FAU Institutional Repository

http://purl.fcla.edu/fau/fauir

This paper was submitted by the faculty of FAU’s Harbor Branch Oceanographic Institute.

Notice: ©1981 Canadian Science Publishing. This manuscript is the post-print of an article published by Canadian Science Publishing, an independent and not-for-profit organization. The final version is published in Canadian Journal of Zoology available at http://www.nrcresearchpress.com/journal/cjz and may be cited as: Jennison, B. L. (1981). Reproduction in three species of sea anemones from Key West, Florida1. Canadian Journal of Zoology, 59(9), 1708-1719. doi:10.1139/z81-235 1708

Reproduction in three species of sea anemones from Key West, Florida 1

BRIAN L. JENNISON Harbor Branch Foundation, RFD No. J Box 196, Fort Pierce , FL, U.S.A . 33450 Received February 16, 1981

JENNISON, B. L. 1981. Reproduction in three species of sea anemones from Key West, Florida. Can. 1. Zool. 59: 1708-1719. Reproduction in three species of sea anemones from Key West, Florida, was studied from October 1977 until July 1978. Condylactis gigantea is dioecious, has a I: I sex ratio, and exhibits an oviparous -+ planktonic -+ lecithotrophic reproductive pattern, based on the scheme of Chia, modified by the author. Phymanthus crucifer is also dioecious and exhibits a sex ratio not significantly different from I:I; however, this species is viviparous. Bartholomea annulata exhibits imperfect gynodioecious hermaphroditism, accompanied by asexual reproduction by pedal laceration; it is oviparous -+ planktonic -+ planktotrophic . It is argued that an oviparous habit with resultant larval dispersal is advantageous to C. gigantea, which is solitary. In contrast , viviparity in P . crucifer and pedal laceration in B. annulata represent two different methods of keeping offspring close, thereby maintaining habitat space. The evolution of three such diverse reproductive modes is discussed through an examination of the morphology, prey, and habitat type of each species .

JENNISON, B. L. 1981. Reproduction in three species of sea anemones from Key West, Florida. Can. 1. Zool. 59: 1708-1719. La reproduction a ere etudiee chez trois especes d'anemones de mer de Key West, en Floride, d'octobre 1977 a juillet 1978. Condylactis gigantea est dioique, a un rapport males.femelles de I: I et son cycle reproducteur suit une sequence ovipare -+ planctonique -+ lecitotrophe, selon Ie modele de Chia rnodifie par I'auteur. Phymanthus crucifer est egalernent dioique et son rapport males.femelles ne differe pas significativement de 1:I; cette espece est par ailleurs vivipare. Bartholomea annulata se reproduit par hermaphroditisme gynodioique imparfait et par reproduction asexuee (laceration du pied); l'espece suit la sequence ovipare -+ planctonique -+ planctotrophe. La reproduction ovipare suivie de la dispersion des larves est avantageuse pour C. gigantea qui est une espece solitaire. En revanche, la viviparite de P. crucifer et la laceration du pied chez B . annulata sont deux rnecanismes destines agarder proche la progeniture, de facon aconserver intact I'habitat. L'evolution de ces trois modes de reproduction differents est examinee ala lumiere de la morphologie de l'habitat et du type de proie de chacune de ces especes . [Traduit par Ie journal]

Introduction reproductive biology of these three anemones both out Although sea anemones occur in south Florida waters of basic interest in their natural history, and in the belief in variety and abundance, surprisingly little is known of that the project might yield information concerning the their reproductive biology. Some information may be ability of these animals to maintain space in their gleaned from systematic works (McMurrich 1889; respective habitats. Duerden 1898, 1902). However, Carlgren (1951) did Materials and methods not discuss the reproductive condition of his specimens Eleven to 17 (usually 12) anemones of each of three species when he reviewed the systematics of Bartholomea were collected on six occasions from October 22, 1977, until annulata and Phymanthus crucifer. Lewis (1960a) June 17, 1978. Collections were made subtidally from ledges briefly described reproduction in several anemones from and Thalassia beds just off the sea wall on the southeast corner Barbados, including a description of larval release and of the island of Key West, Florida. Generally, Bartholomea settling in P. crucifer. Although he found B. annulata annulata was found clustered on vertical ledges at depths of and Condylactis gigantea on coral reefs (1960b), he did 1-3 m, often among dense forests of the chlorophyte Hali­ not deal with their reproduction. meda sp., whereas Condylactis gigantea was collected from Bartholomea annulata and Condylactis gigantea are adjacent Thalassia flats. Phymanthus crucifer was found on important members of their subtidal communities, low rocks and in sandy areas at the edges of both Thalassia providing shelter for a variety of commensals and beds and near the ledges . Following collection, anemones were relaxed in 50:50 serving as "base stations" for fish cleaning activity (see seawater and isotonic MgCIz. Condylactis gigantea was Limbaugh et al. 1961; Mahnken 1972; Herrnkind et al. bisected longitudinally; one-half was weighed. Initially, spec­ 1976). Phymanthus crucifer, which is often found in imens of P. crucifer, B. annulata, and the other longitudinal dense beds in shallow sand surrounding rocks, can also half of C. gigantea were cleaned, fixed in Hollande's fluid, and be a dominant member of the invertebrate assemblages embedded in paraffin for histological examination . In later in which it is found. I undertook this study of the collections, after it was discovered that P. crucifer might contain brooded young, these anemones were dissected from 'Contnbution No. 223 of the Harbor Branch Foundation. the bottom prior to embedding . Sections 8-10 ILm thick were 0008-4301/81/091708-12$01 .00/0 © 1981 National Research Council of Canada/Conseil national de recherches du Canada JENNISON 1709 made on a rotary microtome. These sections were mounted, among males, females, and anemones with no gonads stained with picroindigo carmine, counterstained with basic are significant (F = 3.91 , P < 0.05). fuchsin, and examined and photographed under a compound In males, the first reproductive cells visible are the microscope. spermatogonia, which are 3.5 to 4.0 um in diameter. In order to assess whether the gonads were maturing These first appear in clusters in the gonadal endoderm. synchronously throughout their length, three cross sections were made of each specimen of B. annulata . One was taken Later they occur in vesicles surrounded by mesoglea. just below the oral disc, the second in midcolurnn , and the third The spermatogonia mature and differentiate into sperm, just above the base. Column diameters were measured from with heads approximately J .5 um in diameter. These the basal sections, using a compound microscope equipped cluster in the center of the spermatogenic vesicles (Fig. with an ocular micrometer. Only oral and basal sections were 2). Immature males were found on November 22, 1977, examined in P. crucifer because of its squat shape. Condylac­ and again on June 17, 1978; mature males were found in tis gigantea was so large that histological examination of the all collections (Table 1). entire gonad was impractical; therefore, sections were made In females, the first reproductive cells visible are from the midcolumn region only. However, fresh smears were oogonia, which arise in the endoderm of the gonadal examined during dissection in order to doublecheck the mesentery and range from 10 to 20 urn in diameter. histological results. To assess female maturation in all three species, oogon ia These cells have a relatively large nucleus and a and oocytes from each identifiable female were measured well-defined nucleolus. They migrate into the mesoglea under the microscope from prepared sections, using an ocular where they undergo vitellogenesis. According to Dunn micrometer. At least 50 gametogenic cells were measured (1975) the gametes may be considered oocytes at this from each female ; generally only those cells which had a stage (Fig. 3). Oocytes in C. gigantea are variously distinct nucleolus were measured, in order to standardize the oriented within the mesentery, and different stages of diameter measurements, as well as to avoid counting the same development may occur within a single mesentery . oocyte twice. However, if fewer than 50 cells were present in However, the germinal vesicles are always peripherally the sections, all ofthem were measured, regardless of presence located and appear to be associated during vitellogenesis or absence of a nucleolus. Male maturity was determined by with a cellular structure referred to by Nyholm (1943) the presence or absence of gonial cells and tailed sperm (see Jennison 1979). and Dunn (1975) as a trophonema. Oocytes of C. Oocytes and embryos of P. crucifer were fixed for scanning gigantea may reach a diameter of 500 urn in fixed electron microscopy in 2.5% glutaraldehyde at room temper­ material. ature for I h. The specimens were washed in Millonig's 0.4 M Phymanthus crucifer is also dioecious and exhibits a phosphate buffer, dehydrated in alcohol and amyl acetate, sex ratio not significantly different from I:I. Of 79 passed through a critical point drying apparatus, mounted and anemones examined, 36 were female, 29 were male, sputter coated . Planulae and juveniles were measured and and 14 had no gonads. No hermaphrodites were found. photographed live under a dissecting microscope. Spermatogenesis is similar to that described for C. gigantea except the gametes tend to be slightly smaller. Spermatogonia are approximately 2.5 urn in diameter, Results and mature sperm heads measure 1.0 urn across . Fur­ Gametogenesis ther, ripe testes are more densely packed with sperm Condylactis gigantea is dioecious; males and females than in C. gigantea (Fig. 4). Mature males were found in were found in a 1:I ratio. Of 71 animals examined in six all collections, although all anemones taken on May 18, collections, 33 were females and 32 were males. Five 1978, and half of those collected on June 17, 1978, had had no gonads and one was an hermaphrodite, exhibit­ spawned. Immature males were also found on June 17, ing mature sperm vesicles and developing oocytes in the 1978 (Table J). same mesenteries (Fig. 1). Oogenesis in P. crucifer begins with oogonia arising Although individual collections were not sufficiently in the endoderm. These migrate into the mesoglea and large to yield statistically testable data, it was instructive undergo vitellogenesis. A trophonema is present (Fig. to lump the weight data of C. gigantea over the course of 5). As in C. gigantea, oocytes of different stages of the study. Anemones with no gonads were smallest, development are randomly oriented within a single ranging in weight from 0.5 to 4.5 g (mean = 2.3 g). mesentery, and the germinal vesicle is always at the Males were significantly larger, ranging from 1.3 to periphery of the cell. Unlike C. gigantea, oocytes of P. 28.9 g (mean = 11.9 g), and females were largest, crucifer tend to be flattened out in the gonad in fixed ranging in weight from 2.4 to 63.4 g (mean = 14.6). preparations, and there are many fewer gametes per Even discounting the two largest females (weighing cross-sectional area of mesentery (Fig. 6). Fully grown 63.4 and 36.4 g), the mean weight of the 31 remaining oocytes average approximately 350 urn in diameter, females was 13.1 g. The observed differences in weight although cells of 450 to 500 urn were observed. 1710 CAN. J. ZOOL. VOL. 59, 1981

TABLE I. Numbers of immature and mature male anemones, and percentage which were mature in all collections

Bartholomea annulata, 0- + (( 0- Condyla ctis gigantea , 0- + one (( 0- Phymanthus crucifer, 0-

Immature Mature % mature Immature Mature % mature Immature Mature % mature

10122/77* 0 2 100 2 3 60 0 3 100 12/7/77 0 2t 100 0 4 100 0 5 100 2114/78 0 0 0 8 100 0 6 100 3/29/78 2 2 50 0 6 100 0 6 100 5/18/78 0 4 100 0 4 100 0 3t 100 6117/78 3 7 70 I 5t 83 3 3t 50

'Dales are given in the fonn month/day/year. [ Evidence of residual spenn (spawned).

Bartholomea annulata did not exhibit a 1:1 sex ratio . 9). In collections made on February 14, March 29, and Of 79 anemones examined, 23 were female, 4 were May 18, 1978, oocytes were divided into two size male, 18 were hermaphroditic, and 34 had no gonads. classes: those of approximately 200 urn or less in diam­ Combining basal diameter measurements from six eter, and those of 250 p.m or more. Large cells of collections, the average size ofthe 34 anemones with no 350 um or more made up nearly 50% of the oocyte gonads was 0.74 em. The average basal diameter of the population during these spring collections. By June 17, 23 females was 1.06 em, whereas that of the four males 1978, nearly all ofthese large oocytes were absent. In all was 1.10 em. The 18 hermaphrodites were largest, collections, 60% or more of male C. gigantea contained having an average diameter of 1.42 em. These differ­ mature sperm (Table I) . ences in basal diameter are highly significant (F = 6.41, Phymanthus crucifer females also exhibited a wide P < 0.001). range of oocyte diameters, from 40 um or less to In the few male B. annulata observed, spermatogen­ >450 j.Lm in all collections (Fig. 10). Collections made esis appeared to follow the typical actinian mode as on February 14 and March 29, 1978, yielded large shown by C. gigantea and P. crucifer, with mature numbers of oocytes of 300-400 urn in diameter; these sperm ofapproximately I urn in head diameterclustered were not found on May 18, 1978 . Phymanthus crucifer tightly in the center of mesogleal vesicles which males were all mature in all collections until May 18, measured approximately 150 um across (Fig. 7). Con­ 1978, when only residual sperm were found, and June sidering males and hermaphrodites together, mature 17, 1978, when early spermatogonia were also seen sperm were seen in all except the February 14, 1978, (Table 1). collections (Table 1). In an hermaphrodite collected on Bartholomea annulata females and hermaphrodites December 7, 1977, the sperm vesicles had broken down contained oocytes > 50 urn in diameter only in collec­ and the mesentery was filled with sperm (Fig. 8). Both tions made on October 22, 1977, March 29, 1978, and male and female gametes were found in the same June 17, 1978 (Fig . 11). In males and hermaphrodites, ,.mesentery, but .only oogonia and small oocytes were mature sperm occurred on October 22 , 1977; residual seen in the December 7, 1977, animal. The oocytes of sperm were seen on December 7, 1977. No male B. annulata are considerably smaller than those ofeither gametes of any stages were seen on February 14, 1978, C. gigantea or P. crucifer, reaching a maximum but they reappeared by March 29, 1978, and mature diameter in preserved sections of only 200 urn, and sperm occurred on May 18, 1978, and June 17, 1978 averaging considerably less. (Table 1). Reproductive cycles Brooding Condylactis gigantea females contained oocytes with Phymanthus crucifer broods its young internally. a broad range of diameters throughout this study (Fig. Free eggs, a blastula-stage embryo, planulae, and juve-

FIG. I. Gonad ofhennaphrodite Condylactis gigantea, February 14, 1978.0, oocyte; S, sperm; SG, spermatogonia. Scale = 50 IJ.m. FIG . 2. Spermatogenic vesiciesofC. gigantea, October Zz, 1977. EN; endoderm ; M; mesoglea. Scale = 100 IJ.m. FIG. 3. Oogonia and oocytes of C. gigantea, October 22, 1977. OG , oogonia. Scale = 75 IJ.m. FIG. 4. Testis ofPhymanthus crucifer, December 7, 1977. Scale = 150 IJ.m . FIG. 5. Ovary of P. crucifer, December 7, 1977, illustrating trophonema. RM, retractor muscle; T, trophonema. Scale = 150 IJ.m. FIG. 6. Ovary of P. crucifer with fully grown oocytes , December 7, 1977. GV, germinal vesicle . Scale = 100 IJ.m . FIG. 7. Testis of Bartholomea annulata, March 29, 1978. SV, spermatogenic vesicle. Scale = 200 IJ.m. FIG. 8. Gonad of hermaphrodite B. annulata, December 7, 1977. Scale = 100 IJ.m . JENNI SON 1711

7 8 '. 1712 CAN . J . ZOOL. VOL. 59, 1981

Candy/actis qiqantea f--t--++' N =12 per collect ion o 5 10 % f = Spawning 1= Hermaphrodites ( N) 11

500

450

400

350 E 3 300 ~ (jj'" ~ 250 o ~ 200 c-, u o o 150

100

50

O '---I---+---'---j--+-,---+---+~-f----I---+----'--+-~+-- N'i! 7 5 5 3 8 6 Sep. Oct. Nov. Dec. Jon Feb. Mar. Apr. May Jun. 1977 1978 FIG. 9. Size-frequency histogram of oocyte diameters of Condylactis gigantea ,

niles at 6, 12, and 24-tentacle stages were found in the the parental coelenteron, and those with 12 or 24 coelenterons of females. Planulae and juveniles were tentacles were capable of attaching to gravel, and feeding found in two males, and juveniles were found in an upon Artemia nauplii in a laboratory aquarium . Al­ anemone with no gonads. Oocytes of approximately though histological examination showed a close associa­ 200 urn in diameter were seen being extruded from the tion between the juvenile column and the parental gonadal mesentery of one female on May 18, 1978 (Fig. septum, no tissue connection occurred. One female 12). This anemone also contained a blastula-stage Bartholomea annulata had two planulae in its coelen­ embryo (Fig. 13), the individual cells of which were 100 teron on February 14, 1978; no further evidence of to 150 um in diameter. Planulae measuring 300 u.m in brooding in this species was seen. length were found in one female on December 7, 1977, in one male and one female on February 14, 1978, and in Discussion one female on May 18, 1978 (Fig. 14). In life, extruded Sea anemones reproduce in a variety of ways, both oocytes, the blastula , and planulae were all the same sexual and asexual (Stephenson 1928; Chia 1976). color, a light chocolate brown. Reproduction by both sexual and asexual means within a Juvenile anemones were seen in histological sections single species is quite common (e.g. Anthopleura of one female on December7, 1977, and again on March elegantissima; Metridium senile), and sexual races may 29, 1978. Six anemones from the May 18, 1978, exist, with differences in breeding patterns among collection yielded juveniles upon dissection (Table 2). populations of the same species (Rossi 1975). Chia Four females, one male, and one anemone without (1976) proposed a classification of sexual reproduction gonads bore tentaculate young, either at the 6-tentacle in anemones based on three criteria: the mode of stage (Fig. 15), or with 12 tentacles (Fig. 1'6) or 24 (Fig. fertilization, larval dispersal patterns, and larval feeding 17). Juveniles at these stages were unattached within habits. The most primitive of Chia's reproductive types JENNISON 1713

Phymonlhu5 crucifer N = 12 per collec tio n P = Brooded planulae present (NI J = Brooded juven iles present (N) i = Spawning

P(I) P(Z) J (II P{2 1 J (I) J (6)

500 f 4 50

400

350 E :;c, --... 300 ~ <1> E 250 0 0 <1> 200 >-u 0 0 150

10 0

50

N~ 9 7 4 5 6 5 Sep. Ocl . Nov. Dec. Jan . Feb. Mar. Apr. May Jun. 1977 1978 FIG. 10. Size-frequency histogram of oocyte diameters of Phymanthus crucifer. P(N) and J(N) number of anemon es brooding planulae or juveniles.

B orrtro/omeo annu/ola

~ N = 12 pe r collection o 5 10 P = Br ooded planula e pr esent (N ) % E f = Spawning ~ 200 1 = Hermaphrodi tes (N) ... <1> Q) 150 E l P (I) 0 31 t3{ 9ef 0 100 Ii

<1> +- c-, o 50 0 0 ts N ~ + if 7 2 4 6 6 12 Sep. Oct. Nov Dec Jon. Feb. Mar Apr Ma y J un. 1977 1978

FIG. II . Size-frequency histogram of oocyte diameters of Bartholomea annulata. 1714 CAN. J. ZOOL. VOL. 59. 1981

TABLE 2. Phymanlhus crucifer: brooded embryos and juveniles found in May 18, 1978, collection

Anemone No. Sex Embryos Juveniles

'? I blastula 1.72-4.26 mm basal diameter, N = 9, mean = 2.78 (1.2 mm diameter) I planula (1.07 x 1.48 mm) 2 '? 3.28mm,N = I 3 '? 0.90-2.13 rnm, N = 7, mean = 1.47 4 '? Free eggs in colenteron 1.23 rnm, N = I (200 urn diameter) 5 0 2.46mm, N = I 6 ng 0.74-2.05 rnrn, N = 8, mean = 1.22 is oviparous ~ pelagic ~ planktotrophic. I suggest that then possibly the males observed were in fact hermaph­ it might be more correct to call this type oviparous ~ rodites which had spawned all their oocytes . Perhaps the planktonic ~ planktotrophic. This type is characterized large females observed were hermaphrodites which had by small eggs which are spawned and fertilized exter­ spawned their sperm. In an hermaphrodite collected on nally. The resultant planulae are capable of wide dis­ December 7, 1978, the sperm were mature and in the persal, although because they feed in the plankton they process of being spawned, but the only female gametes face possible starvation and heavy predation pressure. present were oogonia and small, previtellogenic oocytes Sexual reproduction in Bartholomea annulata fits this (Fig. 8). However, the oocyte size-frequency data (Fig. type, although the case is somewhat complicated. 11) indicate that spawning in B. annulata occurred Bartholomea annulata exhibits imperfect gynodi­ between October 22 and December 7, 1978. As residual oecious hermaphroditism. Dunn (1975) reported gyno­ sperm may be present for several months after spawning dioecious hermaphroditism in the extemally brooding in some anemones (Jennison 1979), the co-occurrence anemone Epiactis prolifera, in which female gonads of mature (possibly residual) sperm and immature appear first, followed by male gonads in the same oocytes is not necessarily indicative of a lack of individuals. There are no males. Rossi (1975) also found gametogenic synchrony. A second view is that these this pattern in some populations of Cereus peduncula­ males were in fact hermaphrodites in which the female tus;she called it protogyny. In my study of B. annulata, component of the gonad had been resorbed. It is known males were found, but these represented only 5% of the that anemones may resorb the spent gonad after spawn­ 79 anemones examined. Females accounted for 29% ing (Jennison 1979). However, as the gametes of both and hermaphrodites for 22% of those animals collected. sexes occur in the same gonad in B . annulata (Fig. 8), Further, the smallest male was 6.1 mm in basal diam­ this argues against the selective resorption of the eter, whereas female gonads were found in anemones as gametes of one sex alone. small as 2.6 mm. As the smallest hermaphrodite was According to the oocyte size-frequency data (Fig. 5.8 mm in diameter (most were larger), the data suggest II), B. annulata spawned in November 1977 and again that, as in E. prolifera, female gonads develop first in in April 1978. There was evidence of residual sperm B. annulata, followed by the development in many in December 1977 and immature male gonads appeared animals of male gonads. The occurrence of large in late March and again in June (Table I) . This agrees anemones with only female gonads suggests that not all with the findings of McMurrich (1889), who found that large females necessarily become hermaphroditic. none of the population of Bartholomea (=Aiptasia) The presence of males is difficult to explain under the annulata in Bermuda was mature in July. He did find scheme of gynodioecious hermaphroditism. Ifhermaph­ ". .. immature ova partly embedded in mesoglea, partly rodites of B. annulata lacked gametogenic synchrony, still forming elements of the endoderm layer." How-

FIG. 12. Scanning electron micrograph of oocyte of Phymanthus crucifer, May 18, 1978. Arrow indicates mesoglea . Scale = 40 urn. FIG. 13. Scanning electron micrograph of blastula-stage embryo of P. crucifer, May 18, 1978. Scale = 100 urn. FIG. 14. Planula larva of P. crucifer, May 18, 1978. Arrow indicates mouth. Scale = 100 u.m. FIG . 15. Six-tentacle stage larva of P. crucifer, May 18, 1978. Arrow indicates primordial tentacle. Scale = 150 urn, FIG. 16. Twelve-tentacle stage juvenile of P. crucifer, May 18, 1978. Arrow indicates mouth. Scale = 1.0 mm. FIG . 17. Twenty-four-tentacle stage juvenile of P. crucifer, May 18, 1978. Scale = 2.5 nun. JENNISON 1715 1716 CAN . J. ZOOL . VOL . 59 . 1981 ever, Duerden (1902) found gonads in several of large oocytes in fall 1977 collections and the B. annulata in Puerto Rico during the summer, and continual maturity of males (Table I) support this idea. R. Seaton (personal communication) reports finding McMurrich (1889) reported that all of his specimens fertile specimens in Boca Raton , Florida, in July 1975. of P. crucifer from Bermuda were male s, and all were It is likely that populations spawn variously according to mature in July. Duerden (1898) discussed primarily the local influences , and there is no evidence that spawning external morphology of Jamaican anemones, but he occurs at the same time every year within a population. did find P . crucifer " . .. in great abundance, and of all Further, Cary (19 I I) reported asexual reproduction by sizes ..." in shallow water during summer. Further, he pedal laceration in this species, and evidently reproduc­ reported finding specimens only 1.1 cm in length, and tion occurs by this method year-round. Indeed, most of 0.6 ern in diameter (large specimens, by contrast, were the small infertile anemones collected during this stud y 6cm in length, 1.7cm across midcolumn, and 5.5cm were probably the result of recent pedal lacerations. across the expanded oral disc). I found brooded juve­ Lastly, although one female B. annulata from the niles in May which were only slightly smaller than those February 14, 1978 , collection yielded two planulae free-living specimens of Duerden; it is apparent that when sectioned, none of the other specimens examined Duerden found newly released juveniles. Lewis (1960a), showed any evidence of brooding. It is not known on the other hand , reported that P. crucifer relea sed whether brooding commonly occurs in other popula­ planulae in July in Barbados. These were 0 .24 mm in tions of B. annulata, or whether this was an unusual diameter, had tufts of cilia evenly distributed around the case. edge of the body, and were pale yellow in color with Another reproductive type of Chia's (1976) scheme is scattered darker pigment. Settlement took place within a oviparous ~ pelagic ~ lecithotrophic; I suggest calling day or two ofrelease. This suggests that the reproductive this type oviparous ~ planktonic ~ lecithotrophic. In pattern ofP. crucifer may vary with local conditions, or this case, eggs are large (300-1100 urn in diameter), among populations, as in Cereu s pedunculatus (Rossi and relatively few are produced; gametes are spawned 1975). and fertilization is external; dispersal is thus an adv an­ Brooding in sea anemones has received much atten­ tage , although planktonic life is presumably short; and, tion recently. Loseva (197 I) found that Tealia crassi­ due to lecithotrophy, larval death by starv ation is cornis is viviparous in the Barents Sea, and Spaulding unlikely. Clearly, Condylactis gigant ea fits this repro­ (1971) described the embryonic and larval development ductive type quite well. This species is dioecious with a of Cribrinopsis williamsi, which broods its planulae I: I sex ratio, and there was no evidence of brooding. internally in Puget Sound. Dunn (1975) studied external Spawning is roughly synchronous within the local brooding in Epiactis prolifera in California, and Rossi population, occuning in late May in 1978, although (1975 ) found brooding in Cereus pedunculatus in some low level of gamete release may very well occur Europe. Siebert and Spaulding (1976) reported brood ing throughout the year (Fig. 9; Table 1). in a second species ofCribrinopsis , C .fernaldi, in Puget Of 71 specimens of C. gigantea observed, 66 had Sound. Riemann-Zurneck (19760 , 1976b) discussed gonads, and , although small anemones with gonads "giant larvae" brooded by Actinostola spetsbergensis were occasionally found, in general those animals with from Canada and the Arctic, and Dunn (1978) described gonads were considerably larger than those few without. brooding in Anthopleura handi from Malaysia. This suggests that, although some animals may mature Certainly the genus of brooding anemone that has early (or "shrink" due to lack of sufficient food) , all sparked the most interest is Actinia. Long known to be animals larger than 4 .5 g in body weight are potentially viviparous (Gosse 1860; Stephenson 1928 , 1935), reproductive. Only a single hermaphrodite was found, Actinia equina was studied in detail by Chia and Rostron but it serves to indicate the relative sexual plasticity of (1970 ). In order to account for the presence of brooded actinians as a group. juveniles in males and in anemones without gonads, The third anemone examined here, Phymanthus Chia and Rostron proposed that adults must release their crucifer is in many respects the most interesting. gametes, then swallow and brood the resultant planulae, According to the scheme of Chia (1976), P. crucifer is often altruistically. This theory has sparked great con­ viviparous. Fertilization is apparently internal, embryos troversy. Cain (1974) examined the column color of 43 develop in the coelenteron, and juveniles are brooded adults and their broods, and found that the column colors for some time prior to release. The oocyte size­ of brood s always matched those of the adults which frequency data (Fig. 10) indicate that spawning occurred housed them. He suggested self-fertilization as a mech­ between February and May of 1978, although the anism to explain this phenomenon. Conflicting with presence of brooded planulae and juveniles as early as Cain's findings are those ofRostron and Rostron ( 1978), December 1977 suggests that reproduction may be who studied 480 brooding adults, and reported that there occurring at some low level year-round. The occurrence were 228 incubated young of a different color. JENNISON 1717

Ottaway and Kirby (1975), studying the congeneric alone can explain the presence of brooded embryos in Actinia tenebrosa in Australia, proposed that adult male anemones. As I found a blastula-stage embryo, a anemones might in fact "select" their young from a planula larva, and juveniles all in the same adult female free-living larval pool. However, there is little evidence anemone, there is no need to hypothesize the spawning to support this hypothesis, as Chia and Rostron (1970) of gametes, followed by the swallowing and brooding of found that adults of A. equina would not accept any the resultant planulae, selective or otherwise. It seems postplanular juveniles placed on their oral disc, and likely that the adult which produces specific offspring Carter and Thorp (1979) could not get adults to take their broods those young anemones until they are ready to be own brooded young back after they had been flushed released . Summarizing all brooding information in out. Further, Rostron and Rostron no longer believe in P. crucifer, I found preplanular larvae or planulae in the hypothesis of planula transfer (personal communica­ three females and one male in December, February, and tion). May. I found juveniles in six females, one male, and one Ottaway (1979) has since suggested that adult ane­ anemone with no gonads in December, March, and mones may undergo a sex change while brooding. If May. Clearly further investigation of brooding in complete spawning and gonad resorption were a cyclic Phymanthus, involving a large number of animals such event, and the animals were reproductively plastic with as Rostron and Rostron (1978) examined in Actinia, is respect to sex (or else were sequential hermaphrodites), required before a conclusive statement can be made. then brooded planulae and juveniles from a former However, based on the presence of planulae in one male female adult might well be found in that same animal in February, and juveniles in one male and one anemone following complete gonad resorption, or even as male with no gonads in May, I suggest that the sex-change gonads developed during the following breeding cycle . hypothesis of Ottaway (1979) is the simplest way to Several recent papers have rejected the ideas of brood explain these otherwise anomalous findings . selection and sex change in Actinia (Black and Johnson Evidence presented thus far indicates that the three 1979; Carter and Thorp 1979; Gashout and Ormond species of anemones in this study reproduce in quite 1979). Carter and Thorp (1979) assert that, as females , different ways. The evolution of these different repro­ males and animals without gonads all may brood young , ductive modes can perhaps be explained by an examina­ there is no apparent correlation between sex and tion of the morphology, prey , and habitat-type of each brooding capability. They found all stages of young species. Sebens (1976) described three habitat types for anemones in the coelenterons of adults, ranging from sea anemones from coral reefs in Panama: surface, hole, small motile balls of cells, thought to be preplanular and sand pocket. He classified Condylactis gigantea as a larval forms, up through ciliated planulae to juveniles hole dweller, and Bartholomea annulata and Phyman­ with as many as 48 tentacles. Further, electrophoretic thus crucifer as sand-pocket forms. Hole dwellers are evidence for both A. equina (Carter and Thorp 1979) and characterized morphologically by an elongate column A. tenebrosa (Ottaway and Kirby 1975; Black and and long, slender tentacles; sand-pocket forms may be Johnson 1979) seems to rule out sexual reproduction in similar to hole dwellers, or they may be more squat, with the absence of brood selection. large, flat oral discs and stubby tentacles, thereby Both Carter and Thorp (1979) and Gashout and resembling anemones in the third habitat type, the Ormond (1979) propose that Actinia equina reproduces surface dwellers. Sebens also distinguished these three parthenogenetically. Black and Johnson (1979) modify species by prey type: C. gigantea consumes macro­ this position by suggesting that, as in Cereus peduncu­ scopic prey, depending on wave action to supply food, latus (Rossi 1975), there are distinct sexual races of whereas P. crucifer is a planktivore. Bartholomea A. tenebrosa, some developing parthenogenetically annulata, with its long delicate tentacles can feed on whereas others are sexual. Finally, Carter and Thorp zooplankton, but large individuals can also take macro­ (1979) propose a fertilization-selection-cloning hy­ scopic prey. pothesis, whereby only those juveniles genetically In Key West, Condylactis gigantea occurs solitarily identical to the adult are allowed to survive; these then under ledges or adheres to hard surfaces in Thalassia reproduce asexually. Carter and Thorp state however flats, and grows quite large. That this species can take that ". . . this hypothesis must . . . be considered macroscopic prey is well established (Bursey and tentative and remains to be proven experimentally." Guanciale 1977; author's observations). Reproduction It is difficult to postulate such a complex scheme of in C. gigantea was found to be oviparous ~ planktonic reproduction in Phymanthus crucifer. No electro­ ~ lecithotrophic. There was no evidence of asexual phoretic data were taken during this study, so partheno­ reproduction. Indeed, Chia (1976) gave three criteria genesis certainly cannot be ruled out. However, the which pertain to species that undergo asexual reproduc­ occurrence of a I: 1 sex ratio argues against this mode of tion: they occur in a stressful environment, achieve reproduction. Further, it is unclear how parthenogenesis small adult body size, and have a limited food supply. 1718 CAN. J. ZOOL. VOL. 59, 1981

None of these applies to C. gigantea. Because of the Hand, Susan Jennison, Patricia Linley, Joseph Mur­ large size and prey requirements of this species it is doch, Julianne Piraino, Mary Rice, Robert Seaton, Tom advantageous that individuals be widely dispersed. As Smoyer, Frank Stanton, and Phyllis Thompson. suitable habitat is also widely dispersed (i.e. rock or coral "islands" in a "sea" of sand), a reproductive BLACK, R., and M . S . JOHNSON. 1979 . Asexual viviparity and strategy involving brooding would be of no advantage; population genetics of Actinia tenebrosa. Mar. BioI. 53 : rather, the production of planktonic larvae with resultant 27-31. dispersal is more advantageous, as larval settlement may BURSEY, C. R., and J. M. GUANCIALE. 1977. Feeding thus occur over a wide area . behavior of the sea anemone Condylactis gigantea. Compo Bartholomea annulata may be found under ledges, Biochem. Physiol. 57A: 115-117. CAIN, A . J. 1974 . Breeding system ofa sessile animal. Nature but it is more commonly found in sandy areas and among (London), 247 : 289-290. Halimeda forests. Although individuals may grow quite CARLGREN, O . 1951. Actiniaria from North America. Ark . large (up to 2.3 ern in fixed basal diameter in this study), Zool. 3(30): 373-390. the average anemone was less than half that size. This CARTER, M . A., and C. H. THORP. 1979 . The reproduction of fits the pattern of asexual reproduction discussed above Actinic equina L. var. mesembryanthemum. J . Mar. Biol . (Chia 1976). Further, pedal laceration may help this Assoc. U.K . 59 : 989-1001. species maintain space in the face of shifting sands, CARY, L. 1911. Pedal laceration in actinians. BioI. Bull. siltation , and growth or decline of the Halimeda cover. (Woods Hole, Mass.) , 20: 81-108. The observed gynodioecious hermaphroditism is also CHIA, F.-S. 1976. Sea anemone reproduction: patterns and advantageous, in that large, well-established individuals adaptive radiations. /n Coelenterate ecology and behavior. Edited by G . O . Mackie. Plenum Press, New York . pp . will always have female gonads, thus potentially in­ 261-270. creasing the fecundity of these individuals. This would CHIA, F. -S., and M . A . ROSTRON. 1970. Some aspects of the be especially significant if some populations of reproductive biology of Actinia equina (Cnidaria: An­ B. annulata do in fact brood their young. thozoa). J. Mar. BioI. Assoc. U.K . 50 : 253-264. Phymanthus crucifer occurs primarily in sandy areas, DUERDEN, J. E. 1898. The actiniaria around Jamaica. Inst. adhering to subsurface hard strata and stretching up, Jamaica J. 2(5): 449-465. often several centimetres, to spread its broad, flat oral --- 1902 . Report on the actinians of Puerto Rico . Bull . disc on the surface of the sand. Where the sand is U.S. Fish Comm. 20(2): 321-374. shallow or in surge-scoured areas P. crucifer may be DUNN, D. F. 1975. Reproduction of the externally brooding sea anemone Epiactis prolifera Verrill, 1869. Biol, Bull. found on the surface of flat rocks; indeed, it resembles a (Woods Hole, Mass.), 148: 199-218. surface dwelling type in morphology (Sebens 1976). --- 1978 . Anthopleura handi n. sp . (Coelenterata, Actin­ Phymanthus crucifer was found to brood its young iaria) , an internally brooding, intertidal sea anemone from internally, releasing them as juveniles capable of im­ Malaysia. Wasmann J . BioI. 35: 54-64. mediate settlement and feeding. Under some conditions, GASHOUT, S. E ., and R. F. G . ORMOND . 1979 . Evidence for P. crucifer may release planulae (Lewis 1960a). Asex­ parthenogenetic reproduction in the sea anemone Actinia ual reproduction does not occur. Although the sandy equina L. J. Mar. BioI. Assoc. U.K. 59 : 975-987. habitat may be stressful, a factor which may help to GOSSE, P. H. 1860 . A history of the British sea anemones and explain pedal laceration in Bartholomea annulata, it is corals. Van Voorst, London. important to note that P. crucifer is much better suited HERRNKIND, W . , G . STANTON , and E . CONKLIN. 1976. Initial characterization of the commensal complex associated with morphologically to withstand shifting sands than B. the anemone, Lebrunia danae, at Grand Bahama. Bull. .. annulata. Phymanthus is thick-walled and has a tough Mar. Sci. 26(1): 65-71. ectoderm, whereas Bartholomea is gossamer thin. JENNISON, B. L. 1979 . Gametogenesis and reproductive Further, Phymanthus gets fairly large (up to 6 ern in cycles in the sea anemone Anthopleura elegantissima column length, and 5.5 em across the expanded oral disc (Brandt, 1835). Can. J. Zool. 57: 403-411. according to Duerden (1898)); this would inhibit pedal LEWIS, J. B. 1960a. The fauna of rocky shores of Barbados, laceration according to Chia (1976). In any event, West Indies. Can. J. Zool. 38: 391-435. brooding is probably more effective than pedal lacera­ --- 1960b. The coral reefs and coral communities of tion as a means of maintaining habitat space for future Barbados, West Indies. Can. J. Zool. 38: 1133-1145. generations, as it provides for sexual recombination as LIMBAUGH, C., H. PEDERSON, and F. A. CHACE, JR. 1961. Shrimps that clean fishes. Bull. Mar. Sci. 11(2): 237 -257. well. LoSEVA, L. M. 1971. Observation on the oogenesis in Acknowledgements actiniae. 2. Oogenesis in Tealia crassicornis (Muell.), Metridium senile (L.) and Protanthea simplex (Carlgr.). The author thanks the following for their contribu­ Vestn. Leningr. Univ. BioI. 9: 22-29. tions to this project: Teriann Asami-Oki, Ann Bucklin, MAHNKEN, C . 1972 . Observations of cleaner shrimps of the Daphne Dunn, Kevin Eckelbarger, Linda Eyster, Cadet genus Periclimenes. In Results of the Tektite program: JENNISON 1719

ecology of coral reef fishes . Edited by B. Collette and S. spetsbergensis (Anthozoa: Actiniaria). Helgo!. Wiss . Earle. Nat. Hist. Mus. Los Angeles Cty . Sci. Bull. 14: Meeresunters. 28: 239-249. 71-83. ROSSI, L. 1975. Sexual races in Cereus pedunculatus (Boad .). McMuRRICH,1. P. 1889. A contribution to the actinology of Pubb!. Stn. Zoo!. Napoli (Supp!.), 39: 462-470. the Bermudas. Proc . Acad . Sci. Philadelphia, 1889: ROSTRON , M. A., and 1. ROSTRON. 1978. Fecundity and 102-126. reproductive ecology of a natural population of Actinia NYHOLM, K.-G. 1943. Zur Entwicklung und Entwicklungs­ equina L. (Cnidaria: Anthozoa). J. Exp . Mar . BioI. Eco!. biologie der Ceriantharien und Aktinien. Zoo!. Bidr. 33: 251-259. Uppsala, 22: 87-248. SEBENS, K. P. 1976. The ecology ofCaribbean sea anemones OTTAWAY,1. R. 1979. Population ecology of the intertidal in Panama: utilization of space on a coral reef. In Coelenter­ anemone Actinia tenebrosa , II. Geographical distribution, ate ecology and behavior. Edited by G . O. Mackie. Plenum synonymy, reproductive cycle and fecundity. Aust. J. Zoo!. Press, New York . pp . 67-77. 27: 273-290. SIEBERT, A. E., and J. G. SPAULDING. 1976. The taxonomy, OTTAWAY ,1. R., and G . C. KIRBY . 1975. Genetic relation­ development and brooding behavior of the anemone, ships between brooding and brooded Actinia tenebrosa . Cribrinopsis fernaldi sp. nov . Bio!. Bull. (Woods Hole, Nature (London), 255: 221-223. Mass .), 150: 128-138. RIEMANN-ZORNECK, K. 1976a. A new type of larval devel­ SPAULDING , J. G . 1971. The development and brooding opment in the Actiniaria: giant larvae. Morphological and behavior of the anemone Cribrinopsis williamsi Carlgren. ecological aspects of larval development in Actinostola Am. Zoo!. 11: 71. spetsbergensis, In Coelenterate ecology and behavior. STEPHENSON, T . A. 1928. The British sea anemones. Vo!. I. Edited by G. O. Mackie. Plenum Press, New York. pp . The Ray Society, London. 355-364. --- 1935. The British sea anemones. Vo!. II. The Ray --- I 976b. Reproductive biology, oogenesis and early Society, London. development in the brood- caring sea anemone Actinostola