Invertebrate Biology 121(3): 190-201. 0 2002 American Microscopical Society, Inc.

Aspects of the larval biology of the sea anemones Anthopleura elegantissima and A. artemisia

Virginia M. Weis,a E. Alan Verde, Alena Pribyl,” and Jodi A. Schwarz

Department of Zoology, Oregon State University, Corvallis, OR 9733 1, USA

Abstract. We investigated several aspects of the larval biology of the anemone Anthopleura elegantissima, which harbors algal symbionts from two different taxa, and the non-symbiotic A. artemisiu. From a 7-year study, we report variable spawning and fertilization success of A. elegantissima in the laboratory. We examined the dynamics of symbiosis onset in larvae of A. elegantissima. Zoochlorellae, freshly isolated from an adult host, were taken up and retained during the larval feeding process, as has been described previously for zooxanthellae. In ad- dition, larvae infected with zooxanthellae remained more highly infected in high-light condi- tions, compared to larvae with zoochlorellae, which remained more highly infected in low-light conditions. These results parallel the differential distribution of the algal types observed in adult anemones in the field and their differential tolerances to light and temperature. We report on numerous failed attempts to induce settlement and metamorphosis of larvae of A. elegun- tissima, using a variety of substrates and chemical inducers. We also describe a novel change in morphology of some older planulae, in which large bulges, resembling tentacles, develop around the mouth. Finally, we provide the first description of planulae of A. artemisia and report on attempts to infect this non-symbiotic species with xooxanthellae and zoochlorellae.

Additional key words: Cnidaria, Anthozoa, Actiniaria, larval settlement, melainorphic induccr, symbiosis, zoochlorellae, zooxanthellae

The 4 species of the sea anemone genus Anthopleu- the late summer. Siebert (1 974) and Smith (I 986) ex- ra found along the western coast of North America- amined laboratory-induced spawning, embryological A. ekgantissima (BRANDT 18%), A. soh (PEARSE& development, and morphology and behavior of plank- FRANCIS2000), A. xanthogrammica (BIIANDTI835), totrophic planula larvae of A. elegantissima and A. and A. avtemisia (DANA1846)-are conspicuous mem- xanthogrammica. Finally, Sebens ( 198 1 a, 1982) de- bers of rocky intertidal communities. A. elegantissima scribed recruitment of larvae from both species into in particular, which is common along the coast from mussel beds adjacent to sites with adult anemones. No Alaska to Baja California, can form clonal aggrega- life history studies have focused on either A. sola or tions that occupy large areas within the mid-intertidal A. artemisia. zone, to the exclusion of other competitors (Dayton A. elegantissima and A. xunthogmmmica form sym- 1971; Sebens 1981a; Fitt et al. 1982). Despite their bioses with two different types of unicellular algae importance in these communities, examination of the (Muscatine 197 1 ; Muller-Parker & Davy 200 1 ), di- sexual reproductive cycle and life histories of these noflagellates from the genus Symhiodinium (zooxan- anemones has been restricted to a handful of studies. thellae, LaJeunesse & Trench 2000) and one or more Ford (1 964), Jennison (1 979), and Sebens (1 98 1 b) de- taxonomically undescribed chlorophytes (zoochlorel- scribed gametogenesis and an annual reproductive cy- he). As in other cnidarian-algal symbioses, the algae cle in A. elegantissima and A. xanthogrammica that provide the anemones with photosynthetically fixed or- started with gonad development (both species are di- ganic carbon and the host anemone in turn supplies oecious) in the late fall and ended with spawning in the algae with inorganic nutrients, a stable environ- ment, and a refuge from herbivory (reviewed in Mus- ’’ Author for correspondence. Ph: 541-737-4359. Fax: 541- catine 1990; Muscatine & Weis 1992; Muller-Parker 737-050 1. E-mail: [email protected] & D’Elia 1997). Anemones of both species containing ‘’ Present address: Center for Liinnology, University of Col- these different algal types are differentially distributed orado, Boulder, CO 80309, USA along a latitudinal and tidal-height gradient, with Larval biology of Anthopleura eleguntissimu 191 anemones in more northerly and lower intertidal lo- 1988) and the scleractinian coral Fungia scuturia cations containing predominantly zoochlorellae and (Krupp 1983; Schwarz et al. 1999). anemones in more southerly and higher intertidal lo- What started as the search for an ideal model for cations containing predominantly zooxanthellae (Bates the study of cnidarian larval biology and symbiosis 2000; Secord & Augustine 2000). onset has instead developed into a 7-year love/hate re- This distribution of algal types has been attributed lationship with larvae of A. elegantissima. Despite our to differential physiological tolerance of the algae to best efforts, answers to key questions surrounding de- temperature and light, with zooxanthellae well adapted velopmental processes, such as spawning and fertiliza- to relatively high-light and warm conditions compared tion patterns and settlement and metamorphosis, re- to zoochlorellae, which are better adapted to cooler, main elusive. Here we report results from studies on lower-light environments (Verde & McCloskey symbiosis onset, specifically in the examination of in- 1996a,b, 2001, in press; Saunders & Muller-Parker fection dynamics of the larvae by both zooxanthellae 1997). A. soh, a newly described, solitary sister spe- and zoochlorellae and in the differential infection by cies to A. eleguntissima (Francis 1979; McFadden et zooxanthellae and zoochlorellae as a function of light al. 1997; PEARSE& FRANCIS2000), is restricted to lat- regime. We describe a morphological change in some itudes below the likely southern limit of zoochlorellae mature larvae that may represent the development of (Secord & Augustine 2000) and is symbiotic only with tentacles before settlement. Further, we provide the zooxanthellae. A. artemisia is a non-symbiotic species first description of larvae of A. artemisia and the re- (Hand 1955; PEARSE& FRANCIS2000). Recent molec- sponse of this non-symbiotic species to exposure to ular phylogenetic studies have grouped A. elegantis- zooxanthellae and zoochlorellae. We also report on sima,A. sola, and A. xunthogrummica into an eastern some key failures in our examinations of larval A. ele- Pacific clade, whereas A. artemisia groups with other guntissima. We describe variability in spawning and conspecifics in a distinct western Pacific clade, despite fertilization success through the study period, that has its occurrence in the eastern Pacific (Geller & Walton hampered our ability to continue studies of larvae. In 2001). addition, we describe our numerous failed attempts to Larvae of A. ekgnntissima and A. xanthogrummica induce settlement and metamorphosis of thcsc larvae, are oval planulae, - 150 pm long and 100 p,m wide with the aim of preventing repetition of thesc ap- with a conspicuous apical tuft of numerous elongate proaches by other researchers in future studies. cilia present at the aboral pole (Siebert 1974). They swim actively in the water column (Siebert 1974), feed Methods by extrusion of a mucous thread that traps food (Sie- Animal maintenance, gamete collection, and larval bert 1974; Schwarz et al. 2002), and persist in culture cultures for as long as 3 months (Smith 1986). Despite attempts by many workers to induce settlement by the addition Adults of Anthopleura elegantissima, to be used as of various natural substrates, such as mussel shells, broodstock and as sources of freshly isolated algae, rocks, and macrophytes, to larval cultures (Siebert were collected at various times from 1990 to 2001, 1974; Smith I986), no larvae have ever been observed from the following locations along the Pacific coast of to undergo settlement and metamorphosis. the U.S.: Santa Cruz, California; Strawberry Hill, Nep- Gametes and larvae of A. elegantissima and A. xun- tune Beach, Seal Rock, and Boiler Bay, Oregon; and thogrammica lack symbiotic algae (Siebert 1974). Swirl Rocks, Washington. Zooxanthellat-e animals These anemones must therefore acquire symbionts were collected from all locations, but all zoochlorellate from the environment during their development. We animals were collected at Swirl Rocks. Experimcnts in were initially interested in larvae of A. elegantissima 1995 and 1996 were performed at Long Marinc Lab- as a model for the study of the onset of symbiosis oratory (University of California Santa Cruz) in Santa between sexually produced cnidarian offspring and Crux, California; those in 1997-200 1 were performed their symbiotic zooxanthellae. Symbiosis onset in A. at Hatfield Marine Science Center (Oregon State Uni- elegantissima can begin in the larval phase during the versity) in Newport, Oregon; and those in the summer feeding process, whereby zooxanthellae, freshly iso- of 2000 were performed at Walla Walla College Ma- lated from adults, are taken in along with food rine Station, Anacortes, Washington. All foreign ma- (Schwarz et al. 2002). Algae are phagocytized by host terial adhering to the anemones (e.g., fragments of gastrodermal cells and ultimately persist in the gas- shell and macroalgae) was removed after collection trodermal tissue. A similar process of algal acquisition and individual anemones were allowed to attach to (infection) by feeding during the larval stage has been bricks (3-10 anemones of one clone per brick). The described in the anemone Aiptasia tagetes (Riggs bricks were submerged in outdoor seawater tanks ex- 192 Weis, Verde, Pribyl, & Schwarx posed to ambient sunlight. Animals were fed previ- cut in half transversely and the basal half was discard- ously frozen brine shrimp or minced mussel 1-3 times ed. The oral half was minced with a razor blade, ho- a week. mogenized in filtered seawater (FSW) with a ground- Techniques moditied from Smith (1986) and glass tissue grinder, and centrifuged at 4000 g for 5 Schwarz et al. (2002) were used to induce spawning. min to separate the algal cells from the animal tissue. In the morning, bricks with the anemones were placed A thin animal pellet fraction, layered above the algal into open aquaria containing ambient, static seawater pellet and consisting mostly of nematocysts, was care- and exposed to ambient sunlight for a period of time fully removed and discarded. The algal pellet was re- without changing the seawater. Water temperature was suspended in FSW and re-pelleted 3 times to partially not controlled and rose during this exposure. On over- clean the algae of contaminating animal debris. The cast or foggy days, the seawater in the containers was final pellet was resuspended in FSW and this dense replaced with cold ambient seawater at dusk. On sunny suspension was passed through a 1 SO-Fm mesh screen days, the anemones were exposed for 4-6 h before the to break up large algal clumps. Algal isolates were water was changed. Within 12 h after a water change, used within I h of preparation. the anemones released gametes. In numerous preliminary experiments, we attempted When spawning occurred, thousands to hundreds of to quantify the number of algae that we added to lar- thousands of negatively buoyant eggs settled on the vae. After the 3 rinses in FSW (described above), both oral disc or on the bottom of the aquaria and were zooxanthellae and zoochlorellae were highly clumped subsequently collected with a large plastic pipette (tur- and contaminated with animal debris, making quanti- key bastes) and placed into a large glass container with fication impossible. Additional seawater rinses did not seawater. A small amount of sperm suspension was improve results. In an attempt to obtain more uniform similarly collected and added to the eggs for fertiliza- suspensions, after the rinses, the algal suspension was tion. The mixture was maintained in a cold room at placed in a syringe and forced through a metal SO-pm 14°C (in California and Oregon), or in 10-liter con- mesh screen. Algae in the resulting suspension were tainers suspended in indoor tanks of flow-through am- no longer clumped and could be quantified; however, bient seawater at 12-14°C (in Washington). At 24 h subsequent infection was severely reduced (data not after fertilization, planulae were filtered out of the sea- shown), presumably because the algae were compro- water using a 50-pm mesh screen and placed into fresh mised during preparation. We therefore decided to pro- seawater in glass or plastic containers. For the Cali- vide unquantified, high concentrations of algae to the fornia and Oregon experiments, larvae were kept in larvae (see below) and obtained repeatable infection the incubators mentioned above in a 12-h: 12-h light/ results. dark regime at an irradiance of -40 pmoles of quanta/ m2/sec. For the Washington experiments, larvae were exposed in outdoor tanks to 55% ambient sunlight us- Infection of A. elegantissima and A. artemisia with ing neutral density screens placed over the containers algal isolates of larvae. Tn all locations, every 2 days, the seawater In order to compare the abilities of the two algal was changed throughout the experiments. Larvae kept types to infect A. elegantissima larvae, we measured in unfiltered seawater remained aposymbiotic through- the percentage of larvae infected with zooxanthella out their -30-day lifespan unless experimentally in- and zoochlorella isolates and the den\ity of algae in fected (see below). larvae through time. For each experiment, - 10,000 Field collection of gametes of A. artemisia 4-day-old larvae were divided between two gla\s or At Strawberry Hill and Neptune Beach, Oregon dur- plastic containers. Zooxanthella isolates were added to ing a low tide on 26 June 1998, numerous individuals one container and zoochlorella isolates to the other. of A. artemisia were observed spawning. Eggs and Larvae of A. elegantissima acquire algae during feed- sperms were collected with a Pasteur pipette and ing when algal cells enter through the mouth with food mixed in several 15-ml conical tubes. The tubes were (Schwarz et al. 2002). Therefore, several drops of ho- transported back to the laboratory and the developing mogenized brine shrimps (Artemiu sp.) or mussel tis- planulae were maintained in the same fashion as larvae sue (Mytilus sp.) were added to the bowls to stimulate of A. elegantissirnu. a feeding response. After 5-6 hours, algal iwlates were still present in large quantities at the bottom of Preparation of freshly isolated zooxanthellae and the dish. Larvae were then concentrated with a SO-pm zoochlorellae mesh screen, placed into clean filtered seawater, and Algae were isolated from adults of A. elegantissima returned to incubators (in CA and OR, see above) on (see Schwarz et al. 1999, 2002). Adult anemones were a 12-h:12-h light/dark regime at 40 ymol quanta/m2/ Larval biology of Anthopleura elegantissirnu I93

Table 1. Natural and artificial substrates used in attempts to induce settlement and metamorphosis in larvae of AnthopZeuru eleguntissima.

Age of larvae Location ol Substrates (days) Symbiotic state trcatment Live mussels, shells, and shell debris (MytiZus) 16, 18 aposymbiotic, zooxanthellate, CA, OR zoochlorellate Miscellaneous rocks collected from intertidal 16, 18 aposymbiotic, zooxanthellate, CA, OR zoochlorellate Adult anemone 16, 18 aposymbiotic, zooxanthellate, OR xoochlorellate Plastic Petri dishes roughened with sandpaper" 13 aposymbiotic WA Plastic Petri dishes scored with scalpel" 13 aposymbiotic WA Plastic Petri dishes with light and dark formica tiles 13 aposymbiotic WA on bottom" Plastic Petri dishes with adult anemone (zooxanthel- 13 aposymbiotic WA late and zoochlorellate) in dish" * All Petri dishes and tiles were seasoned in unfiltered running seawater for 10 days before use. sec, or to outdoor seawater tanks in 55% ambient light plastic Petri dishes. All Petri dishes were conditioned (in WA, see above). For some experiments performed in unfiltered running seawater for 10 days before lar- in Oregon in August 1998, symbiotic larvae were in- vae were added. Groups of larvae for each treatment cubated at 120 pmol quanta/m2/sec (see Results). were subdivided into 8 replicates; 4 replicates were Infections of larvae of A. arternisia with algae iso- placed in a refrigerated incubator and 4 in an outdoor lated from A. elegantissima were performed as for lar- flow-through seawater tank at 55% ambient sunlight. vae of A. eleguntissima in Oregon, except that unclum- The dishes were examined daily for - 15 days under ped algae, prepared by forcing them through a 50-pm a dissecting microscope for evidence of larval settle- mesh screen, were used to infect the larvae. ment and metamorphosis. Immediately after feeding, and approximately every Larvae were exposed to several compounds known 2 days thereafter, at least 25 larvae were examined to elicit settlement and metamorphosis in various ma- from each of the larval containers. Larvae suspended rine invertebrates, including some cnidarian species in FSW were dispensed onto a glass microscope slide. (see references in Table 2). Approximately 15 larvac A cover slip was placed on the drop and pressed onto were placed in each well of 24-well culture plates in the slide with a pencil eraser. This pressure was suf- FSW (5 ml per well) with one of the compounds for ficient to squash the larvae without completely dis- the duration of the monitoring period (n = 8 wells for persing their contents. Depending on the experiment, each compound at each concentration). Controls were the number of larvae infected and/or the number of set up under identical conditions without a chemical algal cells per larva were quantified for each sample. inducer. Plates were incubated as described for larval cultures. The plates were examined daily during the Attempts to induce settlement and metamorphosis monitoring period (see Table 2), under a dissecting of A. elegantissima using substrates and microscope, for evidence of larval settlement and metamorphic inducers metamorphosis. Several substrates (Table 1) and compounds (Table Statistical analyses 2) were added to cultures of both symbiotic and apo- symbiotic larvae of A. elegantissima in an attempt to All data were checked for normality and homoge- initiate settlement and metamorphosis. For the exper- neity of variances (Cochran's method, Winer 197 1) be- iments with natural substrates in California and fore conducting statistical analysis. If data were non- Oregon, hundreds of larvae were added to glass bowls normal or heterogeneous, the data were log-transformed containing one of the substrates listed in Table 1. The before analysis. All percentage data were square-root bowls were incubated as described above for larval arcsine transformed before analysis. Analysis of vari- cultures and were monitored daily under a dissecting ance (2-way ANOVA), post-hoc Tukey (HSD) test, microscope for evidence of settlement. For the exper- and linear regression were performed using the statis- iments in Washington, -25-30 larvae were added to tical package STATISTICA\WD (Statsoft, Inc.). 194 Weis, Verde, Pribyl, & Schwarz

Table 2. Metamorphic inducers used in attempts to induce settlement and metamorphosis in larvae of Anthopleura ele- gantissima.

Duration Age of of moni- Location Concentration of larvae toring Symbiotic of Metamorphic inducer inducer (days) (days) state* treatment Reference Metamorphosin A 1, 2, 5, 10, 20 mM 14 5 no record CA Leitz et al. 1994 (peptide: DNPGLW) 1 2-0-tetradecanoyl- 11 3 APO, ZX OR Henning et al. 1996, I998 phorbol- 13-acetate (TPA)

Phorbol 1 2,13- 11 3 APO, ZX OR Henning et al. 1996, 1998 dibutyrate (PDBu)

NH,CI 1s IS APO, ZX, WA Coon ct al. 1990 zc

KCI 15 15 APO, ZX, WA e.g., Pechcnik & Gee 1993; ZC Bryan et al. 1997; Big- gers & Laufcr 1999 y-aminobutyric acid 10-2, 10 7, 10 4, 15 15 APO, ZX, WA e.g., Morse 1985; Bryan et (GABA) 10-5, 10-6, 10 7, zc al. 1997 M 3,4--dihydroxy- 10 2, 10-1, 10 4, IS 1s APO, ZX, WA e.g., Morse 1985; Bryan et phenylalanine 10-5, 10 6, 10-7, zc al. 1997 (DOPA) M

:+ APO = aposymbiotic, ZX = zooxanthcllate, ZC = xoochlorellate.

Results Infection of A. elegantissirna larvae with zooxanthellae and zoochlorellae Spawning and fertilization In 1999 in Oregon and in 2000 in Washington, we Observations of spawning and fertilization of gam- examined the dynamics of infection by zooxanthellac etes of Anthapleuua elegantissirna were recorded and zoochlorellae in larvae of A. elegantissirna, in nu- through the spring and summer of 1995-2001 (Table merous trials using larvae of different ages and from 3). Spawning and fertilization occurred as early as different parents. When unquantified algae were pro- April I in one year to as late as September 4 in an- vided at saturating concentrations, 60- 100% of larvae other. There was a high variability in both spawning acquired algae on day 0 (selected data shown in Fig. and fertilization success throughout the monitoring pe- I). In most experiments, the percentage of infected lar- riod. Between 1995 and 1998, 82% of the attempts to vae decreased over time in both zooxanthellate and induce spawning were successful (does not include zoochlorellate populations to as low as 25% by 14- 16 spontaneous spawning). In contrast, from 1999 days post-infection. In all infection experiments, through 2001, just 69% of attempts yielded gametes, zoochlorellae infected a higher percentage of larvae and 4 of these 9 events resulted in release by only one (usually >90%) on day 0 than did zooxanthellae (60- gender, usually males (6 of 7 such instances). Similar- 90%). ly, in 1995-1998, when both sexes spawned (both Algal infection was strongly influenced by different spontaneous or induced), gametes were successfully light regimes (Fig. 1). Zoochlorellae appear to persist fertilized 100% of the time. In contrast, gametes gen- in larvae incubated in relatively low-light conditions erated from 1999-2001 were successfully fertilized in and zooxanthellae persist in those incubated in rela- only 20% (1 of 5) of events in which both sexes re- tively high-light conditions. In a low-light environ- leased gametes. ment of 40 pmol quanta/m2/sec (OR, June 1998), lar- Larval biology of Anthoplettra elegantissirnu I95

Table 3. Spawning and fertilization of Anthopleura elegantissirnu. Gametes from adults maintained in outdoor seawater tanks.

Spawning Induced or Date of attempt Location spontaneous Males Females Fertilization 199s June 21 CA Induced Y N July 3 CA Spontaneous Y Y Ye5 July 18 CA Spontaneous Y Y Yes August 24 CA Induced N N August 27 CA Induced N N 1996 April 1 CA Spontaneous Y Y Yes April 2 CA Spontaneous Y Y Ye4 June 24 CA Induced Y N July 2 CA Spontaneous Y Y Ye5 July 22 CA Induced Y Y Ye5 1997 May 28 OR Induced Y Y Ye5 September 4 OR Induced Y Y Ye4 1998 June 5 OR Induced Y Y Ye5 June 24 OR Induced Y N July 7 OR Induced Y Y Ye\ August 12 OR Induced Y Y Yc5 I999 June 17 OR Induced N N Junc 24 OR Induced Y N August 3 OR Induced Y Y N 0 August 20 OR Induced Y Y N 0 2000 early June OR lnduced Y N late June OR Induced Y N July 10 WA Induced Y Y Ye5 July 29 WA Induced N Y August 13 WA Induced N N August 22 WA Induced N N August 23 WA Induced N N September 3 WA Induced Y Y NO 200 I August 3 OR Induced Y Y NO vae infected with zoochlorellae remained >90% always the same and is illustrated (Fig. 2) for one ex- infected for the duration of the 16-day experiment periment (OR, August 1998; same larvae a\ in Fig. whereas the proportion of larvae infected with zoo- IB). The average number of algae per larva immedi- xanthellae decreased dramatically from 65% to just ately after feeding was highly variable for both roo- 23%, at a rate of -3.6%/day (Fig. 1A). A similar ex- chlorellae and zooxanthellae, but zoochlorellae were periment, performed in a moderate-light regime of I20 always present at a higher density than were zooxan- kmol quanta/m2/sec (OR, August 1998) resulted in de- thellae. For example, for the experiment shown in Fig. creasing levels of infection for both zoochlorellate 2, the average number of zoochlorellae per larva on (-2.5Wday) and zooxanthellate (-3.7Wday) larvae day 0 was 70 (40% of larvae contained over 100 algal (Fig. IS), rates similar to those shown by zooxanthel- cells), significantly higher than the average of just 20 late larvae at 40 kmol quanta/m2/sec. Finally, in a third algae per larva for zooxanthellate larvae (Tukey, p < experiment in a relatively high-light regime of 55% .OO 1 ; Fig. 2A,B). Average densities of both algal types ambient light (WA, July 2000), proportions of larvae decreased dramatically starting 1 day after infection, infected with Loochlorellae decreased through time at such that by day 3, most larvae contained

40

20 OR, 40 pmol quanta/m*/sec A/ 0 2 4 6 8101214 0 2 4 6 8 10 12 14 16 Days after infection =Q) U t5 100 Day 0 zooxanthellae 80 zoochlorellae 1 60 I 40 I yl . 20 1-, OR, 120 pmol quanta/mZ/sec =o z -B(QO 7- 0 0 2 4 6 810121416 - b 1-10 11-25 25-50 51-75 76-100 >I00 0 .cIc 100 ti a 80 ObQ 60 0 40

WA, 55% ambient sunlight C 20 0 2 4 6 8 10 12 14 16 0 Days after infection 1-10 11-25 25-50 51-75 76-100 >I00 Number of algae / larva Fig. 1. Infection of larvae of Anthopleura elegantissima with moxanthellae or zoochlorellae in 3 different larval co- Fig. 2. Density of zooxanthellae and zoochlorellae in a horts exposed to different light regimes. Larvae from: (A) group of larvae of Anthopleuru elegantissima through time. Oregon in June 1998; (B) Oregon in August 1998; and (C) (A) Changes in average number of zooxanthellae and zooch- Washington in July 2000. Percentage of infection decreases lorellae per larva over time, showing an initial sharp de- through time regardless of symbiont type; however, zooch- crease in density of algae in both zooxanthellate and xooch- lorellae persist in larvae in low light and zooxanthellae per- lorellate larvae and subsequently a gradual decline for the sist in larvae in high light. Larvae were 4 days old when duration of the experiment. Points represent mean t SD, n infected with algal isolates. Each point represents counts of = 25 larvae counted. Frequency distribution of number of 25-30 larvae. zooxanthellae and zoochlorellae per larva at day 0 (B) and day 3 (C) showing a shift from high percentages of' larvae possessing many algae to most larvae possessing <10 algae. A. artemisia larvae and infection with *** = Density of zoochlorellae different from density of zooxanthellae and zoochlorellae zooxanthellae at p < .001 (Tukey). Larval development in A. artemisia was essentially identical to that of A. elegantissima. Within one day of fertilization, barrel-shaped planulae, with an apical Larval biology of Anthopleura elegantissirnu 197

tuft like that of A. elegantissirnu, were actively swim- ming in the water column. By day 4, larvae had de- veloped a mouth. In order to determine if this non- symbiotic species was capable of becoming infected in the larval stage with zooxanthellae and/or zoochlo- rellae, we conducted infection experiments like those described for A. elegantissirnu, providing saturating concentrations of zooxanthellae or noochlorellae iso- lated from adults of A. eleguntissimu. We monitored the larvae on days 0, 2, and 4 only (n = 25 larvae counted daily for each algal type). On day 0, 64% of larvae incubated with zooxanthellae acquired algae whereas only 20% incubated with zoochlorellae ac- quired algae. By day 2, no larvae incubated with LO- oxanthellae contained algal cells and only 12% of those exposed to zoochlorellae still contained any algal cells. By day 4, larvae containing roochlorellae de- creased again to 4%. Regardless of algal type or day, the number of algal cells per larva always remained <10 (data not shown). We did not determine whether algal cells were phagocytized by gastrodermal cells or merely remained inside the gastric cavity without en- tering host cells.

Experiments using substrates and compounds to induce settlement and metamorphosis in A. elegantissirnu larvae No definitive settlement or metamorphosis was ob- served by larvae placed in any of the treatments tried (Tables 1, 2). Some larvae placed in containers with mussels and intertidal rocks (Table 1) did display some early signs of settlement behavior. Larvae were seen attaching to rocks and shells. They would remain for minutes to hours but would be missing later when checked again. Larvae placed with these natural sub- strates were extremely difficult to follow, because of their small size (<200 pm) and the complexity of the substrate. In all cases, within I to 2 days of placing the larvae in these containers, they disappeared. Larvae placed with artificial substrates (Table 1) or Fig. 3. Larvae of Anthopleura elegantissirnu displaying a in metamorphic inducers (Table 2) never displayed any lobed morphology. This morphology was observed in only settlement behavior and remained actively swimming a few groups of larvae, 2-3 weeks old, and occurred re- in the water column for the duration of the exposures. gardless of symbiotic state (see text for details). (A) Planula (aposymbiotic) beginning to change shape, still barrel- Changes in morphology in older larvae shaped but small bulges at oral end around mouth are evi- dent. (B) Larva (zoochlorellate) with more distinct oral bulg- Larvae of A. elegantissirnu from three different es and a shortened oral-aboral axis, causing the larva to take spawns (CA, April 1996; OR, September 1997; and on a rounded appearance. (C) Larva (aposymbiotic) with OR, June 1998) displayed marked changes in mor- fully-developed lobed morphology, with highly pronounced phology at 2-3 weeks of age. Multiple bulges were bulges around the mouth. Open block arrows in all three observed emerging from the oral end (Fig. 3A,B). panels indicate bulges emerging from oral end around Over time, these bulges became pronounced and took mouth. Apical tuft (AT) is evident in larvae in A and C. ZC on the look of developing tentacles (Fig. 3C). At the denotes zoochlorellae. Scale bars, 50 pm. same time, in many larvae, the oral-aboral axis became 198 Weis, Verde, Pribyl, & Schwarz shortened, causing the larvae to lose their barrel shape success over time for these long-lived animals. Other and take on a rounded shape (Fig. 3B). The apical tuft studies have yielded little information about the vari- remained present at the aboral end (Fig. 3A,C). The ability of sexual reproduction in A. elegantissirnu. Pre- change to a lobed morphology occurred synchronously vious studies have reported yearly cycles of gonad in larvae from the same spawn in different bowls, and growth starting in late fall and ending with spawning, occurred in aposymbiotic, zooxanthellate, and zoo- as evidenced by decreased gonad size in late summer chlorellate larvae. In all three groups of larvae in to early fall (Ford 1964; Jennison 1979; Sebens which the lobes were observed, the majority of the 198 lb). Actual spawning or fertilization rate has never larvae assumed this morphology. However not all lar- been monitored. There is one hint from field studies val groups that reached 2-3 weeks of age underwent of larval recruitment in A. elegantissirnu and A. xun- a change to this morphology. Some groups persisted thogrammica that production of larvae may be episod- as barrel-shaped planulae for the duration of their lives ic (Sebens 1981a, 1982). A large, isolated larval re- (-1 month). cruitment event occurred in one year and was not repeated in subsequent years at the same site. This Discussion could be evidence for strong spawning and Fertilization Spawning and fertilization in that one year, but could also have been due to var- ious physical or biotic conditions that favored settle- Spawning in Anthopleura elegantissirnu was ob- ment. served or induced far earlier in the season in this study than had been reported previously, either in field stud- ies from Central California (Ford 1964; Jennison Infection of larval A. elegantissirna with 1979) or from Washington (Sebens 1981b). Spawning zooxanthellae and zoochlorellae was reported in late summer or early fall, whereas we This study demonstrates the ability of larvae of A. observed spontaneous spawning as early as April in to acquire zoochlorel lae, freshly isolated California one year, and repeatedly induced spawning eleguntissirna from an adult, during feeding, in the same manner as in June and July in 1996-1998 (Table 3). larvae acquire zooxanthellae (Schwarz et al. 2002). We report high variability in experimental spawning The proportion of larvae infected with either algal type and fertilization success (Table 3) that we cannot ex- plain by any changes in experimental technique. Like- declined over time in most experiments (Fig. I). No wise, spawning and fertilization variability was not comparable data have been reported lor other symbi- correlated with a potentially aging broodstock. Some otic cnidarian planulae. These declines in percent in- of the animals that spawned (both spontaneously and fection may be due to differential mortality of sym- induced) in Santa Cruz in 1995 and I996 had been in biotic compared to non-symbiotic larvae or, seawater tables for several years. Further, animals were alternatively, to expulsion or digestion of' algae in newly collected from the field before the failed spawn- some symbiotic larvae over time. ing attempts of 24 June 1999, 12 August 1999, and Zooxanthellate larval cultures remained relatively late June 2000 in Oregon and that of 13 August 2000 stable in high-light treatments compared to zoochlorel- in Washington. Spawning did not appear to be related late cultures, which were stable in low-light treatmcnts to the presence of other spawning conspecifics. There (Fig. I). These data are consistent with the variety of' were occasions when anemones spawned when alone studies that have shown differential habitat distribution in a tank and conversely, there were occasions when of A. elegantissirnu and A. xanthogrurnmica harboring a single anemone spawned in a tank with other anem- the two algal types (Bates 2000; Secord & Augustine ones (data not shown). These data are consistent with 2000) and differential tolerance of the two algal types other studies by Smith ( 1986) that found no correlation to temperature and light in adults of A. eleganl between spawning and presence of other spawning (Verde & McCloskey 1996a,b, 2001, in press; Saun- conspecifics. ders & Muller-Parker 1997). Zooxanthellae have been This study provides little insight into the observed shown to out-perform zoochlorellae, as measurcd by variability in spawning and fertilization; the work was growth rates and photosynthetic productivity, in warm- not originally designed to track these parameters. We er, higher-light environments, and have been hypoth- did not monitor gonad size or percent fertility of the esized to outcompete zoochlorellae in anemones found broodstock, nor did we track spontaneous spawning in these habitats. The opposite has been found in events in the field. A more systematic field study, per- anemones in cooler, lower-light environments where formed over a longer period of time, would be required zoochlorellae perform better than zooxanthellae. Our to definitively demonstrate changes in reproductive study suggests that this differential environmental tol- Larval biology of Anthopleura elegantissima 199 erance of the symbionts may be manifest as early as Muller-Parker 2002). If either of these sources of algal the larval stage of these anemones. cells are used by larvae, the high concentrations of The temporal profiles of algal density in A. elegan- cells used in this study could be mimicking these in- tissima larvae were very similar between algal types fection modes. and between different experiments. The high numbers of algae in larvae at day 0 are likely a count of algae Larvae of A. urtemisiu in the gastric cavity, taken in during feeding and not We witnessed a mass spawning of A. artemisia dur- phagocytized by the cells of the larval gastroderm ing a low tide in July of 1998 at Strawberry Hill and (Fig. 2). The much lower numbers of algae on sub- Neptune Beach, Oregon. Virtually all observed indi- sequent days likely represent algae that have been viduals were spawning at both locations we visited. phagocytized by gastrodermal cells. This interpretation Similar events of single species spawning have been is supported by previous work in which transmission observed for A. elegantissima (E. Sanford, Stanford electron micrographs of A. elegantissima larvae 4 days University, pers. comm.) and for A. xanthogrammica after infection showed zooxanthellae only in gastroder- (J. Schwarz, unpubl. obs.). ma1 cells and none in the gastric cavity, although it is A. artemisia, a nonsymbiotic anemone species possible that algal cells remaining in the gastric cavity (Hand 1955; PEARSE & FRANCIS 2000) residing in a could be washed away during fixation (Schwarz et al. separate west Pacific clade from A. eleguntissima, A. 2002). Zoochlorellae average 9.6 pn2 in diameter com- solu, and A. xunthngrammica (Geller & Walton 2001 ), pared to an average of 12.5 km for zooxanthellae in did acquire zooxanthellae and zoochlorellae, but algal A. eleguntissima (Verde & McCloskey 1996b). There- cells occurred at low densities and at very low per- fore, the significantly higher densities of zoochlorellae centages. We did not determine if any algal cells en- compared to 7,ooxanthellae in larvae at day 0 could be tered host gastrodermal cells or if they remained in the due to the gastric cavity being able to accommodate a gastric cavity. The initial numbers and densities of al- greater number of smaller zoochlorellae than larger zo- gae in the larvae at day 0 were much lower than those oxanthellae. By several days after infection, algal den- of A. eleguntissima. We have hypothesized that algae sity stabilized at -1-2 algal cells per larva (Fig. 2) are initially taken into the gastric cavity in A. elegan- and was consistent between experiments and across all tissima and the coral F. scutaria as a result of a non- light regimes (data not shown). In contrast, when ex- specific feeding response (Schwarz et al. 1999, 2002). perimentally infected with zooxanthellae, larvae of the If this were true, then one would predict that larvae of scleractinian coral Fungia scutaria, which are com- A. artemisia would respond in the same way to algae parable in size to larvae of A. elegan in the presence of food; initially they would acquire an average of 20-23 algal cells per larva (Weis et al. large numbers of algal cells that would later be either 200 I). Whether this discrepancy in algal density be- expelled or digested. Instead, far fewer larvae of A. tween F. scutaria and A. elegarztissima is a real bio- artemisia took up algae and far fewer algal cells were logical difference or reflects sub-optimal culturing acquired per larva than in A. elegantissirnu. This sug- conditions of is not known. Riggs A. elegantissima gests that some specificity is involved in the early stag- (1988) did not quantify algal density in experimentally es of symbiont-host contact in A. eleganti.rsimu. Al- infected larvae of Aiptasia tugetes. To our knowledge, ternatively, reduced uptake rates in A. artemisia could no other infection studies have been performed on lar- be explained by the use of freshly isolated algae that vae from any other species of cnidarian. had been passed through a 50-pm mesh screen, a tech- There is no information on the mode of acquisition nique that sometimes reduced infection rates in A. ele- of symbionts by larvae of A. elegantissima or any oth- gantissima larvae (see Methods). er cnidarian, in the field, therefore it is difficult to know how relevant the artificial conditions created in Attempts to induce settlement and metamorphosis this study are to the biology of symbiosis onset. In the in A. eleguntissima laboratory, larvae of A. elegantissirnu can acquire zo- oxanthellae from “black egesta,” which contain high There is a rich literature on inducers of marine in- concentrations of zooxanthellae, released from adult vertebrate larval settlement (see Table 2 for examples anemones (Schwarz et al. 2002). Another possible of references). Much is understood in some species of source of high quantities of algae is the feces of the invertebrates, including some cnidarians, about the nudibranch Aeolidia papillma, which feeds on adults cues and signals, both natural and artificial, that induce of A. elegantissirnu, passes the symbionts unharmed settlement. Borrowing from these studies, we tried to through its digestive tract, and releases them in very induce larvae of A. elegantissima to settle using nu- high concentrations in the fecal pellets (Seavy & merous substrates and chemicals, and Tables 1 and 2 200 Weis, Verde, Pribyl, & Schwarz document our failed attempts. Other researchers have plains in part why progress in this area of cnidarian attempted settlement studies on A. elegantissirnu, using biology and symbiosis lags behind studies on adults. various natural substrates and larvae of approximately Despite the limitations we have encountered during the the same age, also without success (Siebert 1974; course of our 7 years of work on A. elegantissirnu, this Smith 1986). Settlement is a complex signaling pro- study produced some interesting insights into dynam- cess that can vary dramatically between groups of in- ics of infection by both zooxanthellae and zoochlorel- vertebrates (Hofmann et al. 1996). It is possible that lae, and changes in morphology of older planulae. we did not try the right concentration, duration, or combination of cues that the larvae need to initiate Acknowledgments. We thank M. deBocr and N. Covarru- settlement. Furthermore we did not consider environ- bias for help with some of the spawning and W, Reynolds mental parameters, such as light, temperature and even for assistance with the spawned gametes oi‘ A. urtemisia. emersion that could be required for settlement. Anoth- Thanks to members of the Weis lab for comments on the manuscript. This work was supported by a grant from the er possible explanation for the inability to induce set- National Science Foundation (9728405-IBN) to VMW. tlement is that the larvae used in the studies were too young and not yet competent to settle. It has been hy- References pothesiLed that planulae of A. elegantissirnu are long- Bates A 2000. The intertidal distribution of two algal sym- lived planktotrophic larvae capable of long-distance bionts hosted by Anthopleura xanthogramrnicu (Brandt dispersal (Sebens 1981 a,b, 1982). Larvae have been 1835). J. Exp. Mar. Biol. Ecol. 249: 249-262. kept alive in culture for as long as 3 months (Smith Biggers WJ & Laufer H 1999. Settlement and metamorpho- 1986), in our hands, -6 weeks. To date we have been sis of Capitella larvae induced by juvenile hormone-active unable to keep large cultures healthy and axenic for compounds is mediated by protein kinase C and ion chan- long enough to attempt settlement experiments on lar- nels. Biol. Bull. 196: 187-198. vae older than 3 weeks. Bryan PJ, Pei-Yuan Q, Kreider KL, & Chia F-S 1997. In- duction of larval settlement and metamorphosis by phar- Changes in morphology of older A. elegantissima macological and conspecific associated compounds in the serpulid polychaete Hydroides elegans. Mar. Ecol. Prog. larvae Ser. 146: 81-90. Some, but not all, groups of larvae over 2 weeks of Coon SL, Walch M, Fitt WK, Weiner RM, & Bonar DB age, developed a morphology that consisted of several 1990. Ammonia induces settlement behavior in oyster lar- large protuberances situated around the mouth (Fig. 3). vae. Biol. Bull. 179: 297-303. Although these structures resemble tentacles in their Dayton PK 197 I . Competition and community organization: the provision and subsequent utilization of space in a location and shape, histological studies are needed to rocky intertidal community. Ecol. Mono. 41 : 35 1-389. confirm the presence of two cell layers and the gastric Fitt WK, Pardy RL, & Littler MM 1982. Photosynthesis, cavity. These structures were much larger than the respiration, and contribution to community productivity of “collar” described by Siebert (1 974) and observed in the symbiotic sea anemone Anthopleura elegantissimu our cultures, which consists of a small circular ridge (Brandt, 1835). J. Exp. Mar. Biol. Ecol. 01: 213-232. surrounding the mouth of larvae of all ages. Smith Ford CE 1964. Reproduction in the aggregating sca anem- ( 1986) does not report any comparable morphological one, Anthopleura elegunkwima. Pac. Sci. 18: 138- 145. changes in his study of larvae of A. elegantissirnu, and Francis L 1979. Contrast between solitary and clonal life- to our knowledge, no similar accounts of this multi- styles in the sea anemone Anthopleura elegantissima. Am. lobed morphology have been described in other cni- Zool. 19: 669-681. darian planulae. We initially hypothesized that the ap- Geller JB & Walton ED 2001. Breaking up and gelting to- pearance of the morphology signaled competence to gether: evolution of symbiosis and cloning by fission in sea anemones (genus Anthopleura) inferred from a molec- settle; however larvae placed in bowls with natural ular phylogeny. Evolution 55: I78 1-1 794. substrates and the phorbol esters TPA and PDBu (see Hand C 1955. The sea anemones of central California. Part Table 2) did not settle. 11. The endomyarian and mcsomyarian anemones. Was- maim J. Biol. 12: 345-375. Conclusions Henning G, Hofmann D, & Benayahu Y 1996. The phorbol ester TPA induces metamorphosis in Red Sea coral plan- Our ongoing struggles-to obtain A. elegantissirnu ulae (Cnidaria: Anthoxoa). Experirnentia 52: 744-749. planulae predictably, to infect larvae with a quantified 1998. Metamorphic processes in the soft corals Het- number of algal cells, and to induce larvae to settle- eroxeniu fuscescens and Xeniu umbellatu: the effect of illustrate just some of the difficulties encountered in protein kinase C activators and inhibitors. Invertebr. Re- examining the early life history stages of cnidarians prod. Dev. 34: 35-45. and the onset of their symbioses with algae. This ex- Hofmann DK, Fitt WK, & Fleck J 1996. Checkpoints in the Larval biology of Anthopleura elegantissirnu 20 1

life-cycle of Cassiopeia andromeda: control of metagen- ate sea anemone Anthopleura elegantisxima (Brandt, esis and metamorphosis in a tropical jellyfish. Int. J. Dev. 1835). J. Exp. Mar. Biol. Ecol. 21 1: 213-224. Bid. 40: 331-338. Schwarz JA, Krupp DA, & Weis VM 1999. Late larval de- Jennison BL 1979. Gametogensis and reproductive cycles in velopment and onset of symbiosis in the scleractinian cor- the sea anemone Anthopleura eleguntissimu (Brandt, al Fungiu scutaria. Biol. Bull. 196: 70-79. 1835). Can. J. Zool. 57: 403-41 1. Schwarz JA, Weis VM, & Potts DA 2002. Feeding behavior Krupp DA 1983. Sexual reproduction and early development and acquisition of zooxanthellae in larvae OK the sea anem- of the solitary coral Fungia scutaria (Anthozoa: Sclerac- one Anthopleura elegantissima. Mar. Bid. 140: 47 1-478. tinia). Coral Reefs 2: 159-1 64. Seavy BE & Muller-Parker G 2002. Chemosensory and LaJeunesse TC & Trench RK 2000. Biogeography of two feeding responses of the nudibranch Aeolidia pupillo.sa to species of Symbiodinium (Freudenthal) inhabiting the in- the symbiotic sea anemone Anthopleura elegantissima. In- tertidal sea anemone Anthopleura eleguntissima (Brandt). vertebr. Biol. 121: 115-125. Biol. Bull. 199: 126-130. Sebens KP 1981a. Recruitment in a sea anemone population: juvenile substrate becomes adult prey. Science 21 3: 785- Leitz T, Morand K, & Mann M 1994. Metamorphosin A: a 787. novel peptide controlling development of the lower meta- 1981b. Reproductive ecology of the intertidal sea zoan Hydractinia echinata (Coelenterata, Hydrozoa). Dev. anemones Anthopleura xanthogrammica (Brandt) and A. Biol. 163: 440-446. elegantissima (Brandt): body size, habitat, and sexual re- McFadden CS, Grosberg RK, Cameron BB, Karlton DP, & production. J. Exp. M'v. Biol. Ecol. 54: 225-250. Secord D 1997. Genetic relationships within and between 1982. Recruitment and habitat selection in the inter- clonal and solitary forms of the sea anemone Anthopleura tidal sea anemones Anthopleura elegantissima (Brandt) elegantissima revisted: evidence for the existence of two and A. xanthogrammica (Brandt). J. Exp. Mar. Bid. Ecol. species. Mar. Bid. 128: 127-139. 59: 103-124. Morse DE 1985. Neurotransmitter-mimetic inducers of set- Secord D & Augustine L 2000. Biogeography and micro- tlement and metamorphosis of marine planktonic larvae. habitat variation in temperate algal-invertebrate symbio- Bull. Mar. Sci. 37: 697-706. ses: zooxanthellae and zoochlorellae in two Pacific inter- Muller-Parker G & Davy SK 2001. Temperate and tropical tidal sea anemones, Anthopleura elegantissirnu and A. algal-sea anemone symbioses. Invertebr. Biol. 120: 104- xanthogrammica. Invertebr. Biol. 1 19: 139-146. 123. Siebert AE 1974. A description of the embryology, larval Muller-Parker G & D'Elia CF 1997. Interaction between cor- development, and feeding of the sea anemones Anthopleu- als and their symbiotic algae. In: Life and Death of Coral ra elegantissima and A. xanthogrammica. Can. J. Zool. Reefs. Birkeland C, ed., pp. 96-1 13. Chapman Hall, New 52: 1383-1388. York. Smith B 1986. Taxonomy and population genetics of the sea Muscatine L 1971. Experiments on green algae coexistent anemone genus Anthopleura in California. Masters Thesis, with xooxanthellae in sea anemones. Pac. Sci. 25: 13-21. University of California Santa Cruz. 96 pp. 1990. The role of symbiotic algae in carbon and en- Verde EA & McCloskey LR 1996a. Carbon budget studies ergy flux in reef corals. In: Ecosystems of the World: Cor- of symbiotic cnidarian anemones-evidence in support of al Reefs. Dubinsky Z, ed., pp. 75-87. Elsevier, Amster- some assumptions. J. Exp. Mar. Biol. Ecol. 195: 161-171. dam. 1996b. Photosynthesis and respiration of two species Muscatine L & Weis VM 1992. Productivity of zooxanthel- of algal symbionts in the anemone Anthopleura elegantis- lae and biogeochemical cycles. In: Primary Productivity sima (Brandt) (Cnidaria: Anthozoa). J. Exp. Mar. Bid. Ecol. 195: 187-202. in the Sea. Falkowski PG & Woodhead A, eds., pp. 257- 2001. A comparative analysis of' the photobiology of 272. Plenum Press, New York. zooxanthellae and zoochlorellae symbiotic with the tem- Pearse V & Francis L 2000. Anthopleura .sola, a new spe- perate clonal anemone Anthopleura elegantissirnu cies, solitary sibling species to the aggregating sea anem- (Brandt) I. Effect of temperature. Mar. Biol. 138: 477- one, A. elegantissirnu (Cnidaria: Anthozoa: Actiniaria: Ac- 489. tiniidae). Proc. Biol. Soc. Wash. I1 3: 596-608. A comparative analysis of the photobiology of zo- Pechenik JA & Gee CC 1993. Onset of metamorphic com- oxanthellae and zoochlorellae symbiotic with the temper- petence in larvae of the gastropod Crepidula fornicata ate clonal sea anemone Anthopleura elegantissirnu judged by a natural and artificial cue. J. Exp. Mar. Biol. (Brandt) TI. Effect of light. Mar. Biol., in press. Ecol. 167: 59-72. Weis VM, Reynolds WS, deBoer MD, & Krupp DA 200 I. Riggs LL 1988. Feeding behavior in Aiptasia tagetes (Dus- Host-symbiont specificity during onset of symbiosis be- chassaing and Michelotti) planulae: a plausible mechanism tween the dinoflagellate Symbiodinium spp. and the plan- for zooxanthellae infection of aposymbiotic planktotrophic ula larvae of the scleractinian coral Fungiu scutaria. Coral planulae. Carib. J. Sci. 24: 201-206. Reefs 20: 301-308. Saunders BK & Muller-Parker G 1997. The effects of tem- Winer BJ 1971. Statistical Principles in Experimental Dc- perature and light on two algal populations in the temper- sign. 2nd edition. McGraw-Hill, New York. 345 pp.