Bio-foam enhances larval retention in afree-spawning marine tunicateJuan Carlos Castilla*†, Patricio H. Manrıquez*‡, Alejandro P. Delgado*, Ligia Gargallo§, Angel Leiva§, and Deodato Radic§

*Center for Advanced Studies in Ecology and Biodiversity, Facultad de Ciencias Biologicas, and §Departamento de Fısico-Quımica, Facultad de Quımica,Pontificia Universidad Catolica de Chile, Casilla 114-D, Santiago, Chile; and ‡Instituto de Biologıa Marina ‘‘Dr. Jurgen Winter,’’ Laboratorio deRecursos Acuaticos de Calfuco, Universidad Austral de Chile, Casilla 567, Valdivia, Chile

Contributed by Juan Carlos Castilla, September 13, 2007 (sent for review July 27, 2007)

Here we report a mechanism that reduces dispersal of earlydeveloping stages and larvae in a free-spawning intertidal andshallow subtidal tunicate, Pyura praeputialis (Heller 1878), in theBay of Antofagasta, Chile. The spawning of gametes by thetunicate into the naturally turbulent aerated seawater decreasestheir surface tension and induces the formation of a bio-foam.Water collected from foamy intertidal pools and tide channelsshowed a high concentration of P. praeputialis early developingstages and tadpole larvae in the foam. Because gametes aresynchronically spawned for external fertilization and larvae settlenear adults, our results suggest that this bio-foam increases fer-tilization success and effective settlement of their short-livedlarvae in the vicinity of the adults spawning the gametes. Thismechanism reinforces published evidence suggesting that localretention of intertidal and inshore marine invertebrate larvae maybe more common than previously thought, offering, for instance,new perspectives for the design and networking of marine pro-tected and management areas.

Chile � surface tension � tide channels � broadcast spawner � settlement

I t is well known that surfactants can induce the formation ofbio-foams that can ensure egg retention and larval develop-

ment in organisms such as tropical leptodactylid frogs (1–5) andAfrican and Asian freshwater labyrinth fishes (6). However, suchegg-larval retentive mechanisms or ‘‘foam-nests’’ have not beenreported for marine organisms. In Chile, massive populations ofthe intertidal and shallow subtidal tunicate Pyura praeputialisform beds almost exclusively along a 60–70 km stretch of therocky shoreline inside the Bay of Antofagasta (7). This tunicateis an alien species in Chile, and its absence at locations more than3–5 km outside of the Bay of Antofagasta, and along the Chileancoast, suggests the existence of unknown mechanisms preventingits dispersal (8) and maintaining high settlement inside theAntofagasta tunicate beds. In December 2004 and January 2005,midway through the ebb tides, we observed massive co-spawningof both eggs and sperm suspension from P. praeputialis intonaturally turbulent aerated intertidal seawater (Fig. 1A) and thesubsequent formation of localized and conspicuous masses offoams. These bio-foam masses persisted for several hours on topof rocky shores and floating in tidal channels, near where theywere spawned, because of onshore winds (Fig. 1).

Results and DiscussionTo investigate the role of the bio-foam associated with P.praeputialis on the retention of embryos and tadpole larvae ofthe species, seawater samples were collected (December 10 and12, 2004; January 14 and 15, 2005) by using a plankton bilge-pump equipped with a 50-�m elbow end and calibrated with adomestic f low meter (9). The pump was used to collect watersamples at different time intervals through the tide cycle fromthe sea surface just above or near the intertidal rocky shores inBay of Antofagasta that are dominated by P. praeputialis. Eachsampling lasted for 10 min, filtering an average of 140 liters (�25SE, n � 15). The pump, operated with a 12-V battery, was

attached to a PVC float system that restricted sampling to theupper 10 cm of the water column. Water samples were fixed withformaldehyde and stored in labeled plastic bottles for furtherobservations in the laboratory. In December, samples werecollected over two alternate days and within each day in tidechannels with foam. In January, samples were collected over twoconsecutive days and within each day, alternating between tidalchannels with and without foam. Only samples from foamywaters showed the presence of embryos and tadpoles larvae ofP. praeputialis (Fig. 2). In tide channels with foam, developingstages of P. praeputialis were found on the flood tide (Fig. 2). Afew hours later, on the ebb tide, we detected the presence oftadpole larvae of P. praeputialis in the same channels (Fig. 2 A).During the ebb tides, when most intertidal P. praeputialis are

Author contributions: J.C.C. and P.H.M. contributed equally to this work; J.C.C. and P.H.M.designed research; J.C.C., P.H.M., and A.P.D. performed research; L.G., A.L., and D.R.contributed new reagents/analytic tools; J.C.C., P.H.M., A.P.D., L.G., A.L., and D.R. analyzeddata; and J.C.C., P.H.M., L.G., and A.L. wrote the paper.

The authors declare no conflict of interest.

†To whom correspondence should be addressed. E-mail: jcastill@genes.bio.puc.cl.

© 2007 by The National Academy of Sciences of the USA

Fig. 1. Photographic record of P. praeputialis spawning and field sampling.(A) Synchronous release of gametes from intertidal P. praeputialis duringexposure to air. (B) Sperm and egg suspensions retained on the surface ofintertidal specimens. (C) Thick foam (�2 m height) covering the surface of tidechannels. (D) Sampling of sea water from the surface in tide channels. ES, eggsuspensions; SS, sperm suspensions.

18120–18122 � PNAS � November 13, 2007 � vol. 104 � no. 46 www.pnas.org�cgi�doi�10.1073�pnas.0708233104

Dow

nloa

ded

by g

uest

on

Mar

ch 2

7, 2

020

exposed to air, we observed synchronous spawning of both spermand eggs suspensions from hundreds of tunicates (Fig. 1 A andB). This was followed by the formation of thick foam in the watersurface (Fig. 1 C and D).

To assess the potential role of retention associated with thefoam, groups of 50 small plastic ping pong balls filled withseawater were placed in tide channels with and without foammidway through a sequential f lood and ebb tide on December 17,2004. After 2.5 h, all balls placed in the channels without foamwere carried offshore by the currents. In contrast, in the sameperiod 50% (flood) and 60% (ebb) of the balls remained in thefoamy channels.

Foam nests enhance fertilization success and egg and larvalretention in freshwater nest-building fishes (10) and in bothaquatic and terrestrial frogs (5). In frogs, the production ofbio-foams is associated with the production of “ranaspumins”(the surfactant properties of aqueous protein mixtures from thenest of the tropical frog Physalaemus pustulosus) by females (5),and those foams show active surfactant properties at about thesame levels as we report for the tunicate bio-foam. One of theearliest records of coastal bio-foam associated their formationand stabilization with protein debris and alginates derived fromdamaged organism and algae (11). Thereafter, coastal bio-foamshave been mainly associated with exceptionally high cell con-centrations of the foam-forming colonial alga Phaeocystis spp.

and kelp mucilage under specific wind and wave conditions(12–14). However, formation of coastal bio-foam has also beenassociated with detritus or fecal matter in wavy environments(15). Regardless of their source, coastal bio-foams (i.e., shore-cast foam or foam lines) have been described as enclosing marinemetazoan fauna of several taxa and larvae (i.e., polychaete,mussel, and crustacean) (16, 17), and as a food resource forconsumers (11, 16).

Hence, we conclude that the highly active surfactant bio-foamproduced by P. praeputialis gametes, when spawned into turbu-lent aerated seawater, and the high number of tunicate devel-oping embryos and larvae linked to foamy tide channels repre-sent a new type of the global phenomenon of coastal bio-foamsand a mechanism enhancing external fertilization and localretention of larvae in rocky shore environments. We suggest thatin Antofagasta, the described bio-foam also enhances localtunicate settlement, because tadpole larvae of P. praeputialis arein the plankton for �2 h (18), less than the persistence of thebio-foam associated with P. praeputialis. Furthermore, this maybe one of the mechanisms responsible for the as yet unexplainedrestricted distribution of this free-spawning species inside theAntofagasta Bay, as well as for its high settlement success onadult conspecifics inside the bay (19). Additionally, preliminaryfield observations in the intertidal rocky shores of Las Crucesand Pelancura (central Chile) have shown that a similar bio-foamproduction also occurs concurrently with massive gamete spawn-ing of the sunstar Heliaster helianthus and for the chiton Acan-thopleura echinata. This suggests that an active surfactant bio-foam retention mechanism, as described here, may be a morecommon phenomenon among intertidal or near-shore benthicfree-spawning marine invertebrates, enhancing fertilization suc-cess and local rocky-shore retention of embryos and larvae withshort planktonic periods. In the same vein, recent informationindicates that various mechanisms [i.e., upwelling shadows, odorcues, and tidal fronts (7, 20–23)] linked to the retention ofcoastal marine fish and invertebrate larvae may be much morecommon than previously thought. This suggests, for example,that perhaps small no-take zones, or marine protected or man-agement areas, may be indeed effective, and more desirable, forenhancement and recruitment of overfished or critical species;particularly if the areas are numerous and closed enough to allowa stepping-stone like process to operate between them (24, 25),and if the target species present short larval periods. Although

Fig. 3. Comparison of surfactant properties (tension) as a function of theprotein concentration for BSA (open circles) and male gonad tissue (opensquares) and female gonad tissue (filled triangles) of P. praeputialis. Error barsshow 2 SD (n � 3).

Fig. 2. Abundance of developing stages and tadpole larvae of P. praeputialisper 10 liters of water collected from tide channels at different times of a tidecycle. (A) Symbols give the average number and standard error of developingembryos (circles) and tadpole larvae (squares) found in water samples col-lected during the day from three tide channels with (filled symbols) andwithout (open symbols) foam over two alternate days. (B) Symbols depict theaverage number and standard error of developing embryos (squares) andtadpole larvae (circles) found at diurnal hours in three water samples collectedfrom tide channels with (filled symbols) and without (open symbols) foamalong two consecutive days. Dashed lines indicate tidal cycles. Asterisks indi-cate direct observations of gamete spawns in the field.

Castilla et al. PNAS � November 13, 2007 � vol. 104 � no. 46 � 18121

ECO

LOG

Y

Dow

nloa

ded

by g

uest

on

Mar

ch 2

7, 2

020

the potential benefits of bio-foams in terms of increasing fertil-ization success could be quite general in marine free spawnerinvertebrates, their effects on the retention of larvae showinglong planktonic stages need to be further investigated.

Material and MethodsIn the laboratory, we tested the surfactant activity of P. praepu-tialis gonadal tissues. Male and female gonadic tissues weredissected from 15 adults under a stereomicroscope, storedseparately, and frozen for chemical analyses. Tissues were de-frosted, and samples were collected with a pipette to determinetotal protein concentrations by the Bradford method, monitor-ing the adsorption at 595 nm of the Coomassie brilliant blue-G250 bind to the proteins (26). For female gonads, 28.7 mg�ml�1

concentration values were obtained; and for male gonads, 18.8mg�ml�1 concentration values were obtained. The interfacialtensions were measured at 298 K by the Wilhelmy plate method,using a Data Physics DCAT11 interfacial tensiometer. Waterpurified by using the Milli-Q system (Millipore) up a to resistivityof 18.22 M��cm and with a surface tension value of �72.0mN�m�1 was used for dilution. The protein concentration for the

analysis ranged between 0.10 and 410 �g�ml�1. BSA (in phos-phate buffer, pH 7) and male and female P. praeputialis gonadsupernatant were used. Female and male gonadic extractsshowed a strong reduction of water surface tension. At totalprotein concentrations of 18–24 �g�ml�1, the surface tensiondrops sharply from pure water values, �72 mN�m�1 to 66mN�m�1 with BSA and to �57 mN�m�1 with male and femalegonad tissues (Fig. 3). Around this protein concentration level,BSA reduces surface tension by �10%, whereas male and femalegonads reduce surface tension by �21%.

We thank M. E. Jara for valuable assistance during the field samplings,N. Lagos for the use of his bilge pump, and P. Petraitis and J. D. D.Bishop for suggestions on the manuscript. This work was supported byFondo de Financiamiento de Centros de Excelencia en Investigacion–Fondo Nacional de Desarrollo Cientifico y Tecnologico Grant 1501-0001(Project No. 6; J.C.C.) to the Center for Advanced Studies in Ecologyand Biodiversity (CASEB), and by Fondo Nacional de DesarrolloCientifico y Tecnologico Grants 3020035 and 1050841 (to P.H.M.).Additional funding was available from an A. Mellon Foundation grant(to J.C.C. and S. A. Navarrete).

1. Downie JH (1988) Herpetol J 1:302–307.2. Downie JH (1990) Herpetol J 1:501–503.3. Downie JH (1993) Herpetol J 3:35–42.4. Byrne PG, Simmons LW, Roberts JD (2003) Proc R Soc London Ser B

270:2079–2086.5. Cooper A, Kennedy MW, Fleming R, Wilson EH, Videler H, Wokosin DL, Su

T, Green RJ, Lu JR (2005) Biophys J 88:2114–2125.6. Cole B, Tamaru CS, Bailey R, Brown C (1999) A Manual for Commercial

Production of the Gourami, Trichogaster trichopterus, Center for Tropical andSubtropical Aquaculture Publication No 135.

7. Castilla JC, Lagos N, Guinez R, Largier J (2002) in The Oceanography andEcology of the Nearshore and Bays in Chile, eds Castilla JC, Largier JL(Ediciones Universidad Catolica de Chile, Santiago), pp 179–203.

8. Castilla JC, Guinez R, Caro AU, Ortiz V (2004) Proc Natl Acad Sci USA23:8517–8524.

9. Lagos N (2003) PhD Thesis (Pontificia Universidad Catolica de Chile,Santiago).

10. Andrade DV, Abe AS (1997) J Fish Biol 50:665–667.11. Pope EC (1952) Aust Mus Mag 10:265.

12. Southward AJ (1953) Nature 172:1059–1060.13. Velimirov B (1980) Mar Biol 58:311–318.14. Velimirov B (1986) Actas II Congr Algas Mar Chilenas 115–123.15. Lancelot C, Mathot S (1987) Mar Ecol Prog Ser 37:239–248.16. Armonies W (1989) Mar Ecol Prog Ser 18:305–309.17. Shanks AL, McCulloch A, Miller J (2003) J Plankton Res 25:1251–1277.18. Clarke M, Ortiz V, Castilla JC (1999) Bull Mar Sci 65:745–754.19. Alvarado JL, Pinto R, Marquet P, Pacheco C, Guinez R, Castilla JC (2001) Mar

Ecol Prog Ser 224:93–101.20. Swearer SE, Casselle JE, Lea DW, Warner RR (1999) Nature 402:799–802.21. Jones GP, Milicich MJ, Emslie MJ, Lunow C (1999) Nature 402:802–804.22. Zacherl DC, Manrıquez PH, Paradis G, Day RW, Castilla JC, Warner RR, Lea

DW, Gaines SD (2003) Mar Ecol Prog Ser 248:297–303.23. Gerlach G, Atema J, Kingsford MJ, Black KP, Miller-Sims V (2007) Proc Natl

Acad Sci USA 104:858–863.24. Hastings AL, Botsford LW (2006) Proc Natl Acad Sci USA 103:6067–6072.25. Palumbi SR (2004) Annu Rev Environ Resour 29:31–68.26. Bradford MM (1976) Anal Biochem 72:248–254.

18122 � www.pnas.org�cgi�doi�10.1073�pnas.0708233104 Castilla et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 2

7, 2

020