) Same as Figures 2a-2c for
01 May to 31 July 200X (except no mOOTing temperature ill 200X).
co-vary. Especially intense vvaves did not necessarily mean
the highest chI-a levels; for example despite energetic waves
on II June 2007 (events c andj) chl-a peaks were relatively
low compared to other events. Similarly~ weak wave events
vvere sometimes associated with relatively intense blooms.
4.
Discussion
19J \\"inds and waves stand out as the most likely d1ivl-'rs
of ph)10plankton variability in this system. \\11ile some
chI-a events either co-occurred with or immediately followed a vveakening or reversal ofupvvelling-favorable winds,
there was a stronger cOlTesp(mdl-'llce behveen chI-a and
'waves. A majOlity of identified events coi11cided \vith
elevated H;ig events~ while only 7 were associated \vith
vvind relaxation or vveakening. furthermore, five of these
co-occurred vvith energetic \vaves, leaving only 1\vo events
associated \vith \vind relaxation only. These results suggest
that although up\velli11g and relaxatio11 underwlite coastal
ph)10plankton gro\Vih, the building, peaking, and subsiding of
nearshore clll-a often matches the building. peaking. and
subsiding of nearshore vvave energy. \Vhile vve lack the necessary data to rule out influences fium eddies, fl.-onts and othl-'T
convl-'Tgl-'l1t circulation pattl-'lllS 11ear the coast, to-date there
is no evidence of these factors being important in this region
(R. E. fontana and J. L. Largier, personal communication,
2010). The lack of relationship to cold temperatures suggests
that bloom eVl-'l1ts arc not d1iven by pulses of upwelled
nutrients. Further, we do not expect significant terrestrial
nutrient input along this coast in May-June, ,"vhere rainfall
and stornl\vater is absent and groundvvater flow and streamflow is very weak alld mostly trapped withi11 closed estuaries
l Largier and Behrens, 2010 j.
J of 5
L02604
MCPHEE-SHAW ET AL.: NEARSHORE CHLOROPHYLL-A EVENTS
[101 \Vave transport is just one mechanism among many
setting the st1ge for bloom events. However it is a mechanism
that has been overlooked, and during Ibe early upwelling
season it appears to playa dominant rule in setting the timing
of 1.5- to 5-dav chl-a intertidal events. The role of surface
waves in prolDoting blooms has received little previous
consideralion. Although there have been previous studies of
phytoplankton dynamics on sandy bcachcs l Camphell and
Bate, 1997: .HcLachlan and Brawn, 2006J, to our knowledge this srudy is the first to report \vaves as a dominant
forcing on reflective, rocky shorelines. How can surlbce swell
events cause blooms? \Vave resllspension could spur shelf'wiol: productivity via the introdw:tiol1 of benthic microI1Ut1;l:l1tS l Chase et aI., 20071- Or, wave erosion might injcd
benthic diatoms 10 the water column. However, benthic diatoms were not. the dominant taxa observed dill-ing high or low
chi-a periods in preliminary studies of the phytoplankton
communities from thL:sL: sites (K. J. Nielsen and A. Paquin,
unpublished data). One po>sibility is that resuspended sediment dominates the ~ ign<1.l ~ e.g.~ chI-a tluorescence measurements can be contaminated by stray light from co-located
turbidity SL:nSors LOmalld el al., 2009]. However, WL: avoided
SL1,sor intL:I'H.tion by measUling fluoreseem:e only anti, further, our check-samples show a close correspondence
between fluorescence and chi-a (R' ~ 0.78), confIrming thaI
suspended sediments have little direct effect on our chi-a
measurements. Finally, many ,\-vave eVl-'11ts lac.:ked a COlTL:sponding chi-a response, which would be unexpected if
suspended sed iment alone Ikw dominated the fluorescence
signal.
III J Vv'e suggest instead that wave-dtiven mass transpOlt, or
onshore surface StokL:s (ilif\ may be an impOltant mechanism.
Onshore wave-driven tran~port creates a converge nt flux in
nearshore '\-vaters, and resulting "blooms" \vould represent an
accumulation of plankton rather than in-situ reproductive
growth in the local population. The onshore componl-'f1t of
'wave-driven volume transport is given by Qw = K:~:~cos(OI1')'
where g is h'ravitarional accekTation, H,;ig is the significant
'wave height; C is the wave phase speed, and 8\-. is the wave
direction relative to onshore [Fewings el aI., 2008; LonguellIiggilTs, 1953]. The coast angle is fairly similar at bOlh siles,
and accounting for slight variations in incident \vave direction
did not improve the relationship between chI-a peaks and
wave peaks. Direct measurements of wave transpon are rare~
but delailed field studies combined with mod els have found
cross-shelf velocity over the inner shelf to be strongly correl ated witIl \vavc forcing_ with meall values of I to 1.5 cm S- I
•
I
.
[Fewings el al., 20081, and closer to 5 to 6 cm s dUfLllg
higher waves [Lentz et al., 2008; Carce:;: Faria el ai., 20001,
although ~ansport is weakened under high vertical mixing
[Lenlz el al., 200~j. We eonsidcT wave transp0l1 offshore of
the SUIi' zone \ovhL:fe water depth >2 Hsi~ lLentz et al., 200XJ,
or roughly 8 to 1O- m depth based on NOBe significant wave
beights. At Ihis deplh waves have not yet sleepened enough
to break and we can neglect effects of rollers and breakers
LGarcez Faria et al., 2000], yet water is shallo'w enough 10
assume Q" is roughly independent of wave period. Using a
range of onshore transport u - [0.02 to 0.061 m s 1 to characterize wave-driven surface transport offshore of the surf
70111.:, WI.,; can t:haraciclizc the <'}11shore tran spOlt of(,:hl-a and
assess its accumulation rate at the coastal boundary. A simpl!;
2-D model for the rate of increase of chl-a due to a con-
L02604
vergence of chi-li flux between the inner shelf and the shore
is qf == (Q,""_m ~Q"""'i*~ (shoreward positive in the x-direction),
where C is chlorophyll-a concentration and Q is chlorophyll-a
flux . Qcoasl is the flu x thro ugh the coastal boundary, eq ual to
zero, Qoulsidc is the chl -a flux over the hmer-shelf reg iol1~
outside the nearshore zone of interest, and the length scale
L is the distance benveen this region and the coast. If innershelf chl-a pliorto th e initiation ufa wave eV(.,'11t is chl-aouts;ide::
we can estimate the shoreward flux as Qoutside - U • Couts;ide.
Using Coulsidc - 5,1g L -1, based on observed background~
between-peak chl-a values, and using the u range given above,
1
Qoutsitk scales as approximately 0.1 to 0.3 lig L- m s-1. \Ve
assume L to be roughly 500 m, a conservative offshore distam:e over which breaking waVl:S rL:CL:uL: to linear wave
conditions [Careez Faria et ai. , 2000]. Fromlhese scales we
estimate the chi-a accumulation rate at the coastal boundary
to be [2 to 6] x 10-4 ,'gL-I S- l Note Ihat"b loom" rates would
bL: t'wice as fa~t if th(,; initial conc("'f1tration 'Wl-'1'C 10 !I.g L- 1 .
\Vith these estimated accumulation rates, near-shore chl-a
would L'lke approximalely 7 to 21 bours to increase from 5
1020 Ji.g L- 1 These estimales brackel well the chl-a increase
rdtes obs("'1ved at our sites: a l'Cpresentative time scale tor a
bloom eV(,.11t to lis('; from 5 to 20 fJ,g L 1 was about 14 to
22 hours (too short to be explained by reproductive gro\Vth
of a resident population, given typical generation times of
order a day). The agreement bet.ween scaled predictions and
observations suggests that convergent onshore \-vave transpOlt
is indeed a plausible mechanism for near-coast. blooms~ and a
consistent explanation for the fairly immediate «1 day) chl-a
response to rising H sig ' This rapid response may also explain
\\.'hy chI-a usually peaked early in a wave event rather than
l1L:ar it.s end: COnCl-'1Ttrations ncar th(,; coast could quickly (in a
time shorter than the duration of the swell event) build to
levels high enough for an otlSbore mixing loss to balance
the onshore wave-driven flu x.
L12J Stokes drift is not uniqu(,; to shallo w waters, but is a
ubiquitous feature of surface swell. It is the presenCL: of the
impenneable coastal boundary and the buoyancy of phytoplankton that causes the convergent. flux and accumulation
in this simple model. The coaslal boundruy requires that
onshor(,; and ofTshor(,; ma~s transp01t must balanc(,;, yielding
a mean vertical circulation (not unlike estuarine or frontal
circulation). Buoyant phytoplankton "particles" are more
likely to be in the near- surface \vaters that. have a net onshore
flow than ill the ofTshor(,;-tcnding dl:L:pL:r \vatl:r, yiddillg the
observed accumulation. Coastal phytoplankton arc dominated bv diatoms which are ollen seen close to the surface,
and se\~eral taxa have been documented t.o have positive
buoyancy during at least part of their growth cycle [Villareal,
19HX, 1992; Anl/lll et ai. , 2010j. However, phytoplankton
buoyancy must. be better understood before we can assess the
eflectiveness ofthe proposed mechanism. A number of wave
events lacked corresponding intertidal blooms. This highlights th e importance of better determining chl-a Icvds otTshore of the near-eoa')t zone (to properl y quantify Q o utside)
before calculating accumulation rates based on convergent
wave transport, and reminds us that the dynamics determining phytoplankton levels on Ihe shelf can be markedly
dc-coupled ii-om dynamics at thl: shordil1e.
r131 The proposed mechanism of convergent wave transport requires lilrther study, but Ihe cOlmection between wave
events and near-coast blooms has several implications that
4 of 5
L02604
MCPHEE-SHAW ET AL: NEARSHORE CHLOROPHYLL-A EVENTS
challenge our understanding of coastal ecosystem variability.
For example, "\Ie might predict the CO-OCCUlTcm:c of blooms
at sites sl-varatco by great distances, sint:c swell CY(..'lltS Spall
immense scales along the coast. OUf t\vo study sites are
km apm1 and, although not all events coincided
benveen KH and BML, many did co-occur. A chi-square
test n.:jcdco the null hypothesis that bloom CY(..'llts at the
~ 150
two sites are independent of each other (Xl - 7.09, df - I,
P < 0.0078), which strongly suggests co-occurrence at the
event-scale. Another imp0l1ant consideration is that diffen.,'llt plankton taxa may Tcsp<.md ditTcn;ntly to ,,\lave forcing,
implying a t:hangillg TCSPOllSC to the proposed lllct:hanism
as species assemblages shift in time. We speculate that the
diminished relationship behveen \vaves and chI-a events in
late summer may be caused by a seasonal shift in the buoyallCY of the dominant species alld thus ill the ability ofphytoplankton to respond to wave forcing. The role of buoyancy
and species composition in setting the timing of interaction
vvith transport mechanisms and blooms deserves further
study.
r
141 Acknowledgments. \Ve thank N. 'Velschmeyer, M. Fewings.
and S. Moni!'lmith for hclpful convcrsation during thc dcvc:iopmcnt of thi!'l
papcr. A. \-lillcr and an anonymous rcvicwcr kindly providcd hclpful
reviews. Funrling was provided by NSf grant OCE-07276 11, CSU COAST,
NOA.,A., and by the D&L Packard and the G&B :\1oore Foundations.
References
Acuna. J. L. 1-1. Lopez-Alvarez, E. Nogueira, alld F. Gonzalez-Taboada
(201 0), Diatom tlotation at thc onsct ofthc !'lpring phytoplankton hloom:
An in sitv cxpcrimcnt, Mar. Hail. Frng. ,\'er., 4(}(}, 115-125,
doi: 10.3354/meps08405.
Campbell, E. E .• and G. C. Bate (1997), Coast1l feahues associated \vith
diatom discoloration of surf zones. Bol ..Mur. 40,179 186.
doi: 10.1 5 15i hotm.1 'J'J7 .40.1-6.17'J.
Chase, Z., P. O. StruLLon, and Il. Hales (2007), Iron links river runoff and
shelf width to phytoplankton biomass along the U.S. 'Vest Coast.
Geoph.vs. Res. Leil., 34, L04607, doi:1O.l029:'2006GL028069.
Fcwing!'l, \-1., S.. 1. Lcntz, and .1. Frcdc:rick!'l {200N}, Ob!'lcrvation!'l of ero!'lSshdf flow driven by cross-shdf wimls on the inner continental shdf,
J. Phy.I'. Oceanogr., 38. 2358 2378, doi:l0.1175/2008JP03990.1.
Garcez Faria. A. F., E. B. Thornton, T. C. Lipplllallll, alld T. P. Stanton
(2000), Undcrtow ovcr a barrcd bcach,.J. Cieophy,I'. Res., /OJ, 16,'J'J'J17,010, dui: 10.1029:'2000JC900084.
L02604
Kim, II.-J., A. J. Miller, J. :\'icGowan, and M. L Carler {2009}, Coastal
phytoplankton bloolllS in the Southern California Bight. Prog Oceunogr..
82,137 147, doi:l0.10161J.pocean.2009.05.002.
Largicr,.I. L., and D. M. Hchrcn!'l (2010), Hydrography ofthc Rm!'lian
River Estuary, Sunmler-fa1l2009, report, 71 pp., Sonoma County ,"Valer
Agency, Santa Rosa. Calif.
Lentz, S. J., M. Fe\vings, P. Howd, 1. Fredericks, alld K. Hatha\vay (2008).
Ob!'lcrvations and a modcl of undcrtow over thc inner contincntal shclf.
J Phy,I·. Oceal1ogr., 38, 2341-2357, doi:l0.1175/2008JP03986.L
Longuet-Higgins, M. S. (1953), :\1ass transport in water waves, Philo.I'.
Trans. R. Soc. London, Ser. fl, 245, 535 581, doi:l0.l098/
rsta.1 'J5.1.0006.
McLachlan, A., and A. C. I3rown (2006), The Ec%g); 4 Sand)' Shore.I·.
2nd txL Academic. Amsterdam
Omand, M. M., F. Feddersen. D. B. Clark, P. J. S. Franks. 1. J. Leichter.
and R. T. (iu7.a {200'J), Intlucncc ofbuhblcs and !'land on chlorophyll-a
tluorc!'lccncc mca'llrcmcnt!'l in thc survonc, Umno/. Oreanogr. Methodl'.
7,354-362.
Ritov, G .• S. E. Dudas, B. A. :\1enge, B. A. Grantham, J. Lubchenco, and
D. R. Schicl {200N}, Thc !'lurf70nc: A !'lcmi-pcrmcahlc barricr to on!'lhorc
rccruimcnt of invcrtchratc larvac?, J K-rp. Mar. Rio/. Hro/., 3ri/, 5'J-74.
doi: 10.1 0 16iijembe.2008.04.008.
Roughan, M., A. 1. Mace. J. L Largier. S. G. Morgan, J. L. Fisher. and
M. L. Carter (2005), Sub'llrfac·c rccirculation and larval rctcntion in thc
Icc of a !'lmall hcadland: A variation on thc up\\rclling !'lhadow thcmc.
J Geoph)'s. Res., lID, Cl 0027, dui: 10.1029:'2005JC002898.
Ryan, J. P .• J. F. R. Gmver, S. A. King, W. P. Bissett, A. M. Fischer, R.:\1.
Kudela. Z. Kolber. F. Mazzillo, E. V. Rienecker, alld F. P. Chavez
{200N}, A coa!'ltal occan cxtrcmc hloom incubator, CienphY,I'. Res. ',eft..
35, Ll2602, doi:l0.1029:'2008GL0J4081.
Shanks. A. L.. S. G. :\1orgall. J. MacMahall, and 1. H. M. Reniers (2010).
Surf zone physical and morphological regime as determinants of temporal and spatial variation in larval rccruitmcnt, .J. A"xp. Mar. Rio/. Hro/..
392,140-150, doi:l0.10161:i..iembe.2010.04.018.
Villareal, T. A (1988), Positive buoyancy in the oceanic rnatolll Rhizosolenia
debyana H. Peragallo.Deep Seu Ri.~I'., Purl .1,35, 1037 1045, doi:l0.l 016/
01 'IN-014WNX}'J0075-1.
Villarcal, T. A. (l 'J(2), Huoyancy propcrtics ofthc giant diatom Hthmodiscus, J. Plankton Re.I·., 14, 459--463, doi: 10.1093/plallkli14.3A59.
'Vieters, E. A., D. :\1. Kaplan. S. A. Navarrete, A. Sotomayor. J. Largier.
K. .I. Niclscn, and F. V6liz (200.1), Along!'lhoTC and tcmporal variability
in chlorophyll a conccntration in Chilean ncar!'lhorc watcr!'l, Mar. Hro/.
Prog Sa., 249, 93-105, dui:1O.3354/111eps249093.
J. L. Largier, Ilorlega Marine Laboratory, University ofCalifomia. Davis.
PO Box 247. Bodega Bay, CA 94923, CSA.
E. E. McPhee-Sha\v. :\1oss Landing :\1arine Laboratories. San Jose State
Univcr!'lity, X272 \-10!'l!'l Landing Rd., Mos!'l Landing, CA Y50.1'J, LSA.
{eshawq~mI111Lcalstale.txlu)
B. A. :\1enge, Departlllent of Zoology, Oregon State University. 3029
Cordle'v Hall, Corvallis, OR 97331. USA.
K. .I.' ",\ic:i!'lcn, Departmcnt of Hiology, Sonoma Statc Univcrsity, 1N01 E.
Cotali Ave., Rolmert Park, CA 94928, CSA.
5 of 5