Reference: Biol. Bull 180: 112-118. (February, 1991) Settlement, Refuges, and Adult Body Form in Colonial A Marine Invertebrates: Field Experiment LINDA WALTERS' AND DAVID WETHEY12 J. S. 1Department ofBiologicalSciences and-MarineScience Program, University ofSouth Carolina. Columbia, South Carolina 29208 Abstract. We examine the relationship between adult at the time oflarval settlement ordevelop laterasa result body form (sheet vs. arborescent) and larval settlement ofdifferential mortality. The distribution of individuals in colonial animals. Because thin sheet forms are more at the time oflarval settlement has a strong influence on susceptible to overgrowth than arborescent forms, we theirfuturesuccess. Individualsthatsettle neardominant predict that larvae of sheet forms should preferentially competitors are more likely to die quickly, as are those settle in refuges from competitors. On both natural and that settle within the range ofpredators or where distur- artificial substrata, the larvae of the sheet form (Mem- bance events frequently occur. braniporamembranacea)settled moreoftenon highspots, There are a number ofpotential escapes from sources which could serve as refuges from competition. The ar- ofbiotic mortality, including simple avoidance ofsettle- borescent forms (Bitgula ncrilina and Distaplia occiden- ment nearenemies(e.g.. Grosberg, 1981;YoungandChia, talis) settled around the bases ofbumps more frequently 1981) and recruitment to spatial refuges (e.g., Connell, than would beexpectedbychance. Formanyarborescent 1961; Dayton. 1971; Paine, 1974; Wethey, 1983; Walters forms, their most vulnerable periods are the days im- and Wethey, 1986). Organisms located in spatial refuges mediately following settlement, when individuals can be increase their chances of survival against competitors, consumed easily by predators or dislodged by physical predators, and disturbance events. Size can also be pro- disturbances. Settlement in a crevice (base of a bump) tective to colonies once they have grown to certain di- would provide protection from the bulky mouthparts of mensions unaffected bycompetitors; thisisthesize refuge. predators. Moreover, dislodgment would be less likely Potential morphological escapes may also exist. Among than ifsettlement had occurred on flat locations, such as colonial organisms attached to hard substrata, one can the tops ofbumps or the areas between bumps. distinguish a number of morphological types, including sheet and tree forms (Jackson, 1979). The outcomes of Introduction competitive interactions can be strongly influenced by the morphologiesofthe competitors. Tree formsare relatively Strikingpatternsofspatialdistribution arecharacteristic isolated from the substratum-associated competitors ofmany marine invertebrates sessile on algae, rocks, and (Jackson, 1979;Grosberg, 1981), whereassheet formsen- other hard surfaces. Individuals are often found in aggre- crust the substratum and may suffer competitive inter- gations relative to each other (e.g., Knight-Jones, 1951; actions along their edges. Thin sheets tend to lose to Crisp, 1961; Wethey, 1984), relative to topographic fea- thicker forms (Buss, 1980; Seed and O'Connor, 1981; tures ofthe substrata (e.g., Crisp and Barnes, 1954; Ry- Russ, 1982; Sebens, 1985, 1986; Walters and Wethey, land, 1959; Crisp, 1961; Wisely, 1960; Hayward and 1986) unless they have a height advantage in the zone of Harvey, 1974;Keoughand Downes, 1982;Wethey, 1986; contact(Waltersand Wethey, 1986).Therefore,onewould LeTourneuxand Bourget, 1988), orrelativetomicroflora predict that animals with thin, sheet-like growth forms (e.g.. Crisp and Ryland, 1960; Brancato and Woollacott, should preferentially settleon ornearlocationswherethey 1982; Strathmann et ai. 1981). These patterns may arise have a height advantage (Walters and Wethey, 1986). Although tree forms are less likely to be overgrown by Received 27July 1990; accepted 29 November 1990. competitors, they can be more visible to predators and 112 INVERTEBRATE SETTLEMENT REFUGES 113 are more susceptible to total colony mortality than sheet the laboratory, adhesion in Distaplia occurs within 30 s forms. On irregularsubstrata, a potential settlement refuge at 15C (Cloney, 1978). Tail resorption reduces the size location would befoundaround the basesofbumps. Here, ofthe newly settled individual to approximately 650 ^m certain predators may not be able to reach newly settled within 7 min (Cloney, 1978). individuals. Here they are also protected from more dis- Forthe purposes ofthis study, it was important to dis- turbance events than they would be ifthey were located tinguish between newly settled and metamorphosed in- on a flat surface or on the top ofa bump. dividuals. Newly metamorphosed Membraniporacolonies We examined the patterns oflarval settlement in three have only the twin ancestrula skeleton fully formed, and species of encrusting colonial animals with different Bugiila has only the first zooid skeleton completed. Dis- growth forms. We asked whether the settlement patterns laplia colonies were considered new individuals if they mm wereconsistent with our prediction that species with thin occupied less than 1 2. sheet morphologies should choose spatial refuges from competitors, whereas species with tree morphologies Experimentalprocedure should choose refuge locations that would reducethe risk ofpredation and disturbance. The encrustingcheilostome To study larval settlement on natural substrata, weex- bryozoan Membranipora membranaceawasourexample amined the alga Laminaria saccharina. Plants were col- ofathin sheet morphology, andthearborescent bryozoan lected on the floating docks at the Friday Harbor Labo- Bugiila neritina and the pedunculate ascidian Distaplia ratories, San Juan Island, Washington state (48 32' 42" occidentaliswere ourexamples oftree morphologies. We N; 123 0' 39" W) and on the floating public docks at examined two kinds of substrata. The kelp Laminaria Fisherman's Bay on Lopez Island, Washington state(48 saccharina is a substratum commonly colonized by all 30'30"N; 122 54'51"W). Entirebladeswereeitherplaced three species. Settlement platescast from bumpson Lego in running seawater tables and a census taken within 48 toy building blocks and pits created from bubble plastic h, orfrozen immediately foralatercensus. Random pieces served as model topographies ofthe same spatial scale as ofthealga(20 X 20cm)werecut from thecentral portion those found on Laminaria. Our analysis was carried out oflarge ( 1.0-2.0 m in length) Laminaria fronds. All new in two phases: (1) we examined the extent to which set- settlerswere recorded on each algal square. Asthe topog- tlement on our model substrata mimicked that on natural raphies ofthe blades are quite variable, we could not dis- surfaces;and(2)weexamined in detail thespatial pattern tinguish a pit from a bump. Instead, each topographical ofsettlement on the model substrata. feature on the blade was defined as a continuous slope extending from a lowest to a highest point (Fig. 1). The Materials and Methods lowest point on one side ofan algal blade is the highest point on the reverse side. The diameter (base) and the Sludy organisms height of each topographic feature were recorded with The bryozoans Membranipora membranacea and verniercalipers. The slopes ranged in length from 1 to 20 Bugiila neritina have small, ciliated larvae (Membrani- mm. The location ofeach animal wasdetermined bycre- pora: 750 urn; Bugiila: 200 ^m, from Reed, 1987) that have limited swimmingabilities in the ocean (Chia etai, 1984). However, these larvae can choose their settlement Hiqhest point locations. When competent, they move closely over the substrata and test it (Woollacott and Zimmer, 1978, for Bugiila: Atkins, 1955, for Membranipora). During this phase, Bugit/a larvae form temporary attachments using adhesives that are sufficiently strong to prevent the indi- Height of topogrophic vidual from being mechanically dislodged (Loeb and feoture (h) Walker, 1977). Bugiila can quickly dissolve the adhesive orchange itsviscositytodetach from, orrejectthesurface (Reed and Woollacott, 1982). In theplankton, Distapliaoccidentalislarvaearemuch larger tmhman those ofthe other two species, measuring up Distonce from lowest point to animal to 3.2 in length (Cloney and Torrence, 1984). Most encounter a number of surface locations before meta- Diameter of topographic feature (d) morphosingon oneofthem (R. A. Cloney, pers. comm.). Figure I. Each colony was mapped in relation tothe nearesttopo- TorrenceandCloney(1988)suggestthat sensory neurons graphic high and lowpoint. Thedimensionsofthetopographic feature in the adhesive papillae may be common in ascidians. In were measured. 14 L J. WALTERS AND D. S. WETHEY ating a right triangle with the animal location and the rect observation at Lopez Island in 1989. Flash-lit pho- lowest point as two ofthe points (Fig. 1 ). The distance tographs were taken underwater using Kodak Technical from thelowestpoint totheanimal andtheanimal height Pan 2415 film and a Nikonos 5 camera equipped with a above the lowest point were measured (Fig. 1). 5:1 extension tube and focal framer. Negatives were ob- Using the diameter (d) and height (h) of each topo- served under a dissecting microscope equipped with an graphic feature, we calculated: ocular micrometer to determine the specific locations of (1) the radiusofcurvature (re) ofthe topographic feature: newly settled individuals. We distinguished among four h2 + (d/2) kinds of locations on the plates with bumps: (1) top of re = + bump;(2)sideofbump;(3)touchingthebaseofthebump; 2*(1 d)*(d/(2*h)) and (4) on the flat surface not touching the base ofthe (2) the vertical position (vp) ofthe animal, which we use bump. On the pitted surface, we distinguished among todetermine the location (top, side orbase) ofthe organ- three kinds of locations: ( 1) in the pit; (2) touching the ism on the topographic feature: edge ofthe pit, and (3) on the flat surface not touching vp = (animal height/h). the pit. Individual larvae were scored as touching a to- pographic feature ifthey were within 250 ^m ofthe fea- Wilcoxon rank sum testswere usedtodetermine ifthere ture. This distance represents approximately one body twherreeedcioffmemreonncesspiencileso.catWieonsexoacmciunpeieddthbeyelfafrevctaseooffstihzee lenTgothdoeftethremisneettlwehdeltahrevraela(r2v0a0etsoet7t5l0eduprrnefienrelnetnigatlhl)y. in and shape oftopographic features (height, diameter, and relation to topographic features, wecompared ourobser- fhreaoiduginhudts, poaafnidrcwuirvsveearttWiuicralel)copaxososinwteirloalnn)k.assWulahmrevtanelstpdsoiwsfeiftreireoennrcue(nsatnowiemdraele- sveatttiloendsrtaonadormalny,dotmhednistthriebuptrioopno.rtFioorneoxfamlpalrev,aeifseltatrlvianeg in pits should be equal to the proportion of space ac- tferrommintehewFhriicdhayspHeacirebsorweLraebosriagntiofriiceasntalynddifLfoepreenzt.IsDlaatnad counted forby pits. In thiswaywecalculated the number were pooled after Wilcoxon rank sum tests showed that oflarvae expected to settle in each of our classes of lo- there were no differences between the two sites. cations (on or in pits or bumps, touching pits or bumps, To model the kinds and size scales oftopographic fea- away from pits or bumps). Paired simultaneous /-tests tures found on natural substrata, such as the alga Lami- were used tocompare theobserved versusexpected num- pnlaartieass8a.c9cchamriinna,diwameetceorn:st(r1u)cstmeadllthLreeegoty(pLeesgoofSypsltasetmisc Tbehreosfiimnudlitvaindeuaolussi/n-teesatcshwleorceatwieoinghotneda sbeetctaluesmeentthepleasttei.- Imncm.)dbiuaimledtienrg)sbilmouclkatbeudmspmsall(cayllgianldrbiucmalp,s;2(2m)mlargheigLhe,go5 omfatdeisffeorfepntrospiozrest.ioTnhseofesltarivmaaetewepreofalal bparso-epdorotniosnamhpalsesa apbimutiesltde(irhn)egmsibislmpouhclekartibecdualml,aprs2ge(maclymglailnddberueipmca,pls,5;5mamnmdm(d3i)haibmguheb,tbe9lre)mpsmliasmtduii-c- gN20a8ui)ss.sitWahnee dswiaesmitgprhilbteuetdsiioozneurw(ieStsnhteidameacvtaoerrsiabanyncdethCpe(orc1ehcriappnr)o,/caN1l,9o6w7fh:ethripe.s lated small algal pits. These materials were used because variance because we have higher confidence in estimates tshcealieratnopdowgerraephuinciffeoartmulryesspwaecreed.ofWtehepraoppdruocperdiasteettslpeamteinatl wviidtuhaltshewelorweesntotvairnicalnucdee.d.PlaWtees uwsietdh ltehses tBohnafnertrwooniindiin-- plates by pouring polyester resin into silicone rubber equality to make the tests simultaneous (Miller, 1966). molds (Sylgard 184 Silicone Elastomer, Dow Corning For example, when we compared three settlement loca- Corp.). Black resin pigment (Titan Corp.) was added to tions, to maintain an overall error rate of0.05, we used the uncatalyzed resin to make newly settled larvae more an error rate of0.05/3 = 0.016 in each individual com- visible on the plates. parison. The settlement plates were attached to wooden boards To determine whether settlement preference changed with stainless steel screws. These were hung beneath the asspace became occupied, we examined the relationship floating docks with polypropylene rope. The plates were between the proportion of larvae settling in the feature oriented facedown topreventalgal colonization. Six rep- and the proportion ofunoccupied spaceaccounted forby licatesofeachsurfaceweresubmergedineachtrial. Plates thatfeature. Onall dateswecalculatedthespaceavailable werearranged in a Latin squaredesign, with onereplicate for settlement by subtracting from the total the area oc- ofeach type ofplate on each board. Six trials were run cupied by settled individuals. We assumed that all newly during the summers of 1987 and 1989. metamorphosed larvae occupied 1 mm2. We compared Photographs were taken every two days at the Friday settlement in samples with more than theaverage amount Harbor Laboratories and once or twice a week at Lopez offree space, to settlement in samples with less than the Island during 1987. Additional data werecollected by di- average amount offree space. INVERTEBRATE SETTLEMENT REFUGES 115 Stoloniferous hydrozoan (primarily Obelia dichotoma Table I and Obeliageniculata)andentoproct(Barentxiahcncdcni) Settlement location* o/Membranipora membranacea. Bugula neritina colonies were present on all ofthe plates within 10 days, andDistapliaoccidentalismi thealt;a Laminariasacchanna and at leasta fewstolons rapidlycovered theentiresurface ofmost plates. Todeterminewhether the stolonsaffected Species Mean Group settlement ofBugula, Distaplia, and Membranipora, the tops ofthe Lego bumps were divided into ten pie-shaped wedges. Similarly, the bases ofthe Lego bumps were di- vided into ten equal sections. Ifsettlement was random with respect to stolons, then the ratio of wedges where stolonsand larvaeco-occur,towedgeswith larvae, should equal the ratio of wedges with stolons to total wedges. Paired simultaneous /-tests were used to determine whether the observed and expected ratios were equal. Very few individuals of other species settled on our experimental plates. Approximately 75% ofthe plates of each type had no other species settling on them. The re- maining 25% had an average oftwo individuals ofother species on them. These other species included: the bryo- zoans Tegella armifera and Schizoporella itnicornis, the ascidian Diplosoma macdonaldi, the barnacle Balanus crenatus, the serpulid polychaete worm Pseudochitono- ponui occidentalis, and spirorbid polychaetes. Results Natural alga substrata Bugula nentina, Distaplia occidentalis, and Membran- ipora membranacea settled in locations with similar di- ametersand radii ofcurvature (Table I). Bugulaand Dis- taplia settled in significantly lower elevations relative to topographic features than did Membranipora (Table I: Vertical Position). Bugulasettled on topographic features that were significantly tallerthan those on which theother two species settled (Table I). Settlement plate experiments Onthe Legosettlement plates,settlement wasnon-ran- dom forall species (Table II). Distaplia and Bugula were found most often around the bases ofbumps (Table II). These locations covered less than 5% ofthe total surface area ofthe settlement plates, yet more than 50% ofthe larvae ofDistaplia and Bugula settled there. Both arborescent forms, Distaplia and Bugula, were found significantly lessoftenthanexpected on flatsurfaces ofthe largeand small Legosandthe flat surfacesofplates with small pits(Table II). Distaplia settled more than ex- pected by chance in the pits. In contrast, Bugula signifi- cantlyavoided pits(Table II). The sheet form, Membran- ipora, wasfound morethanexpectedon thetopsofbumps andontheflatsurfacesawayfrom thetopographicfeatures in the large Lego treatment, but lessthan expected around the bases of bumps (Table II). On the pitted surfaces. 116 L. J. WALTERS AND D. S. WETHEY Table II Testoifrandomness ofsettlement locations: theresults ofsimultaneous pairedl-lestscomparingthei:\peclcelversus theobservednumberotsettlers Species INVERTEBRATE SETTLEMENT REFUGES 17 where Bitgula and Distaplia settled, are functionally and hydrozoan stolons and the three species, with the equivalent to the bases ofLego bumps and the pits in the laterarrival alwaysgrowingoverthepreviouslyestablished artificial settlement surfaces(Table II). Similarly, the high colony. Neither colony appeared to be affected by these positionson algal slopeswhere Membranipora settledare interactions. Although space was not filled on our settle- functionally equivalent to the elevated locations where ment plates during the time course of this study, little they settled on the settlement plates (Table II). However, bare space existed on the docks from which the plates topography does not fully control settlement pattern, be- were suspended. Because so little free space existed on cause neitherpreviously settled individuals, northe stolon the persistent hard substrata, we believe that competition mats of hydrozoans and entoprocts, affected settlement could act as a selective agent on larval behavior. by the larvae (Table III), even though the presence ofany The resultsofthese studies areconsistent with ourpre- organismson thesubstratum alters the local microtopog- diction that adult body form should be correlated with raphy. larval settlement pattern. The arborescent forms(Bugnla An alternative mechanism that could account for the and Distaplia) settled preferentially in the small amount settlement patterns is passive transport of larvae by hy- ofspacetouching thebasesofthebumps, potentially hid- drodynamic forces. Because of their limited swimming den from predatorsand disturbance events. Thethin sheet abilities (Chia el ai, 1981). larvae are often passively form (Alembranipora) settled most frequently on the transported in boundary layerflows(e.g., Butman, 1987). highest available locations on topographically complex One can model passive larval transport as analogous to surfaces. Thus Membranipora, the adult growth form of sediment transport (e.g., Middleton and Southward, which is most susceptible to overgrowth, had larvae that 1984). The patterns of transport are influenced by the settled in potential refuges from competitors. Adult com- turbulent motion ofthe water and by the topography of petitive ability and susceptibility to predation and distur- the substratum. When the surface topography protrudes bance may be an important influence on selection for beyondthe'viscoussublayer' intotheturbulent overlying larval settlement behavior. water, turbulent eddiescan causeerosion. The roughness Reynolds number. Re*, is a measure of the degree to Acknowledgments which roughness elements protrude above the viscous This study was supported by the University of South sublayer: Carolina, grants from the Office ofNaval Research (Con- Re* = u*Lp/M tract N00014-82-K-0645) and the National Science where u* is the shear velocity ofthe fluid flow regime, L Foundation (Grant OCE86-00531) to D. Wethey and is the height ofthe roughness element, p is the density of grants from the Lerner-Gray Fund for Marine Research, seaIwnataewra,vae-nidnffj.luisentcheeddeynnvaimriocnmveinstc,osuit*yiosfapspearwoaxtiemra.tely sSoicgimaatioXni,toaLn.dWatlhteerIsn.teWrenaatrieongarlateWfoumleton'alslaFtitshheinFgriAdsa-y 10%ofthemaximum watervelocity(Denny, 1988; Denny Harbor Laboratories for providing us with space and fa- and Shibata, 1989; Svenden, 1987). We estimate u* tobe cilities. D. Padilla, D. Pencheff, L. Muehlstein, A. Kettle, in the range of 1.6 to 2.4 cm/s, yielding Re* values of S. Cohen, A. Sewell, and countless others assisted with 30-50forthesmall Legosand 75-120 forthelarge Legos. the field work. J. Sutherland and A. Underwood assisted If the roughness Reynolds number is less than 5, the with the statistical analyses. S. Woodin, J. Sutherland, R. bumps lie within the viscous sublayer. Thus, in all cases, Showman, and two anonymous reviewers made helpful the bumps on our settlement plates are in a potentially comments on the manuscript. erosional regime. Larvaedifferfrom sediment particles in their ability to adhere to surfaces. In tlume experiments Literature Cited with our settlement plates, sediment never accumulated Atkins, 1955. The cyphonautes larvaeofthe Plymouth Areaand the on the tops ofthe Lego bumps, presumably because the metamorphosisofMembraniporamembranacea. J. Mar. Biol.Axsiv. erosional forcesarevery high intheselocations. Therefore, L'K 34:441-449. ifthe pattern were passive, larvae would not have accu- Brancato, M.S.,and R.S.VVoollacott. 1982. Effectofmicrobial films mulated on the tops ofbumps. However, the tops ofthe oBnns/ertrtiulie)m.en.tMaor,fbBrmylo.zo7a1:n5la1r-v5a6e.(Bugidasimplex.B.sloloniferaand bumps are the locations where Membranipora larvae did Buss, L. W. 1980. Bryozoan overgrowth interactions the interde- accumulate. Thus, webelievethatthepassive model can- pendenceofcompetition forspaceand food. Nature281:475-477. not explain our patterns. Butman,C. A. 1987. Larvalsettlementofsoft-sedimentinvertebrates: Competitive interactions were infrequent on these set- thespatialscalesofpatternexplainedbyactivehabitatselectionand tlement plates, because recruitment rates were low and Athnen.emReervg.i2n5g:ro1l1e3-o1f6h5y.drodynamic processes. Oceanogr. Mar Biol space did not become limiting during our experiments. Chia, F-S.,J. Buckland-Nicks, andC. M. Young. 1984. Locomotion Theonlycommon encounterswerebetweentheentoproct ofmarineinvertebratelarvae:areview.Can.J Zoo/.62: 1205-1222. 118 L. J. WALTERS AND D. S. WETHEY Clone),R.A. 1978. Ascidian metamorphosis: reviewandanalysis. Pp. Reed,C.G. 1987. Phylum Bryozoa. Pp.494-510inReproductionand 255-282 in Settlement ami Metamorphosis <>l Mamie Invertebrate Development ol MarineInvertebrates ottheNorthern Pacific (.'oast. Larvae, F-S. Chiaand M. E. Rice.eds. Elsevier. New York. M. F. Strathmann. ed. University ofWashington Press, Seattle. Cloney, R. A., and S. A. Torrence. 1984. Ascidian larvae: structure Reed. C.G., and R. M. \\oollacott. 1982. Mechanismsofrapid mor- andsettlement.Pp. 103-126inMarineBiodeterioration,J.D.Costlow phogeneticmovementsinthemetamorphosisofthebryozoanBugiila and R. C. Tipper, eds. Naval Institute Press. Annapolis. neritina. J Morphol. 172: 335-348. Connell, J. H. 1961. The effects ofcompetition, predation by Than Ryland,J.S. 1959. Experimentsontheselectionofalgalsubstratesby lapillus. and other factors on natural populations ofthe barnacle, polyzoan larvae. J E.\p. Biol. 36: 613-631. Balanusbalanoides. Ecol. Monogr 31: 61-104. Russ, G. R. 1982. Overgrowth in a marine epifaunal community: Crisp, D.J. 1961. Territorial behaviorin barnaclesettlement.J Exp competitive hierarchies and competitive networks. Oecologia: 53: Biol 38: 569-590. 12-19. Crisp, D.J.,and H. Barnes. 1954. Theorientationanddistributionof Sebens, K. P. 1985. Community ecology ofvertical rock walls in the barnaclesat settlement with particularreference tosurfacecontour. GulfofMaine, U.S.A.: small scale processes and alternative stable J. Anim. Eco/. 23: 142-162. states. Pp. 346-371 in The Ecology ojRocky Coasts, P. G. Moore Crisp,D.J.,andJ.S.Ryland. I960. Influenceoffilmingandofsurface and R. Seed. eds. Hodderand Stroughton, London. Dayttoenxt,urPe.o1n97t1h.e seCtotmlpeemteinttioofn,madriisntuerboarngcaen,isamns.dXcaolmnmruen1i8t5y: o1r1g9a.ni- Sebeonrsg,anKi.smPs.in19t8h6e.NeSwpaEtinagllarnedlastuibotnisdhailpszoanme.onEcgole.ncMrounsotgirn.g5m6a:r7i3n-e zation:theprovisionandsubsequentutilizationofspaceintherocky 96. intertidalcommunity. Ecol. Monogr. 41: 351-389. Seed.R..andR.J.O'Connor. 1981. Communityorganizationinmarine Denny, M. \V. 1988. BiologyandMechanics ofthe Have-Swept En- algal epifaunas. Ann. Rev Ecol Sysl. 12: 49-74. vironment Princeton University Press, Princeton, NJ. Snedecor,G.VV.,and VV.G.Cochran. 1967. StatisticalMethods. Iowa Denny, M. VV., and M. F. Shibata. 1989. Consequencesofsurf-zone State University Press. Ames. Iowa. turbulence for settlement and external fertilization. Am. .\al. 134: Strathmann,R.R..K.S.Branscomb,andK.V'edder. 1981. Fatalerrors 859-889. insetasacostofdispersalandtheinfluenceof intertidal floraonset ofbarnacles. Oecologia48: 13-18. Grosmbaerrign,eRe.nKv.ir1o9n8m1e.ntsC.oNmpaettuirtei2v9e0:abi7l0i0ty-7i0n2f.luenceshabitatchoicein Svenden,I.A. 1987. Analysisofsurfzoneturbulence.J. Geophys.Res 92: 51 15-5124. Hayward, P. J., and P. H. Harvey. 1974. The distribution ofsettled Torrence, S. A.,and R. A.Cloney. 1988. Larval sensoryorgansofas- liadrivuamepofoltyheoubmry(oHazsosaanlsl).AlJcyMniadriuBmioh/.irAssustoucm (IF'leKmi5n4g:)6a6n5d-6A7l6c.yn- cniidqiuaenss.ApPppl.ica1b5l1e-1to63theinInMdairainnOeceBaino,deMte-rFi.orTahtioomnp,sAodnv.aRn.ceSadroTgeicnhi-. Jackson, J. B. C. 1979. Morphological strategies ofsessile animals. and R. Nagabhushanam,eds.Oxfordand IBH PublishingCompany keouLPgpa.hr,w4o9Mo9.d-5Ja5.n,5dainnBd.BRiB.o.lRJo.ogsyDeaonwn,ndeedSssy..stA1ec9ma8a2dt.eimcsiRcoefcPCrreouslisot,nmiNeaneltwOroYfgoarmnkai.rsimnse,iGn.- WaltcPevortms.p,eLttIi.d.t...i].v.NeaehnwiderDDae.rhcSlh.ii.Wesetohnemya.r1i9n8e6.hardSusrufbastcreattao:poagfriaelpdheyxipnefrliumeenncte.s vertebrates:theroleofactivelarvalchoicesandearly mortality.Oec- Biol. Bull. 170:441-449. ologia54: 348-352. VVethey,D.S.1983. Geographiclimitsandlocalzonation:thebarnacles Knight-Jones. E. VV. 1951. Gregariousnessand someotheraspectsof Semibalanus(BalanuslandChthamalusinNewEngland. Biol. Bull. thesettlingbehaviorofSpirorbis.J. Mar. Biol.Assoc. U. A".30:201- 165: 330-341. 222. VVethe>, D.S. 1984. Spatialpatterninbarnaclesettlement:daytoday LeTourneux, F., and E. Bourget. 1988. Importance of physical and changesduringthesettlement season./ Mar. Biol. Assoc ( A 64: biologicalsettlementcuesusedatdifferentspatialscalesbythelarvae 687-698. ofSemibalanusbalanoides. Mar. Biol. 97: 57-66. Wethey, D. S. 1986. Ranking ofsettlement cues by barnacle larvae: Loeb, M.J.,andG. Walker. 1977. Origin, composition and function influenceofsurfacecontour. Bull Mar. Sci. 39: 393-400. ofsecretions from pynform organsand internal sacsoffoursettling Wisely,B. 1960. Observationsonthesettlingbehavioroflarvaeofthe cheilo-ctenostome bryozoan larvae. Mar. Biol. 42: 37-46. tubeworm Spirorbis borealts Daudlin (Polychaete). Atixt. J. Mar. Middletnn,G.\'.,andJ. B.Southward. 1984. MechanicsofSediment Freshwat. Res. II: 55. Movement. 2ded. Society of Economic Paleontologists and Miner- Woollacott, R. M., and R. L. Zimmer. 1978. Metamorphosis ofcel- alogists. Tulsa, Okla. lularioid bryozoans. Pp. 49-64inSettlementandMetamorphosisof Miller, R.G. 1966. SimultaneousStatisticalInference. McGraw-Hill. MarineInvertebrateLarvae. F-S.Chiaand M. E. Rice,eds. Elsevier. Inc., NewYork. New York. Paine,R.T. 1974. Intertidalcommunitystructure:experimentalstudies Young,C. M.,and F.S.Chia. 1981. Laboratory evidencefordelayof ontherelationshipbetweenadominantcompetitoranditsprincipal larvalsettlementinresponsetoadominantcompetitor.Int.J Invert. predator. Oeailogia 15: 93-120. Reprod 3:221-226.