M. A. Coroth áH. R. Lasker
Larval paternity and male reproductive success of a broadcast-spawning gorgonian, Plexaura kuna
Received: 1 April 1997 / Accepted: 11 February 1998
Abstract Male reproductive success of the broadcast- spawning gorgonian, Plexaura kuna Lasker, Kim and Coroth, 1996, was measured in June 1994 and June and July 1995 at two sites in the San Blas Islands, Panama in order to determine the environmental and biotic factors aecting individual reproductive success. Developing embryos were collected in the ®eld during natural spawning events, and paternity determined using ran- domly ampli®ed polymorphic DNA markers. Analyses of F1 progeny from de®ned laboratory matings estab- lished that the markers were inherited in Mendelian fashion, and allowed the determination of the zygosities of the markers.P. kunais clonal, but male reproductive success was not strictly proportional to clone size.
Proximity to females appeared to have a greater eect on male reproductive success than clone size, and on both reefs the most successful male clone was the one closest to the spawning female clone. Current direction and transport of gametes by eddies explained variation in paternity assignments between nights. Clonal propa- gation allows clones to grow and spread toward each other, and may enhance male reproductive success.
Introduction
A wide variety of plants and animals reproduce by re- leasing one or both gametes into the environment and relying on air and water currents, or on mutualists, for gamete transport. The ecological and evolutionary consequences of this reproductive strategy are well studied for plants (e.g., Willson 1979; Stephenson and Bertin 1983; Willson and Burley 1983), but they have
only recently been addressed for benthic marine inver- tebrates that broadcast spawn. Starting with Pennington (1985), there is increasing evidence that fertilization processes may be an important source of variation in reproductive success for some benthic species, and sev- eral authors (Yund 1990; Levitan 1993, 1995, 1996;
Yund and McCartney 1994; Levitan and Petersen 1995;
Podolosky and Strathmann 1996) suggest that variation in fertilization may be important in the evolution of broadcast-spawning reproductive systems. Studies demonstrating variance in fertilization rates during natural spawning events support this contention (Pe- tersen 1991; Babcock et al. 1992; Babcock and Mundy 1992; Brazeau and Lasker 1992; Oliver and Babcock 1992; Petersen et al. 1992; Sewell and Levitan 1992;
Lasker et al. 1996; Coma and Lasker 1997a). Variation in reproductive success among broadcast-spawning taxa has not been partitioned between males and females, however, nor between environmental and phenotypic traits. In the present study we examine male reproduc- tive success in the broadcast-spawning gorgonianPlex- aura kuna.
Variance in male reproductive success is highly likely in broadcast spawning because fertilization rates are extremely sensitive to sperm density (e.g., Vogel et al.
1982; Pennington 1985; Levitan et al. 1991; Benzie and Dixon 1994), and because sperm become diluted and density drops rapidly with distance from a spawning male (Pennington 1985; Yung 1990; Grosberg 1991;
Levitan 1991; Brazeau and Lasker 1992; Levitan et al.
1992; Yund and McCartney 1994; Levitan and Young 1995; Coma and Lasker 1997a, b). The eects of sperm dilution are suciently great that fertilization rates of a number of species are sperm limited (see Levitan 1995;
Lasker et al. 1996).
The factors controlling sperm density include current speed, direction, and turbulence and the density and placement of males. Flow has particularly complex ef- fects on fertilization, and water currents can both en- hance and limit reproductive success. Small scale current eects can bring gametes together and mix them (Lasker
Communicated by J.P. Grassle, New Brunswick M.A. Coroth (&)áH.R. Lasker
Department of Biological Sciences, University at Bualo, P.O. Box 1300, Bualo, New York 14260-1300, USA
and Kapela 1997), while turbulent ¯ow quickly dilutes the gametes and, in some instances, makes fertilization improbable (Denny 1988; Denny and Shibata 1989).
Sperm density is also aected by the density and place- ment of colonies, which in turn are aected by many factors. Among clonal taxa, vegetative propagation has marked eects on the density of colonies, the size of clones of a particular genotype, and the dispersion of colonies (Lasker 1990; Coroth and Lasker 1998) all of which should aect the overall fertilization success of the genotype. Local micro-environment (Mead and Denny 1995) and gamete traits should also aect fertilization rates (Levitan 1993, 1995; Lasker unpublished data).
Variation among these dierent physical and biotic parameters leads to variation in sperm density and thus to variation in reproductive success over time, among positions on a reef and among genotypes. In situ ex- periments have detected dierential male success in marine taxa such as tunicates and bryozoans and have related reproductive success to male reproductive out- put, male±female distance, and current direction (Grosberg 1991; Yund and McCartney 1994; Yund 1995; Levitan 1996; McCartney 1997).
We examined patterns of male reproductive success in the broadcast-spawning gorgonian, Plexaura kuna, on reefs in the San Blas Islands, PanamaÂ. We then related those patterns to potential sources of variation in fer- tilization success (e.g., local current regimes and patterns of clonal growth) to determine the eect of these vari- ables on individual reproductive success.P. kunais well suited for an analysis of male success as it exhibits a variety of traits that facilitate obtaining samples and determining paternity. Firstly, paternity analysis of P. kunaplanulae is simpli®ed by the fact thatP. kunais clonal (Lasker 1984, 1990; Coroth et al. 1992; Coroth and Lasker 1998). At most sitesP. kunaabundance is a consequence of vegetative propagation and not recruit- ment of planulae (Coroth and Lasker 1998). As a consequence, there are relatively few genotypes on any reef, and reefs with P. kuna populations of 50 to 1700
colonies have 3 to 19 genotypes (Coroth and Lasker 1998). This limited number of potential parents facili- tates paternity assessment of ®eld-collected larvae. Sec- ond, P. kuna colonies are gonochoric and spawn predictably for four to six nights starting 3 d after the full moon during each of two to four summer months (PanamaÂ, Brazeau and Lasker 1989; Florida Keys, HRL personal observations). Eggs can readily be collected by divers (Lasker et al. 1996), and fertilized eggs can be reared to the planula stage (Lasker and Kim 1996) at which time they are large enough for genetic analysis.
We used the polymerase chain reaction (PCR) to gen- erate randomly ampli®ed polymorphic DNA (RAPD) markers for individual genotypes of the gorgonian P.
kuna. This technique is particularly well-suited for dis- tinguishing larvae as the PCR overcomes the problem of limited quantities of DNA in larvae and RAPDs can be used to resolve individual genotypes (Hadrys et al.
1992). Analysis of the F1progeny from de®ned matings allowed us to determine the zygosities of the parents at each locus and overcome one of the major disadvantages of RAPDs, namely that RAPD markers are dominant, making it impossible to distinguish between the homo- zygotic and heterozygotic state of an allele.
Materials and methods
Study species and site
Male reproductive success in Plexaura kuna Lasker, Kim and Coroth, 1996, was examined on Korbiski Reef and Tiantupo Reef in the San Blas Islands, Panama (Korbiski-1SE and Tiantupo-1E, Fig. 1 in Robertson 1987) in June 1994 and June and July 1995.
The Korbiski population contains ®ve female clones and four male clones (i.e., nine potential parents), and the Tiantupo population contains one female clone and two male clones (i.e., three potential parents) (Coroth and Lasker 1998). Clone sizes and position (Table 1; Figs. 1, 2) were previously determined using DNA mini- satellite ®ngerprinting to distinguish clones (Coroth et al. 1992) and a strati®ed random sampling regime to determine clone dis- tribution on the reef (Coroth and Lasker 1998). Reefs with low genotypic diversity, such as Korbiski and Tiantupo, are common in
Table 1 Plexaura kuna. Popu- lation structure at Korbiski and Tiantupo Reefs, San Blas Is- lands, PanamaÂ. One out of ev- ery 20 colonies was sampled.
Clones not identi®ed in that sample and discovered subse- quently are marked as <20 colonies in the table. Propor- tions of males at Korbiski based on number of colonies actually sampled. This more accurately characterizes numbers of co- lonies than the extrapolation from the population sample
Site,
clone Sex Estimated size Percent
population Percent identi®ed male colonies Korbiski
KA F 740 45
KB F 80 5
KC F 800 48
KD M <20 <1 40
KE F 20 1
KF M <20 <1 20
KG F 20 1
KH M 20 1 20
K809a M <20 <1 20
Tiantupo
TiA M 140 21 54
TiB F 400 61
TiC M 120 18 46
a K809 was a single small colony transplanted to Korbiski in 1982
the San Blas Islands (Coroth and Lasker 1998). Aquarium studies were conducted at the San Blas ®eld station of the Smithsonian Tropical Research Institute (STRI) in PanamaÂ.
RAPD analysis
DNA was extracted from adults and larvae and ampli®ed by the PCR as previously described (Coroth and Mulawka 1995). Re- action products were separated through a 1% Synergel (Diversi®ed Biotech) and 0.7% agarose gel by electrophoresis (Coroth and Mulawka 1995). We screened all potential parents with a total of 120 primers (Operon Technologies Inc. and University of British Columbia Nucleic Acid-Protein Service Unit). In gorgonians each primer typically ampli®es 5 to 15 DNA reaction products that are visualized on the agarose gel. We surveyed each primer and parent for the presence of a product, i.e., marker or polymorphism, that was unique to that parental genotype. Reproducibility of each polymorphism was veri®ed by repeating the ampli®cations on several dierent days with each primer and each individual.
Breeding experiments
De®ned matings were used to establish that the selected clone- speci®c markers were inherited in a Mendelian fashion. Single male and female branches, collected from the ®eld prior to spawning, were paired in aquaria at the STRI ®eld station and allowed to spawn. Developing embryos were removed from the aquaria fol- lowing spawning and reared for 5 to 10 d to competent planula larvae. Fourteen de®ned matings yielded thousands of planulae.
Individual larvae were placed in 0.5-ml tubes and frozen in liquid nitrogen. DNA from each of the parents and a subset of the planulae was ampli®ed with a series of primers, each of which identi®ed a polymorphism unique to one of the two parents used in that particular mating. We used the frequency of the marker among the planulae to test for Mendelian inheritance and to determine whether the parent was homozygotic or heterozygotic for the marker. The zygosity of the parents and pattern of inheritance at each locus were determined from the ratio of progeny with and without the band at that locus. A 1:1 pattern follows the expected Mendelian behavior for segregation of alleles when one parent is heterozygous for the marker (presence of band) and one does not contain the marker (absence of the band). A 1:0 pattern is expected when the parent is homozygous for the marker. Knowing whether the genotype-speci®c markers are homozygotic or heterozygotic allows parental genotype identi®cation using one or several markers. This reduced the number of loci necessary for paternity analysis and eliminated the need for statistical comparisons of band sharing (e.g., Levitan and Grosberg 1993).
Field collections
Eggs and embryos were collected by divers during spawning events at Korbiski on 26 June 1994, 18 June and 16 July 1995 and at Tiantupo on 27 July 1994. Samples were collected randomly along a 10-m transect downstream of the spawning female clone (Figs. 1, 2). Four points on the transect were randomly selected, and two divers rotated among those four positions at 5-min intervals.
Samples were collected in 60-ml plastic syringes and transported back to the ®eld station where they were placed in 130-ml poly- styrene specimen cups or 1-liter polyethylene jars which were then suspended in the water overnight. The following morning, devel- oping embryos were transferred to freshly collected sea water and reared for 5 to 10 d until they became competent planulae. Pla- nulae used for the paternity analyses were randomly selected from the collections, frozen individually in liquid nitrogen, and kept at )70°C until analyzed. DNA was extracted and ampli®ed as before using primers that had been identi®ed to yield genotype-speci®c Fig. 1 Plexaura kuna. Map of
Korbiski Reef. Colony sampled for DNA ®ngerprint analysis (d);
dierent clones (dashed lines); site of S4 current meter (*).Upper left corner: current speed and direc- tion on each night of collection (A26 June 1994;B18 June 1995;
C16 July 1995). Eggs and em- bryos were collected along 10-m transects 1, 2, and 3 (KCfemale clone with multiple colonies;KD, KF,KH,K809male clones)
Fig. 2 Plexaura kuna. Map of Tiantupo Reef. Colony sampled for DNA ®ngerprint analysis (d); dierent clones (dashed lines). Upper left corner: current speed and direction on 27 July 1995. Eggs and embryos were collected along 10-m transect 1 (TiBfemale clone with multiple colonies;TiA,TiCmale clones)
markers for each potential parent. The planula DNA was se- quentially screened with the primers until a marker was found in common with a potential parent. Parentage was assigned to that adult and no further screening was done.
During the collections at Korbiski an InterOcean S4 current meter recorded current speed and direction at 0.5-s intervals. The meter was 1.5 m above the substratum (about 0.5 m above the largestPlexaura kunacolonies) and close to the female clone (KC) whose eggs and embryos were being collected (Fig. 1). The meter was upstream and to one side of the transect so that the divers would not aect the measurements. Current direction at the time of collections was also assessed by diver observations of movement of eggs and rhodamine dye release at the transect positions and used to select the sampling location.
Results
Breeding experiments
We identi®ed one to six polymorphic markers for each adult genotype. Fourteen de®ned matings were con- ducted. In each mating each colony contained at least one unique polymorphism, and the two adults used in any single cross contained two to eight polymorphisms.
The progeny of each cross were scored for inheritance of each of the polymorphic markers. Summed across the matings, 34 dierent markers were tested for Mendelian inheritance: 8 markers were homozygotic, i.e., all of the ospring contained the marker; 26 markers were heterozygotic and inheritance in all cases followed the expected 1:1 ratio of ospring with and without the marker (v2tests, all casesp> 0.10).
Field collections
At Korbiski we collected adjacent to the dominant fe- male clone, KC (Fig. 1), the position of the transect varying with current direction on dierent sampling days, and in one case, over a single night (Table 2). On all nights the KC clone produced all the planulae to which we could assign maternity (85 to 100%, Table 3).
Parental and larval genotypes at each scored locus are summarized in Table 3. Information on all individual genotypes at all assayed loci is available from the au- thors on request. Paternity of 87% of the planulae was determined from analysis of the RAPD data. KD and KH were the only identi®ed sires (Table 3). Parentage was not random between KD and KH. KD was the closer and larger of the two clones, and it sired a total of 79.5 versus 8% of the planulae for KH. The overall proportion of larvae sired by KD was signi®cantly greater than the 2:1 ratio of colonies (v2 18.48, df 1, p<0.001), and the ratio diered signi®cantly from the 2:1 expectation on two of the three nights (v2 = 0.2, df= 1, p> 0.5; v2 8.2, df 1, p< 0.01; v2 15.0, df 1, p< 0.005). Parentage varied among days (v2 15.7,df 3,p< 0.005) with fewer KD planulae on 26 June 1994 (56%) than 18 June 1995 or 16 July 1995 (Table 3).
Some of the adults only contained heterozygous markers, which raised the possibility of producing planula that did not contain any of a parent's marker alleles. The probability of a parent producing an ``un-
Table 2 Summary of sampling dates and current conditions at Korbiski and Tiantupo Reefs, San Blas Islands, PanamaÂ
Reef Date Collection
transect Current
Mean Mean velocity Comments direction (cm s)1)
Korbiski 26 June 1994 1 326° 3.2 1±3 min periods
during which current oscillated over 360°range
Korbiski 18 June 1995 1, 2, 3 112° 3.5 Continuously
oscillating between NE and S
Korbiski 16 July 1995 1 316° 3.5 Single reversal to
SW and back
Tiantupo 27 July 1994 1 90° 2±5 Steady slow ¯ow
throughout sampling period
Table 3 Plexaura kuna. Pater- nity and maternity of larvae collected at Korbiski and Tian- tupo Reefs, San Blas Islands, PanamaÂ
Site,
date Number of
planulae Percent fathered by: Percent mothered by:
Korbiski Reef KD KH Indeterminate KC Indeterminate
26 June 1994 27 56 22 22 85 15
18 June 1995 24 84 4 12 100 0
16 July 1995 32 94 0 6 97 3
Tiantupo Reef TiA TiC Indeterminate TiB Indeterminate
27 July 1994 30 87 0 13 100 0
assignable planula'' was dependent on the number and zygosities of markers used to determine parentage (Table 4). In theory, with a homozygous marker (as in the maternal clone KC), the probability of correctly assigning parentage is 100%. In practice, however, lar-
vae required multiple runs to obtain successful ampli®- cation product, and we would sometimes run out of sample before all of the potential markers were suc- cessfully ampli®ed. The maximum number of markers used for each parent and the probability of successfully identifying one of its ospring are listed in Table 4.
KD and KH were not excluded from paternity in the analysis of the indeterminate planulae (Table 5), i.e., the indeterminate planulae did not contain polymorphisms unique to KD (up to six loci screened) or KH (up to three loci screened). KD and KH were heterozygous at all of the loci that we used as markers, and we were not able to screen all of the indeterminate planulae for large numbers of markers. Some of their progeny could be expected not to have any of the markers at the tested loci (see Table 5), and it is probable that KD and KH fathered a portion of the planulae with indeterminate parentage. No novel bands were observed in the RAPDs of the indeterminate planulae.
Current ¯ow directly over KC and at the current meter was primarily to the northwest on 26 June 1994, but the current was ¯owing in the opposite direction immediately east of the collections, in the slightly deeper channel between Korbiski and Tiantupo. The current plots from the S4 current meter also exhibited ®ve cur- rent reversals at approximately 10-min intervals (Fig. 3).
Those reversals lasted 1 to 3 min. The S4 current meter was positioned near the boundary between the north and south ¯owing currents, and the ¯ow reversals may
Table 4 Plexaura kuna. Prob- ability of assigning parentage to
®eld collected larvae based on the number of polymorphic markers per parent
Parent Probability of detection
(using all primers) Number of homozygous
markers available Number of heterozygous markers available
KC 1 1 4
KD 0.985 0 6
KH 0.875 0 3
TiA 0.75 0 2
TiB 1 1 1
TiC 0.5 0 1
Table 5 Plexaura kuna. Likelihood of parentage by dierent clones for planulae of indeterminate parentage. Probabilities (in par- entheses number of heterozygous markers screened) listed for each planula
Site,
date Planula Potential paternal clone
Korbiski KD KH
26 June 1994 2 0.061 (4) 0.25 (2)
11 0.061 (4) 0.25 (2)
30 0.5 (1) 0.5 (1)
49 0.061 (4) 0.25 (2)
50 0.061 (4) 0.25 (2)
52 0.061 (4) 0.5 (1)
18 June 1995 52 0.031 (5) 0.125 (3)
54 0.031 (5) 0.125 (3)
56 0.031 (5) 0.125 (3)
16 July 1995 27 0.015 (6) 0.125 (3)
30 0.015 (6) 0.125 (3)
Tiantupo TiA TiC
27 July 94 17 0.25 (2) 0.5 (1)
19 0.25 (2) 0.5 (1)
46 0.25 (2) 0.5 (1)
53 0.25 (2) 0.5 (1)
Fig. 3 Current ¯ow at Korbiski Reef on 26 June 1994. Values are 60-s averages of measure- ments made at 0.5-s intervals
correspond to eddies of 2 to 6 m size passing the current meter. Planulae sired by KH and indeterminate planulae were more common on 26 June 1994. Paternity was not determined for six planulae due to the heterozygosity of the markers and/or the lack of sucient DNA to screen more markers. Based on the numbers of markers ana- lyzed for each parent, it is likely that KH sired the ma- jority of the indeterminate planulae (Table 5). The probability that neither KD nor KH sired the indeter- minate planulae ranged from 0.015 to 0.25, depending on the number of markers tested.
Current ¯ow at the collection transect on 16 July 1995 was very similar to 26 June 1994, but there were no indications of eddy-like current structures (Table 2).
Thirty of the 32 planulae were sired by KD. The two planulae with indeterminate fathers were probably sired by KH (Table 5). Currents on 18 June 1995 were highly variable, and the collecting site was continuously ad- justed as the current oscillated (Table 2). Planulae were collected from all three transects. Again most planulae were sired by KD with 4% sired by KH and an addi- tional 12% most probably sired by KH (Tables 3, 5).
Collections at Tiantupo Reef were made downstream of spawning colonies in the female clone TiB (Fig. 2).
All of the planulae contained the marker for TiB. Pa- ternity could be assigned to 83% of the planulae, and all of those planulae were fathered by clone TiA, the male clone closest to TiB (Table 3). None of the planulae had the heterozygous marker for the more distant male clone, TiC.
Discussion
Individual male reproductive success among broadcast- spawning species is dependent on biotic factors con- trolling the numbers and density of gametes and their compatibility, and abiotic factors controlling the trans- port and dispersion of gametes. Factors likely to aect Plexaura kuna reproductive success include current speed and direction, the degree of turbulence, spatial distribution of the colonies (i.e., male±female distance), size of individuals and clones, synchrony of spawning, and gamete motility and longevity. Some factors such as current speed and turbulence may aect all individuals on a reef evenly, while other eects, such as current di- rection, can produce dierential success among colonies, but probably do not result in directional selection on phenotypic traits. There are, however, a suite of traits which can produce heritable dierences in mating suc- cess. These include settlement patterns that might aect male±female distances, size dierences that aect the numbers of gametes produced, the pattern of gamete release, which aects spawning synchrony among indi- viduals, and at the level of gamete interactions, gamete traits such as swimming speed and longevity, which will directly aect fertilization rates.
Genets among clonal species can dier enormously in size, which should have a great eect on the number of
gametes released and on the genet's reproductive suc- cess. McCartney (1997), in a study of reproductive suc- cess of a bryozoan, found that success was directly related to the number of male zooids. Although varia- tion in clone size is one of the most apparent dierences amongPlexaura kuna at Korbiski and Tiantupo, male reproductive success at the two sites was not strictly proportional to clone size (Table 1). On Korbiski Reef, clone KD fathered proportionately more of the larvae than would have been predicted by clone size alone on two of three nights. On the third night, 26 June 1994, the paternity distribution was closer to that predicted by clone size. On Tiantupo Reef the distribution of pater- nity was even more skewed. Clone TiC made up close to a ®fth of the population on that reef and accounted for nearly half of the males (Table 1), yet none of the o- spring collected downstream of the female clone, TiB, contained a TiC-speci®c marker.
On both Korbiski and Tiantupo, male reproductive success was best predicted by the spatial distribution of colonies, i.e., proximity to a compatible mate, and cur- rent direction. At both reefs, the most successful male clone was the one closest to the spawning female. On Korbiski the most successful male, clone KD, is found in two clusters, one adjacent to the female KC clone and one about 10 m distant (Fig. 1). The reproductive suc- cess of clone KH, a single colony located over 10 m from the large female clone that we sampled, was less than predicted based on its relative abundance among males in the population. The primary factor controlling pa- ternity at Tiantupo was also the initial placement of the clones (assuming settlement occurred near the center of each clone). At Tiantupo the successful male clone (TiA) was the closer of the two male clones and was positioned about 10 m upstream of the female clone, TiB. Dilution of sperm would be substantial over that distance (e.g., Coma and Lasker 1997b), but, as the male clone was large (approximately 140 colonies) and many colonies would have spawned, clone size (number of ramets) probably compensated for the distance between male and female. Furthermore, some of those colonies were within several meters of the closest female colonies (Fig. 2). The lone TiC colony in the immediate area was more distant than the closest TiB colonies, and the TiC colony was small (<50 cm). Most TiC colonies were too far away for their sperm to fertilize the sampled TiB eggs.
Position eects dominated male reproductive success forPlexaura kuna, but current movements also aected the observed reproductive success. For instance, varia- tion in current direction explains the higher reproductive success of clone KH seen on 26 June 1994 (Table 3). On most nights current ¯ow was such that sperm released by the KH colony would not have traveled over the female KC clone or would have only reached these female colonies after traversing 10 m. We believe that the greater success of KH in June 1994 was caused by eddies carrying eggs closer to KH and then back to the col- lection transect. On 26 June 1994, the night with the
greatest KH success, there was a south ¯owing current immediately adjacent to KH. That current, which is a common feature at the site, is probably a tidally induced
¯ow transporting water through an 8-m deep channel between Korbiski and Tiantupo Reefs. When the cur- rent runs to the south, it usually weakens and forms a large eddy on the downstream side of the channel where it intersects the longshore ¯ow found along the southern edge of Korbiski Reef (HRL, personal observations).
The eddy forms in the general vicinity of KH. The S4 current data from 26 June indicate that the interaction between the southward tidal current and the northward wind-driven ¯ow produced eddies of 2 to 6 m diameter (Fig. 3). The eddies appear in the current records as reversals in current direction, the duration of which relative to the net current ¯ow suggests they were 2 to 6 m in diameter. The eddies appear to have propagated to the south as the current reversal always started with a change to westward ¯ow. Southward propagating eddies may have carried eggs from KC past KH, and some of those eggs may then have been entrained in the north- ward ¯ow back over KC. On other occasions we have also documented eddy-like structures on Korbiski that could ``recycle'' eggs back onto the reef (Lasker and Kapela 1997).
Vegetative propagation, which is one of the key fea- tures of Plexaura kuna population ecology (Lasker 1990), interacts with male reproductive success via its eects on the size of genetic individuals and their spatial distribution. ForP. kuna, as well as other clonal species, the spread of clones via vegetative propagation may result in clones that are in closer proximity than they might otherwise be following the initial distribution of larvae. Vegetative propagation and the dispersion of ramets may have played a key role in the reproductive success of clone KD on Korbiski. Clone KD is found in two clusters, suggesting clonal spread (Fig. 1). If the more distant and larger cluster of colonies is the original cluster or at least oldest, then clonal spread has brought KD into proximity of a female clone (KC). We do not know with certainty where the clones on Tiantupo originally established themselves, but it is likely that the relative positions of TiA and TiC are a product of clonal propagation. Clonality also may have aected paternity among the TiB colonies that have become established near the center of the TiC clone, which is located ap- proximately 50 m west of TiA and TiB. Those colonies were too far from our transect to have been included in collections, and the number of eggs released by those colonies would have been small relative to the rest of the TiB clone. None-the-less, spread of TiB has provided an opportunity for TiC paternity that would not otherwise have existed. The importance of vegetative propagation in repositioning a clone is again suggested by the pres- ence of one small TiC colony near the Tiantupo female clone (Fig. 2). At the time of our study, sperm produced by that single TiC colony would make only a small contribution compared to that of the approximately 140 adjacent TiA colonies, but establishment of the TiC
colony suggests that clonal growth may eventually alter the pattern of paternity.
Although clone size and current direction had some eect on reproductive success, location, i.e., proximity to a compatible mate, appears to have been the major factor contributing toPlexaura kunamale reproductive success at Korbiski and Tiantupo. This pattern is similar to that found among plant species. Population size, plant density, clumping, and distribution patterns are important determinants of male reproductive success among plants (Handel 1983; Ellstrand and Marshall 1985; Levin 1988). As in P. kuna, proximity to mates plays a key role in plant reproductive success, and when plants are closely spaced paternity is generally assigned to nearest neighbors (Handel 1983, 1985; Levin 1984, 1988; Willson 1984; Ellstrand and Marshall 1985).
Broadcast-spawning in marine systems is analogous to wind-pollination in terrestrial plant systems, where male reproductive success is dependent on the amount of pollen released, the prevailing winds, and the prox- imity of compatible plants (Whitehead 1983). Due to greater dispersion distance and pollen longevity, fertilization may be less frequently pollen limited in wind-pollinated plants than it is among sessile marine invertebrates. Similar factors aect fertilization success in both systems, however, and individual male success among both marine invertebrates and plants may be independent of overall fertilization/pollination rates. In both systems clonal propagation has important, al- though sometimes opposite, eects on male reproductive success. Among clonal, self-incompatible plants, large clone size reduces reproductive success as gametes are less likely to be transported either out of or into the center of the clone (Handel 1985). As with clonal benthic invertebrates, if plant clonemates are dispersed over a wide area, then the probability of fertilization increases for some clonemates. In a study of white spruce,Picea glauca, Schoen and Stewart (1986) found that wide- spread clones contributed to the majority of the pater- nities (60%), while clones with restricted distributions accounted for only 4% of the paternity. If fertilization is dependent on being near a compatible mate, then dis- persal of clonemates over a wide area increases the probability of fertilization for some clonemates. As in our study, clonality gave the clone a limited ``mobility'' and increased access to mates.
Male reproductive success among broadcast-spawn- ing taxa is subject to two strong selective forces. First, fertilization rates of some broadcast-spawning species are sperm-limited (Lasker et al. 1996). In this case, there will be strong selection on males for enhanced fertiliza- tion rates, i.e., fecundity selection in the terminology of Arnold (1994). Second, regardless of overall fertilization rates, there may be competition among males for mating success, i.e., sexual selection. While some traits may be subject to positive fecundity selection and positive sexual selection, others may be more strongly aected by one mode of selection, and still other traits may be subject to opposing selective forces.
In marine systems, numerous studies have shown that fertilization and reproductive success depend upon sperm density and thus the proximity of males and fe- males (e.g. Pennington 1985; Denny 1988; Denny and Shibata 1989; Yund 1990; Grosberg 1991; Levitan 1991;
Brazeau and Lasker 1992; Levitan et al. 1992; Yund and McCartney 1994; Levitan and Young 1995; Coma and Lasker 1997a, b; present study). Given that constraint, clonal propagation enhances male reproductive success in two ways. First, if colonies are not widely separated then large clone size can increase sperm concentration by increasing the total number of sperm released. Sec- ond, clonal spread brings males and females closer to- gether and allows egg and sperm to interact before sperm concentrations decrease. This latter eect was the major determinate of male reproductive success at our two study sites. Although clonality is generally thought of as a growth mechanism and a means of spreading the risk of genet mortality, the present study raises the possibility that clonality can enhance overall fertilization success and should be subject to fecundity selection.
Clonality, spawning synchrony, reproductive eort, and gamete traits such as sperm swimming speed, egg size, and egg receptivity may also enable speci®c males to sire greater proportions of zygotes. Comparisons of the ef- fects of these traits on fertilization in sperm-limited and non-sperm-limited species should enable us to evaluate the roles of fecundity and sexual selection in dierent broadcast-spawning taxa.
Acknowledgements We thank the sta and scientists of the Smithsonian Tropical Research Institute for logistical support, and we are grateful to the Republic of Panama and the Kuna Indians for permission to work in the San Blas Islands. We thank R. Tapia for facilitating our work on Smithsoniantupo, and E. Beiring, R. Coma, T. Goulet, K. Kim, K. Kittle, T. Ross and T. Swain for their assistance with ®eld collections. G. Goldberg, H. Chuang, J. Feng, S. Hurley-Leslie, A. Pai, and T. Goulet assisted in the laboratory. The manuscript was improved by the comments of E. Beiring, R. Grosberg, M. Webster and an anonymous reviewer.
This work was conducted with the support of the National Science Foundation (OCE-92 17014), the National Geographic Society and the Howard Hughes Undergraduate Initiative at the University at Bualo.
References
Arnold SJ (1994) Bateman's principles and the measurement of sexual selection in plants and animals. Am Nat 14: S126±149 Babcock RC, Mundy CN (1992) Reproductive biology, spawning
and ®eld fertilization rates of Acanthaster planci. Aust J mar Freshwat Res 43: 525±534
Babcock RC, Mundy CN, Kessing J, Oliver J (1992) Predictable and unpredictable spawning events: in situ behavioral data from free spawning coral reef invertebrates. Invert Reprod Dev 22: 213±228
Benzie JAH, Dixon P (1994) The eects of sperm concentration, sperm:egg ratio, and gamete age on fertilization success in crown-of-thorns star®sh Acanthaster planciin the laboratory.
Biol Bull mar biol Lab, Woods Hole 186: 139±152
Brazeau DA, Lasker HR (1989) The reproductive cycle and spawning in a Caribbean gorgonian. Biol Bull mar biol Lab, Woods Hole 176: 1±7
Brazeau DA, Lasker HR (1992) Reproductive success in the Ca- ribbean octocoralBriareum asbestinum. Mar Biol 114: 157±163 Coroth MA, Lasker HR (1998) Population structure of a clonal gorgonisn coral: the interplay between clonal reproduction and disturbance. Evolution 52: 316±330
Coroth MA, Lasker HR, Diamond E, Bruenn JA, Bermingham E (1992) DNA ®ngerprinting of a gorgonian coral: a method for detecting clonal structure in a vegetative species. Mar Biol 114: 317±325
Coroth MA, Mulawka JM (1995) Identi®cation of marine invertebrate larvae using PCR-RAPD species speci®c markers.
Limnol Oceanogr 40: 181±189
Coma R, Lasker HR (1997a) Small scale heterogeneity of fertili- zation success in a broadcast spawning octocoral. J exp mar Biol Ecol 214: 107±120
Coma R, Lasker HR (1997b) Eects of spatial distribution and reproductive biology on in situ fertilization rates of a broadcast- spawning invertebrate. Biol Bull mar biol Lab, Woods Hole 193: 20±29
Denny MW (1988) Biology and the mechanisms of the wave-swept environment. Princeton University Press, Princeton, New Jersey Denny MW, Shibata MF (1989) Consequences of surf-zone tur- bulence for settlement and external fertilization. Am Nat 134:
859±889
Ellstrand NC, Marshall DL (1985) Interpopulation gene ¯ow by pollen in wild radish,Raphanus sativus. Am Nat 126: 606±616 Grosberg RK (1991) Sperm-mediated gene ¯ow and the genetic
structure of a population of the colonial ascidian Botryllus schlosseri. Evolution 45: 130±142
Hadrys H, Balick M, Schierwater B (1992) Applications of random ampli®ed polymorphic DNA (RAPD) in molecular ecology.
Molec Ecol 1: 55±63
Handel SN (1983) Pollination ecology, plant population structure and gene ¯ow. In: Real L (ed) Pollination biology. Academic Press, New York, pp 163±211
Handel SN (1985) The intrusion of clonal growth patterns on plant breeding systems. Am Nat 125: 367±384
Lasker HR (1984) Asexual reproduction, fragmentation, and skeletal morphology of a plexaurid gorgonian. Mar Ecol Prog Ser 19: 261±268
Lasker HR (1990) Clonal propagation and population dynamics of a gorgonian coral. Ecology 71: 1578±1589
Lasker HR, Brazeau DA, Calderon J, Coroth MA, Coma RC, Kim K (1996) In situ rates of fertilization among broadcast spawning gorgonians. Biol Bull mar biol Lab, Woods Hole 190: 45±55 Lasker HR, Kapela WJ (1997) Heterogeneous water ¯ow and its
eect on the mixing and transport of gametes. In: Lessios HA, MacIntyre IG (eds) Proc 8th int coral Reef Symp. Vol. 2.
Smithsonian Tropical Research Institute, Balboa, Panama, pp 1109±1114
Lasker HR, Kim K (1996) Larval development and settlement behavior of the gorgonian coralPlexaura kuna. J exp mar Biol Ecol 207: 161±175
Lasker HR, Kim K, Coroth MA (1996) Reproductive and genetic variation among Caribbean gorgonians: the dierentiation of Plexaura kuna, new species. Bull mar Sci 58: 277±288
Levin DA (1984) Immigration in plants: an exercise in the sub- junctive. In: Dirzo R, Sarukhan J (eds) Perspectives on plant population ecology. Sinauer Associates, Sunderland, Massa- chusetts, pp 242±260
Levin DA (1988) The paternity pools of plants. Am Nat 132: 309±
Levitan DR (1991) In¯uence of body size and population density317 on fertilization success and reproductive output in a free- spawning invertebrate. Biol Bull mar biol Lab, Woods Hole 181: 261±268
Levitan DR (1993) The importance of sperm limitations to the evolution of egg size in marine invertebrates. Am Nat 141: 517±
Levitan DR (1995) The ecology of fertilization in free-spawning535 invertebrates. In: McEdwards L (ed) Ecology of marine inver- tebrate larvae. CRC Press, Inc, Boca Raton, Florida, pp 123±156
Levitan DR (1996) Eects of gamete traits on fertilization in the sea and the evolution of sexual dimorphism. Nature, Lond 382:
153±155
Levitan DR, Grosberg RK (1993) The analysis of paternity and maternity in the marine hydrozoan Hydractinia symbiolongi- carpususing randomly ampli®ed polymorphic DNA (RAPD) markers. Molec Ecol 2: 315±328
Levitan DR, Petersen C (1995) Fertilization in the sea. Trends Ecol Evolut 10: 228±231
Levitan DR, Sewell MA, Chia F (1991) Kinetics of fertilization in the sea urchin Strongylocentrotus franciscanus: interaction of gamete dilution, age and contact time. Biol Bull mar biol Lab, Woods Hole 181: 371±378
Levitan DR, Sewell MA, Chia F (1992) How distribution and abundance in¯uence success in the sea urchin Strongylo- centrotus franciscanus. Ecology 73: 248±254
Levitan DR, Young CM (1995) Reproductive success in large populations: empirical measures and theoretical predictions of fertilization in the sea biscuit Clypeaster rosaceus. J exp mar Biol Ecol 190: 221±241
McCartney MA (1997) Sex allocation and male ®tness gain in a colonial, hermaphroditic marine invertebrate. Evolution 51:
127±140
Mead KS, Denny MW (1995) The eects of hydrodynamic shear stress on fertilization and early development of the purple sea urchin Strongylocentrotus purpuratus. Biol Bull mar biol Lab, Woods Hole 188: 46±56
Oliver J, Babcock R (1992) Aspects of the fertilization ecology of broadcast spawning corals: sperm dilution eects and in situ measurements of fertilization. Biol Bull mar biol Lab, Woods Hole 183: 409±417
Pennington JT (1985) The ecology of fertilization of echinoid eggs:
the consequences of sperm dilution, adult aggregation and synchronous spawning. Biol Bull mar biol Lab, Woods Hole 169: 417±430
Petersen CW (1991) Variation in fertilization rate in the tropical reef ®sh, Halichoeres bivattatus: correlates and implications.
Biol Bull mar biol Lab, Woods Hole 181: 232±237
Petersen CW, Warner RR, Cohen S, Hess HC, Sewell AT (1992) Variable pelagic fertilization success: implications for mate choice and spatial patterns of mating. Ecology 73: 391±401 Podolosky RD, Strathmann RR (1996) Evolution of egg size in
free-spawners: consequences of the fertilization±fecundity trade-o. Am Nat 148: 160±173
Robertson DR (1987) Responses of two coral reef toad®shes (Batrachoididae) to the demise of their primary prey, the sea urchin,Diadema antillarum. Copeia 3: 637±642
Schoen DJ, Stewart SC (1986) Variation in male reproductive investment and male reproductive success in white spruce.
Evolution 40: 1109±1120
Sewell MA, Levitan DR (1992) Fertilization success during a nat- ural spawning of the dendrochirote sea cucumber Cucumaria miniata. Bull mar Sci 51: 161±166
Stephenson AG, Bertin RI (1983) Male competition, female choice, and sexual selection in plants. In: Real L (ed) Pollination bi- ology. Academic Press, Orlando, Florida, pp 110±151 Vogel HG, Czihak P, Chang P, Wolf W (1982) Fertilization ki-
netics of sea urchin eggs. Math Biosci 58: 189±216
Whitehead DR (1983) Wind pollination: some ecological and evolutionary perspectives In: Real L (ed) Pollination biology.
Academic Press, Orlando, Florida, pp 97±109
Willson MF (1979) Sexual selection in plants. Am Nat 113: 777±
Willson MF (1984) Mating patterns in plants. In: Dirzo J, Saru-790 khan J (eds) Perspectives on plant ecology. Sinauer Associates, Sunderland, Massachusetts, pp 261±276
Willson MF, Burley N (1983) Mate choice in plants. Princeton University Press, Princeton, New Jersey
Yund PO (1990) An in situ measurement of sperm dispersal in a colonial marine hydroid. J exp Zool 253: 102±106
Yund PO (1995) Gene ¯ow via dispersal of fertilizing sperm in a colonial ascidian (Botryllus schlosseri): the eect of male den- sity. Mar Biol 122: 649±654
Yund PO, McCartney MA (1994) Male reproductive success in sessile invertebrates: competition for fertilizations. Ecology 75:
2168±2184
