Table Of Content

Reference: Bio/. Bull. 191: 367-373.(December, 1996) Chemical Induction of Larval Settlement Behavior in Flow DAMVAIRDIOS. NW.ETTAHMEBYU1RRAIN1D*, RCIHCRHIASRTDOPKH.ERZIMM.MEFRI-NEFLALUIS1,T2 , 1DepartmentofBiologicalSciences, MarineScienceProgram, andBelle l\'. BaruchInstitute forMarineBiologyandCoastalResearch. UniversityofSouth Carolina, Columbia. South Carolina29208:and'Department ofBiology, UniversityofCalifornia, Box 951606. LosAngeles, California 90095-1606 Abstract. The ability ofdissolved chemical cues to in- Gross et ai. 1992). Once a substrate is accepted, larvae duce larval settlement from the water column has long then metamorphose intothejuvenile form (Morse etai, been debated. Through computer-assisted video motion 1979; Burke, 1983;Crisp, 1984; BonaretaL 1990). analysis,wequantifiedthemovementsofindividual oys- Neithersessilenorsedentaryinvertebratestravel far, if ter (Crassostrea virginica) larvae in a small racetrack at all, after metamorphosis. The ability oftheirlarvae to flumeatfree-stream flowspeedsof2.8, 6.2,and 10.4 cm/s. select appropriate substrates is therefore critical to adult In response to waterbornechemical cues, but not to sea- fitness (Grosberg, 1987; Levitan, 1991). Commonly, water (control), oyster larvae moved downward in the substrate discrimination isbased on chemical cuesasso- watercolumn and swam in slow curved paths before at- ciated with predators, competitors, conspecifics, or food taching to the flume bottom. Effective stimuli were (Burke, 1984; Woodin, 1987;Johnson and Strathmann, adult-oyster-conditioned seawater (OCW) and a syn- 1989; Morse and Morse, 1991). These cues are thought thetic peptide analog (glycyl-glycyl-L-arginine) for the to be mostly non-waterborne and detected after larvae natural cue. The chemically mediated behavioral re- contact surfaces (see reviews by Morse, 1990; Pawlik, sponsesofoysterlarvae in flowwereessentially identical 1992). tothose responsespreviously reported in still water. Our Settlement induction by waterborne substances is of- experimental results therefore demonstrate the capacity ten considered to be unlikely for weakly swimming in- ofwaterborne cues to evoke settlement behavior in oys- vertebrate larvae (Crisp and Meadows, 1962; Butman, terpediveligersundervaryinghydrodynamicconditions. 1989). The locomotory speeds ofthese larval forms are typically lower than horizontal water-flow velocities in Introduction benthic boundary layers, even at distances ofless than onelarvalbodylengthabovetheseabed(Butman, 1986). Larval recruitment is a major factor regulating the Yet behavioral responses to chemical cues in the water population dynamics of marine benthic invertebrates column might effectivelyenhance larval settlementonto (Gaines et a/.. 1985; Roughgarden el a!., 1988; Un- the substratum under some circumstances. A recent derwood and Fairweather, 1989; Menge, 1991). Two model suggests that by responding quickly to a water- fundamental steps in the recruitment process are settle- borne cue, weakly swimming larvae might change their mentand metamorphosis. Thetransport anddeliveryof vertical distributionsin thewatercolumn (making more planktoniclarvaefrom thewatercolumn totheseabed is larvaeavailable to be swept nearthe substratum by fluid the settlement stage (Butman, 1987; Eckman, 1990; flow) and thusdramatically alter their patterns ofsettlement(Eckman et a/.. 1994). Received 16May 1996;accepted23September 1996. Oyster larvae provide a valuable tool forassessing the *Currentaddress:MontereyBayAquariumResearchInstitute.P.O. roleofwaterbornechemicalcuesasagentsmediatingset- Box628.MossLanding,California95039. tlement in flow. In assaysconducted in still water, oyster 367 368 M. N. TAMBURRI ET AL. (Crassostrea virginica) pediveligers respond to dissolved Flumedesign chemicalcuesbymovingdownwardinthewatercolumn To minimizethe number oflarvaeand the volume of aetndala,tt1a9c9h2i;ngZitommtheer-bFoatutosmtseolfalt.e.sticnhparmesbse)r.sT(hTeamstbiumrurlii tdeuscttseodluitnioanssmnaelled(e4d0-p1ecraptraicailt,ya)l,lpelxapsetricimreanctetsrwaeckreflcuomn-e asreeawsamtaelrlbbasyicapdeuplttidecson(scpa.ec5i0fi0c-s1,(0Z00imDmae)r-reFlaeuassted ianntdo (sFeimgi.ci1r).cuTlhareefnldusm(e2w0a-scm2r0a-dcimuswaitdei,nnceornwsailsltsi)nganodf ttwwoo cTuaemsbuirsrith,e1t9r9i4-)p.epAtipdoetegnltycaynl-aglloygcyflo-rLt-haergniantiunreal(GcGueR)o.r smteraasiguhrtesdec5ticomn.s(I1n2o0n-ecmstlronagi)g.htWaiwtahy4s0ect1ioofnw,awtaert,erdefpltohw maximally effectiveat 10~8 M. was generated through the use ofa digitally controlled These waterborne chemical cues may also influence motor-driven paddle wheel (48-cm total diameter) with oyster settlement in flowing water. In flume assays, sig- aminimumbottomclearanceof0.3 cm.Thedrivewheel nificantly more oyster larvae accumulate on substrates consisted of 12 paddles (7 cm X 19.2 cm), each tilted so in targetwellsreleasingthesyntheticpeptideanalogthan that it exited vertically from the water on upstrokes. To on substrates in seawater alone (Turner et al., 1994). reduceacross-stream fluid motion, polycarbonate sheet- aSsfputocenrhdcdoitnsottarccihtbeuwmtiiitcohansltescctousueulsbdsetirrteahsteuerlst.inTbeotchfaeuurswteahteterhreesctloaablrluviamsehntrhoeer- iefinnrgsdtwtoaufsrtnphlteoascaeecdctopanasdrfasllltoerwlasittgorhattiahgwehatceynuesrrevsce.tdiAoftn,ltuhameewdoworawlknlissntgirnaetrahemea capacity forsettlement behaviorin flow, wequantify the wasdesignated in which velocity profileswere taken and movements ofoyster larvae suspended in a small race- movementsby larvaewere video-recorded. trackflume.Ourresultsclearlydemonstratechemicalin- Velocity profiles were determined at each of three duction ofsettlement from the water column under hy- speed settings (2.8, 6.2, and 10.4 cm/s free-stream flow drodynamicconditionsapproachingthose in nature. velocity, respectively) usinga hot-film anemometer(TSI 1053-B). At each speed setting, a conical platinum film Materialsand Methods sensor 1 mm diam, TSI model 1231W, 10C over- heat) was used to measure velocity at 16 heights above Larvalcultures the flumebottom. Thesedatawerecollected at 1 Hz, us- Larvae ofCrassostrea virginica were maintained on a ing a data logger (Campbell, Model CR10). Mean flow 12:12 D:Lcycle(lighton: 0700 h)at 1.0/ml in a 1:1 mix- speedateach settingand heightwasdetermined byaver- tureofoceanicandartificial seawatermedium(25Cand aging measurements over 1-min intervals. To establish 25% salinity). Cultures were actively aerated and gently the shear velocity (/*) for flow at each speed setting, stirred by air bubbled through Pasteur pipettes. Marine mean flow velocities were regressed against the natural flagellates (Isochrysisgalbana and Pavlova lutheri) were logarithm ofheight above the flume bottom (Nowell et providedasfoodat 2.5 X 104cells/ml onceeach day. All al.. 1981;JonssontVt//.. 1991). experiments were begun within 6 h after 95% ofthe larvae inculturehaddevelopedeyes, andexperimentswere Experimentalprotocol run for 24 h thereafter (Tamburri et al., 1992). These conditionswereineffectforlarvaeaged 19-21 dayspost- Separatetrialswereperformedintheflumewitheither OCW, GGR, orseawater(control) at a uniform concen- fertilization. tration mixed throughout the water column. In each Chemicalsolutions trial,wesuspended 2000 358 (range)larvaeinabeaker with 200-300 ml of seawater. The larvae were then Procedures for making solutions were similar to ones gently poured from the beaker (over 5-10 s) into the tperre-vcioonudsiltyidoensecdrsiebaewda(teTram(bOuCrrWi)eatnadl..101~9892M).GBGotRhwoeyrse- f1l.0umme autpsatsrietaemraoifsetdheawboourtki4ngcmsecatbioonv.eVtihdeeob-ortetcoomrdainndg usedasteststimuli,andnatural seawater(alone)wasem- was limited to the first pass oflarvae over the working ployed ascontrol. Because a large volume offreshly pre- section. The flume was drained and then cleaned with pared OCW wasrequired in each trial, testswith thisso- several rinses of distilled water between each trial. Six lutionwerelimitedtoaspeedsettingof2.8 cm/s. Incon- trialswere run foreach solution atthe two slowest speed trast, trials with GGR and seawater (control) were settings, and threetrialseach atthe fastest speed. conducted at speed settings of 2.8, 6.2, and 10.4 cm/s. Each test or control solution was prepared and filtered 1'ideo imagingandmotion analysis tcoon1d.u0c^temdwwiitthhinsol1uhtiboenfsorheeludsea.t Aildlenteixcpaelritmeemnptersawteurree Simultaneous video-recordings from two cameras and salinityaslarval cultures. (Panasonic WV-BL-600), each equipped with a 55-mm LARVAL SETTLEMENT IN FLOW 369 370 M. N. TAMBURRI KT AL TableI Mean(SEM)trajectorycalculatedaspercentslopeforeachlarval pathinresponseintestorcontrolsolutionat2.8cm/sflowspeed 3 Solution Slope Significance O U, = 0.420 SWC 39 2.93 1.51 A GGR 43 -3.80 1.81 B 2 OCW 32 -3.42 1.37 B > SWC = seawater control, GGR = 10 8 Mglycyl-glycyl-L-arginine, OCW = adult-oyster-conditioned seawater. Trajectories were calcu- latedastheratioofchangeinvertical positiontochangein horizontal position between beginning and end points ofpaths. Positive values represent net upward motionand negativevaluesrepresent netdown- ward motion. Groupswithdifferent lettersindicatea statistical differ- 10 12 enceatthe = 0.05/3 = 0.017level. Flow speed (cm/s) Figure 2. Velocity profiles for flowing water at each ofthe three The exact height at which each larva swam within flumesettingsusedinexperiments.Fromlefttoright,eachprofileisfor 4 mm ofthe flume bottom could not be precisely identi- f([lo'w)wfoirtheaachfrfeleo-wstirsesahmowsnp.eed of2.8. 6.2. or 10.4cm/s; shear velocity tfioedt.heBebcoatutsoemth(eFisg.pe2e)d, odfetfelruimdinmaottiioonns doefcvreelaosceidtyclaonsde rate ofchange in direction (RCDI) for horizontal move- Results ments by larvae reflected components of both active swimmingandpassiveflow. Wetherefore madecompar- Hydrodynamics isons between test and control treatments separately at Velocityprofiles,aswellasshearvelocity(U*),forflow each speed setting. In thiscontext, slower larval velocity at each speed settingare shown in Figure 2. Theaverage indicatesthatalarvaeitherswimsclosertotheflumebot- velocity (U) for the slowest flow was 2.75 cm/s, for the tom or indeed swims more slowly. In either case, a sig- intermediate flow was6.21 cm/s, and forthe fastest flow nificant difference between stimulus and control treat- was 10.35 cm/s. Regressions used to calculate shear ve- mentsdemonstratesasignificantchange in larval behav- locities accounted for more than 95% ofthe variation in ior. With these caveats in mind, we found that paths thedata in each case. traveledbylarvaeinseawater(control)weresignificantly straighter (lower mean RCDI) and faster than those in Settlementandbehavioroflarvaeinflow OCW and GGR (Table III, Kruskal-Wallis P < 0.05 experimental-wise error, all comparisons). Remarkably, The behavior ofoyster larvae in response to seawater larvae often turned and swam upstream in response to O(cConWtroaln)dwaGsGmRar(kKerdulsykadli-ffWearlelnitsfxr:om=t1h2a.t7i4n.rdefs=po2n.sePt=o GonGlyRfaltow2.s8peaendda6t.2whcimc/hs athnids tsoolOutCioWn watas2.t8esctme/d)s,(btuhte 0.002). In seawater, the average trajectory was upward neverin responseto seawater(Fig. 3). (2.9% slope), indicating that the larvae remain suspended above the flume bottom (Table I). An opposite OCW GGR effectwasfoundin responseto and theav- TableU erage trajectory was downward (3.4% and 3.8% slope. Mean(SEM)numberoflan'aesettlingonthe/tunicbottom TvaaeblmeoIv).edMeiannressppeoendssetaotOwhCicWhadnodwnGwGarRd-wdeirreec1t.7edmlmar/- inresponsetocitherlestorcontrolsolution s (0.6 mm/s SEM) and 1.8 mm/s (0.7 mm/s SEM), respectively. After traveling downward in the water column, larvae approached the flume bottom and settled. Solution OCW Hence, the numbers ofsettlers in at 2.8 cm/s flow speed and in GGRat 2.8 and 6.2 cm/sweresignificantly higher than those in seawater (control) at the same flow speeds (Table II, Kruskal-Wallis test). Although not sta- tistically significant, there was a trend in the same direction for GGR at the fastest flow speed (10.4 cm/s. Ta- ble II). LARVAL SETTLEMENT IN FLOW 371 TableIII Mean(SEM)speedandraleo/changeindirection(RCOllforhorizontalmovementsbylarvae, aminumberoflarvae turningupstreaminresponsetoeithertestorcontrolsolution Movementbylarvae Numberoflarvae Flowspeed(cm/s) Solution Speed(mm/s) RCD1(degrees/s) Observed Movingupstream 2.8 372 M. N. TAMBURRI ET AL. Flow swimminginvertebratelarvae hasbeen proposed by But- man and Grassle (1992). In this model, larvae are hy- 2.8 cm/s pothesized to move up and down in the water column closetothe bottom,whilebeingtransported by flow,and SWC to test substrates on contact. Selection is thus accom- A. plished by active acceptance or rejection oftouchdown sites. We suggest, at least foroyster pediveligers, that behavioral response toawaterbornechemical cue isan ad- o o o ditional factor in habitat selection. Oysterlarvae rapidly respond (within 5 s ofinitial exposure) to dissolved substances produced byjuvenile and adult conspecifics and 00000 by bacteria in biofilms on the surface of oyster shell (Tamburri et ill., 1992; Zimmer-Faust et ai, in press). 1 mm These stimulatorycompoundsare low molecularweight peptides with arginine at the C-terminal; at concentra- tions of 10 "' M, or lower, they effectively cause down- B. GGR ward-directed movements by oyster larvae in the water column (Zimmer-Faust and Tamburri, 1994). Because thisbehavioral responsedoes not requireachemicalgra- dient, induction of settlement should be possible even inturbulent-flowenvironments. Presumably,downward movements increase the probability that oyster larvae will be swept into contact with preferred benthic sub- strata by near-bed turbulent eddies. Juvenile and adult oysters commonly form extensive aggregations ( 10,000 m:, and beyond) at high densities (15,000 individuals/in2 in benthic estuarine habitats ) C. OCW (Bahr, 1981). Waterbornecompoundsproducedbypost- metamorphic oysters will thus be distributed over large areas, rather than confined to isolated point sources. Field data further indicate that flow velocities measured 2-10 cm above oyster reefs rarely exceed 7 cm/s (Breit- burg et ul.. 1995). This means that oyster larvae drifting near the bottom will likely pass over cued substratum for several minutes, long enough for the induction ofa settlement response. We suggest that the effectiveness ofwaterborne settle- Figure3. Computer-digitizedvideorecordsshowingthehorizontal ment cues may be greatest in environments similar to pathsoflarvaemovingwithin4mmoftheflumebottomandentering those in which oyster larvae settle. Perhaps the capacity the 1.2- x0.7-cmfieldviewedbytheoverheadcamera.Thepathsgen- ofdissolved compoundsto influencesettlement islinked erated for four larvae appear in each box. Each path ofdots presents to cue production and, therefore, is most important in ctohlelepcotseidtiaotn3o0ffaralmaersv/as.atWceonsseelceucttievde,thoenpea-tfhrsamtoeiilnltuesrtvraatlestwhiethvavriidoeuos habitats that have large aggregations of post-metamor- movementpatterns(includingupstreamswimming)observed in flow- phic conspecifics or food resources. Waterborne cues ing waterat 2.8cm/s. Thesedata are for larvae exposed toeither(A) might also function as general chemical signatures indi- seawatercontrol(SWC),(B) I0~8Mglycyl-glycyl-L-arginine(GGR).or catingthe presence ofpreferred substrates, such asthose (inC)roeyssptoenrs-ectoondSiWtiContehdasnetaowaetietrhe(rOGCGW)R.oPraOthCsWa.restraighterandfaster pmruodvidfleadts.wiAtlhtihnocuogrhaltrheeefse,ffkeectlspofofrewsatts,erabnodrnoregacnhiecm-irciaclh cuesmay be profound, the extent towhich such cues influencesettlement remainslargely unexplored. velocities > 3 mm/s) can overcome turbulent transport. Remarkably, these speedsare common in water flowing Acknowledgments above oyster reefs in tidal estuaries (Breitburg el ai, 1995). This research was sponsored by awards from the Na- A model for benthic-habitat site selection by weakly tional Science Foundation (OCE 94-16749). South Car- LARVAL SETTLEMENT IN FLOW 373 olina Sea Grant Consortium (P-M-2F, P-M-2K, R-92- elingoflarval settlement in turbulent bottom boundary layers. J. Mar Re* 50:611-642. cR88eu)sm,e-aOLrfcufhin,czeaFonofduNnSadivagatmilaonRX.eis.eLaeWrrecnhetrh(-aNGn0rk0a1yK4.-FKuK-nu8dr2k-fo6o4wr6s)Mk,airSailnnode- JohnsHsuiobolsn.t,rEaLct.oalE.p.,r1ea2vn8id:o8uR7s.l-y1R0.o3Sc.tcurpaitehdmabnyn.a1m9o8b9i.leSpertetdlaitnogrb.arJnaEc.l\re>s.aMvaori.d V. L. Shaffer for generously providing larvae, C. C. Gee Jonsson, P.R., C. Andre, and M. Lindegrath. 1991. Swimmingbe- forassistancein illustratinglarval paths, and N. D. Pent- haviorofmarinebivalvelarvaeinaflumeboundarylayerflow:ev- chefT and Drs. A. W. Decho, M. W. Luckenbach, and idencefornear-bottomconfinement.Mar. Ecol. Prog. Ser. 79:67- 76. S. E. Stancyk for their comments, which greatly im- Levitan, D. R. 1991. Influence ofbody size and population density provedearlierdraftsofthis manuscript. onfertilizationsuccessandreproductiveoutputinafree-spawning invertebrate.Bioi Bull. 181:261-268. Mcnge, B. A. 1991. Relative importance ofrecruitment and other LiteratureCited causesofvariationinrockyintertidalcommunitystructure.J. E.\p Mar Bioi Ecol 146:69-100. Bahr, L.M. 1981. Theecologyofintertidal oysterreefsoftheSouth Morse,A.N.C.,andD.E.Morse. 1991. Enzymaticcharacterization Atlanticcoast: a community profile. U.S. Eixh II';/<//. Sen'. FWS/ ofthe morphogen recognized by Agancia humilix (Scleractinian BonaOrB,S-D8.1B/.1,5S..L.Coon,M.\Valch,R.M.Weiner,andW.Fitt. 1990. Morcsoer,alD).laEr.va1e9.9B0i.oiRBeulcle.nt18p1r:og1r0e4s-s12i2n.larval settlement: closingthe Control ofoyster settlement and metamorphosis by endogenous gap between molecular biology and ecology. Bull. Mar. Sci. 46: andexogenouschemicalcues.Bull. Mar.Sci.46:484-498. 465-483. Breitburg, D. L.,M.A.Palmer,and I. Loner. 1995. Larvaldistribu- Morse, D.E., N. Hooker, H. Duncan, and L.Jensen. 1979. y-Ami- tions and the spatial patterns ofsettlement ofan oyster reeffish: nobutyricacid,aneurotransmitter,inducesplanktonicabalonelar- responsestoflowandstructure.Mar. Ecol. Prog.Set: 125:45-60. vaetosettleandbegin metamorphosis.Science204:407-410. Burke,R. D. 1983. Theinductionofmetamorphosisofmarineinver- Mullineaux, I,.S., andC.A. Butman. 1991. Initial contact,explora- tebratelarvae:stimulusandresponse.Can../ Zoo/.61: 1701-1719. tionandattachmentofbarnaclecypridssettlinginflow.Mar. Bioi Burke,R.D. 1984. PheromonalcontrolofmetamorphosisinthePa- 110:93-103. cificsanddollar,Dendraxlerexcenlricns.Science225:442-443. Nowell, A. R.M., P.A.Jumars, and J.E. Eckman. 1981. Effectsof Butman, C.A. 1986. Larval settlement of soft-sediment inverte- biological activity on the entrainment ofmarine sediments. Mar. brates:somepredictionsbasedonananalysisofnear-bottomveloc- GVo/.42: 133-153. ity profiles. Pp. 487-513 in MarineInterlace* Ecohydrodynamics, Pawlik, J. R. 1992. Chemical ecology ofthe settlement ofbenthic J. C.J. Nihoul.ed. Elsevier,Amsterdam. marine invertebrates. Oceanogr. Mar Bioi Annit. Rer. 30: 273- Butman, C.A. 1987. Larval settlement of soft-sediment inverte- 335. brates: the spatial scales ofpattern explained by habitat selection Pawlik,J. R., and C. A. Butman. 1993. Settlement ofa marinetube andtheemergingroleofhydrodynamicprocesses. Occam>K>'- Mar wormasafunctionofcurrentvelocity:interactingeffectsofhydro- Biol.Annu.Rev.25: 113-165. dynamicsandbehavior. Limnol. Oceanogr.38: 1730-1740. Butman, C. A. 1989. Sediment-trapexperimentson the importance Pawlik, J.R., C.A. Butman, and V.R. Starczak. 1991. Hydrody- ofhydrodynamical processes in distributing settling invertebrate namic facilitation ofgregarious settlement ofa reef-building tube larvaeinnear-bottomwaters.J. E.\p. Mar. Bioi Ecul. 134:37-88. worm.Science251:421-424. Butman, C.A., and J.P. Grassle. 1992. Active habitat selection by Roiighgarden, J., S. D. Gaines, and H. Possingham. 1988. Recruit- Capilella sp. larvae: I. two-choice experiments in still water and mentdynamicsincomplexlife-cycles.Science241: 1460-1466. flumeflow.J.Mar. Res.50:669-715. Tamburri, M.N., and R. K. Zimmer-Faust. 1996. Suspension-feed- Crisp, D.J. 1984. Overview ofresearch on marine invertebrate lar- ing:basicmechanismscontrollingrecognitionandingestionoflar- vae. 1940-1980.Pp. 103-126inMarineBiode/eriora/ion:AnInter- vae.Limnol. Oceanogr.41: 1188-1 197. disciplinaryStudy, J. D. Costlowand R.C. Tipper, eds. Naval In- Tamburri,M. N.,R.K.Zimmer-Faust.andM.L.Tamplin.1992. Nat- stitute,Annapolis, Maryland. ural sourcesand propertiesofchemical inducers mediatingsettle- Crisp,D.J.,and P.S.Meadows. 1962. Thechemicalbasisofgregar- mentofoysterlarvae:are-examination.Bioi Bull 183:327-338. iousnessincirripedes.Proc. Roy. Sin: Loud. B156:500-520. Turner,E.J.,R.K.Zimmer-Faust,M.A.Palmer,M.Luckenbach,and Eckman.J.E.1990. Amodelofpassivesettlementbyplanktoniclar- N.D. PentchefT. 1994. Settlement ofoyster (Craxxox/rea virgin- vae onto bottoms ofdiffering roughness. Limnol. Oceanogr. 35: ica) larvae: effectsofwaterflowandawater-solublechemical cue. 887-901. Limnol. Oceanogr.39: 1579-1593. Eckman,J.E.,F.E.Werner,andT. F.Gross. 1994. Modelingsome Underwood, A.J.,and P.G. Fairweather. 1989. Supply-sideecology effects of behavior on larval settlement in a turbulent boundary andbenthicmarineassemblages. Trend.',Ecol.Evoi4: 16-20. layer.Deep-SeaRes.41: 185-208. Woodin, S. A. 1987. Recruitment ofinfauna positive or negative Gaines,S.D.,S.Brown,andJ.Roughgarden. 1985. Spatialvariation cues?Am. Zoo/.31:797-807. inlarvalconcentrationsasacauseofspatial variationinsettlement Wright, L. D., R.A. Gammisch, and R.J. Byrne. 1990. Hydraulic forthebarnacle.Balamixglanditla. Oecologia67:267-272. roughnessand mobility ofthreeoyster-bed artificial substrate ma- Grassle,J. P.,C.A. Butman, and S. VV. Mills. 1992. Active habitat terials.J CoaxialRex.6:867-878. selection by Capile/lasp. 1 larvae: II. multiplechoice experiments Zimmer-Faust, R. K.,and M. N.Tamburri. 1994. Chemical identity instillwaterandflumeflows.J Mar. Ret 50:717-743. andecological implicationsofawaterborne, larval settlementcue. Grosberg, R.K. 1987. Limited dispersal and proximity-dependent Limnol. Oceanogr.39: 1075-1087. matingsuccessinthecolonial ascidian Bolryllusschlosseri. Evolu- Zimmer-Faust, R. K., M.N. Tamburri, and A.W. Decho. In press. twn41:372-384. Chemosensory ecology ofoyster larvae: benthic-pelagic coupling. Gross,T.F.,F.E.Werner,andJ. E.Eckman. 1992. Numericalmod- Mar. Fri'slimit Bcliav. I'liyxiol.