Flexibility in timing of molting of fiddler crab ... - Inter Research

MARINE ECOLOGY PROGRESS SERIESMar Ecol Prog SerPublished January 30Flexibility in timing of molting of fiddler crabmegalopae: evidence from in situmanipulation of cuesNancy J. onnor nor'^', Michael L. ~udge~'Department of Biology, University of Massachusetts Dartmouth, N. Dartmouth, Massachusetts 02747-2300, USA'Department of Biology, Manhattan College, Riverdale, New York 10471, USAABSTRACT: Rather than relying on chance delivery of larvae to experimental field substrata, we developeda field caging method to expose lab-reared crab megalopae to natural cues that could affect timingof molting to the first crab stage. Uca pugnax megalopae were placed in plastic mesh cages containingfreshly collected sediments from a marsh occupied by U. pugnax, or similar sediments that hadbeen combusted to remove organic material. Cages of each sediment type were placed at identical tidalelevations in a Spartina alterniflora marsh for 1 or 3 d of exposure to natural seawater. Significantlymore megalopae molted in cages containing fresh than combusted marsh sediments (90% vs 32%.ANOVA F,,,= 53.95, p < 0.001) after a 3 d penod of exposure. Three days after the 1 d exposure period.nearly all megalopae from cages of both sediment types had molted. Furthermore, all sibling megalopaeremaining In the lab during the field experiment fa~led to molt. High rates of molting of megalopaecaged in the marsh w~th natural and combusted sediments ind~cate that chemical substancesboth in the water overlying the marsh and associated with marsh sediments stimulated molting ofmegalopae.KEY WORDS: Fiddler crab . Uca pugnax. Larvae . Molting . Chemical cuesINTRODUCTIONResearch on settlement and metamorphosis of decapodpostlarvae has documented a behavioral responseof postlarvae to cues in the field, evidenced by preferentialsettlement in certain areas, and a physiologicalresponse to cues in the lab, evidenced by earlier moltingto the benthic stage in the presence of certain cues.During settlement in the field, decapod postlarvaeoften respond to particular habitat features. For example,spiny lobster postlarvae prefer to settle in clumpsof branched red algae (Herrnkind & Butler 1986), Dungenesscrab megalopae settle in higher numbers ontobivalve shell debris than onto sediments (Fernandez etal. 1993, Eggleston & Armstrong 1995), and newly settledblue crab megalopae are found in greater abun-dance within seagrass beds or intertidal marshes thanin unvegetated areas (Orth & van Montfrans 1987,Mense & Wenner 1989). Fiddler crab megalopae settlein salt marshes but not adjacent tidal flats, and arefound in greater abundance in areas inhabited byolder conspecifics (O'Connor 1993).Habitat characteristics may also affect the timing ofmetamorphosis. In the laboratory, fiddler crab megalopaemolt to the first crab stage sooner in the presenceof particular habitat components. Uca pugilator megalopaerespond (by molting earlier) to sediments collectednear adult burrows (Christy 1989) or to naturalsediments maintained with adult crabs (O'Connor1991). Uca pugnax megalopae molt in response tochemicals released by adults, but not to mock sediments(silt-sized glass beads) that were maintained inthe presence of adult crabs (O'Connor & Greggunpubl.). Yet we do not know how megalopae respondphysiologically, by molting, in response to natural cuesin the field.0 Inter-Research 1997Resale of full article not permitted

56 Mar Ecol Prog Ser 146: 55-60, 1997We have developed and tested a method to placeequal numbers of sibl~ng, laboratory-reared larvae intothe field, enclosed in mesh cages, for exposure to differenttypes of potential habitat cues. The methodfacilitates easy retrieval of cages and subsequent examinationof enclosed animals. Other advantages ofour method are that the larvae have had no prior exposureto natural cues in the field, they have been rearedunder identical conditions, they are of known ages andcompetent to settle and metamorphose, and we do nothave to rely on chance transport of larvae to experimentalsubstrata. In addition, the responses of larvae tocomplex cues in the field can be compared directlywith those of larvae exposed to single cues in the laboratory.The present study examined the effects of 2 potentialhabitat cues on timing of molting of Uca pugnax megalopaein the field. The first cue was the natural seawateroverlying the surface of a salt mdrsh inhabited by Lipugnax (representing waterborne cues). The secondcue was the sedirnents freshly collected from the marshsurface (= benthic cues). Although both cues ~ndlviduallyaffected the timing of molting, the greatest responseto these complex habitat cues occurred whenwaterborne and benthic cues were combined.MATERIALS AND METHODSLarval culture. Larvae used in the experiments werereleased by ovigerous female Uca pugnax collectedfrom the Slocums River estuary in S. Dartmouth, Massachusetts,USA (salinity approx. 30) in June 1995.Additional larvae were obtained from crabs that weremaintained in the laboratory and periodically checkedfor the presence of eggs. Ovigerous crabs were placedindividually in 11.5 cm glass culture dishes with-200 m1 filtered (0.45 pm) seawater diluted with deionizedwater to a salinity of 25 (= FSW), to which theantibiotics sodium penicillin G (21.9 mg I-') and streptomycinsulfate (36.5 mg 1-l) were added. Each morningthe dishes were checked for the presence of larvae.If larvae had not hatched, the ovigerous females weretransferred to a clean bowl with fresh FSW and antibiotics.The crabs were not fed.Newly released sibling zoeae were placed in groupsof -40 in each of 3 or 4 culture dishes (diam. 11.5 cm).Each dish contained -175 m1 of FSW and ant~blotics.The zoeae were offered the rotifer Brachionuspllcatilisand newly hatched brine shrimp nauplii (Arternla sp.)as food. As the zoeae molted to later stages, the proportionof rotifers in their diet was decreased, and thatof Artemia was increased. By the time they reachedthe fourth zoeal stage, the larvae were fed brineshrimp only. Each day the larvae were transferred to aclean dish with fresh FSW and antibiotics, and foodwas added. The number of deaths was noted, and theproportion of Individuals molting (evidenced by thepresence of exuviae) was approximated Mortal~ty wasgenerally very low (15%), and larvae tended to moltsynchronously. Both ovigerous females and larval cultureswere maintained in an environmental chamber(26°C k 1°C, 14 h light: l0 h dark photoperiod).After molting to the megalopa stage (usually after 11to 13 d of larval life), the larvae were cultured ingroups of 25 to 80 in 20 cm glass culture dishes containing600 rnl FSW, antibiotics, and Arternja nauplii.Megalopae were transferred and fed each day. Sixdays after they had molted to the megalopa stage,megalopae were transferred to the field site.Cage design. Cylindrical cages (10 cm height X8.8 cm width) were used for field deployment of labrearedlarvae. The base of the cage consisted of thebottom portion (8.8 cm width x 1.4 cm height] of a polystyrenepetri dish (Baxter D1906). The sides and top ofthe cages were made of 300 pm nylon mesh. The topcontained a flap that opened halfway to allow additionof megalopae in the field. All seams of the cage wereglued with 100% silicone sealant (G.E. #012). A 15 cmmild steel (high carbon) stake was glued to the outsideof the bottom of the petri dish to secure the cage inbenthic sediments in the field.Field site. The field site was a marsh consisting ofSpartina alterniflora and Salicornia europaea growingin muddy sand sediments (75Y0 particles >63 pm),located in Dernarest Lloyd State Park near the mouthof the Slocums River estuary in Dartmouth, ~Massachusetts,USA (41" 31' N, 70" 59' W). The marsh containednumerous burrows occupied by the fiddler crab Ucapugnax.Treatments and cage deployment. Cages contained15 m1 of sediments of 1 of 2 types: (1) surface (-0.5 cmdepth) marsh sediments freshly collected near fiddlercrab burrows (= fresh marsh sediments treatment), and(2) identical marsh sediments that had been dried andcombusted in a muffle furnace (550 to 600°C for 2 h) tooxidize organic material (= muffled marsh sedimentstreatment). Large detrital fragments and pebbles wereremoved from both sediment types. Both sedimenttypes were placed in cages in the field just before additionof larvae.Within the marsh, a transect parallel to the shore wasestablished at a tidal height of 0.9 m, the elevation ofthe marsh where crabs were abundant. Ten cages,each separated by 50 cm, were placed along the transect.Cages alternated between fresh and muffled sedimenttypes. All cages in the marsh were 23 m into thevegetated habitat.Experiments began 6 d after the fiddler crab larvaehad reached the rnegalopa stage. In the first experi-

O'Connor &Judge: Molting of fiddler crab megalopae 5 7ment, cages were deployed on 13 July and retr~evedon 14 July 1995 (20 h in the field with-10 h of submersion; = l-day deployment, Fig. 1).The second experiment took place from 14 July to17 July 1995 (70 h in the field with -28 h of submersion;= 3-day deployment).Six-day old megalopae pooled from 3 differentbroods were divided into groups of 10 and placedin 12 m1 of FSW (salinity = 25) in polypropylenecentl-ifuge tubes for transport to the fleld siteSome Artemia naupl~i were added to prov~de foodfor the megalopae. Megalopae were introducedinto each cage during late afternoon -3 h prior tosubmersion by the incoming tide. Megalopaewere rinsed from the centrifuge tube into the cageusing sieved, field-collected seawater. The openingat the top of the cage was then sealed w~thsilicone.Cages were monitored during daylight hours.One of the fresh sediment cages in the marsh wasdis~laced -0.5 m landward and reanchored on thesecond day of the 3-day deployment. B~~~~~~ airtemperatures ranged from 20°C at night to 32°Cduring the day, additional seawater was added tocages of the 3-day deployment after 48 h. Daytimewater temperature in the vicinity of the field site, however,never exceeded 26°C. Salinity of water at the sitewas -28 to 30.Laboratory observations. Sixteen siblings of themegalopae used in the 3-day deployment were maintainedsimultaneously in the laboratory In a 11.5 cmculture dish containing FSW (salinity = 25) with antibiotics.Each day the number of individuals that hadmolted to the first crab stage was noted. The remainingmeglaopae were transferred to a clean dish with freshFSW and Artemia.Subsequent to the field deployments, the effect onmoltlng time of muffled sediments as a potential tactilecue was examined in the laboratory. Four daysafter molting to the megalopa stage, larvae wereplaced individually into 60 X 15 mm sterile polystyrenepetri dishes containing 10 m1 FSW or 10 m1FSW plus 0.5 m1 muffled sediments from the samestock used in the field experiments. Twenty megalopaewere tested in each treatment. Several Artemianauplii were added as food for the megalopae. Eachday the number of individuals molting to the first crabstage was noted, and megalopae were transferred tonew dishes containing fresh treatments every otherday.Data analysis. The proportion molting to the firstcrab stage within each replicate cage was the responsevariable. After arcsine transformation, the varianceswere found not to differ significantly from homogeneity(Bartlett's test = 0.47, p = 0.52). Data from each fieldPredicted Tidal Series, July 1995l -day3-day1.8 deployment deploymentl3 Jbly 14 July 15 July 16 July 17 JulyDateFig. 1. Tidal series at the field site dunng the l-day (vert~cal dashedlines) and the 3-day (vertical solid lines) deployment study periods.Tidal heights were calculated from the predicted tidal levels forNewport. Rhode Island, USA (NOAA 1995) and adjusted frommeasurements conducted at the site. The horizontal dotted lineindicates the tidal height of the experimental cages. Megalopaewere submerged by the tide 47 5% of the time during the l-dayand 39.7 % during the 3-day deploymentdeployment were analyzed via l-way ANOVA withsediment as the main effect.The effect of the presence of muffled sediments ontiming of molting in the lab was analyzed by comparingthe median day of molting of megalopae in eachtreatment using the non-parametric Mann-Whitney U-test, due to the non-normal distribution of the data.RESULTSVery few megalopae molted to the first crab stageduring the l-day deployment in the marsh habitat: anaverage of 2% molted in cages with the muffled sedimentand 0% in the fresh sediment treatments (Fig. 2).However, after 3 d in the marsh, significantly moremegalopae molted in cages containing the fresh sedimenttreatment (Fig. 2; mean proportion + SD = 0.90 +0.07) than in the muffled sediment treatment (0.32 +0.11) (ANOVA, = 53.95, p < 0.001). Survivorship ofmegalopae in cages in the marsh habitat was veryhigh: 90 to 100% in each cage survived the l-day and3-day deployment periods.Megalopae from the fresh and muffled sedimentcages placed in the marsh for the l-day deploymentwere removed from the cages at the end of the deployment,and held in a plastic container containing FSWin the laboratory. One day after removal from themarsh, only -10% had molted to first crab. Three daysafter field exposure, however, 80 to 90% of the mega-

Mar Ecol Prog Ser 146: 55-60, 1997E4l-day3-dayIn the laboratory, the presence of muffled marsh sediment~had no effect on timing of molting (Mann-WhitneyU-test, p = 0.43). The median day of molting ofmegalopae in dishes with FSW was 12 0 (mean = 13.8,SD = 6.9), and in dishes with both FSW and muffledsediment it was 12.5 d (mean = 14.6, SD = 10.1). Thissuggests that the muffled sediments did not provide atactile cue for molting in the field.MuffledSediment TypeFreshFig. 2. Uca pugnax. Proportion (mean i SD) of fiddler crabmegalopae molting to first crab stage within marsh (Spartinaalterniflora) habitats during 2 field deployments. Althoughproportion molting during the l-day deployment did not differbetween sediment treatments, the proportion moltingwithin 5 replicate cages during the 3-day deployment wassignificantly greater (ANOVA on arcsine-transformed data,F,,, = 53.95, p < 0.001) for sediment freshly collected fromnearby fiddler crab burrows than similar sediment that hadbeen previously combusted in a muffle furnace (2 to 3 h at550 to 600°C). Survivorship of ~ndividuals was greater than90% during both deploymentsDays Since Experiment BeganFig. 3. Uca pugnax. Cumulative molting frequency of fiddlercrab megalopae maintained in the laboratory without fieldcues. Sixteen 6-day-old megalopae from the same broods utlllzedIn the field expenment were scored daily for numbermoltlng to crab 1, commencing with ~nitiation of the 3-daydeployment. The correspond~ng mean molting percentage Inthe field during the 3-day deployment is indicated by 'F'(fresh sediment) and 'M' (muffled sediment)lopae molted to the first crab stage, which is comparableto the percentage molting during the 3-day periodof exposure to natural cues in the field.Megalopae maintained in the laboratory molted farmore slo~vly than their siblings in the field. At the endof the 3-day field deployment, none of the specimensin the lab had molted to the first crab stage (Fig. 3).Even 1 wk after the field experiment ended, only 19%of the megalopae in the lab had molted.DISCUSSIONOur short-term field experiment provides strong evidencethat natural cues greatly affect metamorphosis(molting) of crab megalopae, by inducing moltingbetween 1 and 3 d after initial exposure. Both naturalseawater overlying a marsh and natural marsh sedimentsdramatically affected timing of molting, comparedto molting of sibling megalopae maintained inthe laboratory. In fact, the magnitude of the response,as indicated by proportion rnolting, was greater thanthat seen in laboratory experiments (O'Connor 1991,O'Connor & Gregg unpubl.). However, Christy (1989)also observed that sediments collected adjacent toadult burrows stimulated substantially earlier moltingof Uca pugilator megalopae in the lab. The dramaticresponse of larvae to cues in the field compared withmolting of siblings in the lab (Fig. 3) strongly suggeststhat larvae in the lab were in fact delaylng metamorphosis(Pechenik 1990), rather than simply developingat a slower rate. Even though daytime air temperaturesin the field became rather high, average water temperaturewas similar to that in which megalopae in thelaboratory were maintained. Therefore, temperaturedifferences between the laboratory and the field werenot great, and probably did not accelerate molting ofrnegalopae in the field.The question arises as to the identity of the cues innature. Two types of cues were important in stimulatingmolting: cues in the water and cues associated withbenthic sediments. Water-soluble cues could havebeen derived from benthic sediments, marsh grass, oradult fiddler crabs inhabiting the marsh. Marsh grass isan unlikely cue, based on testing of leachates fromSpartina in the laboratory (O'Connor & Greggunpubl.). It is likely that soluble cues from either conspecificadults or natural marsh sedirnents, or both,were effective in stimulating molting. In laboratorystudies, cues derived from adult crabs significantlyaffect timing of molting of fiddler crab larvae (Christy1989, O'Connor 1991, O'Connor & Gregg unpubl.). Inaddition, substances associated with natural marshsedirnents, rather than sediment particles themselves,may represent a contact-dependent cue for megalopae.Sediments without natural organisms and mole-

O'Connor & Judge- Molting of f~ddler crab megalopae 59cules (i.e. muffled marsh sediments in FSW) were noteffective at stimulating molting in the laboratory.Blue crab megalopae also molt sooner when exposedto potential habitat cues in the laboratory. Megalopalmolting is advanced by exposure to estuarine seawatercompared with offshore seawater (Wolcott & De Vries1994), and soluble materials leached from seagrassesare particularly effective at stimulating molting (Forwardet a1 1994, 1996). The present study providesadditional evidence that crab megalopae molt inresponse to chemical cues derived from adult habitats.Benthic invertebrate larvae are known to utilize avariety of environmental cues to locate and metamorphosein suitable adult habitat. Hydrodynamic processesinfluence larval delivery and behavior at a varietyof spatial scales (Eckman et al. 1983, Butman 1987).Surface characteristics may also serve as importanttactile cues for larvae (Rittschof et al. 1984, Wethey1986). There is ample evidence that larvae activelyexplore such surfaces after contact (Mullineaux & Butman1991, Walters 1992). In addition to their welldocumentedrole in crustacean b~ology, chemlcal cueshave been shown important in many benthlc groups,including polychaetes (Pawlik et al. 1991) and bivalves(Turner et al. 1994), even under simulated natural flowregimes. While we are not able to rank the relativeimportance of these cues in the present study, thestrong differential response under field cond~tions toour manipulations suggests that chemical cues are criticalto the settling larvae of benthic invertebrates.The present study represents a new approach to thestudy of cues stimulating metamorphosis of invertebratelarvae. The caging technique allows control oflarval age at exposure to cues, and manipulation of thesource of the cues to which they are exposed. Our technlquealso ensures that equal numbers of larvae areexposed to different potential habitat cues. Other fieldexperiments, primarily dealing with larval choice of settlementsite, rely on chance transport of planktonic larvaeto experimental sites, and must assume that larvaehave full access to different experimental substrata.Even in studies in which larvae are pelaglc for short periodsand can be tracked directly by human observers,e.g. relatively large ascidian larvae (Olson 1985, Stoner1990), researchers are unable to control larval abundanceor delivery to f~eld substrata The success of thepresent caging study should open new avenues of fieldexper~mentation on larval metamorphosis.Acknowledgenlents We are very grateful to our studentsAmanda Gregg, Kerry Cudmore McCarthy David Cnno, andDanlelle Cigliano for theii assistance in the laboratory and thefield We thank Richard B Foiward Jr, Jan Pechenik, and ananonymous reviewer for their comments on the manuscriptWe also thank the staff of Demarest Lloyd State Park forallow~ng us to set up our odd-looking expeliment in the parkLlTERATURE CITEDButman CA (1987) Larval settlement of soft-sediment invertebrates,the spat~al scales of pattern explained by actlvehabitat selection and the emerglng role of hydrodynamicalprocesses. Oceanogr Mar Biol A Rev 25:113-165Chr~sty JH (1989) Rapld development of megalopae of the fiddlercrab Uca pugilatorreared over sediment: implicationsfor models of larval recruitment. Mar Ecol Prog Ser 57.259-265Eckman JE (1983) Hydrodynamic processes affecting benthicrecruitment. Limnol Oceanogr 28:241-257Eggleston DB. Armstrong DA (1995) Pre- and post-settlementdc,tei-minants of estuarine Dungeness crab recruitmentEcol Monogr 65:193-216Fernandez M, Inbarne 0, Armstrong D (1993) Habitat selectionby young-of-the-year Dungeness crab Cancer maglsterand predation risk In intertidal habitats Mar Ecol ProgSer 92:171-177Forward RB Jr, De Vries MC, Rittschof D, Frankel DAZ,Bischoff JP, Fisher CM, Welch JM (1996) Effects of environmentalcues on metamorphosis of the blue crab Calllnectessapidus Mar Ecol Prog Ser 131:165-177Forward RB Jr, Frankel DAZ, Rittschof D (1994) Molting ofmegalopae from the blue crab Callinectes sapldus: effectsof offshore and estuarine cues Mar Ecol Prog Ser 113:55-59Heri-nkind WF, Butler MJ IV (1986) Factors regulating postlarvalsettlement and juvenile microhab~tat use by spinylobsters Panullrus argus. Mar Ecol Prog Ser 34:23-30hlense DJ, Wenner EL (1989) Distribution and abundance ofearly life h~story stages of the blue crab, Calllnectessapidus, in tidal marsh creeks near Charleston, South Carol~na.Estuanes 12 157-168Mull~neaux LS, Butman CA (1991) lnitial contact, explorationand attachment of barnacle (Balanus amphltrite) cypndssettling in flow. Mar Biol 110.93-103NOAA (1995) T~de tables, high and low water predictions,east coast of North and South Amerlca, includ~ng Greenland.US Dept of Commerce. Washington, DCO'Connor NJ (1991) Flexib~lity in timlng of the metamorphicmolt by fiddler crab megalopae Uca pugilator. Mar EcolProg Ser 68:243-247O'Connor NJ (1993) Settlement and recruitment of the f~ddlercrabs Uca pugnax and U pug~lator In a North Carolina,USA, salt marsh Mar Ecol Prog Ser 93-227-234Olson RR (1985) The consequences of short-distance larvaldispersal in a sesslle marine invertebrate Ecology 66-30-39Orth RJ, vdn Montfrans J (1987) Ut~lization of a seagrassmeadow and t~dal marsh creek by blue crabs Callinectessapidus. I. Seasonal and annual vanations in abundancewith emphas~s on post-settlementjuveniles. Mar Ecol ProgSer 41 283-294Pabvhk JR. Butman CA, Starczak VAR (1991) Hydrodynamicfacilitation of gregarious settlement of a reef-budding tubeworm Science 251 421-424Pechenik JA (1990) Delayed metamorphosis by larvae of benthicmarine invertebrates: Does it occur? Is there a price topay? Ophelia 32.63-94R~ttschof D, Branscomb ES, Costlo\v JD (1984) Settlement andbehavior in relation to flow and surface in larval barnacles,Balanus arnphltrite Darwin J Exp Mar Biol Ecol 82.131-146Stoner DS (1990) Recruitment of a troplcal colonial ascidian:relatlve ~mportance of pre-settlement vs post-settlementprocesses. Ecology 71.1682-1690

Mar Ecol Prog Ser 146: 55-60, 1997Turner EJ, Zimmer-Faust RK, Palmer IMA, Luckenbach M,Pentcheff ND (1994) Settlement of oyster (Crassostrea virginica)larvae: effects of water flow and a water-solublechemical cue. Limnol Oceanogr 39:1579-1593Walters U (1992) Field settlement locations on subtidalmarine hard substrata. Is active larval explorationinvolved? Limnol Oceanogr 37:1101-1107This article was submitted to the editorWethey DS (1986) Ranking of settlement cues by barnaclelarvae: influence of surface contour Bull Mar Sci 39:393-400Wolcott DL, De Vries MC (1994) Offshore rnegalopae of Callinectessapidus: depth of collection, molt stage andresponse to estuarine cues. Mar Ecol Prog Ser 109:157-163Manuscript first received: July 17, 1996Revised version accepted: November 15, 1996